Haywood County Landslide Data

5/9/2018 Haywood County Landslide Data - slidepdf.com

http://slidepdf.com/reader/full/haywood-county-landslide-data 1/178

ABSTRACT

Witt, Anne Carter. Using a GIS (Geographic Information System) to model slopeinstability and debris flow hazards in the French Broad River watershed, North Carolina.

(Under the direction of Dr. Michael M. Kimberley.)

Catastrophic, storm-generated mass wasting is a destructive erosional process in the

portion of the southern Appalachians that extends through western North Carolina. Steep slopes,

a thin soil mantle, and extreme precipitation events all increase the risk of slope instability, slope

movement and failure. Since the late 1800’s, several intense storms and hurricanes have tracked

through the French Broad watershed initiating thousands of debris flows and causing severe

flooding. Studying the history of debris flows has identified triggering mechanisms that are

particular to North Carolina and the recurrence interval of these events.

This study was initiated to investigate and predict the spatial distribution of

regional slope instability within the French Broad watershed by comparing the results of

two GIS-based modeling applications: SINMAP (Stability Index Mapping) and

SHALSTAB (Shallow Landsliding Stability Model). As extensions to ArcView® 3.x,

SINMAP and SHALSTAB use a modified form of the infinite slope equation to compute

and map slope-instability by calculating either a factor of safety (SINMAP) or the critical

steady-state rainfall intensity necessary to trigger slope instability (SHALSTAB). In both

models, topographic slope is derived from digital-elevation data while parameters for soil

and climate are considered more variable and can be adjusted to better match existing

conditions. An inventory of actual debris flow locations, collected from aerial



using the program’s default parameters were compared with those for four recharge

events (50, 125, 250, and 375 mm/d). In the latter, parameters for soil density, cohesion,

internal soil friction angle, and transmissivity were adjusted to better match existing

watershed conditions.

As with the SINMAP model, SHALSTAB was used to model instability using

both a 10-meter and 30-meter DEM. Limitations in the SHALSTAB program only allow

smaller (county-size) DEMs to be processed. Because of these limitations Haywood

County was chosen for several model runs in SHALSTAB for comparison to the

SINMAP results. Parameters for soil density, soil depth, cohesion, and soil friction angle

were adjusted and results were compared to 23 mapped debris flow locations.

The modeled results for the default SINMAP and SHALSTAB parameter values

underestimate the extent of instability in the study area. By adjusting soil parameters,

SINMAP calculated 88% -to- 94% of the inventoried landslides would fall into the

“lower threshold”, “upper threshold”, and “defended” stability classes. Generally,

predicted areas of unstable land did not change, even as recharge increased.

SHALSTAB calculated 91% to 100% of the mapped debris flows would occur in the

three most unstable stability classes for low values of soil friction angle (26°) and

cohesion (0). Overall, SHALSTAB seems to over-predict areas of instability for these

values. An increase in cohesion or soil friction angle decreases the amount of land

predicted as unstable (88%) but increases the landslide density. When the results of the



groundwater flow in the watershed and failure tends to occur along these planes of

weakness. Neither model takes into account either antecedent moisture or the effect that

geologic structure can have on concentrating groundwater flow.



USING A GIS (GEOGRAPHIC INFORMATION SYSTEM) TO MODEL SLOPE

INSTABILITY AND DEBRIS FLOW HAZARDS IN THE FRENCH BROAD

RIVER WATERSHED, NORTH CAROLINA

by

ANNE CARTER WITT

A thesis submitted to the Graduate Faculty of North Carolina State University

in partial fulfillment of the

requirements for the Degree of

Master of Science

MARINE, EARTH AND ATMOSPHERIC SCIENCES

Raleigh

2005

APPROVED BY:

Chair of Advisory Committee ______________________________________ Dr. Michael M. Kimberley

Committee Member ______________________________________

Dr. Jeffery C. Reid



BIOGRAPHY

Anne Carter Witt was born in Lynchburg, VA on August 10, 1977. She grew

up in Forest, VA and graduated from Brookville High School in 1995. In 1999, she

graduated from Mary Washington College in Fredericksburg, VA with an undergraduate

degree in Geology. Before being accepted into the Masters degree program at North

Carolina State University, she worked for three years as a GIS analyst with Dewberry and

Davis, LLC in Fairfax, VA.



ACKNOWLEDGEMENTS

I would like to express sincere thanks and appreciation to the following

individuals: Rick Wooten with the North Carolina Geological Survey who provided me

with landslide data and an open discussion on the complexities of SINMAP. His

guidance, patience and knowledge of slope movements and the geomorphology of

western North Carolina are greatly appreciated. Thanks are also extended to Jody Kuhne

from the North Carolina Department of Transportation whose tattered, long-forgotten

landslide map was really the foundation for a digital landslide inventory in North

Carolina.

In addition, I would like to thank Drs. Helena Mitasova and Lonnie Liethold for

their advice and counsel is serving as members on my committee. I would also like to

thank Dr. Mary Schweitzer for being a last minute substitution during my thesis defense

on one of the busiest days of her life. Many thanks are also extended to Dr. Jeff Reid for

his time, guidance, critical comments, and dogged determination to see me finish. Thanks

are also extended to Dr. Michael Kimberley for his editorial expertise, assistance, and

support. He generously donated both a computer and his own office space to help with

my research.

I would also like to express my sincere gratitude to all of my family and friends.

Thanks go out to my fellow graduate students for their sometimes backwards but well-

meaning encouragement and enlightened conversations. I also would like to acknowledge



best friend, Jamie Gibson. Finally, I would like to thank my parents and brother who

provided the foundation for me to be the best that I could be and instilled a lifelong love

of learning. Thank you for your love and support.



TABLE OF CONTENTS

Page

LIST OF TABLES .......................................................................................................... vii

LIST OF FIGURES ....................................................................................................... viii

CHAPTER 1: INTRODUCTION.................................................................................... 1

1.1 R ESEARCH AIM AND OBJECTIVES ........................................................................ 3

1.2 SIGNIFICANCE AND SCIENTIFIC APPLICATIONS .................................................... 41.3 LANDSLIDE CLASSIFICATION ............................................................................... 6

1.3.1 Classification of Sharpe (1938) ...................................................................... 7 1.3.2 Classification of Varnes (1978) and of Cruden and Varnes (1996) ............... 8

1.4 COMMON CAUSES OF SLOPE MOVEMENTS .......................................................... 91.4.1 Precipitation ................................................................................................. 10 1.4.2 Human Interference ...................................................................................... 10

1.4.3 Tectonic Activity............................................................................................ 11 1.4.4 Geologic Material and Structures ................................................................ 11

1.5 DEBRIS FLOWS................................................................................................... 12

1.6 DEBRIS FLOWS WITHIN THE BLUE R IDGE AND WESTERN NORTH CAROLINA ..... 14

1.7 SINMAP AND SHALSTAB.............................................................................. 161.8 LIMITATIONS OF R ESEARCH............................................................................... 18

CHAPTER 2: PROJECT SETTING ............................................................................ 21

2.1 I NTRODUCTION .................................................................................................. 21

2.2 GEOLOGY........................................................................................................... 212.3 SOILS ................................................................................................................. 23

2.4 CLIMATE ............................................................................................................ 252.5 VEGETATION...................................................................................................... 26

CHAPTER 3: PRE-HISTORIC AND HISTORIC DEBRIS FLOWS IN WESTERN

NORTH CAROLINA..................................................................................................... 33

3.1 QUATERNARY DEBRIS FLOWS ........................................................................... 33

3.2 MODERN FLOODING AND DEBRIS FLOWS .......................................................... 373.2.1 June, 1876..................................................................................................... 38 3.2.2 May, 1901 ..................................................................................................... 39

3.2.3 July, 1916...................................................................................................... 40 3.2.4 August, 1940 ................................................................................................. 42 3 2 5 N b 1977 44



4.2.2 Debris Flow Inventory.................................................................................. 68

4.2.3 Soil Data ....................................................................................................... 68

4.2.4 Soil Density ................................................................................................... 69 4.3 SINMAP PARAMETERS ..................................................................................... 69

4.3.1 T/R (Ratio of Transmissivity to Effective Recharge)..................................... 71

4.3.2 Dimensionless Cohesion ............................................................................... 72 4.3.3 Internal Soil Friction Angle.......................................................................... 72

4.4 SHALSTAB PARAMETERS ............................................................................... 73

CHAPTER 5: RESULTS AND DISCUSSION ............................................................ 80

5.1 I NTRODUCTION .................................................................................................. 805.2 DEBRIS FLOW I NVENTORY................................................................................. 80

5.3 SINMAP R ESULTS ............................................................................................ 81

5.3.1 Results – 30-meter DEM............................................................................... 83

5.3.2 Results – 10-meter DEM............................................................................... 84 5.3.3 SINMAP Interpretation................................................................................. 85

5.4 SHALSTAB R ESULTS....................................................................................... 87

5.4.1 Results – 30-meter DEM............................................................................... 89

5.4.2 Results – 10-meter DEM............................................................................... 90 5.4.3 SHALSTAB Interpretation ............................................................................ 92

5.5 SINMAP VS. SHALSTAB................................................................................ 935.6 GEOLOGY AND SOILS ......................................................................................... 94

5.7 JOINTING, FRACTURING AND FOLIATION ........................................................... 97

CHAPTER 6: CONCLUSIONS .................................................................................. 116

REFERENCES.............................................................................................................. 122

APPENDICES............................................................................................................... 133

APPENDIX A: GEOLOGIC U NITS ................................................................................... 134APPENDIX B: GENERAL SOIL DATA............................................................................. 139

APPENDIX C: SLOPE MOVEMENT DATA FOR THE 2004 HURRICANES FRANCES AND IVAN

..................................................................................................................................... 148

APPENDIX D: DEBRIS FLOW I NVENTORY..................................................................... 152

APPENDIX E: SINMAP R ESULTS................................................................................. 157APPENDIX F: SHALSTAB R ESULTS ........................................................................... 161



LIST OF TABLES

Page

Table 2.1: Table of the average, median, minimum, and maximum precipitation totals

(mm/month) from 1895 to 2001 for the mountains of North Carolina

(NCDC Climate Data Online, 2003)................................................................31Table 3.1: Prehistoric debris flow studies in the southern Blue Ridge and the age-dating

techniques utilized. ..........................................................................................53

Table 3.2: Major storms within the French Broad Watershed and their minimum,average, and maximum precipitation amounts. ...............................................63

Table 4.1: Table of the hydraulic conductivity (K, m/hr), transmissivity (T, m²/hr),and T/R (m) values used for each precipitation threshold (50 mm/d, 125mm/d, 250 mm/d, and 375 mm/d) in the SINMAP analysis. The numbers

in blue are the lower bound values while the numbers in red are the upper

bound values. ...................................................................................................78Table 5.1: The parameters used in all of the SINMAP model runs...................................99

Table 5.2: SINMAP stability index definitions (Pack et al., 1998b). ..............................100

Table 5.3: Mapped instability classes used in the SHALSTAB model analysis. ............104

Table 5.4: Table comparing q/T and log (q/T) values and the precipitation raterequired to initiate instability for soils with a transmissivity of 65 m²d and

17 m²/d (after Dietrich and Asua, 1998). .......................................................105

Table 5.5: Parameters used in the SHALSTAB model runs............................................105Table 6.1: The advantages and disadvantages of SINMAP and SHALSTAB. ...............121



LIST OF FIGURES

Page

Figure 1.1: The morphology of a typical debris flow found in the Southern Appalachians

(courtesy of the North Carolina Geological Survey, 2003). ............................19Figure 1.2: Threshold precipitation values necessary for producing debris flows in the

southern Appalachian Mountains. Storms likely to start debris flows occur

above the 125 mm/d threshold. Storms with precipitation values higher than250 mm/d are deemed “rare” but do occur in North Carolina (after Eschner

and Patric, 1982). .............................................................................................20Figure 2.1: Location Map of the French Broad Watershed in western North Carolina. ...27Figure 2.2: General geologic map for the French Broad Watershed. Individual geologic

unit descriptions can be found in Appendix A (adapted from North Carolina

Geological Survey, 1985). ...............................................................................28Figure 2.3: General soil map for the French Broad Watershed. Individual soil descriptions

can be found in Appendix B (adapted from U.S. Department of Agriculture,

1998). ...............................................................................................................29

Figure 2.4: The U.S. Department of Agriculture guide for the textural classification of soils. This guide is only for soils with a particle size of less than 2 mm in

diameter. A rock fragment modifier (gravelly, cobbly, stony, bouldery)

prefaces the textural name if particles larger than 2mm compose more than15% of the soil (Buol et al., 2003)...................................................................30

Figure 2.5: Graph based on the data from Table 2.1 (NCDC Climate Data Online,

2003). ..............................................................................................................31Figure 2.6: Average annual precipitation in inches within the French Broad Watershed.

(Adapted from data provided by North Carolina Center for GeographicInformation and Analysis map server (http://204.211.135.111)). ...................32

Figure 3.1: Old slope movement deposit along a private drive in Maggie Valley, North

Carolina (age not determined). There is significant soil development in this poorly sorted colluvial deposit. It is located along a chute where a modern

debris flow occurred. .......................................................................................52

Figure 3.2: Areas of major debris flows and landslides in western North Carolina (after

Scott, 1972)......................................................................................................54Figure 3.3: A sketch map of debris flows that occurred along Gouges Creek in Mitchell

County, North Carolina in May, 1901 (Myers, 1902). ....................................55

Figure 3.4: Map showing some of the hurricane and storm paths that have affectedwestern North Carolina as reported by the U.S. National Hurricane Center and

the U.S. Geological Survey – Water Resources Branch (1949). .....................56



moving northwestward over the watershed. This is the greatest recorded

streamflow for the gauge at Asheville. ............................................................58

Figure 3.7: Total storm precipitation for August 14-15, 1940 adapted from U. S.Geological Survey, 1949). ...............................................................................59

Figure 3.8: Total storm precipitation for August 28-31, 1940 (adapted from U. S.

Geological Survey, 1949). ...............................................................................59Figure 3.9: Total storm precipitation for November 2-5, 1977 (adapted from Neary and

Swift, 1987)......................................................................................................60

Figure 3.10: Total storm precipitation (inches) for the remnants of Hurricane Frances(National Weather Service, 2004b)..................................................................61

Figure 3.11: Total storm precipitation (inches) for the remnants of Hurricane Ivan(National Weather Service, 2004c)..................................................................61

Figure 3.12: A debris flow that blocked the westbound lanes of Interstate-40 near Old

Fort Mountain in McDowell County (North Carolina Geological Survey,

2004a). .............................................................................................................62Figure 3.13: The initiation zone of the debris flow that occurred on Fishhawk Mountain

and devastated the Peeks Creek area of Macon County on September 17, 2004

(Wilett, 2004)...................................................................................................62

Figure 4.1: The infinite slope equation as defined by Hammond et al., (1992) and Pack etal., (1998b) where C r is root cohesion, C s is soil cohesion, θ is slope angle, ρ s

is soil density, ρ w is the density of water, g is acceleration due to gravity, D is

the vertical soil depth, Dw is the vertical height of the water table, and Φ is the

internal soil friction angle. In the SINMAP model, the ratio of the vertical soildepth to the vertical soil height is simplified so that depth is measured

perpendicular to the slope (h). (Diagram after Hammond et al., 1992 and

Otteman, 2001) ................................................................................................76Figure 4.2: Default parameters used in the SINMAP model analysis. The values for the

gravitational constant and the density of water were not adjusted in thisstudy.................................................................................................................77

Figure 4.3: Default values used for SHALSTAB..............................................................79

Figure 5.1: Landslide inventory map for the French Broad Watershed. The cluster of location points in southern Buncombe County is due to the extensive mapping

of 1977 debris flows done by Pomeroy (1991) and Otteman (2001). .............98

Figure 5.2: A comparison of 30-meter and 10-meter DEMs in Haywood County, NorthCarolina. In the coarser 30-meter DEM, there is a great deal of pixelation.Lower resolution can lead to an underestimation of both the slope and

instability in the study area. .............................................................................99

Figure 5.3: SINMAP results for a 30-meter DEM and using default parameters............101Figure 5.4: SINMAP results for 125 mm/d recharge and using a 30-meter DEM. .........102



In the legend, the first number is the degree if soil friction angle and the

second number is the amount of cohesion (N/m³). ........................................107

Figure 5.9: Cumulative percent of debris flows for each log (q/T ) instability category for a variety of soil parameters for a 30-meter DEM (after Dietrich et al.,

2001). .............................................................................................................107

Figure 5.10: Cumulative percent of the area if Haywood County for each log ( q/T )instability category for a variety of soil parameters for the 10-meter DEM

(after Dietrich et al., 2001).............................................................................108

Figure 5.11: Cumulative percent of debris flows for each log (q/T ) instability category for a variety of soil parameters for a 10-meter DEM (after Dietrich et al.,

2001). .............................................................................................................108Figure 5.12: Mapped log (q/T ) results for a 10-meter DEM of Haywood County. The

parameters used for this model run are essentially that same as those used in

the SINMAP lower bounds: 26° soil friction angle, and zero cohesion. .......109

Figure 5.13: SHALSTAB results for a soil friction angle of 35° and cohesion of 2000 N/m³. ..............................................................................................................110

Figure 5.14: Comparison of the output of SINMAP and SHALSTAB for 125 mm of

recharge, 35° soil friction angle, 1922 kg/m³ soil density, and zero soil

cohesion for a location in Haywood County. The red areas are calculated asunstable by both programs whereas the grey areas are calculated to be stable.

Visually the SHALSTAB results seem to cluster better while the SINMAP

results are more scattered. Overall, the results are very similar. ..................111Figure 5.15: Results for SINMAP and SHALSTAB. Red areas are the areas calculated by

both models as being unstable. The purple areas were only calculated by

SHALSTAB and the green areas were only calculated by SINMAP............112Figure 5.16: The mean stability index for each geologic unit in the French Broad

Watershed. The most unstable unit, Zchs, is located in the northwestern portion of Haywood County. The most stable units, bz and Ctzp, are located in

the southeastern portion of the study area......................................................113

Figure 5.17: The mean stability index for each soil unit in the French Broad Watershed.The most unstable soil unit in the watershed is NC104, which is located in

western Haywood County. The most stable soils are located around the

floodplains of the Pigeon and French Broad Rivers. .....................................114

Figure 5.18: Picture taken in October, 2003 along the Blue Ridge Parkway, near Mt.Mitchell. Even during this light rain event, water is pouring out from fractures

in the rock. .....................................................................................................115



CHAPTER 1: INTRODUCTION

Catastrophic, storm-generated mass wasting is a destructive erosional process in

the portion of the Southern Appalachians that extends through western North Carolina.

This study was initiated to investigate and predict the spatial distribution of regional

slope instability within the French Broad watershed by comparing the results of two GIS-

based modeling applications: SINMAP (Stability Index Mapping) (Pack et al., 1998b)

and SHALSTAB (Shallow Landsliding Stability Model); (Dietrich and Montgomery,

1998).

Mass wasting is a general term used to describe the “dislodgement and downslope

transport of soil and rock material under the direct application of gravitational body

stresses” (Jackson, 1997). Mass wasting can range from very slow creep to a rock

avalanche. The term is often synonymous with the term “mass movement” (Crozier,

1986). Both Crozier (1986) and Varnes (1978) advocate the use of the term “slope

movements” for mass movement restricted to slopes, as “slope movement” appears to be

a suitably neutral, all-encompassing term. Throughout this paper the term “slope

movement” will be used interchangeable with the terms “mass wasting” and “landslide”

to describe the movement of a mass of rock, soil and debris downslope. Of the several

types of slope movements that occur in the Appalachians, rapid mass movement,

particularly debris flows, are considered the most dangerous and will be the focus of this



early 1900’s, several intense storms and hurricanes have tracked through western North

Carolina, initiating hundreds of debris flows and causing severe flooding. In the

Appalachian Mountain chain, it has been estimated that as many as 1,700 debris flows

occurred in the 20th century, killing at least 200 people and destroying thousands of acres

of farm and forested land (Scott, 1972).

As extensions to ArcView®

3.x GIS software, both SINMAP and SHALSTAB

compute and map areas of potential slope instability based upon digital-elevation data

and observed landslide locations. These models combine steady-state hydrologic

concepts and an infinite-slope-stability analysis with a digital elevation model (DEM) to

calculate either a factor of safety (SINMAP) or the critical steady-state rainfall intensity

necessary to trigger slope instability (SHALSTAB).

As in any landslide investigation using modeling software, model results should

be compared with mapped landslide locations whenever possible. In this study,

preliminary field investigations were completed in three key locations in Haywood,

Yancey, and Transylvania Counties, initially modeled as having a high probability for

slope instability and failure. Aerial photography was used to search for debris-flow scars,

in the development of a landslide database for the region, and to complete more-detailed

SINMAP and SHALSTAB model runs for those locations. During field work in these

areas, soil and geologic data was collected for each location; debris-flow locations were

more precisely located with a hand-held global positioning system, and the overall



1.1 Research Aim and Objectives

Few comprehensive research studies comparing slope-stability-modeling software

have been completed within a specific drainage basin in North Carolina. Most of these

studies have tended to focus on identifying landslide hazard areas on the U.S. West Coast

or in other countries (Appt et al., 2002; Dietrich et al., 2001; Montgomery et al., 2001;

Zaitchick et al., 2003). Other studies have focused on identifying relict landforms and

debris aprons in the Appalachians formed during periglacial conditions (Gryta and

Bartholomew, 1987; Clark, 1987; Kochel, 1987; Jacobson et al., 1989; Liebens and

Schaetzel, 1997). This study focuses on identifying debris-flow hazard areas and the

factors that affect this instability. The objectives of this research are as follows:

1. To identify areas within the French Broad watershed prone to debris flows.

2. To identify triggering mechanisms particular to the watershed that promoteinstability and failure.

3. To determine which soil and geologic units are the most prone to instability andthose that are the most stable, in general.

4. To compare the results, reliability, and effectiveness of two widely used, andtheoretically similar, GIS-based programs that model shallow landslide potentialin a watershed in North Carolina.

5. To study the effect of parameter variation on each of the modeling programs.

As part of this study, preliminary debris-flow hazard maps have been produced,

for comparison, using SINMAP and SHALSTAB. Results from both computer programs

were compared to an inventory of recent debris-flow and landslide locations determined



1.2 Significance and Scientific Applications

The French Broad river basin is located in western North Carolina, in the central

portion of the Appalachian Mountain chain. The French Broad River itself flows through

the City of Asheville, a major commercial and manufacturing center, and a popular

mountain resort area. According to data collected by the U.S. Census Bureau (2000),

approximately 426,000 people live within the French Broad watershed and this

population is predicted to increase, particularly in and around the City of Asheville.

Debris flows hazards are a major concern in mountainous areas as debris fans are favored

areas of development due to their flat building surface and location above the floodplain

(Ritter et al., 1995). With continued development and tourism in the forested areas of the

Blue Ridge, the risk to people and property will increase because of debris flows,

especially during periods of high precipitation.

The major hazard to human life and property from debris flows is from burial or

impact by boulders and other debris. Usually starting on steep hillsides, debris flows can

accelerate to speeds between 15-55 kph (10-35 mph) (Highland et al., 2004) and often

strike without warning. Because of their relatively high density and viscosity, debris

flows can move and even carry away vehicles, bridges and other large objects (Cruden

and Varnes, 1996).

The economic costs of slope movements are can be both “event-related” and

“preventative” (Crozier, 1986). Event-related costs may involve the removal of debris



mapping, and laboratory testing. After an initial analysis has been completed, policies for

the establishment, management, and inspection of preventative measures should be

completed (Crozier, 1986). This may also include the restriction of development in areas

considered landslide-prone or the removal or conversion of existing developments.

Control measures in prone areas may also be designed and implemented including

controlled drainage, planting, slope-geometry modification, and structures such as rock

fences or other barriers (Schuster and Kockelman, 1996).

Continued poor land-management practices and deforestation add to the risk of

soil mass movement by increasing runoff, erosion, and flooding. Human activity also

disturbs large volumes of geologic materials with the construction of housing

developments, commercial buildings, mines and quarries, dams and reservoirs, and

particularly the emplacement of transportation systems along steep slopes (Schuster,

1996). Roadcuts and other altered or excavated areas along slopes are particularly

susceptible to debris flows. Repeated landslides and rock falls have plagued the Interstate

40 corridor as it winds through the mountains of Western North Carolina and the Pigeon

River Gorge (Leith et al., 1965; Glass, 1981; Kaya, 1991; Wieczorek, 1996).

By combining a theoretical model of slope instability with actual debris-flow

locations, landslide hazard areas may be efficiently identified and mapped for regulatory

purposes. Identifying areas already at risk for slope failure could prevent loss of property

and lives. This would be a valuable resource to government agencies, as well as city and



download from the World Wide Web at no cost, making them economical for both

government and businesses.

Debris flows are not just an agent of destruction, but also play a critical role in the

processes of erosion and sediment transport in the Blue Ridge Mountains. They provide a

major means of removing weathered material from steep areas that normally experience

little concentrated surface drainage (Scott, 1972). By studying the causes, characteristics,

and effects of debris flows in the Southern Appalachians, we can better understand their

erosional importance, as well as their destructive influences. Improved knowledge of the

Blue Ridge may also be extended to comparable sub-tropical mountainous areas in other

parts of the world.

1.3 Landslide Classification

There have been several attempts to create a suitable landslide classification

system that is useful for both scientific and engineering purposes. Early attempts

occurred in Europe during the late 1800’s and early 1900’s (Sharpe, 1938). These

schemes tended to be incomplete; in that slow movements such as creep were ignored,

they were developed for a single specific purpose or were developed based on only

regional observations. The more successful classification systems are those that provide

clear and unambiguous terminology that can be applied in a number of different

situations (Crozier, 1986). Generally, these systems classify landslides on the basis of the

kind and quality of material moved, its water content, and the type, size, and rate of



in Denver, CO. But this classification system, like many before it, was influenced by the

classification system proposed by C.F.S. Sharpe (1938).

1.3.1 Classification of Sharpe (1938)

In the late 1930’s, C. F. S. Sharpe became one of the first American scientists to

create a widely accepted landslide classification system. Sharpe created his classification

based primarily on the kind and rate of movement (making a distinction between slides

and flows) and the forms of the resulting deposits. Other important factors included the

moisture content in the moving mass and the type of material involved. Based on these

factors, mass movement was separated into four groups, i.e. slow flowage, rapid flowage,

sliding, and subsidence.

Sharpe (1938) defined the mass-wasting process that typically occurs in the

Southern Appalachians as debris-avalanche, a type of rapid flowage. A debris avalanche

is a rapidly-moving, sliding flow with a long, narrow track that occurs on steep terrain in

humid mountainous areas with significant vegetative cover. This type of slope movement

tends to have higher water content than does a slow flow. They are usually preceded by a

period of heavy precipitation “which increases the weight of the unadjusted material and

aids in its lubrication” (Sharpe, 1938, p. 61). The material of a debris avalanche is

variable and can consist of soil, rock, vegetation and ice. Initial failure typically occurs

between the soil-rock interface or within loose debris on slopes between 20 and 40

degrees. When there is less water present in the same type of material, the movement



1.3.2 Classification of Varnes (1978) and of Cruden and Varnes (1996)

The landslide classification system of D. J. Varnes (1978) was well-received and

has been repeatedly updated, the latest update having been published in Cruden and

Varnes (1996). The goal of the current version of this classification was to provide

definitions and vocabulary that allow an investigator to observe and describe a landslide

in the field succinctly and unambiguously.

The Cruden and Varnes (1996) classification scheme emphasizes the type of

material and the type of movement in a slide. Two terms are needed to describe any

landslide, i.e. one that describes the material (rock, debris, earth) and one that describes

the movement (fall, topple, slide, spread, flow). Other descriptors can then be added in

front of the two-term classification as more information about the movement becomes

available. These include the water content, rate of movement, the current activity of the

slide (reactivated, inactive, etc.), distribution, and overall style (complex, composite,

etc.).

Like the classification scheme of Sharpe (1938), Cruden and Varnes (1996) uses

the terms debris avalanche, debris slide, and debris flow to describe rapidly-flowing

mass-wasting events. The terms “debris avalanche” and “debris slide” in this

classification are varieties of the more general debris flow. “Debris avalanche” is an older

term used to in the literature to describe extremely large, rapidly-moving mass

movements that involve large amounts of ice, snow, and soil. A “debris slide” generally



Driving forces cause material to move downslope and can increase depending on the

mass of the material involved in the movement and the slope angle (Easterbrook, 1999).

Frictional resistance opposes deformation or motion and is caused by the friction between

grains within the material or by the material at the base of the slope. The resistance of a

material to shear along a slip surface due to an applied external force can be referred to as

the materials shear strength. Shear strength is determined by analyzing the amount of

cohesion, effective normal stress, and internal friction between material particles (Ritter

et al., 1995). When the gravitational driving force is greater than the frictional resistance

the slope will fail (Easterbrook, 1999).

While a single trigger may actually cause a slope to fail, a number of other

destabilizing factors are often needed to contribute to a reduction in a material’s shear

strength in order to ultimately initiate movement (Crozier, 1986). First, though, the

topography must have a sufficient slope. Usually slopes of 35° or greater are prone to

instability simply because of the effects of gravity (Sidle et al., 1985). But often other

factors, such as soil and root cohesion, increase the shear strength of materials on a slope.

1.4.1 Precipitation

A number of slope movements are generated by an excessive amount of water,

usually due to heavy precipitation or a moderate rainfall lasting several days. Slope

movements in soil and weathered rock are usually produced on steep slopes during the

most intense part of a storm (Wieczorek, 1996). Intense rainfall results in a higher



material or along an existing plane of weakness. Thin, loose soils on hillslopes with

sparse vegetation are particularly prone to failure during an intense rainfall.

Soils with a high rate of hydraulic conductivity (the amount of water that will move

through a porous medium in unit time) have the ability to transmit water more quickly

downslope. This decreases soil pore pressure and may help to increase shear strength.

Typically, an unconsolidated, coarse grained, well- rounded, and well-sorted soil will

have a higher value of hydraulic conductivity (Fetter, 1994).

1.4.2 Human Interference

Road construction is a major contributor to slope failure and their mitigation can

often incur enormous public cost. Excavation of the toe of a hillslope by emplacing a

road, quarry, canal, or other type of cut, removes support and may induce anthropogenic

slope moment (Cruden and Varnes, 1996). Road fill and traffic also increases weight on a

hillslope, increasing shear stress on materials (Sidle et al., 1985). In developed areas,

slope saturation may occur, even during moderate recharge events, because of

concentrated run-off from rerouting of drainage systems during road construction and

from man-made structures such as drainpipes, buildings, and paved impervious surfaces.

Vegetation often acts to stabilize slopes and increases shear strength and root

cohesion. Foliage intercepts rainfall whereas roots and stems extract moisture from the

soil (lowering pore-water pressure), increase surface roughness (Greenway, 1987), and

provide an interlocking network that strengthens unconsolidated sediment (Easterbrook,



1.4.3 Tectonic Activity

In tectonically-active areas, volcanic eruptions and earthquakes have often caused

slope instability and failure by causing uplift or tilting (Cruden and Varnes, 1996).

Volcanic ash deposited on steep volcanic peaks, combined with the rapid melting of snow

or heavy rainfall, can initiate deadly lahars, debris flows, and mudflows. Slope

movements involving loose, saturated, low-cohesion soils commonly occur as a result of

earthquake-induced liquefaction, a process in which shaking temporarily raises pore-

water pressure and reduces shear strength. Fault zones may also contain fractured,

crushed, or low-metamorphic-grade rock that contain inherent weakness and may be

susceptible to failure (Sidle et al., 1985).

1.4.4 Geologic Material and Structures

Some particular rock and soil types may be inherently weak and can influence

slope instability. Organic soils and clays naturally have low shear strength and are

particularly prone to weathering processes (Cruden and Varnes, 1996). A predominantly

clay soil layer can prevent vertical infiltration of water and cause a buildup of pore-water

pressure, while also providing a smooth, low-friction surface for failure. In contrast,

sandy soils commonly erode quickly despite rapid infiltration and sandy slopes may be

undermined by rivers and run-off. A permeable shallow soil overlying a hard,

impermeable rock layer may also be susceptible to mass movement. As water builds up

and travels along the rock-regolith interface, cohesion between the layers diminishes and



for groundwater during rain events and reducing cohesion between layers or bedding

(Sidle et al., 1985). Structures parallel to the ground surface and downslope-dipping beds,

particularly those that separate two distinct lithologic units, may also act as a viable

failure plane.

1.5 Debris Flows

Although several types of slope movements have been described in the high-relief

portions of the French Broad watershed, this study focuses on debris flows, i.e., rapid

downslope movement of regolith. It was noted during field work that a number of the

regularly-occurring mass-wasting events were actually simple rock falls or slides along

over-steepened roadcuts. Debris flows seem to be a less common mass-wasting event

than those mentioned above, but may be devastating when they occur in populated areas.

Unfortunately, there is a considerable amount of inconsistency in terminology for

modern rapid channelized downslope movement of poorly sorted sediment. The terms,

debris torrent, debris avalanche, debris flow, debris slide, mudflow, and mud flood are

occasionally used interchangeably (Pierson and Costa, 1987). These terms all have

similar descriptive characteristics for grain-size and the rate of downslope movement, but

individually are often difficult to distinguish from one another in the field (Ritter et al.,

1995). Other authors also use the terms, alluvial fan, debris fan, hillslope, toe-slope or

foot-slope deposit to refer to relict fan-shaped features found throughout the

Appalachians produced by repeated depositional episodes by debris flow processes



The term “debris flow” is used herein to describe swift-moving mass-wasting

events that occur predominantly in shallow, silty-to-gravelly soil on steep slopes during

periods of exceptionally heavy precipitation (Cruden and Varnes, 1996) (Figure 1.1).

“Debris” defines a material that contains 20-80 percent coarse-grained particles larger

than 2mm. These materials may include boulders to clay with varying amounts of water

(Ritter et al., 1995). The “flows” begin in depressions or hollows on steep slopes and

tend to move downslope following preexisting drainage channels. The most common

movement interface is between the bedrock-soil contact, but slippage may also occur

within deep soils (Clark, 1987). Debris flows from several different sources often

converge into one main drainage channel, increasing the flow’s overall volume of water

and material (Highland et al., 2004). Debris flows can travel for several kilometers before

releasing their suspended load and coming to rest upon reaching an area of low gradient

(Ritter et al., 1995).

Debris flows may be triggered in a variety of ways. The most common trigger is

an abundant amount of moisture, either from intense rainfall or rapid snowmelt, or a

combination of heavy precipitation and antecedent soil moisture. Topography also

influences debris-flow initiation by concentrating subsurface flow and determining slope.

Soil thickness, conductivity, soil strength, bedrock-fracture flow, and root strength also

influence the spatial distribution of debris flows and other types of shallow landslides

(Montgomery and Dietrich, 1994). During a heavy rain event, piezometric head in the



laden slurry that gains material as it travels downslope or as a shallow-slope movement

that is mobilized into a flow (Ritter et al., 1995).

1.6 Debris Flows within the Blue Ridge and western North Carolina

Debris flows have been reported to be the most common form of rapid slope

movement in the Blue Ridge (Mills et al., 1987). Debris flows have occurred throughout

the Blue Ridge province and have been specifically documented in Virginia (Woodruff,

1971; Williams and Guy, 1973; Gryta and Bartholomew, 1987; Gryta and Bartholomew,

1989; Mazza and Wieczorek, 1997), North Carolina (Holmes, 1917; Gryta and

Bartholomew, 1983; Pomeroy, 1991; Wooten et al., 2001), West Virginia (Jacobson et

al., 1989) and Tennessee (Clark et al., 1987). Mills et al. (1987) suggest that the

abundance of this type of slope movement may be due to the amount of thick,

unconsolidated colluvium derived from crystalline rock. This colluvium is highly

permeable and susceptible to weathering, both of which contribute to the generation of

slope movements.

In western North Carolina, debris flows are activated primarily by either localized

severe storms that produce intense rainfall for several hours or by more regional

moderate storms that may last for several days (Wieczorek, 1996). Most debris-flow-

producing storms can be linked to the incursion of warm, tropical air masses over the

mountains between May and November, or the remnants of hurricanes and tropical

storms (Kochel, 1990).



Precipitation rates that readily induce debris flows in western North Carolina range from

125 mm/day (Neary and Swift, 1987) to the upper end of observed precipitation (560

mm/day). These thresholds may vary due to lithology, vegetation, and topography but

generally catastrophic rainfall is required to initiate debris flows in heavily vegetated

areas (Kochel, 1990). Under these conditions, rapid infiltration and a corresponding

increase in soil saturation brings the soil mantle to field capacity. This tends to occur in

shallow (< 1 m thick) mountain soils on slopes averaging 25-40 degrees, overlying an

impermeable horizon of metamorphic rock or saprolite (Eschner and Patric, 1982). A

temporary rise in piezometric pressure within slope sediment causes an increase in shear

stress while decreasing shear strength. This, combined with a decrease in soil cohesion,

reduces the shear resistance force enough to lessen the stability of the soil and eventually

induce failure (Neary and Swift, 1987).

Typically, debris flows have a characteristic long narrow shape. In the Southern

Appalachians, the width of a debris flow, or the “chute,” may range from only a few

meters to 60 m wide, although most tend to be narrow (Mills et al., 1987). Runout lengths

rarely exceed 600 m, but a few have been known to extend as far downslope as 1600 m

(Scott, 1972). Although single linear chutes are common, larger tracks are often produced

by coalescing debris flows that form a dendritic pattern upslope (Scott, 1981). This shape

is due to the formation of a majority of debris flows in topographic hollows (Scott, 1972).

At their bases, debris accumulates as a washout fan of colluvium or may merge with



1.7 SINMAP and SHALSTAB

There are many approaches to the problem of predicting shallow slope movements,

and almost as many predictive models. Simple models, based on identifying and

classifying high-hazard areas on the basis of critical slope angle, do not take into account

the effects of topographic form or position and lithology. More complex approaches to

prediction should consider a wide range of variables such as drainage area, bedrock

geology, soil thickness and cohesion, precipitation, vegetation and land use. Two such

models, which take many of these variables into account, are SINMAP and SHALSTAB.

SINMAP was designed as an extension to ArcView®

GIS, a product of

Environmental Systems Research Institute, Inc. SINMAP is applied to shallow

transitional landsliding phenomena controlled by shallow groundwater convergence

(Pack et al., 2001), as is generally found in western North Carolina. The SINMAP

methodology is based on the infinite-slope equation, an equation that has been found to

be adequately accurate in the analysis of debris flows for planning purposes in the U.S.

Western Cordillera (Hammond et al., 1992) as well as in western North Carolina

(Otteman, 2001).

A digital elevation model (DEM) is used to determine inputs of topographic slope

and catchment area. Landslide point data is added to compare locations of predicted

instability with areas of actual instability and to evaluate the accuracy of the model

results. Other parameters for both natural material and climatological properties are



ranges are reached and yet stability is still retained, the stability index (defined as the

factor of safety) is calculated as greater than one (Pack et al., 2001). The default output of

a SINMAP session is a series of map grids that define areas of potential terrain

instability; shaded green areas are considered stable whereas dark red areas have a high

probability of failure, based on parameter inputs (Pack et al., 1998b).

Like SINMAP, SHALSTAB is an extension of ArcView

®

and is theoretically

similar to SINMAP. The SHALSTAB model is also based on the infinite-slope equation

and uses a DEM for the input of topographic grid information. SHALSTAB assumes

steady-state, saturated flow parallel to the slide surface and uses Darcy's law to estimate

the spatial distribution of pore pressures (Dietrich and Montgomery, 1998).

The authors of SHALSTAB intended the model to be as simple as possible and

nearly “parameter free” (Dietrich et al., 2001). In other words, most of the input variables

are derived directly from the slope and area grids created from the input DEM and do not

require the user to calculate specific parameters based on known soil properties which,

without direct sampling and laboratory investigations, are often difficult to derive

(Dietrich et al., 2001). Although SHALSTAB does allow for adjustments to certain soil

parameters to better match existing conditions, the model does not provide upper and

lower boundaries for parameters like SINMAP. Moreover, the model does not allow for

an adjustment to recharge, but rather calculates the effective recharge (precipitation

minus evapotranspiration) needed to induce failure for each cell within the DEM grid.



should be compared with mapped landslide features whenever possible (Dietrich and

Montgomery, 1998).

1.8 Limitations of Research

The prime limitation to both SINMAP and SHALSTAB is the availability of

high-quality digital-elevation data i.e. a 10-meter scale or smaller. Surface topography,

determined from digital-elevation data, has great bearing on the location and frequency of

shallow landsliding predicted by both models. Currently, only 30-meter and 10-meter

DEMs are available for the western portion of North Carolina. Both SINMAP and

SHALSTAB require the input of high-quality digital-elevation data to identify areas of

steep slope. As more accurate DEMs and LIDAR data are released for this portion of the

state, more accurate slope-instability and landslide-hazard maps can be developed.

Limitations to SINMAP and SHALSTAB include the lack of high-quality soil

data necessary to parameterize the models, the size of the study area, and the need to

generalize soil and climate parameters over such a large area. Better quality results can be

derived from larger-scale studies covering towns or individual neighborhoods. Such

studies should collect and analyze soil samples as well as create a detailed inventory of

all present and past debris flows for the study area.

The authors of SINMAP and SHALSTAB did not intend for these models to

provide a complete forecast of all types of landslide potential. These models were

designed to predict only the potential for shallow landsliding, not deep-seated slides, rock



Figure 1.1 - The morphology of a typical debris flow found in the Southern Appalachians (courtesy of

the North Carolina Geological Survey, 2003).



Figure 1.2: Threshold precipitation values necessary for producing debris flows in the southern

Appalachian Mountains. Storms likely to start debris flows occur above the 125 mm/d threshold.

Storms with precipitation values higher than 250mm/d are deemed “rare” but do occur in North

Carolina (after Eschner and Patric, 1982).



CHAPTER 2: PROJECT SETTING

2.1 Introduction

The French Broad watershed is approximately 7330 km² and includes the counties

of Buncombe, Haywood, Henderson, Madison, and Transylvania, and portions of Avery,

Mitchell, and Yancey counties (Figure 2.1). Major tributaries of the French Broad River

include the Nolichucky, Toe, and Pigeon Rivers. The watershed includes large portions

of Great Smoky Mountains and Pisgah National Parks. Two major interstates, Interstate-

40 and Interstate-26, cross the basin, as does the Blue Ridge Parkway. The French Broad

River actually begins in Rosman, North Carolina (35 miles southwest of Asheville, NC)

where four tributaries converge.

Topography within the French Broad basin ranges greatly, from the relatively

gently sloping floodplains along the banks of the French Broad River to steep slopes in

the mountains and along roadcuts. The highest point in the watershed, and the entire

Appalachian mountain chain, is Mount Mitchell at an elevation of 2073 m.

2.2 Geology

Given the large size of the French Broad River basin, 40 geologic units and 17

general soil types have been mapped within the watershed by North Carolina agencies.

The watershed lies within both the Blue Ridge Belt and, to the east, a small portion of the

Inner Piedmont Belt. Bedrock generally consists of sedimentary, metasedimentary, and

intrusive igneous rock of Proterozoic and Paleozoic age (North Carolina Geological



The Blue Ridge geologic province reaches its greatest east-west extent in the

Carolinas, Tennessee and Georgia. The province is bounded on the northwest by the Blue

Ridge fault systems (Holston-Iron Mountain, Great Smoky, and Cartersville Faults) and

on the southeast by the Brevard fault zone, while the Hayesville and Greenbrier fault

zones intersect in the middle of the French Broad basin (Hatcher and Goldberg, 1994).

These faults transported a series of large crystalline thrust sheets over the Paleozoic rocks

of the Valley and Ridge province, each with different tectonic histories and degrees of

metamorphism (Hatcher, 1987). The western Blue Ridge is composed of a rift-facies

sequence of clastic sedimentary rocks deposited on basement rock. The eastern Blue

Ridge records a series of slope-and-rise sequences associated with rifting and continental

collision (Hatcher and Goldberg, 1994).

The Brevard Fault zone separates the Blue Ridge province from the Inner Piedmont

block to the east. The block is a composite stack of thrust sheets containing a variety of

metamorphic rocks, intrusive granitoids, and sparse ultramafic bodies (Horton and

McConnell, 1994). Within the French Broad Watershed, only a few units from the Inner

Piedmont block are represented i.e., the Henderson gneiss and middle-to-late Paleozoic

granite gneiss. The Brevard fault zone itself represents a linear, southeast-dipping belt of

cataclastic and mylonitic rock, 1-to-2 km wide (Horton and McConnell, 1994).

According to Scott (1972), geologic structure and bedrock orientation play a more

important role in slope stability than rock type in the Southern Appalachians. When soils



particularly true when fracture surfaces are parallel to the dip surface. It was observed in

the study area that even during a light precipitation event, groundwater flow through

fracture zones was swift. This concentration of groundwater could quickly cause an

increase in pore-water pressure in soils on a slope or create ephemeral channels for debris

flows to follow. A similar correlation between joint orientation, direction of groundwater

flow, and debris-flow initiation was noted in the Coweeta Basin, an experimental forest

and research station just south of the watershed (Grant, 1987). In the SINMAP and

SHALSTAB models, groundwater flow is incorporated by using a value for soil

transmissivity (m²/s).

2.3 Soils

The types of soil in the French Broad watershed reflect the regional geology

because variation in bedrock mineralogy partly controls soil mineralogy (Figure 2.3).

Herein soil will be defined as “unconsolidated mineral or organic material on the

immediate surface of the earth that has been subjected to and shows effect of genetic and

environmental factors” (Jackson, 1997). Steep relief, broad ridges, and humid

temperatures allow for a wide range of soil-forming conditions. On steep side-slopes,

Inceptisols are common whereas Ultisols are found on gently sloping areas (Graham and

Buol, 1990). Inceptisols, particularly Dystrocrepts, are soils that have not developed

many diagnostic features due to rapid erosion rates and downslope movement. Ultisols

develop from acidic parent materials, such as granite, and have an increase in clay



Mitchell and at very high elevations. Mesic soils have mean soil temperatures of 8° to

15° C (46° to 59° F) during June, July, and August, whereas frigid soils have mean

temperatures less than 8° C (46° F) during the summer months (Buol et al., 2003).

Generally, soils with a high susceptibility of failure tend to have a large mica content and

develop over micaceous schist, slate, and phyllite (Scott, 1972).

Soil cover varies in thickness and development depending upon slope and

weathering and can range from less to one meter to several meters in depth (Clark, 1987).

Steep slopes create a shallow soil veneer held together by plant roots that may overlie a

rock or saprolite layer. More moderate slopes, i.e., those between 30 and 35 degrees,

develop thicker soil profiles and are more prone than shallow soils to debris flows. This

may seem contradictory, but steep slopes are not able to develop and retain soil cover and

can easily be washed away, even during moderate rainfalls. Thus, an increase in soil

depth and a decrease in slope increase the risk of slope instability (Scott, 1972).

In the unglaciated portions of the Blue Ridge, chemical weathering plays an

important role in the breakdown of rock into soil regolith (Grant, 1987). According to

Graham et al. (1990), three types of regolith are found in the Southern Appalachians i.e.,

saprolite, colluvium, and soil residuum. Thick sequences of saprolite, or “rotten rock,”

develop on gentle slopes composed of metasedimentary and metaigneous rock,

promoting stability and chemical weathering (Jacobson, et. al., 1989). Typically, saprolite

is covered by soil residuum. Residuum is separated from saprolite where the visual



colluvium and alluvium are deposited in valley bottoms and along rivers (Hatcher, 1987;

Otteman, 2001).

2.4 Climate

Due to the variation of altitude (460-to-2073 m) within the French Broad

watershed, temperature and moisture regimes vary greatly from one place to another

within this area. In fact, the mountains have some of the wettest and driest weather in

North Carolina (Daniels et al., 1999). The greatest 24-hour rainfall total in the State (565

mm) was measured in the watershed at Altapass in Mitchell County on July 15-16, 1916

when a hurricane passed through the area. In contrast, the station with the driest weather

on average is located in downtown Asheville in Buncombe County (State Climate Office

of North Carolina, 2003).

Mean annual rainfall in the southern Appalachians ranges from 1000 to 2700 mm

(1-to-2.7 m) with snowfall only contributing 5 percent of the total precipitation (Neary

and Swift, 1987). Rainfall occurs frequently as small, low-intensity rains in all seasons

but localized heavy precipitation is uncommon. Precipitation is usually greater during the

winter and spring, with March being the wettest average month (Table 2.1 and Figure

2.5). The highest maximum precipitation amounts have been recorded in the summer

months when localized, high-intensity thunderstorms and hurricanes are more common.

As a result, a majority of debris-forming rainfall events in the Blue Ridge occur in June,

July, and August (Clark, 1987). Recently, in September 2004, several debris flows were



Orographic influences generate extremely heavy rainfall in localized mountainous

areas, even in storms with weak pressure gradients and gentle air circulation (Scott,

1972). Generally, rainfall increases with elevation at a rate of 5 percent per 100 m (Swift

et al., 1988), but altitude is not as important as orographic boundaries. The Blue Ridge

produces an elongate area of high values of mean precipitation (Jacobson et. al., 1989).

As can be seen in Fig. 2.6, there is a marked increase in the amount of annual

precipitation in northern Transylvania County and southern Haywood County.

2.5 Vegetation

Like rainfall, vegetation within the watershed varies with the topography. Slope

aspect and shading by adjacent higher mountains also influences the distribution of major

tree species (Daniels et al., 1999). At lower elevations (below 1400 m) hardwoods, oak,

hemlock and pine forests dominate. Hardwoods such as yellow poplar, ash, and black

cherry are found in coves and along steep slopes whereas several varieties of pine and

oak thrive in open areas (Scott, 1972). Except for the most rugged terrain, the region’s

forestland has been cut or burned at least once since European settlement (Clark, 1987).

In the very high mountainous areas of the watershed (above 1400 m) distinctive

ecological systems have been established as a result of the cool year-round temperatures.

Areas are often wind-swept and trees are damaged by ice and winter wind. Red spruce,

mountain ash and Fraser fir are common with the latter dominating above 1890 m

(Daniels et al., 1999). Grass balds and areas dominated by low shrub like rhododendron



Figure 2.1: Location Map of the French Broad Watershed in western North Carolina.



28

Figure 2.2: General geologic map for the French Broad Watershed. Individual geologic unit descriptions can be found in Appendix A (adapted from

North Carolina Geological Survey, 1985).



29

Figure 2.3: General soil map for the French Broad Watershed. Individual soil descriptions can be found in Appendix B (adapted from U.S. Department

of Agriculture, 1998).



Figure 2.4: The U.S. Department of Agriculture guide for the textural classification of soils. This

guide is only for soils with a particle size of less than 2 mm in diameter. A rock fragment modifier

(gravelly, cobbly, stony, bouldery) prefaces the textural name if particles larger than 2 mm compose

more than 15% of the soil (Buol et al., 2003).



Table 2.1: Table of the average, median, minimum, and maximum precipitation totals (mm/month)

from 1895 to 2001 for the mountains of North Carolina (NCDC Climate Data Online, 2003).

North Carolina Climate Division 01 - Southern Mountains

Average Total Precipitation (mm/month) 1895-2001

JAN FEB MAR APR MAY JUN JUL AUG SEP OCT NOV DEC TOTALAverage 117.86 115.82 142.49 114.81 110.74 121.67 136.14 133.60 99.06 92.20 92.71 116.08 1393.18Median 109.98 112.52 131.83 111.76 100.58 118.36 131.06 120.65 91.95 82.55 90.93 110.74 1312.93Minimum 8.38 13.21 20.83 23.11 29.21 37.34 49.02 19.81 8.89 1.27 24.13 9.14 244.35Maximum 279.15 254.00 282.96 208.79 271.02 245.87 315.21 398.27 249.43 257.56 303.02 272.54 3337.81

Monthly Precipitation for the NC Southern Mountains -

1895-2001

0

50

100

150

200

250

300

350

400

J A N

F E B

M A R A P

R M A Y J U

N J U L

A U G

S E P

O C T

N O V

D E C

Month

M i l l i m e t e r s ( m m )

Average

Median

Minimum

Maximum

Figure 2.5: Graph based on the data from Table 2.1 (NCDC Climate Data Online, 2003)



32

Figure 2.6: Average annual precipitation in inches within the French Broad Watershed. (Adapted from data provided by North Carolina Center for

Geographic Information and Analysis map server (http://204.211.135.111)).



CHAPTER 3: PRE-HISTORIC AND HISTORIC DEBRIS FLOWS IN WESTERN

NORTH CAROLINA

The Appalachian Mountains have a long history of producing destructive debris flows.

Through the Pleistocene, temperature and moisture fluctuations associated with the transition

from glacial to interglacial ages, destabilized exposed soil and rock. These pre-historic debris

flows have formed prominent modern landforms and a rolling topography (Jacobson et. al.,

1989). Records of flooding in western North Carolina associated with hurricanes and other

strong storms exist back into the late 1700’s. Recorded instances of debris flows and other slope

movements during major rain events began in the late 1800’s, but their generation and mechanics

have been poorly understood until recently. This study is only a part of the ongoing natural

hazards research being conducted by the North Carolina Geological Survey (Wooten et al., 2004)

and North Carolina Department of Transportation. Continued study of the history of debris flows

can help identify triggering mechanisms that are particular to North Carolina and the recurrence

interval of these events.

3.1 Quaternary Debris Flows

The Quaternary history of the Appalachian Mountains consists of slow uplift and

subsequent denudation. Although extensive terrain is mantled with colluvial deposits, individual

ancient mass movements have only recently been identified in the Blue Ridge Province. Studies

of these deposits have elucidated the rate of soil development and erosion, catastrophic debris-

flow frequency and triggering events, and the possible role of periglacial processes in the



poorly sorted, but may be either matrix-or-clast-supported (Figure 3.1). Typically, fans are

composites of several mass-wasting events with a weathered surface on each colluvial unit in the

sequence. This indicates that there may be great differences in age between the units and

upwards several thousand years have elapsed between debris-flow-forming events (Kochel,

1990). In the Great Smoky Mountains, characteristic recurrence intervals for major debris flows

are on the order of 400 to 1600 years (Kochel, 1990) whereas catastrophic debris flows have

been estimated to occur every 3000-6000 years in Nelson County, Virginia (Kochel, 1984;

Kochel and Johnson, 1984).

Catastrophic geomorphic events, such as debris flows, have been the principal means of

erosion of the central Appalachian mountain chain during the Quaternary. The long-term

denudation rates in the Appalachians average 40mm/kyr (Hack, 1980). In the short term, a

single debris flow can remove enough material to account for 1 kyr of erosion during a single

storm (Mills, et. al., 1987).

Numerous studies (Table 3.1) hypothesize that extreme precipitation events, associated

with major Quaternary climate change and periglacial environmental conditions have encouraged

the formation of a number of debris flows in the Southern Blue Ridge during the Pleistocene

(Liebans and Schaetzel, 1997). During the last 850 kyr there have been at least ten major ice

advances that have brought glaciation to much of the northern Appalachians and periglacial

conditions to the southern Appalachians (Braun, 1989). The term “periglacial” refers to the area

near a glacier that is affected significantly by cold weather and has a landscape appreciably



During glacial periods, North Carolina experienced a greater frequency of freeze-thaw

cycles and physical weathering. Rock exposed at high elevations decomposed to a thin loose soil

mantle (Mills, 2000). A dry polar climate dominated the region. In modern polar climates,

monthly temperatures average below 10°C (50°F) year-round, resulting in little to no tree

growth (Lydolph, 1985). Although a polar climate can create a ready supply of sediment through

erosion and physical weathering, the lack of precipitation inhibits the formation of debris flows.

In contrast, slow mass movements, such as solifluction and creep, are common (Ritter et al.,

1995). In Virginia, slope wash of material may have proliferated more than debris flows during

the Pleistocene (Eaton et al., 1997; Eaton, 1999). In the Great Smoky Mountains National Park,

block field and slope deposits produced by Pleistocene frost wedging have been identified Clark

and Torbett (1987).

After the late Wisconsin glacial maximum, near the end of the Pleistocene, the northward

migration of the polar front would have allowed tropical moisture to reenter the area during the

summer months (Kochel, 1990). Previously undisturbed soil and rock then became exposed to

heavy precipitation. Slopes that were still sparsely vegetated (due to cold winter temperatures)

apparently became saturated and unstable, creating numerous large debris flows. Repeated

intervals of glacial and interglacial climate on a periglacial landscape probably created episodic

sequences of catastrophic mass-wasting during the Pleistocene and early Holocene (Kochel,

1990).

Hillslopes remained generally unstable during the late Wisconsin glaciation, but

6 3



high elevations (>1100 m), produced large volumes of colluvium (>106 m3) and may have

transported material as far as 8 km in a single event (Hatcher et al., 1996). Giant Pleistocene rock

and block slides also have been identified in both southwestern Virginia (Shultz, 1986) and the

Great Smoky Mountains (Hadley and Goldsmith, 1963). These slides indicate that large sections

(100-400 million m³) of sandstone detached along bedding planes and joints and slid downslope

(Schultz, 1986). If so, these debris flows and slides were significantly larger than those produced

during modern times.

Studies of pre-historic debris flows and fans tend to use three major dating techniques,

i.e., relative-age dating of soil-horizon development and clast weathering, carbon-14 dating of

organic material, and thermoluminescence dating of colluvium and related sediments (Table 3.1).

These studies quantitatively provide evidence of repeated debris-flow activity during the

Quaternary (Clark, 1987).

Relative dating, when used as the only means for quantifying the age of a colluvial or fan

deposit, may be less reliable than the other two techniques. Nonetheless, relative-age

relationships have been used to correlate the development of debris flows from several different

sites in western North Carolina (Mills, 1982; Mills and Allison, 1995b; Liebens and Schaetzl,

1997). Generally, these studies use weathering-rind development, Munsell soil color, clay

content, and superposition to differentiate between older and younger debris-flow deposits.

Although these studies reach reasonable conclusions, none of them provide a probable date for

the surface deposits being studied, only a generalized time period. Results could be improved if



unless the weathering conditions in both areas are very similar. Nonetheless, relative dating is a

useful descriptive tool and a good first approximation for modern debris flow hazard areas.

These studies also provide evidence of periods of episodic mass-wasting that was more extensive

than the processes occurring today.

Few radiocarbon-dating studies have been completed in North Carolina due to a lack of

datable organic material in the stratigraphic record and the presumption that some deposits are

too old (Table 3.1; Jacobson et al., 1989; Mills and Allison, 1995b). Existing dates in nearby

Virginia range from as recent as 2,200 kyr to as old as 50,800 kyr (Eaton et al., 1997). Dates for

North Carolina range from 16,000 – 25,000 kyr. Kochel (1990) believes that these older dates

indicate that while debris-flow activity may have been impeded during glacial maxima, tropical

moisture occasionally has invaded the area.

Further radiocarbon dating, where possible, would greatly improve our knowledge of the

Pleistocene environment and pre-historic Holocene mass-wasting in western North Carolina and

the Appalachians. Few historical accounts of debris flows have actually been recorded and this is

a field that could fruitfully be explored in more detail, particularly with new dating techniques

using longer-half-life techniques such as K-Ar dating.

3.2 Modern Flooding and Debris FlowsThe first recorded instance of a major flood in the French Broad watershed occurred in

April 1791, six years before Asheville, NC was incorporated with its present name (Tennessee

3 2 1 J 1876



3.2.1 June, 1876

The first recorded instance of debris flows affecting the N.C. Blue Ridge occurred on

June 15-17, 1876. The debris flows accompanied flooding that is often called the “June Freshet,”

one of the greatest floods in the upper reaches of the French Broad watershed (Tennessee Valley

Authority, 1960). At the time, a debris flow was generally attributed to a “waterspout”, i.e., a

sudden funnel-shaped cascade of water falling from the sky during a torrential rain event

(Clingman, 1877). It was believed that the force of the falling water ripped away the soil from

the side of the mountain, leaving only solid bedrock. The term “waterspout” was used not only to

describe a meteorological event but also a geomorphic feature. The closest modern term to this

"waterspout" is “microburst.”

Clingman (1877) reports that at least 40-60 waterspouts were reported in portions of

Macon and Jackson counties during the June, 1876 storm. Although Clingman (1877) did not

provide a mechanism for the “waterspouts”, his detailed descriptions of the erosion and

deposition are excellent, e.g., that of two debris flows that occurred in Macon County near the

crest of Fishhawk Mountain and the Tessantee River on the afternoon of June 15, 1876

(Clingman, 1877). There were no known fatalities but the Conley family witnessed the debris

flow across the river from their home:

“They saw a large mass of water and timber, heavy trees floating on the

top, which appeared ten or fifteen feet high, moving rapidly towards them, as if itmight sweep directly across the Tessantee and overwhelm them. Fortunately,however, sixty or seventy yards beyond the creek the ground becamecomparatively level, and the water expanded itself, became thus shallower, andleaving many of the trees strewn for a hundred yards along the ground, entered

ti Fi hh k M t i i th h f l h kill d d 15 h



noting as Fishhawk Mountain is the same area where four people where killed and 15 houses

destroyed in a debris flow that occurred on Sept. 16, 2004 (see below).

3.2.2 May, 1901

From May 18-to-23, 1901 a series of low-pressure systems passed through western North

Carolina and brought heavy rain, with the heaviest precipitation occurring on May 21-22. The

storm was centered near the Black Mountains of North Carolina. Total precipitation amounts

ranged from 8.99” (22.8 cm) in Marion, North Carolina to 5.04” (12.8 cm) in Asheville, North

Carolina (Myers, 1902). Extreme flooding affected portions of the Nolichucky, Watauga, Little

Tennessee, and Catawba Rivers systems (Myers, 1902; Scott, 1972). Later flooding in the spring

and summer only added to the destruction. Total damage to farms, bridges, highways, and

buildings in the French Broad watershed was estimated to be four million dollars (U. S.

Department of Agriculture, 1902).

Most of the debris flows associated with the 1901 storm occurred in Buncombe,

Henderson, Mitchell and McDowell Counties (Scott, 1972) (Figure 3.2). The Southern Railroad

was particularly affected as a number of slides buried tracks for hundreds of meters or had

portions of track washed away in the associated flooding. A resident of Marion, George Bird,

reported that a number of slides affected the hills near his home and generated large piles of

timber (Holmes, 1917). Landslides and “waterspouts” seemed to have been particularly

prevalent in Mitchell County where as many as 17 slides were observed on one hill by Myers

(1902) (Figure 3.3). Myers (1902, p. 104) describes in detail one of the largest slides he

end of the cavity a sharp and well defined channel led down the hill to the stream



end of the cavity a sharp and well-defined channel led down the hill to the streamat the base, this channel being from 5 to 6 ft. wide and from 4 to 5 ft. deep withside walls practically vertical cut down though a gravelly clay. …It is estimated

that the excavation has a total content of about 2,500 cu. yds. of earth whichseems to have disappeared utterly.”

The particular slide described by Myers (1902) destroyed a log house that was in the flow

path. Other accounts by area residents describe cloudbursts of extreme intensity accompanying

the “waterspouts” and that water bubbled and then burst from the ground at the head of many

smaller slides (Myers, 1902). It can be assumed from these descriptions that the mass movements

in Mitchell County were debris flows, given their high water-and-debris content, their

characteristic flow path, and their rupture surface.

3.2.3 July, 1916

During July of 1916 two tropical cyclones moved through the French Broad watershed

causing extensive flooding and numerous debris flows. On July 5-6, 1916 a weak hurricane

passed over the Mississippi and Alabama coast and moved northeast, eventually deteriorating

into a tropical depression by the time that it reached western North Carolina (Southern Railway,

1917). This storm produced 4 to 10 inches (10 to 25 cm) of rain but did not create any known

debris flows. On July 14, a hurricane made landfall in Charleston, South Carolina and traveled

rapidly northwest into the mountains of North Carolina (Figure 3.4). By the morning of July 15,

the center of the powerful storm had already reached western North Carolina. The flood of July

14-16, 1916 was the largest recorded flood ever to have affected western North Carolina. It also

where 22 22 inches (56 4 cm) fell in a 24-hour period (Hudgins 2000) This is also the greatest



where 22.22 inches (56.4 cm) fell in a 24-hour period (Hudgins, 2000). This is also the greatest

24-hour rainfall total ever recorded in North Carolina. Generally, the storms of 1916 produced

two distinct regions of exceptionally heavy precipitation, i.e., one in Mitchell, Avery, and

Caldwell counties, and the other in Transylvania and Henderson counties (Figure 3.5). Runoff

from the second storm was estimated to be as high as 80-90 percent (Southern Railway, 1917).

The first storm had already thoroughly soaked the soil, increasing antecedent moisture

conditions, and filled most streams nearly to flood stage (Scott, 1972). Rainfall from the second

storm exacerbated flood conditions (Figure 3.6).

The July, 1916 storms killed about 80 people and caused $22M in damages (Southern

Railway, 1917). In Asheville, several homes and buildings were destroyed and four of the main

river bridges were washed away (Tennessee Valley Authority, 1960). The Southern Railway

Company suffered extreme financial losses and transportation within western North Carolina

was disrupted for several days. Many railway lines were covered by debris flows, trapping

freight and passenger trains between terminals. The Southern Railway (1917) reported that

almost every mile of track between Asheville and Statesville was covered by debris or washed

out. At some places, track was suspended in mid-air after the fill below was washed away

(Southern Railway Company, 1917).

Generally, debris flows were reported along the Blue Ridge Mountains to the east,

southeast, and south of Asheville (Holmes, 1917; Scott, 1972) (Figure 3.2). Most slides occurred

between 5 pm., July 15 and 7 a.m., July 16. The flows began before dark and could be heard

0 6 to 6 1 m (2 to 20 ft) and averaged 1 5 to 1 8 m (5 to 6 ft) Bedrock was seldom exposed



0.6 to 6.1 m (2 to 20 ft) and averaged 1.5 to 1.8 m (5 to 6 ft). Bedrock was seldom exposed

anywhere along any slide (Holmes, 1917).

3.2.4 August, 1940

In August of 1940, a pair of storms caused significant flooding and numerous debris

flows in the western mountains of North Carolina; the first occurred from August 11-17 and the

other from August 28-31. These storms also brought record flooding to portions of Virginia,

Tennessee, and South Carolina. Approximately 30-40 lives were lost and there were at least

$30M in damages (U.S. Geological Survey, 1949). The situation was similar to that of 1916,

with two large storms occurring in the same month. The 1940 mid-August storm was strikingly

similar to the to the second 1916 storm in terms of rainfall intensity and storm path (Figure 3.4).

However, unlike the 1916 storm, the antecedent moisture conditions in 1940 were relatively dry,

allowing for increased infiltration, hence lower overall flood discharge levels (U.S. Geological

Survey, 1949).

The first storm in 1940, an unnamed hurricane, made landfall between Beaufort, South

Carolina and Savannah, Georgia on August 11, 1940. Although no wind speeds were recorded,

damage reports indicate that trees were uprooted and broken, many buildings were damaged or

destroyed, and 20 coastal residents were killed. An unusually high tide was reported, reflecting

the storm surge. The storm then moved inland and curved northward following the Savannah

River Valley, weakening significantly. It followed a semi-circular path through Georgia,

Tennessee and Virginia, and then back into North Carolina before it moved offshore on August

slow rate of movement allowed for heavy precipitation for several days over the North Carolina



y p p y

Blue Ridge, resulting in high rainfall totals. Maximum precipitation totals ranged from 13 to 16

inches (33 to 41 cm) at to as little as 5 inches (13 cm) in Asheville (Tennessee Valley Authority,

1960). A series of well-defined storms centers over the Appalachians Mountains extended

toward the northeast from Blue Ridge, Georgia to Luray, Virginia (Figure 3.7), apparently due to

an orographic influence on the storm precipitation (U.S. Geological Survey, 1949).

The second storm in 1940 occurred during the period of August 28-31 but intense rainfall

did not begin until the morning of August 29. Rainfall continued to fall until August 30 when it

abruptly ended around noon. Only passing showers remained by August 31 (U.S. Geological

Survey, 1949). This storm was a relatively local meteorological disturbance that only affected

the French Broad and Little Tennessee watersheds (Figure 3.8). Precipitation was of both shorter

time duration and aerial extent than the mid-August storm, but of higher intensity. Rainfall

amounts ranged from 8 to 13 inches (20 to 33 cm) on the western slopes of the Blue Ridge in 20

to 30 hours (U.S. Geological Survey, 1949). Given the antecedent moisture conditions due to the

earlier storm, flooding was more severe near the storm center but overall was not as widespread.

The 200-300 debris flows associated with both storms of 1940 attributed greatly to the

devastation wrought by the floods (Scott, 1981). These slides occurred near the centers of both

storms in shallow saturated soils on steep slopes (Figure 3.2). They originated on slopes 300 to

400 feet (91.5 to 122 m) from the tops of mountains, near the slope break. The debris flows

ranged in width from 6 to 8 feet (1.8 to 2.4 m) and 40 to 50 feet (12 to 15 m) long to 200 to 300

During the mid-August storm of 1940, debris flows mainly occurred in the Blue Ridge



g g y g

Mountains, from the North Fork of the Catawba River northward into Watauga County near the

North Carolina – Virginia border (Figure 3.2). During the late August storm, debris flows

occurred mainly in the Upper Pigeon and Tuskasgee River basins (Figure 3.2). Because of the

concentration of high-intensity rainfall within a small area, more than 200 debris flows occurred

in an area of only 150 mi2 (388.5 km2) (U.S. Geological Survey, 1949).

3.2.5 November, 1977

In early November 1977, a storm system that had formed as a low-pressure system in the

Gulf of Mexico moved northwestward into the Appalachian Mountains (Neary and Swift, 1987).

Rainfall began in western North Carolina in the early morning of November 2 and continued at a

steady rate (20-50 mm/day) until November 5. This steady rain was followed by intense

downpours (102 mm/hr) on the night of November 5-6 during which most of the debris flows

were initiated (Nearly and Swift, 1987). This heavy precipitation, as in 1916 and 1940, was

produced by convection associated with orographic lifting over the southern Appalachians. Four

areas of exceptionally heavy precipitation (200-320 mm) were produced along the southeast

ridges of the North Carolina Blue Ridge. Two of these areas were within the French Broad

Watershed (Neary and Swift, 1987) (Figure, 3.9).

Although the heaviest rainfall in 1977 occurred in the vicinity of Mt. Mitchell, the best

information about debris flows and flooding came from the Bent Creek watershed, located about

15 km southwest of Asheville. A survey was conducted here immediately following the storm

ephemeral creekbeds or along hillslope depressions (Pomeroy, 1991). Scarps occurred in shallow



residual soils less than 1 m deep over gneissic bedrock (Neary et al., 1986). All of the flows

occurred in undisturbed, forested areas (Neary et al., 1986).

Topography in the Bent Creek watershed is at least partially controlled by the underlying

concentration of tension joints in the bedrock. Where there is a greater amount of jointing,

topographic hollows tend to develop. These joints allow for the infiltration of groundwater,

enhancing breakdown of the rock. This accelerates weathering, providing loose material for mass

wasting (Pomeroy, 1991). Debris flows seem to originate on the bedrock-soil or bedrock-

colluvium interface within these hollows.

The November 1977 flood killed at least thirteen people; sixteen counties in western

North Carolina were declared disaster areas. The most serious flooding occurred along the

French Broad River downstream from Asheville and in Yancey County where nearly every

bridge was washed out (Eshner and Patric, 1982; Stewart et al., 1978). Flooding destroyed 384

homes, 389 miles (622 km) of highway, and 12 dams. In total there was over $50M in damages

associated with this storm (Stewart, 1978).

In 1977, precipitation, slope, and topography all contributed to the initiation of debris

flows southeast of Asheville (Pomeroy, 1981). In comparison to the historical range of rainfall

intensities for the entire Tennessee Valley area, the maximum intensities associated with the

1977 storm were in the middle-to-low range but antecedent moisture was exceptionally high

(177% above normal) for the two months preceding the storm (Neary and Swift, 1987). The

towards the northeast, east, and southeast (Pomeroy, 1991). These slopes would only receive



sunlight in the morning and thus would have higher soil moisture.

3.2.6 September, 2004

The 2004 Atlantic hurricane season was exceptionally brutal. Fifteen tropical or

subtropical storms formed in the North Atlantic. Nine of these storms became named hurricanes

and six of these struck the United States (National Weather Service, 2004a). In North Carolina,

the remnants of three tropical systems, i.e., hurricanes Frances, Ivan and Jeanne, impacted the

western part of the state in September. Frances and Ivan caused extreme flooding in Asheville

and several debris flows and rockslides in the mountains, affecting Interstate-40. Rainfall totals

for the month over much of western North Carolina ranged from 10 to 25 inches (25 to 64 cm).

This was 2-to-5 times greater than normal (Badgett et al., 2004).

Hurricane Frances struck the east coast of Florida early on Sept. 5, 2004 and quickly

weakened into a tropical storm (National Weather Service, 2004a). The storm then rapidly

moved across the state, through the panhandle of Florida, and northeastward across the eastern

United States (Figure, 3.4). The effects of hurricane Frances could first be felt in North Carolina

on September 6 around 6:00 p.m. (Boyle, 2004) but most of the flooding and mass wasting

occurred on September 8 (Appendix C).

In North Carolina, the heaviest precipitation in 2004 occurred slightly east of the French

Broad Watershed but heavy rain also fell in the eastern portions of Tennessee and Kentucky, and

through most of West Virginia, Virginia, Maryland, Ohio and southwestern Pennsylvania

Sixty miles southwest of Asheville, Lake Toxaway received 14 inches (35.6 cm) of rain (Nowell,



2004). In total, 17 western counties were affected by flooding (Anonymous, 2004). Hundreds of

people were evacuated from their homes and several had to be rescued from the rising water

(Nowell, 2004). Areas of Asheville located near the Swannanoa River were flooded, particularly

the shopping center near the entrance to the Biltmore Estate, were water stood as much as 5 feet

(1.5 m) deep (Nowell, 2004). In Haywood County, flooding along the Pigeon River also

inundated downtown Canton and Clyde.

The remnants of hurricane Frances caused at least 21 reported incidents of mass wasting

(Appendix C) along several major roadways in seven western North Carolina counties. However,

only three counties within the boundaries of the French Broad Watershed experienced debris

flows, i.e., Avery, Henderson, and Transylvania. The largest reported debris flow occurred east

of Asheville on Interstate-40, near Old Fort Mountain in McDowell County (Figure 3.12). This

slide crossed the westbound lane and the median to block four of the six lanes of a five-mile

stretch of Interstate-40 (Nowell, 2004). In Watauga County, one house was destroyed and eight

others condemned when a debris flow moved through a subdivision (North Carolina Geological

Survey, 2004a). Portions of the Blue Ridge Parkway were closed when at least six debris flows

destroyed the roadway in four areas between Linville Falls and Waynesville (Ball, 2004). About

250 roads became impassable or were closed due to flooding and mass wasting (Barrett, 2004).

Most of that road damage was in Buncombe County (Ball, 2004).

Ivan was an unusually long-lived hurricane that made landfall along the United States

southwestward, crossed over Florida and then into the Gulf of Mexico. By September 23, the



remnants of Ivan had strengthened into a tropical storm and the “re-born” storm made landfall

for the second time on September 24 over southwestern Louisiana (National Weather Service,

2004a).

The remnants of Ivan moved into western North Carolina early on September 16.

Although Ivan had weakened to a tropical storm by the time it reached North Carolina, it still

packed powerful winds and heavy rain. Rainfall was not as heavy as that which fell during

Frances, mainly because the storm moved rapidly northeastward, but the western portion of the

state still received between 4 to 8 inches (10 to 20 cm) of rain. The heaviest precipitation fell in

Transylvania, Jackson and McDowell Counties at high elevations. Black Mountain (near

Asheville) received 11.5 inches (29.2 cm) of precipitation and Sapphire (in Transylvania County)

reported 15 inches (38 cm) (Figure 3.11) (Badgett et al., 2004).

Although Ivan produced less rain than Frances, high antecedent-moisture conditions and

saturated soils allowed for more slope movements to be produced. A total of 53 slope

movements have been attributed to hurricane Ivan (Cabe, 2004). Those recorded in Appendix C

occurred along major roadways (20 slope movements), but several other slope movements may

also have occurred in undisturbed or rural areas and were not reported by either the North

Carolina Department of Transportation (NCDOT) or major news agencies. Further work will

have to be conducted to obtain a complete record of these slope movements.

Slope movements, downed trees, and flooding obstructed several roads throughout

Further west, near the North Carolina-Tennessee border, a large portion of the eastbound lane of



Interstate-40 collapsed due to undercutting by the swollen Pigeon River. A major debris flow

also destroyed a home in Candler in Buncombe County (Cantley-Falk, 2004).

The worst damage occurred in the community of Peeks Creek in Macon County. At

around 10:10 p.m. on September 16, a debris flow originated near the peak of Fishhawk

Mountain (Figure 3.13) destroying at least fifteen houses, injuring several people, and resulting

in the deaths of four people (and an unborn baby). The debris flow traveled approximately 2.25

miles (3.6 km), possibly in pulses, dropping nearly 2200 feet (670 m) in elevation as it

progressed down a mountain cove and into the north fork of Peeks Creek (Cabe, 2004). The

velocity of the flow was estimated to be 20.3 mi/hr (32.7 kph) near the scarp and 33.2 mi/hr

(5305 kph) just upstream of the area of major damage (Cabe, 2004). The force of the flow

scoured the streambed, ripped trees down and left other striped of bark; houses were removed

from their foundations (North Carolina Geological Survey, 2004b; Ostendorff, 2004). The flow

probably originated as a debris slide; a slab of cohesive rock, debris and earth the size of a

football field detached from the side of the mountain and quickly disintegrated into a debris flow

as more water mixed in with the slide material (Cabe, 2004).

What is remarkable about the Peeks Creek disaster is that this location is the same area

where a two large debris flows occurred in 1876 (Clingman, 1877). Observations of residents

living in the area were strikingly similar in both incidents. Clingman (1877) describes trees

stripped of bark and limbs, a “clean, broad furrow more than two miles” long carved into the side

transformer, one resident described seeing debris spinning and flying around in the air in a



circular motion above their house (Biesecker and Shaffer, 2004). In 1876, residents described

seeing funnel-shaped spinning masses of water near the crest of the mountain (Clingman, 1877).

Tornadoes are fairly rare in mountainous areas, but do occasionally develop. While there was

wind damage throughout the Peeks Creek area after the passage of Ivan, this damage was more

consistent with wind shear. So far, the National Weather Service has not been able to conclude if

a tornado actually did touch down on Fishhawk Mountain, but they do not discount the

eyewitness accounts of local residents (Cabe, 2004).

The question remains, "Why did mass movements occur in these Macon County areas as

opposed to elsewhere?" In the Peeks Creek flow, fracture planes in the rock, sloping 35-55

degrees, provided a smooth slip surface. Soil layers over this bedrock were thin, generally less

than three feet (1 m) deep (Cabe, 2004). These physical properties of the terrain probably

facilitated the debris flows on Fishhawk Mountain. Meteorologically, the rainfall rates from the

remnants of hurricanes Frances and Ivan were not unusually intense for either event. However,

the combined rainfall totals were exceptionally heavy. The rainfall produced by Frances initially

saturated mountain soils and slopes. Before the soil had a chance to drain sufficiently, Ivan

moved through the area, bringing even more rain to already soaked areas. Rainfall from Ivan

may have caused even higher soil-water pressure on slopes than during Ivan, explaining why

there was more mass wasting during the second storm. Nonetheless why, exactly, debris flows

occurred in one area and not another, even under similar meteorological and physical conditions

3.2.7 Precipitation Thresholds and Recurrence Intervals

Generally historical rainfall totals from 1901 to 2004 are well within the 125 mm/d to



Generally, historical rainfall totals from 1901 to 2004 are well within the 125 mm/d to

250 mm/d precipitation thresholds suggested by Eschner and Patric (1982) as necessary for

debris-flow generation (see Figure 1.2). Average rainfall amounts were greater than 125 mm/d in

all cases, i.e., greater than the minimum amount of precipitation necessary to saturate soil and set

the stage for debris flows (Table 3.2). While storms with rainfall totals of 250 mm/d are

described as extremely rare, all but one storm produced maximum precipitation amounts that

exceeded 250 mm/d (Eschner and Patric, 1982). Extreme precipitation does not necessarily

guarantee that debris flows will occur, as during the July 5-6, 1916 storm, but extreme

precipitation certainly increases the risk of slope instability.

Escher and Patric (1982) suggested a return interval of 100 years or less for storms

producing debris flows in western North Carolina. Nonetheless, historical rainfall data from 1901

to 2004 indicates that recurrent intervals may be smaller. According to Cabes (2004), there have

been 14 storms or hurricanes that have triggered slope movements in western North Carolina

since 1901. Based on the occurrence of major debris-flow triggering storms from the historical

record (i.e. those that have extensive amounts of precipitation and debris-flow location data), a

preliminary recurrence interval for slope instability can be estimated. The return interval for

mass wasting from 1876 to 2004 (for six events) averages 25.6 years. This return interval has

varied from as few as 15 years to as many as 37 years. Based on Cabes’ data (2004), the

recurrence interval for slope movements from 1901 to 2004 may be as few as 7.4 years. On

traveled over the area within 10 to 20 days of each other. Antecedent moisture seems to



play a crucial role in predisposing slope to debris-flow generation. Geoscientists,

emergency management, and citizens living in these mountainous areas, must be vigilant

in monitoring weather conditions, particularly with repeated sequences of heavy rain

events.

Table 3.1: Prehistoric debris flow studies in the southern Blue Ridge and the age-dating techniques

utilized.



Reference Year Dating Technique Location Age of Features

Kochel 1987 Radiocarbon Davis Creek, VA > 11,000 BP

Jacobson et al. 1989 Radiocarbon West Virginia10,000 - 12,000 BP; 315BP

Behling et al. 1993 Radiocarbon West Virginia 17,000 - 22,000 BP

Kochel 1990 RadiocarbonAppalachian Mountains, NC

16,000 - 25,000 BP

Eaton et al. 1997 RadiocarbonUpper Rapidan River Basin,VA

2,200-50,800 BP

Shafer 1984 Thermoluminescence Flat Laurel Gap, NC Late QuaternaryMills 1982 Relative-age North Carolina ?

Mills and Allison 1995aRelative-age/paleomagnetism

Watauga County, NC 780 ka - 1Ma

Mills and Allison 1995b Relative-age Haywood County, NC ?

Liebens andSchaetzl

1997 Relative-age Macon and Swain Co., NC ?

Mills 2000 Relative-age Appalachians ?



54

Figure 3.2: Areas of major debris flows and landslides in western North Carolina (after Scott, 1972).



Figure 3.3: A sketch map of debris flows that occurred along Gouges Creek in Mitchell County,

North Carolina in May, 1901 (Myers, 1902).



Figure 3.4: Map showing some of the hurricane and storm paths that have affected western North

C li t d b th U S N ti l H i C t d th U S G l i l S W t



Figure 3.5: Total storm precipitation for July 14-16, 1916.



Figure 3.6: U. S. Geological Survey hydrograph from the river gauge located on the French Broad

River in Asheville for the month of July, 1916. In early July, there is a large hydrographic spike

generated by rainfall from the remnants of a hurricane that passed through the area. There is also a

second large peak between July 16 and 17 generated by the flooding associated with a hurricane

moving northwestward over the watershed. This is the greatest recorded streamflow for the gauge at

Asheville.



Figure 3.7: Total storm precipitation for August 14-15, 1940 adapted from U. S. Geological Survey,

1949).



Figure 3.9: Total storm precipitation for November 2-5, 1977 (adapted from Neary and Swift, 1987).



Figure 3.10: Total storm precipitation (inches) for the remnants of Hurricane Frances (NationalWeather Service, 2004b).



Figure 3.12: A debris flow that blocked the westbound lanes of Interstate-40 near Old Fort Mountain

in McDowell County (North Carolina Geological Survey, 2004a).

Figure 3.13: The initiation zone of the debris flow that occurred on Fishhawk Mountain anddevastated the Peeks Creek area of Macon County on September 17, 2004 (Wilett, 2004).

Table 3.2: Major storms within the French Broad Watershed and their minimum, average, and

maximum precipitation amounts.

STORM DATE MIN (in) AVG (in) MAX (in) MIN (mm) AVG (mm) MAX (mm)



( ) ( ) ( ) ( ) ( ) ( )

1901 5 7 8.9 128 177.8 228.41916 (1)* 4 7 10 101.6 177.8 254

1916 (2) 1 10 22.2 25.4 254 564.4

1940 (1) 4 13 16 101.6 330.2 406.4

1940 (2) 3 8 13 76.2 203.2 330.2

1977 2 8 14 50.8 203.2 355.6

2004 - FRANCES 4 10.3 16.6 101.6 261.6 421.6

2004 – IVAN 4 9.5 15 101.6 241.3 381

*no debris flows produced

CHAPTER 4: METHODOLOGY

In order to quantify debris-flow potential in the French Broad Watershed, two



deterministic process-based computer-modeling programs have been chosen, i.e.,

SINMAP (Stability INdex MAPping) (Pack et al., 1998) and SHALSTAB (Shallow

Landsliding Stability Model) (Dietrich and Montgomery, 1998). Both use similar

numerical models and steady-state hydrologic assumptions to quantify the influence of

topography on pore pressure. These computer programs, instructions, and examples are

freely available for download from their respective websites.

4.1 Infinite-slope Equation

SINMAP and SHALSTAB are based on the infinite-slope form of the Mohr-

Coulomb failure law, an equation commonly applied as part of a slope-stability model

within the GIS environment (Hammond et al., 1992; Montgomery and Dietrich, 1994;

Wu and Sidle, 1995; Pack et al., 1998). The infinite slope equation is:

where C r is root cohesion, C s is soil cohesion, θ is slope angle, ρ s is soil density, ρ w is the

density of water, g is acceleration due to gravity, D is the vertical soil depth, Dw is the

vertical height of the water table, and Φ is the internal soil friction angle. Ultimately, the

infinite-slope equation calculates a dimensionless stability index, or factor of safety (FS),

by comparing the destabilizing components of gravity with the stabilizing components of

This assumption is reasonable because the drainage barrier, e.g., top of bedrock, and the

ground surface are often parallel on colluvial slopes. The contrast in hydraulic



conductivity between the overlying soil and the material forming the drainage barrier

causes groundwater to flow nearly parallel to the ground surface, providing a surface for

slope failure (Hammond et al., 1992). Another assumption is that the failure plane is

assumed to be infinite in extent and the resistance to movement along the sides and ends

of the slope movement are so insignificant that they may be ignored (Hammond et al.,

1992). Finally, the soil is assumed to be of uniform thickness.

Although SINMAP and SHALSTAB share similar physical assumptions and

equations, they use different indices to quantify instability. SINMAP uses the infinite-

slope equation to calculate a “factor of safety”, i.e., an estimated potential for instability

whereas SHALSTAB quantifies terrain instability in terms of the effective recharge

required to trigger instability. A comparison of the results and relative performance of

these programs is difficult due to their dissimilar output. To simplify comparison, model

parameters have been kept as similar as possible.

4.2 Parameterization of the Models

SINMAP and SHALSTAB share specific parameters that must be input into the

program to optimize the models for local soil and meteorological conditions. The models

have three necessary components in common, i.e., digital terrain data in the form of a

digital elevation model (DEM) or some other type of grid-based digital topographic data,

4.2.1 Digital Elevation Models

Both SINMAP and SHALSTAB assume that the dominant control on debris-flow



occurrence is surface topography, specifically the interplay between slope and shallow

subsurface flow convergence (Montgomery and Dietrich, 1994). To calculate topography,

both computer programs require digital elevation data. In this study, DEMs at 10-meter

and 30-meter scales were obtained from the United States Geologic Survey National

Elevation Dataset (http://seamless.usgs.gov/). Both SINMAP and SHALSTAB are highly

sensitive to the accuracy of the DEM. Verification of data quality is advisable prior to

running these models (Zhang et al., 1999; Dietrich et al., 2001; Guimaraes et al., 2003).

As explained by Michael Oimoen, a scientist from the USGS National Center for

Earth and Resources Observations and Science (pers. comm., 2005), USGS DEMs used

for this study were created from the digitized contours of scanned 1:24,000-scale

topographic quadrangle sheets. The contours from these maps are digitized and either a

10 x 10 meter or 30 x 30 meter mesh grid is overlaid on top of the digitized contours. For

each grid cell, inverse distance weighting interpolation is used to calculate the elevation

value of each cell (Maune et al., 2001). Since the 10-meter and 30-meter DEM are

created from the same scale map and source data, neither is more “accurate” than the

other. But as the cell size is smaller for the 10-meter DEM, the 10-meter spacing better

captures the detail in the contours. Given both 10- and 30-meter DEM data are available

for the watershed, SINMAP and SHALSTAB runs were completed in both resolutions

catchment or drainage area, and compute the slope for each grid cell (Dietrich and

Montgomery, 1998; Pack et al., 2001).



Both programs have procedures for increasing accuracy and eliminating potential

errors inherent in the DEM grid. Pit-filling corrections are automatically executed within

both programs to eliminate artificial sinks or depressions in the DEM. Given that such

grid elements are rare in natural topography, they are assumed to be errors created during

the initial preparation of the DEM (Pack et al., 2001). Both programs use a “flooding”

approach where pits are filled, raising the elevation to that of the lowest neighboring grid

cell.

The processing time for DEMs with large spatial areas, such as a DEM for the

entire watershed, can be extremely long, depending on the speed and memory availability

of the user’s personal computer. Both SINMAP and SHALSTAB seem to have a DEM

size limitation. Although SINMAP could process the entire watershed at the 30-meter

scale, the 10-meter DEM had to be broken down into county-size pieces. SHALSTAB

cannot process DEMs larger than a county at a 30-meter scale without creating run-time

errors and premature termination. At the 10-meter scale, SHALSTAB can run a

maximum of four 1:24,000-scale quad-sized DEM blocks before shutting down. This

significantly increases process time. For this reason, the original 30-meter DEM had to be

clipped to county-size when run in SHALSTAB. In the interest of time, only Haywood

County was chosen for several model runs in SHALSTAB for comparison to the

4.2.2 Debris Flow Inventory

In order to complete the slope-stability model runs for SINMAP and



SHALSTAB, basic digital-line and polygon coverages were collected and formatted for

use in ArcView 3.x and ArcGIS 8.3. Both programs require the input of landslide

location data to verify the model results; SINMAP requires point data whereas

SHALSTAB requires polygonal shapes of each landslide feature.

Landslide location data for this study was obtained through a combination of

aerial-photography interpretation, previous studies, field investigation, and data provided

by both the North Carolina Geological Survey and the North Carolina Department of

Transportation. Most of this data was not in a digital (e.g., GIS) format and was digitized

into a single landslide-location database for the French Broad Watershed (Appendix D).

Other GIS data, e.g., roads, hydrology, and geology, and digital orthophoto quadrangles

(DOQs) were obtained through the North Carolina State University Geodigital Library

website (http://www.lib.ncsu.edu/stacks/gis/) and the North Carolina Center for

Geographic Information and Analysis map server (http://204.211.135.111).

4.2.3 Soil Data

Various soil parameters, such as cohesion, density, and hydraulic conductivity,

have been estimated from the general soil maps of North Carolina and from the advice of

experienced geologists who routinely work with soils in the study area. Basic soil maps

were downloaded from the State Soil Geographic database (STATSGO) available

that give the proportional extent of the component soils and their properties. There are

eighteen general soil types found within this watershed (see Figure 2.2). The STATSGO



soils coverage for the State of North Carolina was downloaded as a polygon shapefile,

imported into ArcGIS, and clipped to the extent of the French Broad Watershed.

4.2.4 Soil Density

Both programs require the input of a value for soil density. This value is used to

represent the total bulk density of the soil over the entire study area. In SINMAP the

default value for soil density is 2000 kg/m3

while in SHALSTAB the value is 1700

kg/m3. Otteman (2001) used a value of 1450 kg/m3 for her study area in the Bent Creek

Experimental Forest in Buncombe County, North Carolina whereas in Madison County

Virginia, a value of 1200 kg/m3

was used by Morrissey et al. (2001).

A geologist with

the North Carolina Department of Transportation estimates typical density values of

1441.6 kg/m3to 2082 kg/m

3in soils of western North Carolina (Jody Kuhne, pers.

comm., 2004). Although no independent soil-density analysis was completed, a value of

1922 kg/m³ was chosen as representative value for the entire watershed and was used in

all model runs. This number was suggested by Richard Wooten, a geologist with the

North Carolina Geological Survey who has done extensive slope-stability work in

western North Carolina.

4.3 SINMAP Parameters

SINMAP has simplified the infinite-slope equation (1) originally proposed by

combined into one dimensionless cohesion factor. The SINMAP stability-index equation

is given by:



where the variables a and θ are the specific catchment area and slope, respectively, and

are derived from the topography determined using a DEM. The other parameters, C

(cohesion), Φ (soil friction angle), R/T (recharge divided by transmissivity), and r (the

ratio of water and soil density) are manually entered into the model. These parameters are

considered more uncertain and are specified in terms of upper and lower boundary values

(Pack et al., 1998b).

The default values for the foregoing parameters are provided in Figure 4.2. Values

may be adjusted to represent local conditions, to identify landslide-prone areas more

precisely, and to assure that a significant amount of the landslides are captured in areas

considered “unstable”. The following text rationalizes the specific input values and

distributions used in this study (T/R, C , and Φ).

For each cell of the DEM, SINMAP calculates a factor of safety (FS), a ratio of

stabilizing-to-destabilizing forces found in equation (2). This is a dimensionless index

number with a value between zero and 10. If the index falls below one, there is a high

probability that the area is unstable whereas high index values (greater than one) indicate

4.3.1 T/R (Ratio of Transmissivity to Effective Recharge)

The ratio of transmissivity of the soil (m2/hr) to the effective recharge (m/hr) has a



default range of 2000 to 3000 (m). When multiplied by the sine of the slope, the T/R

value can be interpreted as the length of the hillslope (in meters) required to develop

saturation (Pack et al., 1998b).

The recharge rates used in this study have been derived from four precipitation

thresholds, i.e., 50 mm/d, 125 mm/d, 250 mm/d, and 375 mm/d. Two of these

precipitation thresholds, 125 mm/d and 250 mm/d, are described by Eschner and Patric

(1982) as necessary for debris-flow initiation in the Appalachians (see Figure 1.2). The

50 mm/d rate was chosen as a minimum rate, i.e., one where SINMAP should not

indicate a large mapped area of instability. The last threshold amount, 375 mm/d, is used

as a maximum, i.e., an extreme example of the precipitation that can produce debris flows

in the French Broad Watershed. Only a few times have recorded rainfall totals actually

exceeded this rate (see Table 3.2).

The transmissivity rate (m²/hr) was calculated using the basic equation:

T = Kb (3)

where K (m/hr) is the permeability or hydraulic conductivity of the soil and b (m) is the

soil depth. Permeability rates within each soil unit are available in the STATSGO soil

attribute database. The permeability of each of the eighteen soil units within the

watershed was calculated by determining the average permeability of each of the

the recharge rate to find the upper and lower T/R (m) values. The final T/R values used

were an average of each of the eighteen soil units (Table 4.1).



4.3.2 Dimensionless Cohesion

In SINMAP, root and soil cohesion is combined with soil density and thickness to

calculate a dimensionless cohesion factor, C (Pack et al., 1998b). Conceptually, this is the

ratio of the cohesive strength of the soil and roots relative to the weight of a saturated

thickness of soil, or the contribution of cohesion to the stability of a slope (Pack et al.,

1998b). The equation used to determine dimensionless cohesion is:

C = (Cr + Cs)/(hρsg) (4)

where C r is root cohesion (N/m2), C s is soil cohesion (N/m2), h is the soil thickness (m),

ρ s is the wet soil density (kg/m3), and g is the acceleration due to gravity (9.81 m/s

2).

Various values for cohesion have been suggested in diverse SINMAP studies. The

default values used for cohesion in SINMAP are 0 and 0.25 but Morrissey et al. (2001)

used values between 0 and 1.28. Wooten uses values of 0.6 to .96, suggesting that zero

cohesion is unlikely in the Southern Appalachians due to the considerable amount of low

vegetation growing on mountain slopes (e.g., rhododendron), which would add to the

overall root cohesion found in the soil. Because of the lack of more precise soil cohesion

data, the SINMAP default values were deemed reasonable for the watershed and used in

all of the SINMAP model runs.

S

shearing angle. The angle is constant for a specified material and depends on the size,

shape, and surface roughness of the grains, the density of the soil, moisture content, and



material saturation (Easterbrook, 1999).

Values for the soil friction angle for certain soil types can be estimated from

tables provided by Hammond et al., (1992). These tables require that the soil be classified

according to the Unified Soil Classification (USC) system (ASTM D-2487-85 and D-

2488-84) which are part of the STATSGO attribute data. Although no independent soil

analysis was completed, the 26° and 45° soil-friction angles used to calibrate the model

were considered realistic for the study area (Hammond et al., 1992). Given that this study

requires that parameters be generalized over large areas, a wide variation encompassing

the properties of several different soil types, from clayey to sandy and gravelly soils, is

more realistic than a small range of values.

4.4 SHALSTAB Parameters

Like SINMAP, SHALSTAB is an extension of ArcView©

and is theoretically

based on the infinite-slope equation. SHALSTAB assumes steady-state, saturated flow

parallel to the slide surface and uses Darcy’s Law to estimate the spatial distribution of

pore water pressure. However, the equation that SHALSTAB uses to formulate a slope-

stability map is quite different from that used by SINMAP. The SINMAP equation solves

for a factor of safety whereas SHALSTAB calculates the ratio of log q/T (effective

precipitation over transmissivity) (Dietrich et al., 2001). Essentially, this is the amount of



The two hydrologic-slope stability equations solved by SHALSTAB are shown

above. Equation (5) solves for the hydraulic ratio while equation (6) solves for the area

per outflow boundary length. The equations have three topographic terms defined by the

DEM: drainage area (a), outflow boundary length (b), and hillslope angle (θ ). Of the

other four parameters, soil bulk density ( ρ s) and internal friction angle (Φ) can be

assigned by the user. The ratio of transmissivity (T ) and effective precipitation (q) are

solved by SHALSTAB and are given as the final output of the model (Dietrich et al.,

2001).

The authors of SHALSTAB made even more simplifications to the infinite-slope

equation, and their slope-stability model, than did the authors of SINMAP. First, they

decided to set the cohesion to zero and eliminate the variable of root strength altogether.

Although this approximation is clearly incorrect in some situations, the elimination of

root cohesion makes parameterization easier for the user. Root strength varies widely,

both spatially and over time. It is especially hard to estimate over large areas, such as at

the watershed-scale. They also consider setting the cohesion to zero to be a very

“conservative” thing to do as this maximizes the extent of possible instability within the

be “parameter free” when run with the default parameter values. In other words, they

wanted a slope stability model where a user did not have to have any available soil data

t t lib t th d l Wh th d f lt i f SHALSTAB i l



parameters to calibrate the model. When the default version of SHALSTAB is run, only

the soil density and the internal soil friction angle can be adjusted. Later versions of

SHALSTAB have included an option where the cohesion value and soil depth can be

adjusted if necessary (Figure 4.3).

The SHALSTAB model needed to be calibrated in a way in which it could be

directly comparable with SINMAP. Some of the variables, such as soil density (1922

kg/m³) and soil thickness (2 m), could remain the same between the two models. But

other soil property values, such as cohesion and soil friction angle, vary widely over the

study area. This problem is overcome by SINMAP, which uses a range of values for

cohesion, soil friction angle, and transmissivity over recharge. SHALSTAB only uses a

single input value for cohesion and soil friction angle. As specific soil data is not

available at the watershed-scale, a variety of parameters were tested during several

different model runs to find which values best-approximated landslide susceptibility.



Figure 4.1: The infinite slope equation as defined by Hammond et al., (1992) and Pack et al., (1998b)

where C r is root cohesion, C s is soil cohesion, θ is slope angle, ρ s is soil density, ρw is the density of

water , g is acceleration due to gravity, D is the vertical soil depth, Dw is the vertical height of the

water table, and Φ is the internal soil friction angle. In the SINMAP model, the ratio of the vertical

soil depth to the vertical soil height is simplifed so that depth is measured perpendicular to the slope

(h). (Diagram after Hammond et al., 1992 and Otteman, 2001)



Figure 4.2: Default parameters used in the SINMAP model analysis. The values for the gravitational

constant and the density of water were not adjusted in this study.

Table 4.1: Table of the hydraulic conductivity ( K , m/hr), transmissivity (T , m²/hr), and T/R (m) values used for each precipitation threshold (50 mm/d,

125 mm/d, 250 mm/d, and 375 mm/d) in the SINMAP analysis. The numbers in blue are the lower bound values while the numbers in red are the upper

bound values.



78

SOILMAPPING

UNIT

HYDRAULICCOND (m/hr)

LOWER

HYDRAULICCOND (m/hr)

UPPER

T (m2 /hr)

LOWERT (m

2 /hr)

UPPER

T/R 50 (mm/d)LOWER

T/R 50 (mm/d)

UPPER

T/R 125(mm/d)

LOWER

T/R 125(mm/d)

UPPER

T/R 250(mm/d)

LOWER

T/R 250(mm/d)

UPPER

T/R 375(mm/d)

LOWER

T/R 375(mm/d)

UPPER

NC005 0.024 0.331 0.048 0.662 24.0 331.0 9.6 132.4 4.6 63.7 3.1 42.4

NC006 0.014 0.152 0.028 0.304 14.0 152.0 5.6 60.8 2.7 29.2 1.8 19.5

NC088 0.005 0.038 0.010 0.076 5.0 38.0 2.0 15.2 1.0 7.3 0.6 4.9

NC089 0.003 0.027 0.006 0.054 3.0 27.0 1.2 10.8 0.6 5.2 0.4 3.5

NC090 0.005 0.115 0.010 0.230 5.0 115.0 2.0 46.0 1.0 22.1 0.6 14.7NC091 0.009 0.053 0.018 0.106 9.0 53.0 3.6 21.2 1.7 10.2 1.2 6.8

NC092 0.009 0.091 0.018 0.182 9.0 91.0 3.6 36.4 1.7 17.5 1.2 11.6

NC093 0.01 0.032 0.020 0.064 10.0 32.0 4.0 12.8 1.9 6.2 1.3 4.1

NC094 0.005 0.075 0.010 0.150 5.0 75.0 2.0 30.0 1.0 14.4 0.6 9.6

NC095 0.003 0.021 0.006 0.042 3.0 21.0 1.2 8.4 0.6 4.0 0.4 2.7

NC096 0.008 0.027 0.016 0.054 8.0 27.0 3.2 10.8 1.5 5.2 1.0 3.5

NC097 0.025 0.156 0.050 0.312 25.0 156.0 10.0 62.4 4.8 30.0 3.2 20.0

NC098 0.004 0.015 0.008 0.030 4.0 15.0 1.6 6.0 0.8 2.9 0.5 1.9NC099 0.01 0.081 0.020 0.162 10.0 81.0 4.0 32.4 1.9 15.6 1.3 10.4

NC100 0.01 0.099 0.020 0.198 10.0 99.0 4.0 39.6 1.9 19.0 1.3 12.7

NC102 0.007 0.035 0.014 0.070 7.0 35.0 2.8 14.0 1.3 6.7 0.9 4.5

NC103 0.005 0.025 0.010 0.050 5.0 25.0 2.0 10.0 1.0 4.8 0.6 3.2

NC104 0.012 0.046 0.024 0.092 12.0 46.0 4.8 18.4 2.3 8.8 1.5 5.9

AVERAGE 3.0 331.0 1.2 132.4 0.6 63.7 0.4 42.4



Figure 4.3: Default values used for SHALSTAB.

CHAPTER 5: RESULTS AND DISCUSSION

5.1 Introduction

The purpose of this chapter is to provide model results for SINMAP and



The purpose of this chapter is to provide model results for SINMAP and

SHALSTAB for a variety of parameter ranges, to describe the effectiveness of each

model, and to identify the most sensitive model parameters. SINMAP and SHALSTAB

results are compared herein to identify weaknesses and strengths inherent to each model.

Finally, results will be compared with respect to slope, aspect, geology, and soil type.

5.2 Debris Flow Inventory

A total of 142 debris flows have been mapped in the 7330 km² study area (Figure

5.1 and Appendix D). All of these slope movements have been located using historical

documents and maps, aerial photography, field identification, and data provided by both

the North Carolina Geological Survey and the North Carolina Department of

Transportation. Landslides were identified in all counties except for Avery County. Of

the 42 geologic units that occur within the watershed, debris flows have developed on

fourteen. Most of these units are gneiss with a large component of biotite or muscovite

(coded Ybgg, Ymg, Zabg, Zatm, Zatb, Zybn on the geologic map). Debris flows also

have occurred in quartz diorite (Dqd), metagraywacke (Zatw, Zgs), quartzite (Zsl),

siltstone (Zsp), sandstone (Zsr), and slate (Zwc).

The greatest number of debris flows (52) have occurred in soil map unit NC095,

generally a stony fine-to-sandy loam, and NC093 (46), a stony silt-to-sand loam. The

coarse-grained soils may be attributed to rapid infiltration through these soils. In the thin

soils of the Southern Appalachians, soil pore pressure can quickly increase as

groundwater builds above the impermeable bedrock, increasing the risk of slope failure.



g p , g p

The gradient and aspect of the debris-flow locations have been calculated using

the 10-meter DEM (with ArcGIS Spatial Analyst extension), given that the 30-meter

DEM seems to underestimate slope (Zhang et al, 1999). For the entire watershed, slope

varies from 0 to 74 degrees. Debris flows have occurred on slopes ranging from 10 to 50

degrees, although 88% of the slopes are at least 20 degrees. The average slope on which

debris flows have initiated is 28 degrees.

The aspect slope angle, as calculated in ArcGIS, is the compass direction towards

which a slope faces. The "aspects" of the debris-flow headscarps occur in diverse

directions, but show an affinity toward the east (32), southeast (23), southwest (21), and

south (21). East-facing slopes receive only morning sunlight during the winter and thus

have higher soil moisture than south- and west-facing slopes. During high rainfall events,

this leads to higher levels of antecedent moisture, faster saturation, and greater soil pore-

water pressure, leading to an increased incidence of slope failure. The incidence of debris

flows on south-facing slopes may be due to the direction of storm tracks over the

watershed. Several of the hurricanes that have tracked over western North Carolina have

come from south, out of the Gulf of Mexico (see Figure 3.4).

5 3 SINMAP R lt

summarized in Table 5.1. Although there is an option in SINMAP to create multiple

calibration regions, usually based on the properties of individual soil types, only one

general calibration region has been used in this study. This was deemed appropriate due



to the general nature of the soil data and the regional scale of the study.

SINMAP defines six different stability-class definitions based upon the stability

index (SI) (Table 5.2). The terms “stable”, “moderately stable”, and “quasi-stable”, are

used to represent areas that should not fail under the most conservative input parameters.

The areas modeled as having “lower threshold” and “upper threshold”, are those that

respectively have a <50% or >50% probability of instability. Areas defined as “defended

slopes” are unstable throughout the range of the specified parameter. Where these slopes

occur in the field, they are only stable due to factors not modeled by SINMAP. For

example, they may be bedrock outcrop (Pack et al., 1998b). Parameters are adjusted so

that the resulting map “captures” the maximum amount of observed landslides in regions

with a low SI (stability index), while minimizing the spatial extent of low-stability

regions (Pack et al., 2002).

Initially, SINMAP model runs for this study involved using 30-meter DEMs,

default parameters, and the four aforementioned precipitation thresholds. At the outset of

this study, only 30-meter DEMs were publicly available for western North Carolina. In

the summer of 2004, 10-meter DEMs also became available. As a result, the SINMAP

model runs were completed at both scales to see if the coarseness of the 30-meter DEM

5.3.1 Results – 30-meter DEM

The default SINMAP values seem to underestimate instability in the watershed.

For the 30-meter DEM, the model predicts that 52% (73) of the observed debris flows



would have occurred in the lower threshold for instability whereas none were found in

the “upper” and “defended” stability classes (Figure 5.3). All of other 67 debris flows

(48%) occurred in the more stable stability classes.

In the next SINMAP calculation, a precipitation threshold of 50 mm/d was used

with the 30-meter DEM. Soil friction angle, soil density, cohesion and T/R were adjusted

(Table 5.1). Under these conditions, 93.6% (131) of the inventoried debris flows are

predicted to occur in the unstable classes, but only one of these was predicted to fail

unconditionally. In this model run, 62.7% (4431.8 km²) of the study area is predicted to

be unstable.

Thresholds of 125 mm/d, 250 mm/d, and 375 mm/d were subsequently used with

SINMAP. For all three simulated rain events, 131 observed debris flows occurred in

unstable zones whereas 9 slides (6.4%) were predicted to form in stable zones. Both the

total area of instability and the area predicted to be “defended” increased slightly as

recharge increased. For a precipitation event of 125mm/d, the unstable zones accounted

for 4418.1 km² (62.6%) of the study area while 28.2 km² (0.4) of the watershed was

considered unconditionally unstable (Figure 5.4). The defended class contains four

(2.9%) of the inventoried debris flows.

defended zone increases to 61.4 km² (0.9%) and includes eight (5.7%) of the debris

flows.

In the final calculation, the low value for dimensionless cohesion was increased



slightly from zero to 0.1 for a recharge event of 125 mm/d, with all other properties

remaining the same. This should better model the effects of root cohesion in the

watershed, a factor which often has a strong effect on slope stability, even during heavy

precipitation. With cohesion increased, the area of predicted unstable land decreased to

3170.6 km² or 44.6% of the watershed, but still contained 124 (88.6%) of the inventoried

debris flows. Increased cohesion also increased the stability zones so that 35.7% of the

study area (2534.7 km²) became classified as unconditionally stable.


When comparing the area of the 10-meter grid calculated by SINMAP to the 30-

meter grid, the total area of the 10-meter grid is less than the 30-meter grid (Appendix E).

The 10-meter SINMAP model run seems to have assigned “NO DATA” values instead of

a stability index to many cells in the raster. Moreover, the stability index classes captured

only 122 of the 142 debris flows used in the parameter run. No explanation or repair has

yet been discovered for this problem.

The default results for the 10-meter DEM differ slightly from 30-meter results but

still underestimate instability (Figure 5.5). The same number of debris flows (73)

occurred in unstable stability classes, accounting for 18.4% of the study area (1274.3

As with the 30-meter DEM model, the results for the four precipitation thresholds

are very similar. The extent of the “stable”, “moderately stable”, and “quasi-stable” zones

(2409 km²) and the number of debris flows (7) occurring within these classes are the



same for all four recharge events (Figure 5.6). Only the “lower”, “upper”, and “defended”

stability classes change slightly for each precipitation threshold, shifting to a more

unstable class as the recharge increases. The number of debris flows that fall in the

unstable classes (115) also remains the same for all four recharge events.

Compared to the 30-meter DEM, a larger portion of the watershed is predicted to

be unconditionally unstable in the 10-meter DEM. For the 50 mm/d threshold, 24.8 km²

of the area is predicted to be “defended” although no debris flows fall into this zone. For

125 mm/d of recharge, the “defended” zone increases to 70.8 km² (1.0%) and includes

eight (6.6%) slides. In the calculation for 250 mm/d and 375 mm/d recharge, the

unconditionally unstable area increases to 126.4 km² and 138.8 km², respectively, and

contains 12 (9.8%) of the debris flows.

As with the 30-meter DEM, when the value for dimensionless cohesion is

increased (0.1), the areas of instability decrease while stability increases. Of the 122

inventoried debris flows, 108 (88.5%) fall within the three unstable zones. This is

equivalent to 3278.7 km² (47.7%) of the watershed. More of the watershed falls within

the “defended” stability zones (50.2 km²) than in the 30-meter DEM, whereas 33.6% of

the area is predicted to be unconditionally stable (2308.1 km²).

parameters used. The areas of greatest instability also match well with the steepest terrain

and generally follow steep ridgelines. Overall, the model does well, accurately modeling

approximately 94% of the inventoried debris flows in unstable zones with an SI less than



1.0 for both DEM scales and for all four precipitation thresholds. With a slightly

increased cohesion value, accuracy decreases to about 88%, but landslide density

increases as the model minimizes the extent of the areas mapped as unstable.

The 10-meter DEM models a slightly greater percentage of instability in the

watershed than the 30-meter DEM. Slopes that are derived from DEMs vary with the

spatial resolution, becoming lower at larger pixel sizes (Zhang et al., 1999). Given the

coarser resolution of the 30-meter DEM, the elevation data is more generalized and

slopes are typically underestimated. Using the 30-meter DEM, SINMAP could not

predict small areas of increased instability along narrow ridges and valleys. The greater

resolution of the 10-meter DEM allows for better prediction of unstable areas (2.7%

improvement) and a slightly greater percentage of inventoried debris flows within the

three unstable classification zones (an 0.7% improvement).

In this study, SINMAP is not particularly sensitive to changes in the value for

recharge. Even with significant changes in the precipitation threshold, there was little

change in the predicated areas of instability. This is unexpected because one of the most

important factors in triggering debris flows surely is the rate of recharge. SINMAP was

more sensitive to changes in the values for dimensionless cohesion, a value increased

British Columbia so the default parameters reflect representative soil data for that cool-

climate area (Pack et al., 1998b). Model calibration was based on the thick packages of

coarse subangular till and colluvium found in British Columbia. Even with low values of



hydraulic conductivity ( K ), a large value for soil thickness (b) would result in a greater

transmissivity value. Deeper soils occur in the study areas of Pack et al. (1998b) than in

the present study area, hence their use of T/R default values of 2000 m and 3000 m.

Given these default values with a 2-meter soil thickness and a moderate (125 mm/d)

recharge rate, the hydraulic conductivity lies between 5.21 m/hr and 7.81 m/hr. These

rates are representative of well-sorted sand and gravel but are unrealistic for the French

Broad Watershed (Fetter, 1994). Clearly, there are other factors at work in the Southern

Appalachians that trigger debris flows, factors that are not yet taken into account by

SINMAP.

5.4 SHALSTAB Results

As with the SINMAP model, SHALSTAB was used to model instability using

both a 10-meter and 30-meter DEM. Due to limitations in the SHALSTAB program, only

Haywood County was run in the model (see discussion in Chapter 3). Haywood County

was chosen for its rugged topography and the dispersed nature of the inventoried debris

flows in the county. Slopes range from 0° to 67° with a median of 20°. Some of the

steepest terrain in the French Broad Watershed is located in the northwestern corner of

the county, within the Great Smoky Mountains National Park.

condition. Consequently, if the tangent of the slope (θ ) is greater than or equal to the soil

friction angle (Φ), the slope will be unstable, even under dry conditions (Dietrich and

Real de Asua, 1998). High values of log (q/T ) are the SHALSTAB lower bounds, the



unconditionally “stable” condition. Shallow slopes do not allow for high enough pore

pressure in the soil, even during full saturation, and are rarely unstable (Table 5.3). For

every grid cell, SHALSTAB calculates the amount of critical effective precipitation

necessary to trigger pore-pressure-induced instability, at a constant rate of transmissivity.

Areas with lower values are considered more unstable because they require less

precipitation to cause them to fail than do areas with higher values (Table 5.4).

SHALSTAB requires that landslide locations be supplied to verify the model

results. Instead of simple point locations, polygonal shapes of each landslide scar are

necessary for this model. In the case of Haywood County, 23 landslide polygons were

interpreted and digitized from digital orthophotos and DEMs using ArcGIS 3-D Analyst.

The landslide location has been placed within a stability class based upon the lowest q/T

value that it intersects. The authors of SHALSTAB consider the model to be successful if

the majority of landslides occur in grid cells with low values of log (q/T ) (Dietrich and

Montgomery, 1998).

The four parameters that can be adjusted in SHALSTAB are soil friction angle,

soil density, soil depth, and cohesion. Whereas SINMAP uses a range of input

parameters, SHALSTAB allows for only a single variable to be adjusted and run at a

upper and lower bounds. The parameters used in each SHALSTAB run are summarized

in Table 5.5.




The results for all parameters run in SHALSTAB with a 30-meter DEM are

summarized in Appendix F. Even though 23 landslide polygons were digitized for the

county, only 22 of the landslides were actually calculated by SHALSTAB and used in the

final results.

As with SINMAP, the SHALSTAB default parameters seem to underestimate

instability (Figure 5.7). Seven of the mapped debris flows occur in the unconditionally

stable zone, with land characterized by gradients too low to fail even when saturated.

Collectively, this comprises 974.9 km² (31.82% of the county). All of the other debris

flows occur in stability classes between –3.1 and –2.2, approximately 38% (421 km²) of

the county. No landslides fall into the “chronic instability” zone, the category where areas

are defined as potentially unstable even without the addition of significant rainfall

(Dietrich et al., 2001).

To estimate which parameter values best predict instability in the study area, the

number of cells in each log (q/T ) category have been determined for each parameter run.

The resulting cumulative frequency (percent area) of the county falling into each

instability class is plotted on Figure 5.8 (Dietrich et al., 2001). The cumulative percent of

debris-flows-per-instability category also is calculated (Figure 5.9).

(20) assigned to the “chronic,” the “<–3.1,” and the “-3.1 - -2.8” categories. This is

equivalent to 979.3 km² or 59.7% of the county.

The next test involved increasing cohesion slightly to 2000 (N/m³); this value is



equivalent to a dimensionless cohesion value of 0.1 as used in SINMAP. For soil friction

angles of 26° and 35°, this increase in cohesion decreased the cumulative percentage of

the area and number of landslides predicted to occur in unstable areas. For log (q/T )

values less than –2.8, the area decreased by 13% given a soil friction of 26° and 12%

with a soil friction angle of 35°. Strangely, for a soil friction angle of 45°, there was no

change in the calculated results for values of cohesion between zero and 2000 (N/m³).

This seems to indicate that SHALSTAB is more sensitive to changes in soil-friction angle

than soil cohesion.

The most stable parameters were equivalent to the SINMAP upper bound, i.e., a

soil friction angle of 45° and cohesion equal to 9427.41 N/m³. This high cohesion number

is equal to a dimensionless cohesion value of 0.25 when the soil depth is 2 m and soil

density is 1922 kg/m². All of the landslides and nearly the entire area (99.92%) fell into

the “stable” category.


A benefit of the SHALSTAB program is that the model does not “lose” grid cells,

like SINMAP, during individual calculations. There was a slight loss of grid-cell area

from the 30-meter to the 10-meter DEMs, but not nearly as much as with the SINMAP

the clipping procedure in ArcInfo where some pixels were excluded from the final

clipped DEM in the 30-meter scale due to the coarse nature of the data.

Appendix F summarizes results for a 10-meter DEM and each of the tested

h d f l d i i bili i h



parameters. The default parameters seem to underestimate instability in the county, even

though more of the mapped landslides are captured in higher stability classes. Only two

landslides (8.8%) fall into the “stable” class but the area is slightly smaller than in the 30-

meter model run, i.e., 870.1 km² or 53% of the county. The rest of the debris flows (21)

are contained in the stability classes between –3.1 to –2.2 that encompass 37% (615.7

km²) of Haywood County.

As with the 30-meter DEM, a variety of parameters were tested with the 10-meter

DEM and both the cumulative percent of the county area and the area of landslides were

calculated (Figure 5.10 and Figure 5.11). The smaller grid size of the 10-meter DEM

increases the predicted extent of unstable ground in these latter calculations. The lower

bound values (26, 0) were found to contain the most debris flows in the category of

“chronic instability”: 20 slides were assigned to an area of 534.8 km² (32.6% of the

county). Slightly less of the area, 23%, is predicted to be unconditionally stable. This is a

difference of only 3% from that defined by the 30-meter DEM.

As with the 30-meter DEM, the influence of cohesion on the model was tested by

increasing the value slightly from zero to 2000 N/m³ for each soil friction angle (26°, 35°,

45°). For log (q/T ) values less than –2.8, the area decreased by 12.7% with a soil-friction

` The upper bound parameters, 45° and 9427.41 N/m³ calculate the most stable

land. A majority of the debris flows (21) fell into the “stable” category that comprises

1633.7 km² of the county. But the areas of instability were also increased by 7.2% from

th f d b th 30 t DEM T lid d i th “ 3 1 2 8” d “ 2 5



those found by the 30-meter DEM. Two slides occurred in the “-3.1 – -2.8” and “-2.5 - -

2.2”instability classes and unstable zones with a log (q/T ) of greater than –2.2 make up

5.4 km² of the county (8.5%).

5.4.3 SHALSTAB Interpretation

Like SINMAP, SHALSTAB is more sensitive to changes to some parameters than

others. A significant increase in cohesion causes more of the study area to be designated

as unconditionally stable, but overall the model is less sensitive to this parameter.

SHALSTAB seems to be most sensitive to changes in soil-friction angle. With only a 9

degree increase in friction angle from 26º to 35º, the total extent of predicted instability,

for a log (q/T ) less than –2.8, decreases by approximately 29%. Unconditionally stable

areas increase by around 20%. SHALSTAB calculates the unstable terrain depending on

the tangent value of slope. Given that a greater percentage of a county has a lower

gradient when lower soil-friction angle values are used in the model runs, larger regions

will be calculated as having a potential for instability. Accurate topography, as

determined from the DEM, plays an important part in the SHALSTAB calculation.

The authors of SHALSTAB based their interpretation of the success or failure of

the model on the mapped log (q/T ) results. The model is considered successful if the

the interpretation of model effectiveness is based upon the potential for future mass-

wasting in areas that have not yet failed (Guimaraes et al., 2003). This interpretation is

necessary with the lowest tested values, a soil-friction angle of 26o

and cohesion of zero.

Using these parameters SHALSTAB accurately models 90 9% to 100% of the mapped



Using these parameters, SHALSTAB accurately models 90.9% to 100% of the mapped

debris flows in the upper three instability categories (log (q/T ) less than –2.8). Despite the

fact that these low values could be interpreted as a worst-case scenario for debris-flow

initiation, they seem to overestimate the extent of instability in Haywood County (Figure

5.12).

If one compares the debris-flow density for each instability category, mid-range

parameters capture the most landslides in the smallest area. Using the 10-meter DEM

with a 35° soil-friction angle and soil cohesion value of 2000 N/m³, six debris flows are

captured in the “chronic instability” class within an area of 51.9 km² (3.2% of total area).

This is a landslide density of 0.116 slides per square kilometer, but an accuracy of only

26% for the “chronic instability” region, and 60.9% for the three highest instability

classes (Figure 5.13).

All of these debris-flow scars have been interpreted from DOQs, but field

verification of the landslides may improve the modeled results. SHALSTAB is very

sensitive to the accuracy of mapped debris-flow locations, requiring that the standard for

mapping be higher than normal, particularly when 10-meter or higher-resolution DEMs

are used (Dietrich and Montgomery, 1998b).

comparison of the relative performance of the programs is difficult because they calculate

completely different indices to quantify instability. In order to compare them, the

SINMAP stability index and the SHALSTAB log (q/T ) must be transformed into an

analogous format This can be accomplished using the Spatial Analyst extension of



analogous format. This can be accomplished using the Spatial Analyst extension of

ArcGIS.

For direct comparison, the parameters used in SINMAP and SHALSTAB must be

identical. A moderate soil-friction angle of 35°, a zero value for cohesion, and a soil

density 1922 kg/m³ were chosen. Transmissivity and recharge were also kept at a

constant rate of 1 m²/d and 125 mm/d, respectively, which set the SINMAP T/R value

equal to 200 m. The SHALSTAB log (q/T ) value for these parameters is equal to –2.28.

This procedure is similar to the one used by Dietrich et al. (2001). Figure 5.14 shows a

comparison of the two programs for the same area in Haywood County using identical

parameters. Both programs calculated 705.4 km² (49%) of Haywood County to be

unstable. SHALSTAB calculated slightly more of the county as unstable (870.3 km² or

61%) than SINMAP (740.5 km² or 52%). Visually, the SHALSTAB results appear more

clustered, while the SINMAP results seem more scattered. These differences are due to

the different ways that the models calculate area and slope (Dietrich et al., 2001). Overall,

the calculated results for both programs are strikingly similar i.e., both programs

predicted 81-95% of the same area to be unstable (Figure 5.15).

5 6 G l d S il

“Zchs”, i.e., slate of the Copperhill Formation (Figure 5.16). This unit is a graphitic to

sulfidic slate-to-phyllite found in the Great Smoky Mountains National Park in Haywood

County (North Carolina Geological Survey, 1985). Rock descriptions of the Copperhill

Formation are scarce but in the geologic map of the Great Smoky Mountains of Hadley



Formation are scarce but in the geologic map of the Great Smoky Mountains of Hadley

and Goldsmith (1963), "Zchs" corresponds to a map unit called “pЄa”, i.e., the Anakeesta

Formation. According to Merschat (pers. comm., 2005), a geologist with the North

Carolina Geologic Survey, the slate of the Copperhill and the Anakeesta are separate

formations but are nearly identical. The Anakeesta is part of the Smoky Mountain Group

and the Ocoee Supergroup. It is a pyritic “dark, fine-grained argillaceous rock” with

interbedded metasiltstone and coarse sandstone (Hadley and Goldsmith, 1963). Both the

Anakeesta and the Copperhill formations represent repetitive turbidite sequences at

different stratigraphic levels (Merschat, pers. comm., 2005). Topography in this portion

of the Great Smoky Mountains National Park is exceptionally narrow and steep with

“serrate crests and craggy pinnacles,” and thin residual soils (Hadley and Goldsmith,

1963).

Near the bottom of the Anakeesta, sulfidic argillaceous rock commonly

intertongues with sandstone like that within the underlying Thunderhead sandstone. The

Thunderhead, a member of the Smoky Mountain Group (Zgs), also can be extremely

unstable based on the SINMAP and SHALSTAB results (Figure 5.16). It is also known to

produce boulder fields and slope in western North Carolina and Tennessee.

chemical weathering and could account, in part, for the steep weathered topography and

instability associated with this unit. Transverse and strike jointing also cause angular cliff

exposures in the Anakeesta and Thunderhead sandstone in the Great Smoky Mountains

National Park (Hadley and Goldsmith, 1963).



National Park (Hadley and Goldsmith, 1963).

The most stable geologic units in the French Broad Watershed were found to be

that coded “bz ” on the Geologic Map of North Carolina (1985), i.e., the metamorphosed

rocks of the Brevard Fault Zone, and that coded “CZtp”, i.e., porphyroblastic gneiss of

the Sauratown Formation. Both of these units are located in the southeastern portion of

the watershed and lie within the geomorphologic region known as the Western Piedmont.

The Brevard Fault zone seems to form a dividing line between moderately unstable rock

and stable rock (Figure 5.16).

Both SINMAP and SHALSTAB predict the most unstable soil unit in the

watershed to be that coded NC104, i.e., a general soil group that occurs in northern

Haywood County, generally overlying the Copperhill Formation. This soil unit is

principally composed of the Tanasee (21%), Burton (19%), Oconaluftee (16%), and

Porters (14%) soil series (U.S. Department of Agriculture, 1998) (see Appendix B). The

Tanasee soil series is sandy loam derived from colluvium and often forms toe slopes, fans

and benches on coves in the high elevations of the Appalachians. The Burton and Porters

series are a sandy-to-fine loam derived from residuum on ridge and side slopes. The

Ocanaluftee is a channery loam, also derived from residuum, or from slate, phyllite, or

(U.S. Department of Agriculture, 2005). All of the stable soil units are located on the flat

topography along the floodplain of the French Broad and Pigeon Rivers (Figure 5.17).

5.7 Jointing, Fracturing and Foliation

One factor that neither SINMAP nor SHALSTAB takes into consideration is the



influence of geologic structure on groundwater flow. Groundwater flow through a

fracture is notoriously difficult to model because the rate of transmissivity is strongly

controlled by the size of the fracture aperture and the connectivity of the fracture network

(Renshaw, 2000). Nonetheless, flow through joints and fractures, and over bedrock

foliation planes parallel to the dip slope, may play a significant role in triggering debris

flows in western North Carolina (Figure 5.18).

In the Great Smoky Mountains National Park, it was noted that the heads of

debris flows originate at the intersection of cleavage and joints or beds, or on the

cleavage or bedding plane (Southworth et al, 2003). Intersecting joints and fractures can

cause groundwater flow to become concentrated in topographic hollows. During a heavy

rainfall event, these hollows could periodically be “flushed out” of weathered colluvial

material, triggering a debris flow. Foliation planes provide a smooth surface on which

sliding may initiate, particularly if the dip angle for bedding is close to the dip of the

foliation.



Figure 5.1: Landslide inventory map for the French Broad Watershed. The cluster of location pointsin southern Buncombe County is due to the extensive mapping of 1977 debris flows done by Pomeroy

(1991) and Otteman (2001).



Figure 5.2: A comparison of 30-meter and 10-meter DEMs in Haywood County, North Carolina. In

the coarser 30-meter DEM, there is a great deal of pixelation. Lower resolution can lead to an

underestimation of both the slope and instability in the study area.

Table 5.1: The parameters used in all of the SINMAP model runs.

Precipitation

Threshold

Soil

Density

T/R

Max.

T/R

Min.

Cohesion

Max.

Cohesion

Min.

ΦMax. ΦMin.

Default 2000 3000 2000 0 .25 30 4550mm 1922 331 3 0 .25 26 45

125mm 1922 132.4 1.2 0 .25 26 45

125mm 1922 132 4 1 2 10 25 26 45

Table 5.2: SINMAP stability index definitions (Pack et al., 1998b).

Condition Class Predicted

State

Parameter

Range

Possible Influence

of Factor Not

Modeled

Map

Color

SI > 1.5 1 Stable SlopeZone

Range cannotmodel instability

Significantdestabilizing factorsrequired for

green



required for instability

1.5 > SI >1.25

2 Moderatelystable slopezone


Moderatedestabilizing factorsrequired for instability

blue

1.25 > SI >1.0

3 Quasi-stableslope zone


Minor destabilizingfactors could lead toinstability

yellow

1.0 > SI >0.5

4 Lower thresholdslope zone

Pessimistic half of range requiredfor instability

Destabilizing factorsare not required for instability

pink

0.5 > SI >

0.0

5 Upper

thresholdslope zone

Optimistic half of

range required for stability

Stabilizing factors

may be responsiblefor stability

red

0.0 > SI 6 Defendedslope zone

Range cannotmodel stability

Stabilizing factorsare required for stability

tan



Figure 5.3: SINMAP results for a 30-meter DEM and using default parameters.



Figure 5.4: SINMAP results for 125 mm/d recharge and using a 30-meter DEM.



Figure 5.5: SINMAP results for a 10-meter DEM and using default parameters.



Figure 5.6: SINMAP results for 125 mm/d recharge using a 10-meter DEM.

Table 5.3: Mapped instability classes used in the SHALSTAB model analysis.

Log (q/T ) Interval Color

Chronic Instability Dark red

< - 3.1 Red

Table 5.4: Table comparing q/T and log (q/T ) values and the precipitation rate required to initiate

instability for soils with a transmissivity of 65 m²d and 17 m²/d (after Dietrich and Asua, 1998).

q/T (1/m) log (q/T ) (1/m) Precip for T =

65 m²/d (mm/d)

Precip for T =

17 m²/d (mm/d)

.00079 -3.1 52 14

.00158 -2.8 103 27



.00316 -2.5 206 54

.00833 -2.2 410 103

.01266 -1.9 818 214

Table 5.5: Parameters used in the SHALSTAB model runs.

Soil Friction

Angle

Density

(kg/m³)

Soil Depth (m) Cohesion

(N/m²)

*45 1700 1 0**45 1922 2 9427.41

45 1922 2 2000

45 1922 2 0

35 1922 2 2000

35 1922 2 0

26 1922 2 2000

***26 1922 2 0

* SHALSTAB default values** SINMAP upper bounded values*** SINMAP lower bounded values





Figure 5.8 Cumulative percent of Haywood County found in each log (q/T ) category for a variety of soil parameters for the 30-meter DEM (after Dietrich et al., 2001). In the legend, the first number is

the degree if soil friction angle and the second number is the amount of cohesion (N/m³).



Figure 5.10: Cumulative percent of the area if Haywood County for each log (q/T ) instability

category for a variety of soil parameters for the 10-meter DEM (after Dietrich et al., 2001).







Figure 5.14: Comparison of the output of SINMAP and SHALSTAB for 125 mm of recharge, 35° soil

friction angle, 1922 kg/m³ soil density, and zero soil cohesion for a location in Haywood County. The

red areas are calculated as unstable by both programs whereas the gray areas are calculated to be

stable. Visually the SHALSTAB results seem to cluster better while the SINMAP results are more

scattered. Overall, the results are very similar.





Figure 5.16: The mean stability index for each geologic unit in the French Broad Watershed. The

most unstable unit, Zchs, is located in the northwestern portion of Haywood County. The most stable

units, bz and Ctzp, are located in the southeastern portion of the study area.



Figure 5.17: The mean stability index for each soil unit in the French Broad Watershed. The most

unstable soil unit in the watershed is NC104, which is located in western Haywood County. The most

stable soils are located around the floodplains of the Pigeon and French Broad Rivers.



Figure 5.18: Picture taken in October 2003 along the Blue Ridge Parkway, near Mt. Mitchell. Even

during this light rain event, water is pouring out from fractures in the rock.

CHAPTER 6: CONCLUSIONS

Debris flows and other forms of mass-wasting in western North Carolina are a

destructive force that deserves more attention from both local and federal-level planners

and citizens living in these mountainous areas. Debris flows are difficult to predict and

have such a low frequency that it is rare for them to recur in the same area within the



q y

average life span of area residents. Unlike a flood-hazard map, a landslide-hazard map is

inherently imprecise with regard to the extent, location, and timing of future mass-

wasting. After the tragic results of the September 2004 hurricane season, North Carolina

Senate Bill 7, the Hurricane Recovery Act of 2005 was written and passed (available

online at http://www.ncga.state.nc.us/). It recognizes the need for better understanding of

debris flows and provides funds for disaster relief and to identify areas of potential slope

instability.

This study has incorporated historical rainfall data, geology, soil types, and

geomorphology to help predict debris flows in the French Broad Watershed. A

combination of factors have been found to trigger a debris flow:

1. Topography is the most important factor in generating a debris flow. Slopes

with historic incidences of mass-wasting have averaged 28 degrees in this

study area but range up to 50 degrees.

2. Intense precipitation, generally greater than 125 mm/d, tends to increase the

probability of debris flows. This is especially true when a second large storm

pressure, decreases cohesion, and may induce slope failure. Steeply sloping

areas may be of potential concern under these certain weather conditions and

should be closely monitored.

3. Both SINMAP and SHALSTAB predict that certain geologic units are more

prone to failure than others. Sulfidic shale beds of the Copperhill Formation in



the Great Smoky Mountains National Park are particularly susceptible, largely

due to their steep topography, thin soil cover, and vulnerability to chemical

weathering.

4. Both SINMAP and SHALSTAB predict that certain soil units are more prone

to failure than others. Particularly vulnerable to mass-wasting are those that

form at high elevations on steep slopes, are well drained, have a moderate to

rapid rate of permeability, and develop over less permeable bedrock.

5. The movement of groundwater may be concentrated through joints, fractures,

and foliation, quickly increasing pore pressure and decreasing cohesion and

friction between the soil and underlying bedrock. Areas with thin soils

overlying fractured bedrock may be more prone to slope instability.

SINMAP and SHALSTAB have both proven adequately accurate in the

prediction of slope instability on the regional scale of the French Broad Watershed. Both

ArcView and ArcGIS provide useful visual representations of landslide-hazard maps that

may readily be updated as new information becomes available. The performance of both

both slope and relief. A finer 10-meter grid maximizes the overall extent of

predicted instability, but significantly increases computer-processing time.

Extremely large 10-meter DEMs, covering more than one county, may cause

either computer program to abort.

2. The modeled results for the default SINMAP and SHALSTAB parameter



values underestimate the extent of instability in the study area.

3. A SINMAP stability index of less than 1.0, designating instability, has been

calculated for 88% -to- 94% of the inventoried landslides whereas the

corresponding stability threshold in SHALSTAB has designated 91% to 100%

of the mapped debris flows. The SHALSTAB threshold is a log (q/T ) value

less than –2.8 for a soil-friction angle of 26° and cohesion of zero. Overall,

these results seem to over-predict the areas of instability.

4. The results from both models are similar, calculating 81-95% of the same area

of Haywood County as unstable. Visually, the SINMAP results seem

dispersed whereas the SHALSTAB results are more clustered.

The advantages and disadvantages of SINMAP and SHALSTAB are discussed

below and summarized in Table 6.1. Neither model was conclusively found to predict

instability better than the other and each should be used with caution and with the most

accurate input data available.

1. Some of the advantages of the SINMAP model are as follows:

• It can run large DEMs, e.g., watershed size, without aborting the program

or causing errors.

• Recharge values are used in the calculation of the T/R values so these

precipitation thresholds can be tested.

2. Some of the disadvantages of the SINMAP model are as follows:



• During some model runs, some grid cells calculated as “NO DATA,”

effectively changing the area calculated by the model.

• The model was created for study areas that had thick soil packages and

high transmissivity rates, neither of which characterize the French Broad

Watershed.

3. Some of the advantages of the SHALSTAB model are as follows:

• When run with only the default values, the user needs little knowledge of

the actual soil parameters, or even mapped landslides, to produce a map of

relative instability.

• No grid cells are “lost” in the SHATSTAB model calculation, making a

direct comparison of total watershed area between the 10-meter and 30-

meter DEM easier and more accurate.

4. Some of the disadvantages of the SHALSTAB model are as follows:

• To obtain more accurate results, specific soil-parameter data must be

introduced into the model however the model does not allow for a range

5. Neither model takes into account either antecedent moisture or the effect that

geologic structure can have on concentrating groundwater flow. Nonetheless,

both of these factors probably have a significant effect on instability.

Deterministic slope-stability analytical methods, e.g., SINMAP and SHALSTAB,

are generally more useful in areas where ground conditions are fairly uniform throughout



the study area (Dai et al., 2002). Due to the scale of this study, assumptions and

generalizations about soil parameters had to be made, thus forcing these parameters to be

uniform over a large area. From this broad reconnaissance study, the French Broad

Watershed offers several avenues for further study:

1. Subsequent studies should use a smaller study area, so that a site-specific soil

and geologic analysis may be completed. Precise measurements of hydraulic

conductivity (permeability), soil density, soil-friction angle, and soil cohesion

increase the accuracy of the predicted SINMAP and SHALSTAB results.

2. An accurate and detailed debris flow inventory should be constructed when

using these models. Verification of the modeled results using actual debris-

flow locations is essential.

3. Mapping the extent of “ancient” debris-flow deposits in the study area could

be used to better understand the return interval for such catastrophic events.

4. High-quality (finer-resolution) precipitation data must be acquired. Although

generalized rainfall data is useful, debris flows are often generated by intense,

topographic information. This kind of data would include 10-meter DEMs or

smaller and LIDAR data.

Table 6.1: The advantages and disadvantages of SINMAP and SHALSTAB.

SINMAP SHALSTAB

Advantages Disadvantages Advantages Disadvantages



Advantages Disadvantages Advantages Disadvantages

Allows for a rangeof input values/

recharge thresholdscan be tested

Requires knowledgeof specific

soil/recharge parameters

Can be run withoutknowing soil

parameters

Does not allow for arange of input

values

Factor of safety easyto interpret

Must adjust parametersrepeatedly tocapture landslides

Does not requirerecharge values, can back-calculate

Log (q/T ) valuedifficult to interpret

Can run large DEMs “NO DATA” gridcells/shifting

Retains calculatedgrid cell area

Aborts when usinglarge DEMs

--

Does not take intoaccount geologicstructure or antecedent moisture

--

Does not take intoaccount geologicstructure or antecedent moisture

REFERENCES

Appt, J., Skaugset, A., Pyles, M., 2002, Discriminating between landslide sites andadjacent terrain using topographic variables: Geological Society of America, Abstracts

with Programs, v. 34, no. 5, p. 93.

Associated Press, 2004, Frances’ Remnants Cause Flooding in N.C., viewed September 9, 2004 at http://www.abcnews.com.

Badgett P Locklear B Blaes J 2004 September 2004 NC Weather Review 8 p



Badgett, P., Locklear, B., Blaes, J., 2004, September 2004 NC Weather Review, 8 p.,viewed January 27, 2005 athttp://www.erh.noaa.gov/rah/climate/data/MonthlySummary.Sep.2004.pdf .

Ball, J., 2004, Collapsed sections of parkway force travelers to detour: Asheville Citizen-Times, viewed September 14, 2004 athttp://www.citizentimes.com/cache/article/print/61363.shtml.

Barrett, M., 2004, Old Fort mudslide could be clear by noon; many roads still closed:Asheville Citizen-Times, viewed September 9, 2004 athttp://www.citizentimes.com/cache/ article/print/61068.shtml.

Behling, R. E., Kite, J, S., Cenderelli, D.A., Stuckenrath, R., 1993, Buried organic-richsediments in the unglaciated Appalachian Highlands: A Stratigraphic model for finding pre-late Wisconsin paleoenvironmental data: Geological Society of America,Southeastern Section, Abstracts with Programs, v. 25, no. 6, p. 60.

Biesecker, M. and Shaffer, J., 2004, Ivan starts fatal slide: The News & Observer, viewedSeptember 20, 2004 at http://www.newsobserver.com/print/sunday/front/v-

printer/story/1652067p-7880420c.html.

Boyle, J., 2004, WNC on alert for floods: Asheville Citizen-Times, viewed September 8,2004 at http://www.citizen-times.com/cache/article/print/60934.html.

Braun, D. D., 1989, Glacial and periglacial erosion of the Appalachians: Geomorphology,v. 2, p. 233-256.

Buol, S.W., Graham, R.C., McDaniel, P.A., Southard, R.J., 2003, Soil Genesis andClassification: Iowa, Iowa State Press, 494 p.

Chinnayakanahalli, K., Tarboton, D.G., Pack, R.T., 2003, An objective method for theintercomparison of terrain stability models: San Francisco, Eos Transactions, AGU FallMeeting, 8-13 December.

Clark, G.M., 1987, Debris Slide and debris flow historical events in the Appalachians

south of the glacial border in Costa J.E. and Wieczorek G.F., eds, Debrisflows/avalanches: process, recognition, and mitigation: Geological Society of America,Reviews in Engineering Geology, v. 7, p. 125-138.

Clark, M.G., Ryan, P.T., Drumm, E.C., 1987, Debris slides and debris flows on



, , y , , , , ,Anakeesta Ridge, Great Smoky Mountains National Park, Tennessee in Schultz, A. P andSouthworth, S.C. eds., Landslides of Eastern North America: Geological Survey Circular 1008, U.S. Dept. of the Interior, U.S. Geological Survey, Denver, CO, p. 18-19.

Clark, M.G., and Torbett, C.A., 1987, Block fields, block slopes, and block streams in theGreat Smoky Mountains National Park, North Carolina and Tennessee in Schultz, A. Pand Southworth, S.C. eds., Landslides of Eastern North America: U.S. Dept. of theInterior, Geological Survey Circular 1008, U.S. Geological Survey, Denver, CO, p. 20-21.

Clingman, T.L., 1877, Selections from the speeches and writings of the HonorableThomas L. Clingman of North Carolina with Additions and Explanatory Notes: John Nichols, Raleigh, N.C.

Crozier, M.J., 1986, Landslides: causes, consequences, and environment: Croon Helm,Dover, N.H., 252 p.

Cruden, D.M. and Varnes, D.J., 1996, Landslide Types and Processes, in Turner A.K.,

and Schuster, R. J., eds., Landslides: Investigation and Mitigation - Special Report 247:Washington, D.C., Transportation Research Board, National Research Council, p. 36-75.

Dai, F.C., Lee, C.F., Ngai, Y.Y., 2002, Landslide risk assessment and management: anoverview: Engineering Geology, v. 64, p. 65-87.

Daniels, R.B., Buol, S.W., Kleiss, H.J., Ditzler, C.A., 1999, Soil Systems in NorthCarolina: North Carolina State University, Soil Science Department, Raleigh, NC, 118 p.

Dietrich, W. E. and Real de Asua, R., 1998, A validation of the shallow slope stabilitymodel, SHALSTAB, in forested lands of Northern California,

Dietrich, W.E. and Montgomery, D.R., 1998, SHALSTAB: A digital terrain model for mapping shallow landslide potential, viewed October 23, 2003 athttp://socrates.berkeley.edu/ ~geomorph/shalstab/.

Eaton, L.S., Kochel, R.C., Howard, A.D., Sherwood, W.C., 1997, Debris flow and

stratified slope wash deposits in the central Blue Ridge of Virginia, Geological Society of America, Abstracts with Programs, v. 29, no. 6, p. 410.

Easterbrook, D.J., 1999, Surface Processes and Landforms: Macmillan PublishingCompany, New York, NY, 546 p.



p y, , , p

Eschner, A.R. and Patric, J.H., 1982, Debris Avalanches in Eastern Upland Forests:Journal of Forestry, v. 80, p. 343-347.

Fetter, C.W., 1994, Applied Hydrogeology: Prentice Hall, Englewood Cliffs, NJ.

Glass, F.R., 1981, Unstable Rock Slopes Along Interstate 40 Through Pigeon River Gorge, Haywood County, North Carolina in Proceedings, 32nd Annual HighwayGeology Symposium, v. 32: Atlanta, GA, Highway Geology Symposium, 15 p.

Graham, R.C., Daniels, R.B., Buol, S.W., 1990, Soil-Geomorphic Relations on the BlueRidge Front: I. Regolith Types and Slope Processes: Soil Science Society of AmericaJournal, v. 54, p. 1362-1367.

Graham, R.C. and Buol, S.W., 1990, Soil-Geomorphic Relations on the Blue RidgeFront: II. Soil Characteristics and Pedogenesis: Soil Science Society of America Journal,v. 54, p. 1367-1377.

Grant, W.H., 1988, Debris Avalanches and the Origin of First-Order Streams in Swank,W.T. and Crossley, D.A., eds., Forest hydrology and ecology at Coweeta: New York, N.Y., Springer-Verlag, p. 103-110.

Greenway, D.R., 1987, Vegetation and Slope Stability, in Anderson, M.G., and Richards,K.S., eds., Slope Stability: geotechnical engineering and geomorphology: Chichester, New York, John Wiley & Sons Ltd., p. 187-517.

Gryta, J.J. and Bartholomew, M.J., 1983, Debris-Avalanche type features in WataugaCounty North Carolina, in Lewis, S.E., ed., Geological Investigations in the Blue Ridgeof Northwestern North Carolina: Carolina Geological Society Field Trip Guidebook:

Gryta, J.J. and Bartholomew, M.J., 1989, Factors influencing the distribution of debrisavalanches associated with the 1969 Hurricane Camille in Nelson County, Virginia inLandslide Processes of the Eastern United States and Puerto Rico, Special Paper -Geological Society of America, no. 236, p. 15-28.

Guimaraes, R.F., Montgomery, D.R., Greenberg, H.M., Fernandes, N.F., Gomes, R.A.T.,Junior, O.A., 2003, Parameterization of soil properties for a model of topographiccontrols on shallow landsliding: application to Rio de Janeiro: Engineering Geology, v.69, p. 99-108.



Hack, J. T., 1980, Rock control and tectonism; Their importance in shaping theAppalachian Highlands: US Geologic Survey Professional Paper 1126-B, 17 p.

Hadley, J.B., and Goldsmith, R., 1963, Geology of the Eastern Great Smoky Mountains, North Carolina and Tennessee: U.S.Geological Survey Professional Paper 349-B, p.

Hammond, C., Hall, D., Miller, S., Swetik, P., 1992, Level 1 Stability Analysis (LISA)documentation for version 2.0 Gen. Tech. Rep. INT-285: Ogden, UT, Department of Agriculture, Forest Service, Intermountain Research Station, 190 p.

Hatcher, R.D., Jr., 1987, Bedrock geology and regional geologic setting of CoweetaHydrologic Laboratory in the eastern Blue Ridge in Swank, W.T. and Crossley, D.A.,eds., Forest Hydrology and Ecology at Coweeta, p. 81-92.

Hatcher, R.D., Jr., Carter M. W., Clark, G.M., Mill, H.H., 1996, Large landslides inwestern Blue Ridge of TN & NC; normal mass-wasting phenomena, products of latePleistocene climates or smoking gun for earthquake(s) in East TN?, Geological Societyof America, Abstracts with Programs, v. 28, no. 7, pg. 299.

Hatcher, R.D., Jr. and Goldberg, S.A., 1994, The Blue Ridge Geologic Province in Horton, J.W. Jr. and Zullo, V.A., eds., The Geology of the Carolinas: The University of Tennessee Press, Knoxville, TN, p. 11-35.

Highland, L.M., Ellen, S.D., Christian, S.B., Brown, W.M., 2004, Debris-Flow Hazardsin the United States, U.S. Geological Survey Fact Sheet 176-97, viewed November 10,2003 at http://geohazards.cr.usgs.gov/factsheets/html_files/debrisflow/fs176-97.html.

Holmes, J.S., 1917, Some notes on the occurrence of landslides: Journal of ElishaMitchell Society, v. 33, p. 100-105.

Administration, National Weather Service, Eastern Region Headquarters, ScientificServices Division, Bohemia, N.Y., 83 p.

Jackson, J.A., 1997, Glossary of Geology: Alexandria, VA, American GeologicalInstitute, 769 p.

Jacobson, R.B., Cron, E.D., McGeehin, J.P., 1989, Slope movements triggered by heavyrainfall, November 3-5, 1985, in Virginia and West Virginia, U.S.A.: Geological Societyof America, Special Paper 236, 13 p.



Jacobson, R.B., Miller, A.J., Smith, J.A., 1989, The Role of catastrophic events in CentralAppalachian landscape evolution, in Gardner, T.W., and Sevon, W.D. eds. AppalachianGeomorphology, v. 2, p. 257-284.

Kaya, A., 1991, Rock slope stability analysis along the Pigeon River, Interstate Highway40 in Western North Carolina: M.S. Thesis, Purdue University, West Lafayette, IN, 127 p.

Kochel, R.C., 1984, Quaternary debris avalanche frequency, sedimentology, andstratigraphy in Virginia, Geological Society of America, Abstracts with Programs, v. 16,

no. 6, p. 562.

Kochel, R.C., 1987, Holocene debris flows in Central Virginia: Geological Society of America, Reviews in Engineering Geology, v. 7, p. 139-155.

Kochel, R.C., 1990, Humid fans of the Appalachian Mountains in Rachocki, A.H., andChurch, M. eds., Alluvial Fans: A field approach: New York, NY, John Wiley & SonsLtd., p. 109-129.

Kochel, R.C., and Johnson, R.A., 1984, Geomorphology and sedimentology of humid-temperate alluvial fans, central Virginia in Koster, E.H. and Steel, R.J., eds.,Sedimentology of gravels and conglomerates: Canadian Society of Petroleum Geologists,v. 10, p. 109-122.

Leith, C.J., Fisher, C.P., Deal, C.S., Meyer, H.D., 1965, An investigation of the stabilityof highway cut slopes and embankments in North Carolina: Raleigh, NC, Engineering

Research Department of North Carolina State University, 30 p.

Liebens, J. and Schaetzl, R.J., 1997, Relative-age relationships of debris flow deposits in

Mazza, N.A. and Wieczorek, G.F., 1997, A volumetric sediment analysis for Kinsey Rundebris flow, triggered during an extreme storm event, Madison County, Virginia:Geological Society of America, North Eastern Section, Abstracts with Programs, v. 29,no. 1, p. 64.

Mills, H.H., 1982, Piedmont-cove deposits of the Dellwood quadrangle, Great SmokyMountains, North Carolina, USA: Morphometry: Zeitschrift fuer Geomorphologie, v. 26,no. 2, p. 163-178.

Mill H H 2000 O h di ib i f ld ll i l d i i h A l hi



Mills, H.H., 2000, On the distribution of old colluvial deposits in the Appalachians:Abstracts with Programs – Geological Society of America, v. 32, no. 7, p. 182.

Mills, H.H. and Allison, J.B., 1995a, Controls on the variation of fan-surface age in theBlue Ridge Mountains of Haywood County, North Carolina: Physical Geography, v. 15,no. 5, p. 465-480.

Mills, H.H. and Allison, J.B., 1995b, Weathering rinds and the evolution of Piedmontslopes in the Southern Blue Ridge Mountains: The Journal of Geology, v. 103, p. 379-394.

Mills, H.H., Brakenridge, G.R., Jacobson, R.B., Newell, W.L., Pavich, M.J., Pomeroy,J.S., 1987, Appalachian mountains and plateaus, in Graf, W.L, ed., Geomorphic systemsof North America: Boulder, CO, Geological Society of America, Centennial SpecialVolume 2, p. 5-50.

Montgomery, D.R. and Dietrich, W.E., 1994, A physically based model for thetopographic control on shallow landsliding: Water Resources Research, v. 30, no. 4, p.

1153-1171.

Montgomery, D.R., Greenberg, H.M., Laprade, W.T., Nashem, W.D., 2001, Sliding inSeattle: Test of a Model of Shallow Landsliding Potential in an Urban Environment inWigmosta, M.S. and Burges, S. J. eds., Landuse and Watersheds: Human Influence onHydrology and Geomorphology in Urban and Forest Areas: Washington, D.C., AmericanGeophysical Union, v. 2, p. 59-73.

Morrissey, M.M., Wieczorek, G.F. and Morgan, B.A., 2001, A comparative analysis of hazard models for predicting debris flows in Madison County, Virginia: U.S. GeologicalSurvey Open-File Report 01-0067, 15 p.

National Weather Service, 2004b, Preliminary storm total rainfall amounts for thewestern Carolinas and northwest Georgia…associated with Frances, viewed September 23, 2004 at http://www.erh.noaa.gov/gsp/localdat/cases/6-8Sep2004/frances.htm.

National Weather Service, 2004c, Rainfall totals from around the area associated with theremnants of Ivan, viewed September 23, 2004 athttp://www.erh.noaa.gov/gsp/localdat/cases/16-17Sep2004/ivan.htm.

Neary, D.G. and Swift, L.W., 1987, Rainfall thresholds for triggering a debrisl hi t i th S th A l hi M t i G l i l S i t f



avalanching event in the Southern Appalachian Mountains: Geological Society of America, Reviews in Engineering Geology, v. 7, p. 81-92.

Neary, D.G., Swift, L.W., Jr., Manning, D.M., Burns, R.G., 1986, Debris avalanching inthe Southern Appalachians: An Influence on Forest Soil Formation: Soil Science Societyof America Journal, v. 50, p. 465-471.

NCDC Climate Data Online, 2003, Monthly surface data, selected climate divisions in North Carolina, viewed October 23, 2003 at http://cdo.ncdc.noaa.gov/pls/plclimprod.

North Carolina Geological Survey, 1985, State Geologic Map of North Carolina: Raleigh, North Carolina Geological Survey, scale 1:500,000.

North Carolina Geological Survey, 2004a, Landslides from Hurricane Frances, viewedJanuary 23, 2005 athttp://www.geology.enr.state.nc.us/Geologic_hazards_Landslides_Hurricane_Frances_1/Landslides_Hurricane_Frances.htm.

North Carolina Geological Survey, 2004b, Landslides from Hurricane Ivan, viewedJanuary 23, 2005 athttp://www.geology.enr.state.nc.us/Geologic_Hazards_Landslides_Hurricane_Ivan/Landslides_Hurricane_Ivan.htm.

Nowell, P., 2004, Far-stretching Frances disrupts N.C. from mountains to coast, viewedSeptember 6, 2004 athttp://abclocal.go.com/wtvd/news/print_090904_Apstate_francesroundup.html.

Ostendorff, J., 2004, Flood work in Peeks Creek moves from rescue to recovery; one manremains missing: Asheville Citizen-Times.

Terrain stability and forest management in the interior of British Columbia: WorkshopProceedings, May 23-25, 2001, Nelson, British Columbia, Canada, no. 3, p. 209-210.

Pack, R.T., Tarboton, D.G., Goodwin, C.N., 1998a, The SINMAP approach to terrainstability mapping: 8

thCongress of the International Association of Engineering Geology,

Vancouver, British Columbia, Canada.

Pack, R.T., Tarboton, D.G., Goodwin, C.N., 1998b, Terrain stability mapping withSINMAP, technical description and users guide for version 1.00: Terratech ConsultingLtd., Salmon Arm, B.C., Canada, Report Number 4114-0, 68 p. (Report and softwareavailable online: http://moose cee usu edu/sinmap/sinmap htm)



available online: http://moose.cee.usu.edu/sinmap/sinmap.htm)

Pack, R.T., Tarboton, D.G., Goodwin, C.N., 2001, Assessing terrain stability in a GISusing SINMAP: Presented at the 15th Annual GIS Conference, GIS 2001, February 19-22, Vancouver, B.C., Canada, 9 p.

Pierson, T.C. and Costal, J.E., 1987, A rheologic classification of subaerial sediment-water flows: Geological Society of America, Reviews in Engineering Geology, v. 7, p.1-12.

Pomeroy, 1991, Map showing late 1977 debris avalanches southwest of Asheville,western North Carolina: U.S. Geological Survey Open File Report 91-334, scale1:24,000.

Renshaw, C.E., 2000, Fracture spatial density and the anisotropic connectivity of fracturenetworks in Faybishenko, B., Witherspoon, P.A., Benson, S.M., eds., Dynamics of fluidsin fractured rocks: Washington, D.C., American Geophysical Union, p. 203-211.

Ritter, D.F., Kochel, R.C., and Miller, J.R., 1995, Process geomorphology: ThirdEdition, Wm. C. Brown Publishers, Dubuque, Iowa, 546 p.

Schuster, R.L., 1996, Socioeconomic significance of landslides in Turner A.K., andSchuster, R.J., eds., Landslides: Investigation and Mitigation - Special Report 247:Washington, D.C., Transportation Research Board, National Research Council, p. 12-35.

Schuster, R.L. and Kockelman, W.J., 1996, Principles of Landslide Hazard Reduction in

Turner A.K., and Schuster, R.J., eds., Landslides: Investigation and Mitigation - SpecialReport 247: Washington, D.C., Transportation Research Board, National ResearchCouncil, p. 91-105.

Sharpe, C.F.S., 1938, Landslides and related phenomena: a study of mass-movements of soil and rock: New York, Columbia University Press, 137 p.

Sidle, R.C., Pearce, A.J., O’Loughlin, C.L., 1985, Hillslope stability and landuse: Water

Resources Monograph 11 Edition, American Geophysical Union, p. 1940.

Southern Railway, 1917, The floods of July 1916: how the Southern Railwayorganization met an emergency: Washington, D.C., Southern Railway Company, 131 p.

Southworth S Schultz A Denenny D Triplett J 2003 Surficial geologic map of the



Southworth, S., Schultz, A., Denenny, D., Triplett, J., 2003, Surficial geologic map of theGreat Smoky Mountains National Park region, Tennessee and North Carolina, U.S.Geological Survey Open File Report 03-381: Reston, VA, U.S. Geological Survey, 44 p.

State Climate Office of North Carolina, 2003, Aspects of NC Climate Extreme weather records, viewed October 17, 2003 at http://www.nc-climate.ncsu.edu/climate/extremes.html.

Stewart, J.M., Heath, R.C., Morris, J.N., 1978, Floods in Western North Carolina, November 1977: Water Resources Research Institute of the University of North Carolina,

25 p.

Swift, L.W., Cunningham, G.B., Douglass, J.E., 1988, Climatology and hydrology atCoweeta in Swank, W.T. and Crossley, D.A., eds., Forest hydrology and ecology atCoweeta: New York, N.Y., Springer-Verlag, p.

Tennessee Valley Authority, Division of Water Control Planning, 1960, Floods onFrench Broad and Swannanoa Rivers in vicinity of Asheville, North Carolina: Knoxville,

TN, Tennessee Valley Authority, 88p.

U.S. Census Bureau, 2000, County and city data books (electronic resource):Washington, D.C., viewed November 16, 2002 athttp://www.census.gov/statab/www/ccdb.html.

U.S. Department of Agriculture, 1902, Message from the President of the United Statestransmitting a report of the Secretary of Agriculture in relation to the forests, rivers, and

mountains of the southern Appalachian region: Washington, Government Printing Office,210 p.

U.S. Department of Agriculture – NRCS Soil Survey Division, 2005, Official Soil SeriesDescriptions, viewed March 1, 2005 at http://ortho.ftw.nrcs.usda.gov/cgi- bin/osd/osdname.cgi.

U. S. Geological Survey - Water Resources Branch, 1949, Floods of August 1940 in the

Southeastern States: Geological Survey Water – Supply Paper 1006: U.S. GovernmentPrinting Office, Washington D.C., 554 p.

U.S. Geological Survey, 1999, National Elevation Dataset: United States GeologicalSurvey (USGS), EROS Data Center: United States Geological Survey, Sioux Falls, SD,Available Online: http://gisdata usgs net/ned/



Available Online: http://gisdata.usgs.net/ned/.

Varnes, D.J., 1978, Slope movement types and processes in Schuster, R., and Krizek, R.,eds., Landslides: Analysis and Control: Washington, D.C., National Academy of Science, p. 12-33.

Wieczorek, G.F., 1996, Landslide triggering mechanisms, in Turner, A.K. and Schuster,R.J., eds., Landslides: Investigation and Mitigation: Washington, D.C., TransportationResearch Board Special Report 247, National Research Council, p. 76-90.

Wieczorek, G.F., Mossa, G.S., Morgan, B.A., 2004, Regional debris-flow distributionand preliminary risk assessment from severe storm events in the Appalachian Blue RidgeProvince, USA: Landslides, v. 1, pp. 53-59.

Williams, G.P. and Guy, H.P., 1973, Erosional and depositional aspects of HurricaneCamille in Virginia, U.S.A: U.S. Geological Survey Professional Paper 804, 80 p.

Woodruff, J.F., 1971, Debris avalanches as an erosional agent in the Appalachian

Mountains: The Journal of Geography, v. 70, p. 399-406.

Wu, W. and Sidle R.C., 1995, A distributed slope stability model for steep forested basins: Water Resources Research, v. 31, n. 8, p. 2097-2110.

Wooten, R.M., Clark, T.W., Bateson, J.T., 2001, Geologic features related to recentdebris flows and weathered-rock slides in the Blue Ridge and Piedmont of NorthCarolina: Abstracts with Programs - Geologic Society of America, v. 33, no. 2, p. 14-15.

Wooten, R. M., Reid, J.C., Clark, T.W., Medina, M.A., Bateson, J.T., Davidson, V.A.,2004, Selected geologic hazards in North Carolina – An overview (poster), Northeastern

Zhang, X., Drake, N.A., Wainwright, J., Mulligan, M., 1999, Comparison of slopeestimates from low resolution DEMs: Scaling issues and a fractal method for their solution: Earth Surface Processes and Landforms, v. 24, 763-799.





Appendices

Appendix A: Geologic Units

GeologicUnit

Rock Type Formation Name (if given)

Geologic Description Composition

CZab Metamorphic Amphibolite and Biotite Gneiss

Interlayered; minor layers and lenses of hornblende gneiss, metagabbro, micaschist, and granitic rock

Inequigranular, locally abundant potassicfeldspar and garnet; interlayered and



134

CZbg Metamorphic Biotite Gneiss and Schist

feldspar and garnet; interlayered andgradational with calc-silicate rock,sillimanite-mica schist, mica schist, and

amphibolite. Contains small masses of granitic rock

CZgms Metamorphic SAURATOWN MOUNTAIN Garnet-Mica Schist Interlayered with amphibole

CZtp Metamorphic SAURATOWN MOUNTAIN Porphyroblastic gneiss

Ccl Sedimentary LOWER CHILHOWEE Arenite

Feldspathic arenite, white to yellowishgray. Minor silty shale, feldspathicsiltstone, and conglomerate in lower part

Ccu Sedimentary UPPER CHILHOWEE AreniteVitreous quartz arenite, white to lightgray; interbedded sandy siltstone andshale

Chg Metamorphic HENDERSON GNEISS GneissMonzonitic to granodioritic,inequigranular

Cr Sedimentary ROME FORMATION Shale and SiltstoneShale and siltstone, variegated red to brown; interbedded fine-grainedsandstone

DSc Metamorphic

CAESARS HEAD GRANITE

GNEISS Granite Gneiss

Equigranular to porphyritic, massive to

well foliated; contains biotite andmuscovite

Dqd Igneous INTRUSIVE Quartz Diorite and GranodioriteContains biotite, muscovite, andxenocrysts

Dsc ? ? ? ?

PzZu Metamorphic INTRUSIVE Meta-Ultramafic Rock

Metamorphosed dunite and peridotite;serpentinite, soapstone, and other alteredultramafic rock

SOgg Metamorphic INTRUSIVE Granite GneissPoorly foliated; interlayered with biotiteaugen gneiss

Ybam Metamorphic UNCONFORMITY Amphibolite

Equigranular, massive to well foliated,interlayered, rarely discordant,metamorphosed intrusive and extrusivemafic rock; may includemetasedimentary rock



135

y

Ybgg Metamorphic UNCONFORMITY Biotite Granite Gneiss

Pinkish gray to light gray, massive towell foliated, granitic to quartz

monzonitic; includes variablymylonitized orthogneiss and para-gneiss,interlayered amphibolite, calc-silicaterock, and marble

Ygg Metamorphic UNCONFORMITY Granodioritic Gneiss

Greenish gray to pinkish gray, porphyroclastic to mylonitic; epidote,sericite, and chlorite common

Ymam Metamorphic UNCONFORMITY Amphibolite

Equigranular, massive to well foliated,interlayered, rarely discordant,metamorphosed intrusive and extrusivemafic rock; may includemetasedimentary rock

Ymg Metamorphic UNCONFORMITYMigmatitic Biotite-HornblendeGneiss

Layered biotite-granite gneiss, biotite-hornblende gneiss, amphibolite, calc-silicate rock; locally contains relictgranulite facies rock

Ytg Metamorphic TOXAWAY GNEISS GneissPoorly foliated to well foliated,

equigranular to inequigranular, granitic

Zaba MetamorphicALLIGATOR BACK FORMATION

Amphibolite

Equigranular, massive to well foliated,interlayered, rarely discordant,metamorphosed intrusive and extrusivemafic rock; may include

metasedimentary rock

Zabg MetamorphicALLIGATOR BACK FORMATION

Gneiss

Finely laminated to thin layered; locallycontains massive gneiss and micaceousgranule conglomerate; includes schist, phyllite and amphibolite

Z M hiASHE METAMORPHICSUITE AND TALLULAH A hib li

Equigranular, massive to well foliated,interlayered, rarely discordant,

h d i i d i



136

Zata Metamorphic SUITE AND TALLULAHFALLS FORMATION

Amphibolite metamorphosed intrusive and extrusivemafic rock; may include

metasedimentary rock

Zatb MetamorphicASHE METAMORPHICSUITE AND TALLULAHFALLS FORMATION

Biotite gneiss

Interlayered with biotite-garnet gneiss, biotite-muscovite schist, garnet-micaschist, and amphibolite

Zatm MetamorphicASHE METAMORPHICSUITE AND TALLULAHFALLS FORMATION

Muscovite-biotite gneiss

Locally sulfidic, interlayered andgradational with mica schist, minor amphibolite, and hornblende gneiss

Zatw Metamorphic

ASHE METAMORPHIC

SUITE AND TALLULAHFALLS FORMATION

Metagraywacke

Foliated to massive. Locallyconglomeratic; interlayered andgradational with mica schist, muscovite- biotite gneiss, and rare graphitic schist

Zch Metamorphic COPPERHILL FORMATION Metagraywacke

Metagraywacke, massive, graded bedding common; includes dark-grayslate, mica schist, and nodular calc-silicate rock

Zchs MetamorphicSLATE OF COPPERHILLFORMATION

Slate

Slate to phyllite, dark gray, graphitic,sulfidic; includes metagraywacke with

local graded bedding

Zgma MetamorphicGRANDFATHER MOUNTAINFORMATION

Meta-ArkoseSericitic, conglomeritic, locally cross- bedded interlayered metasiltstone andslate

Zgmg MetamorphicGRANDFATHER MOUNTAINFORMATION

Greenstone

Schistose to massive, amygdaloidal;interlayered with metasedimentary rock

Zgs MetamorphicGREAT SMOKY GROUP,UNDIVIDED

Metagraywacke andmetasiltstone

Thick metasedimentary sequence of

massive to graded beds of metagraywacke and metasiltstone withinterbedded graphitic and sulfidic slateand schist

Zlm Metamorphic Lincolnton Metadacite Metadacite

Zm Igneous MAX PATCH GRANITE Granite

Mottled pink and light green, coarsegrained to porphyritic, massive; containsbi tit



137

biotite

Zrb Sedimentary RICH BUTT SANDSTONE SandstoneFeldspathic; interbedded with dark argillaceous layers and laminae

Zs MetamorphicSNOWBIRD GROUP,UNDIVIDED

Metasiltstone and sandstone

Feldspathic metasiltstone,metasandstone, and phyllite. Basal schistcontains lenses of quartz-pebbleconglomerate

Zsl Metamorphic LONGARM QUARTZITE Quartzite

Cross-bedded. feldspathic, locallyconglomeratic; includes dark slate andmetasiltstone

Zsp Sedimentary PIGEON SILTSTONE Siltstone

Thin bedded to laminated, commonlycross-bedded, metamorphosed; locallyincludes argillite and calcareous andankeritic metasiltstone grading to siltymetalimestone

Zsr SedimentaryROARING FORK SANDSTONE

Sandstone

Greenish gray, fine to medium grained,locally cross-bedded, metamorphosed;interbedded metasiltstone and phyllite

Zss Metamorphic SANDSUCK FORMATION Slate

Slate and metasiltstone, dark green to black. Metaconglomerate lentils in upper part; calcareous metasandstone, sandymetalimestone, and quartzite in lower part

Zsw MetamorphicWADING BRANCHFORMATION

Slate

Sandy slate to coarse-grained pebblymetagraywacke with local graded bedding. Basal quartz-sericite schist or phyllite

Zwc Metamorphic WALDEN CREEK GROUP,UNDIVIDED

Slate Slate to metasiltstone, local limy bedsand pods; interbedded with quartz-pebblemetaconglomerate and metasandstone

ZYbn Metamorphic Biotite gneiss

Migmatitic; interlayered and gradationalwith biotite-garnet gneiss andamphibolite; locally abundant quartz andalumino-silicates. Stratigraphic positionuncertain



138

bz Metamorphic BREVARD FAULT ZONE Schist and phyllonite

Fish scale schist and phyllonite,

graphitic; interlayered with feldspathicmetasandstone, marble lenses

Appendix B: General Soil Data

MUSEQ NAME S5ID

% OF

MAPUNIT

SURFACE

TEXTURE UNIFIED AASHTO

DEPTH

LOW

DEPTH

HIGH

PERM

LOW

PERM

HIGH

HY COND

LOW

HY COND

HIGH

NC005 1 TOXAWAY NC0021 35 SIL CL A-4 0.0 1.0 0.60 20.00 0.21 7.00

NC005 2 ROSMAN NC0024 17 L ML A-4 2.5 5.0 2.00 20.00 0.34 3.40

NC005 3 DELANCO MD0155 26 SIL ML A-4 1.0 2.5 0.20 2.00 0.05 0.52

NC005 5 COMUS MD0050 9 FSL ML A-2 6.0 6.0 0.60 6.00 0.05 0.54

NC005 6 HATBORO PA0016 7 L ML A-4 0.0 0.5 2.00 6.00 0.14 0.42



139

NC005 7 BRADSON NC0028 4 GR-L SM A-2 6.0 6.0 0.60 20.00 0.02 0.80

NC005 8 SUNCOOK CT0001 2 LS SM A-2 3.0 6.0 6.00 20.00 0.12 0.40

0.94 13.08 (IN/HR)

0.024 0.331 (M/HR)

NC006 1 CLIFTON NC0015 15 L ML A-4 6.0 6.0 0.60 6.00 0.09 0.90

NC006 2 BRADDOCK VA0054 23 L CL A-2 6.0 6.0 0.60 6.00 0.14 1.38

NC006 3 EVARD SC0083 19 L ML A-4 6.0 6.0 0.60 2.00 0.11 0.38

NC006 4 BRADDOCK VA0231 17 GR-L SM A-2 6.0 6.0 0.60 6.00 0.10 1.02

NC006 5 TOXAWAY NC0021 9 L CL A-4 0.0 1.0 0.60 20.00 0.05 1.80

NC006 6 URBANLAND

DC0035 8 VAR 2.0 2.0


0.55 6.02 (IN/HR)

0.014 0.152 (M/HR)

NC088 1 CHESTER MD0001 23 L CL A-4 6.0 6.0 0.60 2.00 0.14 0.46 NC088 2 ASHE NC0186 15 ST-FSL SM A-2 6.0 6.0 2.00 6.00 0.30 0.90

NC088 3 CHESTER MD0001 15 L CL A-4 6.0 6.0 0.60 2.00 0.09 0.30

NC088 4 CHESTER MD0001 6 L CL A-4 6.0 6.0 0.60 2.00 0.04 0.12 NC088 5 CODORUS PA0015 6 L ML A-4 1.0 2.0 0.60 20.00 0.04 1.20

NC088 6 ASHE NC0019 5 FSL SM A-4 6.0 6.0 2.00 6.00 0.10 0.30

NC088 7 CHANDLER NC0263 6 ST-FSL SM A-4 6.0 6.0 2.00 6.00 0.12 0.36

NC088 8 ASHE NC0019 3 FSL SM A-4 6.0 6.0 2.00 6.00 0.06 0.18



NC088 11 SUNCOOK CT0001 3 LS SM A-2 3.0 6.0 0.60 20.00 0.02 0.60

NC088 12 TATE NC0025 3 L ML A-4 6.0 6.0 2.00 6.00 0.06 0.18 NC088 13 WATAUGA NC0091 3 L SM A-4 6.0 6.0 0.60 6.00 0.02 0.18

NC088 14 WATAUGA NC0091 3 L SM A-4 6.0 6.0 0.60 6.00 0.02 0.18

NC088 15 FANNIN NC0020 2 SIL ML A-4 6.0 6.0 0.60 6.00 0.01 0.12

0.21 1.52 (IN/HR)

0.005 0.038 (M/HR)

NC089 1 CHESTER MD0001 32 L CL A-4 6 0 6 0 0 60 2 00 0 19 0 64



140

NC089 1 CHESTER MD0001 32 L CL A 4 6.0 6.0 0.60 2.00 0.19 0.64

NC089 2 CLIFTON NC0264 13 ST-L SM A-4 6.0 6.0 0.60 6.00 0.08 0.78

NC089 3 CHESTER MD0001 11 L CL A-4 6.0 6.0 0.60 2.00 0.07 0.22 NC089 4 CLIFTON NC0015 9 L ML A-4 6.0 6.0 0.60 6.00 0.05 0.54


NC089 6 CODORUS PA0015 8 L ML A-4 1.0 2.0 0.60 20.00 0.05 1.60






NC089 12 PORTERS NC0152 3 ST-L ML A-2 6.0 6.0 0.60 6.00 0.02 0.18


NC089 14 TATE NC0025 1 L ML A-4 6.0 6.0 0.60 6.00 0.01 0.06

0.11 1.08 (IN/HR)

0.003 0.027 (M/HR)

NC090 1 HAYESVILLE NC0013 25 L SM A-4 6.0 6.0 0.60 6.00 0.15 1.50


NC090 3 BRADSON NC0028 13 GR-L SM A-2 6.0 6.0 0.60 20.00 0.08 2.60 NC090 4 CODORUS PA0015 11 L ML A-4 1.0 2.0 0.60 20.00 0.07 2.20

NC090 5 BRADSON NC0028 9 GR-L SM A-2 6.0 6.0 0.60 20.00 0.05 1.80

NC090 6 EVARD SC0083 7 SL SM A-2 6.0 6.0 0.60 6.00 0.04 0.42

NC090 7 ROSMAN NC0024 6 L ML A-4 2.5 5.0 0.60 20.00 0.04 1.20

NC090 8 DELANCO MD0155 4 FSL ML A-4 1.0 2.5 0.60 6.00 0.02 0.24

NC090 9 EDNEYVILLE NC0023 4 FSL SM A-2 6.0 6.0 0.20 6.00 0.01 0.24


NC090 11 TOXAWAY NC0021 2 SIL CL A-4 0.0 1.0 0.60 20.00 0.01 0.40 0.20 4.54 (IN/HR)

0.005 0.115 (M/HR)

NC091 1 DITNEY TN0075 27 CB-SL ML A-4 6.0 6.0 0.60 6.00 0.16 1.62

NC091 2 UNICOI TN0054 18 STV-L GM A-2 6.0 6.0 2.00 6.00 0.36 1.08

NC091 3 JUNALUSKA NC0181 12 CN-FSL SM A-4 6.0 6.0 0.60 6.00 0.07 0.72

NC091 4 BRASSTOWN NC0206 8 CN-FSL SM A-4 6.0 6.0 0.60 6.00 0.05 0.48



141

NC091 5 LONON NC0203 10 CB-SL SM A-2-4 6.0 6.0 0.60 6.00 0.06 0.60

NC091 6 NORTHCOVE NC0204 7 STV-FSL GM A-2-4 6.0 6.0 0.60 6.00 0.04 0.42 NC091 7 SOCO NC0180 10 CN-FSL SM A-4 6.0 6.0 2.00 6.00 0.20 0.60

NC091 8 JUNALUSKA NC0181 4 CN-FSL SM A-4 6.0 6.0 0.60 6.00 0.02 0.24

NC091 9 BRASSTOWN NC0206 2 CN-FSL SM A-4 6.0 6.0 0.60 6.00 0.01 0.12

NC091 10 LONON NC0203 1 CB-SL SM A-2-4 6.0 6.0 0.60 6.00 0.01 0.06

NC091 11 NORTHCOVE NC0204 1 STV-FSL GM A-2-4 6.0 6.0 0.60 6.00 0.01 0.06

0.35 2.10 (IN/HR)

0.009 0.053 (M/HR)

NC092 1 SYLCO TN0014 32 SIL GC A-4 6.0 6.0 0.60 2.00 0.19 0.64

NC092 2 DITNEY TN0075 41 CB-L ML A-4 6.0 6.0 0.60 6.00 0.25 2.46

NC092 3 TUSQUITEE NC0158 3 ST-L SM A-2-4 6.0 6.0 2.00 6.00 0.06 0.18



NC092 6 CATASKA TN0133 14 STV-L CL-ML A-4 6.0 6.0 2.00 20.00 0.28 2.80

NC092 7 SPIVEY TN0109 2 STV-L GM A-2 6.0 6.0 0.06 6.00 0.00 0.12

NC092 8 HAYWOOD NC0095 2 ST-L SM A-2-4 6.0 6.0 0.06 20.00 0.00 0.40

NC092 9 SYLCO TN0014 1 L GC A-4 6.0 6.0 0.60 2.00 0.01 0.02 NC092 10 UNICOI TN0054 1 STV-L GM A-2 6.0 6.0 0.60 6.00 0.01 0.06

NC092 11 UNICOI TN0054 1 STV-L GM A-2 6.0 6.0 2.00 6.00 0.02 0.06

0.35 3.58 (IN/HR)

0.009 0.091 (M/HR)

NC093 1 ASHE NC0186 53 ST-FSL SM A-2 6.0 6.0 2.00 6.00 1.06 3.18


NC093 3 CHESTER MD0001 12 L CL A-4 6.0 6.0 0.60 2.00 0.07 0.24 NC093 4 ASHE NC0019 9 FSL SM A-4 6.0 6.0 0.60 6.00 0.05 0.54

NC093 5 ROCK OUTCROP

DC0015 8 UWB 6.0 6.0 2.00 0.00 0.16 0.00

NC093 6 ASHE NC0019 1 FSL SM A-4 6.0 6.0 2.00 6.00 0.02 0.06



0.41 1.26 (IN/HR)



142

0.010 0.032 (M/HR)






NC094 6 ASHE NC0186 5 ST-FSL SM A-2 6.0 6.0 0.60 6.00 0.03 0.30



NC094 9 CLIFTON NC0015 4 L ML A-4 6.0 6.0 0.60 6.00 0.02 0.24 NC094 10 TATE NC0025 3 L ML A-4 6.0 6.0 0.60 6.00 0.02 0.18

NC094 11 CHANDLER NC0017 4 L ML A-2 6.0 6.0 0.60 6.00 0.02 0.24



NC094 14 CODORUS PA0015 9 SIL ML A-4 1.0 2.0 0.60 20.00 0.05 1.80

0.18 2.98 (IN/HR)

0.005 0.075 (M/HR)

NC095 1 EVARD SC0083 17 SL SM A-2 6.0 6.0 0.60 6.00 0.10 1.02



NC095 4 BREVARD NC0012 9 L ML A-4 6.0 6.0 2.00 20.00 0.18 1.80



NC095 7 ASHE NC0186 8 ST-SL SM A-2 6.0 6.0 2.00 6.00 0.16 0.48


NC095 9 ASHE NC0186 5 ST-SL SM A-2 6.0 6.0 2.00 6.00 0.10 0.30 NC095 10 TUSQUITEE NC0026 5 L ML A-4 6.0 6.0 2.00 6.00 0.10 0.30

NC095 11 EDNEYVILLE NC0023 3 FSL SM A-2 6.0 6.0 2.00 6.00 0.06 0.18

NC095 12 ROCK OUTCROP

DC0015 1 UWB 6.0 6.0 2.00 0.00 0.02 0.00


NC095 14 CHANDLER NC0263 2 ST-L SM A-4 6.0 6.0 0.60 6.00 0.01 0.12




143

NC095 16 EVARD SC0083 1 SL SM A-2 6.0 6.0 0.60 6.00 0.01 0.06

NC095 17 EVARD SC0083 1 SL SM A-2 6.0 6.0 0.60 6.00 0.01 0.06 NC095 18 TUSQUITEE NC0026 1 L ML A-4 6.0 6.0 0.60 6.00 0.01 0.06

0.13 0.82 (IN/HR)

0.003 0.021 (M/HR)



NC096 3 CODORUS PA0015 10 SIL ML A-4 1.0 2.0 0.60 20.00 0.06 2.00


NC096 5 TUSQUITEE NC0158 7 ST-L SM A-2-4 6.0 6.0 0.60 6.00 0.04 0.42 NC096 6 CHESTER MD0001 6 L CL A-4 6.0 6.0 0.60 2.00 0.04 0.12


NC096 8 PORTERS NC0022 4 L ML A-4 6.0 6.0 2.00 6.00 0.08 0.24

NC096 9 PORTERS NC0022 4 L ML A-4 6.0 6.0 2.00 6.00 0.08 0.24

NC096 10 TUSQUITEE NC0026 3 L ML A-4 6.0 6.0 2.00 6.00 0.06 0.18





0.33 1.08 (IN/HR)

0.008 0.027 (M/HR)

NC097 1 EDNEYVILLE NC0023 73 L SM A-2 6.0 6.0 0.60 6.00 0.44 4.38

NC097 2 EDNEYVILLE NC0023 11 L SM A-2 6.0 6.0 2.00 6.00 0.22 0.66


NC097 4 ASHE NC0019 3 GR-FSL SM A-2 6.0 6.0 2.00 6.00 0.06 0.18

NC097 5 TUSQUITEE NC0026 3 L ML A-4 6.0 6.0 2.00 6.00 0.06 0.18 NC097 6 TOXAWAY NC0021 1 L CL A-4 0.0 1.0 2.00 20.00 0.02 0.20

0.99 6.17 (IN/HR)

0.025 0.156 (M/HR)

NC098 1 WAYAH NC0188 43 GR-L SM A-2-4 6.0 6.0 0.60 6.00 0.26 2.58

NC098 2 TANASEE NC0197 9 ST-L SM A-2-4 6.0 6.0 2.00 6.00 0.18 0.54

NC098 3 PORTERS NC0152 8 ST-FSL ML A-2 6.0 6.0 2.00 6.00 0.16 0.48



144

NC098 4 WAYAH NC0188 8 GR-L SM A-2-4 6.0 6.0 2.00 6.00 0.16 0.48

NC098 5 WAYAH NC0188 8 GR-L SM A-2-4 6.0 6.0 2.00 6.00 0.16 0.48 NC098 6 TANASEE NC0197 7 ST-L SM A-2-4 6.0 6.0 2.00 6.00 0.14 0.42


NC098 8 CULLASAJA NC0237 1 STV-L SM A-1-B 6.0 6.0 2.00 6.00 0.02 0.06

NC098 9 EDNEYVILLE NC0115 3 ST-FSL SM A-4 6.0 6.0 2.00 6.00 0.06 0.18


NC098 11 CHESTNUT NC0242 1 ST-FSL SM A-2-4 6.0 6.0 2.00 6.00 0.02 0.06


NC098 13 TUSQUITEE NC0158 1 ST-FSL SM A-2-4 6.0 6.0 0.60 6.00 0.01 0.06

NC098 14 CHESTNUT NC0242 1 ST-FSL SM A-2-4 6.0 6.0 2.00 6.00 0.02 0.06 NC098 15 TUCKASEGE

E NC0226 1 ST-L ML A-2 6.0 6.0 2.00 6.00 0.02 0.06

0.17 0.60 (IN/HR)

0.004 0.015 (M/HR)



NC099 3 TALLADEGA GA0037 17 SIL SM A-4 6.0 6.0 0.60 2.00 0.10 0.34

NC099 4 IOTLA NC0140 9 L SM A-2 1.5 3.5 2.00 20.00 0.18 1.80

NC099 5 EVARD SC0083 6 SL SM A-2 6.0 6.0 0.60 6.00 0.04 0.36

NC099 6 EVARD SC0083 6 SL SM A-2 6.0 6.0 0.60 6.00 0.04 0.36

NC099 7 TATE NC0025 3 FSL ML A-4 6.0 6.0 0.60 6.00 0.02 0.18



0.39 3.22 (IN/HR)

0.010 0.081 (M/HR)

NC100 1 EVARD SC0083 37 L ML A-4 6.0 6.0 0.60 2.00 0.22 0.74

NC100 2 OTEEN NC0107 18 L ML A-4 6.0 6.0 0.60 6.00 0.11 1.08


NC100 4 SALUDA SC0082 25 L SM A-2 6.0 6.0 0.60 6.00 0.15 1.50


NC100 6 TATE NC0025 1 ST-L ML A-4 6.0 6.0 2.00 6.00 0.02 0.06


NC100 8 HAYESVILLE NC0151 12 ST L SM A 4 6 0 6 0 0 60 6 00 0 07 0 72



145

NC100 8 HAYESVILLE NC0151 12 ST-L SM A-4 6.0 6.0 0.60 6.00 0.07 0.72

0.39 3.92 (IN/HR) 0.010 0.099 (M/HR)

NC102 1 JUNALUSKA NC0181 21 CN-L SM A-4 6.0 6.0 0.60 6.00 0.13 1.26

NC102 2 TSALI NC0179 11 CN-L SM A-4 6.0 6.0 2.00 6.00 0.22 0.66


NC102 4 SANTEETLAH

NC0208 11 CN-FSL SM A-4 6.0 6.0 0.60 6.00 0.07 0.66

NC102 5 STECOAH NC0184 6 CN-L SM A-4 6.0 6.0 2.00 6.00 0.12 0.36

NC102 6 SOCO NC0180 5 CN-L SM A-4 6.0 6.0 2.00 6.00 0.10 0.30 NC102 7 JUNALUSKA NC0181 6 CN-L SM A-4 6.0 6.0 0.60 6.00 0.04 0.36


NC102 9 CHEOAH NC0190 8 CN-L SM A-4 6.0 6.0 0.60 6.00 0.05 0.48


NC102 11 SANTEETLAH

NC0208 3 CN-FSL SM A-4 6.0 6.0 0.60 6.00 0.02 0.18

NC102 12 JUNALUSKA NC0181 1 CN-L SM A-4 6.0 6.0 0.60 6.00 0.01 0.06


NC102 14 BRASSTOWN NC0206 1 CN-L SM A-4 6.0 6.0 0.60 6.00 0.01 0.06 0.28 1.38 (IN/HR)

0.007 0.035 (M/HR)




NC103 4 SOCO NC0180 5 CN-L SM A-4 6.0 6.0 2.00 6.00 0.10 0.30

NC103 5 STECOAH NC0184 6 CN-L SM A-4 6.0 6.0 2.00 6.00 0.12 0.36 NC103 6 SOCO NC0180 4 CN-L SM A-4 6.0 6.0 2.00 6.00 0.08 0.24




NC103 10 COWEE NC0171 6 GR-L SM A-2-4 6.0 6.0 2.00 6.00 0.12 0.36

NC103 11 EVARD SC0135 8 GR-L SM A-2 6.0 6.0 0.60 6.00 0.05 0.48


NC103 13 CHESTNUT NC0242 2 ST FSL SM A 2 4 6 0 6 0 2 00 6 00 0 04 0 12



146


NC103 14 PORTERS NC0152 6 ST-FSL ML A-2 6.0 6.0 2.00 6.00 0.12 0.36 NC103 15 SPIVEY TN0109 6 STV-L GM A-2 6.0 6.0 2.00 6.00 0.12 0.36


NC103 17 SANTEETLAH

NC0208 2 CN-L SM A-4 6.0 6.0 2.00 6.00 0.04 0.12

NC103 18 SAUNOOK NC0195 3 GR-L SM A-2 6.0 6.0 0.60 6.00 0.02 0.18


NC103 20 DELLWOOD NC0183 1 GR-FSL SM A-2-4 2.0 4.0 2.00 20.00 0.02 0.20

NC103 21 DILLARD GA0061 1 L ML A-4 2.0 3.0 0.60 2.00 0.01 0.02

0.22 1.00 (IN/HR) 0.005 0.025 (M/HR)


NC104 2 BURTON NC0114 19 ST-L SM A-2 6.0 6.0 2.00 6.00 0.38 1.14

NC104 3 OCONALUFTEE

NC0192 16 FL-L SM A-4 6.0 6.0 2.00 6.00 0.32 0.96


NC104 5 SPIVEY TN0109 7 CB-L GM A-2 6.0 6.0 2.00 6.00 0.14 0.42

NC104 6 CHEOAH NC0190 10 CN-L SM A-4 6.0 6.0 0.60 6.00 0.06 0.60 NC104 7 STECOAH NC0184 2 CN-L SM A-4 6.0 6.0 2.00 6.00 0.04 0.12


NC104 9 BURTON NC0114 2 ST-L SM A-2 6.0 6.0 2.00 6.00 0.04 0.12

NC104 10 SANTEETLAH

NC0208 5 L SM A-4 6.0 6.0 2.00 6.00 0.10 0.30

NC104 11 SOCO NC0180 1 CN-L SM A-4 6.0 6.0 2.00 6.00 0.02 0.06

0.46 1.80 (IN/HR)

0.012 0.046 (M/HR)



147

Appendix C: Slope Movement Data for the 2004 Hurricanes Frances and Ivan

HURRICANE FRANCES - SEPTEMBER 6-8, 2004County Location Date Reported Source Description

Avery County, Near Crossnore US-221 Both Directions NCDOT Road is one lane in certain locations

Avery County, Near Crossnore SR-1504 Both Directions,Pineola Rd.

NCDOT HIGH WATER, MUDSLIDE

Blue Ridge Parkway Mileposts: 322, 345, 348,349, 413, and 429

9/9/2004 Asheville Citizen-Times

Most of those between mile 322 and 349 took out major portions of the motor road, from just south of Linville Falls tosouth of Buck Creek Gap at NC 80 near Marion.

Henderson County SR-1710 Both Directions,Bald Rock Rd.

9/10/2004 17:06 NCDOT Trees down and Mudslide, Road has collapsed along with amudslide near waterfall.



148

Henderson County SR-1799 Both Directions,Deep Gap Rd.

9/8/2004 NCDOT Deep Gap has many landslides blocking the road

Henderson County SR-1194 Both Directions,Patterson Rd.

9/8/2004 NCDOT Two mudslides into road

Henderson County SR-1613 Both Directions,Slick Rock Rd.

9/8/2004 NCDOT Road is washed out. Mudslide covering entire road

Henderson County, Near Hendersonville

SR-1706 Both Directions,Little Creek Rd.

NCDOT Sinkhole - road washed away about 1 mile off Sugarloaf Mountain Road (SR 1868).

Henderson County Shepard Street portion of

N.C. 191

9/8/04 11:00AM NCDOT One lane closed due to mudslide

Jackson County NC-281 (Mile Marker 13to 17) Both Directions,Little Canada

9/8/2004 NCDOT Road is closed due to slide from SR 1140 (Fanny Mae BrownRoad) to Rock Bridge near Tannassee Gap. Detour is signed.Detour distance is 41 miles.

McDowell County, Near OldFort

I-40 (Mile Marker 72 to67or 69-67) BothDirections

early morning,9/8/04

NCDOT A slide on I-40 at Old Fort Mountain has blocked 2 of 3 laneson westbound I-40 as well as 2 of the 3 eastbound lanes.Emergency crews are working to reopen the rest of the lanesas quickly as possible. Expect delays.

McDowell County, Near Woodlawn

US-221 Both Directions NCDOT There are several mudslides and dangerous areas at the top of US 221 North (Linville Mtn.). The road is closed and will bereopened as soon as the crews can complete the work andmake the road passable again. US 221 North (Linville Mtn.)is closed due to Hurricane Frances. Approximately 70 feet of

the road is washed away at the Mountain ParadiseCampground. It will take a couple of months to rebuild this portion.

Polk County Pearson Falls Rd., near Saluda

9/9/2004 Asheville Citizen-Times

A mudslide closed Pearson Falls Road near Saluda.

Polk County U.S. 176 9/9/2004 Asheville Citizen-Times

between Saluda and Tryon, one lane closed from mudslides

Polk County U.S. 176 9/9/2004 Asheville Citizen- Temporarily closed Tuesday near the Henderson-Polk County



149

y

Times

p y y y

line when mud and trees slide down a bank Transylvania County Old Highway 64/ N.C.

281N9/8/04 11:00AM NCDOT Portion of road closed due to mudslides near the Jackson

County line

Watauga County White Laurel subdivision,near Boone

9/8/2004 Rick Wooten, personalcommunication

This appears to have been an embankment failure. One homewas destroyed, and eight condemned for occupancy

HURRICANE IVAN - SEPTEMBER 16-18, 2004County Location Date Reported Source Description

Avery County N.C. 184 and 194 Asheville CitizenTimes

There was so much rain that high water and mudslidescovered N.C. 184 and 194, shutting Banner Elk off from therest of the region.

Buncombe County Mink Farm Road 9/18/2004 Asheville CitizenTimes

mudslide

Buncombe County Hookers Gap 9/18/2004 Asheville CitizenTimes

mudslide

Buncombe County Gibbs Road 9/18/2004 Asheville Citizen

Times

mudslide

Buncombe County Freedom Farm Road 9/18/2004 Asheville CitizenTimes

mudslide

Buncombe County Newfound Rd 9/18/2004 Asheville Citizen mudslide

Times

Buncombe County North Turkey Creek Rd 9/18/2004 Asheville CitizenTimes

Mudslide at Early's Mtn. Road

Buncombe County Sluder Branch Road - #99 9/18/2004 Asheville CitizenTimes

Mudslide & Pvt. Bridge damaged

Buncombe County N.C. 151 Asheville CitizenTimes

Closed due to major slide near Blue Ridge Parkway access

Buncombe County Arrowood Rd. near Starnes Cove Rd.

9/17/2004 2:00 Asheville CitizenTimes

debris flow occurred on side of mountain near the town of Enka-Candler, destroyed at least one home

Haywood County I-40 mudslide in thewestbound lane at MM 35

9/17/2004 13:44 NCDOT Interstate 40 is closed in Haywood County from Exit 451 inTennessee to Exit 20 in North Carolina due to a slope failure.

Haywood County U.S. 19-23 9/17/2004 0:00 Asheville CitizenTimes

mudslides



150

Haywood County Dutch Cove at Turnpike 9/17/2004 Asheville CitizenTimes mudslide

Haywood County I-40, between the TN stateline and MM 20

9/17/2004 NCDOT I-40 is closed in Haywood County between the TennesseeState Line and mile marker 20 due to the road being washedaway.

Haywood County U.S.276 9/17/2004 Asheville CitizenTimes

rockslide south of Waynesville

Henderson County Middle Fork (one laneclosed)

9/17/2004 Asheville CitizenTimes

mudslide

Jackson County, Near Cashiers NC-107 (Mile Marker 18to 18) Both Directions

9/16/2004 23:02 NCDOT mudslide - N.C. 107 is closed form the Thorpe Power Plantsouth to Cashiers. Downed trees, boulders and mudslides are blocking the road, the main access the area.

Jackson County, Near Cashiers US-64 Both Directions 9/16/2004 23:19 NCDOT US 64 is closed between Cashiers and the Transylvania Co.Line due to slides and debris. US 64 is closed betweenCashiers and the Macon Co. Line due to slides and debris.Hwy 64 at Spring Forest Road, three-fourths of highway hasslid off

Macon, near Franklin SR-1310, Wayah Rd. 9/17/2004 7:46 NCDOT Down Trees and Slides

McDowell County, Near Woodlawn

NC-226 ALT BothDirections

9/17/2004 6:21 NCDOT Rockslide on NC226-A will probably take all day to move.

Swain, Near Bryson City SR-1195, Hwy 19a 9/17/2004 8:02 NCDOT Mud Slide and shoulder broke off

Transylvania County, near Brevard

SR-1540, Wilson Rd. 9/18/2004 3:35 NCDOT Due to high water, trees, power lines, and slides road isclosed.

Watauga, near Boone SR-1130, Lee GualtneyRd.

9/17/2004 17:16 NCDOT Slide blocking road



151

Appendix D: Debris Flow Inventory

MATERIALDATA

SOURCEREMARKS LAT LONG COUNTY YEAR TYPE GEO UNIT SOIL UNIT SLOPE ASPECT

NCDOT SR 1607, .5mi S of Madison County line 35.7300 82.7000 Buncombe 1990 Slump Ymg NC090 10 N

Cv Scott, R Brevard slide 35.1900 82.8700 Transylvania 1970 Debris Flow Dqd NC093 11 SE

NCGS BRP Mile 357.7 35.7305 82.3082 Buncombe 1995 Debris Flow Zatw NC098 11 SW

Pomeroy 35.4800 82.6500 Buncombe 1977 Slump Zatm NC093 13 E

Rd NCDOT MP 17.6, E of SR 1336WBL

35.6300 82.9900 Haywood 1968 Debris Flow Zch NC006 13 SE

Rd NCDOT Intersection of BRP, 694,SR 2053

35.6500 82.4900 Buncombe 1982 Slump Zatm NC095 13 N

Pomeroy 35 4800 82 6700 Buncombe 1977 Slump Zatm NC093 15 E



152

Pomeroy 35.4800 82.6700 Buncombe 1977 Slump Zatm NC093 15 E

Cv NCDOT .6 mi E of SR 1130 WBL 35.4700 82.9500 Haywood 1990 Slump ZYbn NC090 15 SW

Rd NCDOT .2mi N of BuncombeCounty line, US 25-70

35.7600 82.6100 Madison 1989 Debris Flow Ybgg NC100 16 NW

Rd NCDOT Balsam Gap Landslide 35.4400 83.0800 Haywood 1989 Slump ZYbn NC095 17 E

Rd NCDOT Rt. 206 35.6000 82.9400 Haywood 1996 Slump ZYbn NC006 17 W

NCDOT NBL NC 226, .6mi S of 1253

35.9900 82.1600 Mitchell 1990 Slump Zatm NC093 17 SW

Lambe Balsam Gap II Landslide 35.4500 83.0700 Haywood Debris Flow ZYbn NC103 18 NE

Cv NCDOT 600 ft. S of SR 1319 35.8900 82.7500 Madison 1986 Slump Zsr NC092 18 W

Rd NCDOT Fines Creek Slide 35.6700 82.9900 Haywood 1978 Debris Flow Ybgg NC006 19 N

Pomeroy 35.4800 82.6500 Buncombe 1977 Slump Zatm NC093 19 S

NCDOT NC 262, 1.8mi S of TNline

36.0900 82.0900 Mitchell Slump Ymg NC098 19 S

Dockal Off FR475, W of DavidsonRiver, elev. 2500 feet

35.2800 82.8100 Transylvania 1993 Slump Dqd NC094 20 S

FS Possible landslide -Richland Ridge Rd.

35.1800 82.8900 Transylvania 2003 Debris Flow Zata NC006 20 S

Cv? NCGS Maggie Valley debris flow 35.5014 83.0944 Haywood Debris Flow Zgs NC103 21 NE

Dockal Cove Creek Slide 35.2800 82.8100 Transylvania Slump Zatb NC094 21 SE

Pomeroy Number 50 35.4500 82.6600 Henderson 1977 Debris Flow Zatm NC095 21 E

Rd NCDOT N-side of NC19 35.9200 82.0600 Mitchell 1976 Slump Zabg NC093 21 N

Cv? USFS Microburst; dual slides near Dry Branch

35.7700 83.0400 Haywood 1994 Debris Flow Zsl NC103 22 SE

Rd Pomeroy Number 24 35.4700 82.6600 Buncombe 1977 Debris Flow Zatm NC093 22 SE


Lambe Waterville Landslide 35.6900 83.0100 Haywood Debris Flow Ybgg NC102 23 SW

Rd NCDOT SBL NC 25-70, S of NC208

35.9100 82.7500 Madison 1986 Slump Zsr NC092 23 SW

Rd NCGS Deadly debris flow (12/11)- Maggie Valley Debris

35.5048 83.8500 Haywood 2003 Debris Flow Ybgg NC103 23 NE

Pomeroy Number 15 35.5000 82.6500 Buncombe 1977 Debris Flow Zatm NC093 23 NE


Pomeroy Number 41 35.4300 82.6100 Henderson 1977 Debris Flow Zatm NC093 23 SE

Pomeroy Number 43 35.4300 82.6400 Henderson 1977 Debris Flow Zatm NC093 23 S

Pomeroy Number 35 35.4600 82.6400 Buncombe 1977 Debris Flow Zatm NC093 23 SE



153


Pomeroy Debris Torrent 35.4800 82.6800 Buncombe 1977 Debris Flow Zatm NC093 23 W

Dockal Near Maxwell Cove, east of drainage

35.3200 82.7400 Transylvania 1996 Slump Zatb NC095 24 SW

Dockal Cove Creek Slide 35.2800 82.8100 Transylvania Slump Zatb NC094 24 SE

Rd NCDOT SBL NC 215, 200' N of SR 1216, approximate

35.4000 82.9400 Haywood 1989 Debris Flow Zatm NC095 24 E

rk NCDOT I-40, World's Fair Slide 35.7700 83.0900 Haywood 1982 Slump Zsp NC102 24 SW

Cv? NRCS Several debris flowstriggered by trop.

depression

35.7647 82.2656 Yancey 1977 Debris Flow Zatw NC098 25 W


Pomeroy Number 37 35.4700 82.6100 Buncombe 1977 Debris Flow Zatm NC093 25 S

NCDOT SBL SR 1137 35.9300 82.5000 Mitchell 1990 Slump Zabg NC093 25 SW

NCDOT SR 1318, .55mi W of SR 1334

35.9200 82.6700 Madison 1990 Debris Flow Ybgg NC093 25 W

Rd Pomeroy Number 16 35.4966 82.6504 Buncombe 1977 Debris Flow Zatm NC093 26 E

Pomeroy 35.4800 82.6800 Buncombe 1977 Slump Zatm NC093 26 SW


Pomeroy 35.4800 82.6500 Buncombe 1977 Slump Zatm NC093 27 S

Dockal Cove Creek Slide 35.2800 82.8100 Transylvania Slump Zatb NC094 27 E

NCDOT WBL SR 1334, 0.2 mi NWof SR 1425

35.9100 82.6900 Madison 1990 Debris Flow Ybgg NC093 27 N


Pomeroy Number 39 35.4400 82.6400 Buncombe 1977 Debris Flow Zatm NC093 27 N



Pomeroy 35.4800 82.6700 Buncombe 1977 Debris Flow Zatm NC095 28 SE

Pomeroy 35.4900 82.6600 Buncombe 1977 Debris Flow Zatm NC095 28 S

Rd NCDOT W of SR 1395 35.8900 82.6600 Madison 1980 Slump Ybgg NC095 28 E


wr NCDOT 35.5400 82.8100 Haywood 1984 Slump ZYbn NC006 28 S

Cv NCDOT MP 8.8, E of Harmon Den 35.7200 83.0400 Haywood 1979 Slump Zsl NC102 28 SW

NCDOT NC 1318, .25 W of 1334 35.9200 82.6700 Madison 1990 Debris Flow Ybgg NC093 28 N

Cv NCGS Hickey Fork Debris Flow 36.0068 82.6977 Madison 1999 Debris Flow Zwc NC092 28 SE

Cv NCGS Allen Stand Debris Flow 35.9865 82.7611 Madison 1999 Debris Flow Zwc NC092 28 E



154

Rd Pomeroy No. 23, Runnout length of 2820 ft. 35.4700 82.6600 Buncombe 1977 Debris Flow Zatm NC090 29 S




NCGS Blackstone Knob WestSlide

35.7353 82.3174 Buncombe 2004 Debris Flow Zatm NC095 29 S


Pomeroy No. 31, Chute 1800 ft. long,

into Laurel Branch

35.4600 82.6500 Buncombe 1977 Debris Flow Zatm NC095 30 E

Rd Pomeroy Number 30 35.4600 82.6700 Buncombe 1977 Debris Flow Zatm NC095 30 NE

Pomeroy Number 25 35.4800 82.6700 Buncombe 1977 Debris Flow Zatm NC095 30 E

Cv NCDOT MP 1.0, I-40 35.7700 83.0800 Haywood Debris Flow Zsr NC102 30 SW

Pomeroy Number 10 35.4900 82.6800 Buncombe 1977 Debris Flow Zatm NC093 30 SW



Rd/Cv NCDOT MP 7.95 EBL, N of singletunnel

35.7200 83.0300 Haywood 1973 Debris Flow Zsl NC102 30 N

Pomeroy 35.4900 82.6500 Buncombe 1977 Debris Flow Zatm NC093 31 SWWitt Rt. 215 35.2900 82.9100 Transylvania Slump Zatm NC095 31 E



Pomeroy Number 44 35.4300 82.6400 Henderson 1977 Debris Flow Zatm NC093 31 S

Pomeroy No. 8, Debris Torrent 35.4800 82.6800 Buncombe 1977 Debris Flow Zatm NC093 31 SW

Rd NCDOT NC 63, 5.1mi N of Buncombe County line

35.7000 82.8200 Madison 1980 Slump Ybgg NC095 31 NE

Rd Otteman 35.4967 82.6431 Buncombe 1977 Debris Flow Zatm NC093 31 S

Cv? USFS Microburst; dual slides near Dry Branch

35.7700 83.0400 Haywood 1994 Debris Flow Zsl NC102 32 SE


Rd Otteman Number 29 35.4600 82.6800 Buncombe 1977 Debris Flow Zatm NC095 32 NE

Pomeroy 35.4900 82.6500 Buncombe 1977 Debris Flow Zatm NC095 32 S

Witt identified in 3D Analyst 35.7200 86.0400 Haywood Slump Zsl NC102 32 W

Rd Pomeroy Number 28/29 35.4700 82.6600 Buncombe 1977 Debris Flow Zatm NC095 33 E

rk DOQQ identified from DOQQ 35.7700 83.0900 Haywood Slump Zsp NC102 33 SW




155

y




Cv NCGS Blackstone Knob East Slide 35.7338 82.3154 Buncombe 2004 Debris Flow Zatm NC095 33 SW


Pomeroy Number 14 (second head) 35.4900 82.6600 Buncombe 1977 Debris Flow Zatm NC095 34 NE


rk NCDOT I-40, East of Twin Tunnels 35.7600 83.0400 Haywood 1985 Slump Zsl NC102 34 SE

NCGS Pounding Mill Branch Slide 35.9951 82.7325 Madison 2001 Debris Flow Zwc NC092 34 SPomeroy Number 33 35.4600 82.6500 Buncombe 1977 Debris Flow Zatm NC095 35 E

Cv Pomeroy Chute 920 ft. long, debristorrent

35.4600 82.6500 Buncombe 1977 Debris Flow Zatm NC095 35 SE

MGW Pomeroy Number 28/29 35.4700 82.6600 Buncombe 1977 Debris Flow Zatm NC095 35 E

Wooten Mt. Mitchell Slide 35.7300 82.3000 Buncombe 2003 Debris Flow Zatw NC098 35 S

Rd NCDOT SBL NC 25-70, 250 ft. N of SR 1319

35.8900 82.7500 Madison 1986 Slump Zsr NC092 35 W


Pomeroy Debris Torrent 35.4900 82.6800 Buncombe 1977 Debris Flow Zatm NC095 35 Wwr NCDOT Lesser Fines Creek Slide 35.6700 82.9900 Haywood 1950 Slump Zbgg NC006 35 S


Witt Rt. 215 35.2900 82.9100 Transylvania Slump Zatm NC095 36 E

Pomeroy Number 8 35.4800 82.6800 Buncombe 1977 Debris Flow Zatm NC093 36 NW

Pomeroy Debris Torrent 35.4900 82.6600 Buncombe 1977 Debris Flow Zatm NC095 36 SW

Rk NCDOT MP 0.4, I-40 35.7700 83.0900 Haywood 1997 Slump Zsp NC102 36 S

Pomeroy 35.4468 82.6474 Henderson 1977 Debris Flow Zatm NC095 36 S

Otteman number 27 35.4600 82.6700 Buncombe 1977 Debris Flow Zatm NC095 37 E


Cv NCDOT SBL NC 261 36.1000 82.1000 Mitchell 1977 Debris Flow Ymg NC098 37 SW



Cv Otteman number 28 35.4600 82.6700 Buncombe 1977 Debris Flow Zatm NC095 38 NE

Rd NCDOT NBL NC 25-70, S of NC208

35.9100 82.7500 Madison 1989 Slump Zsr NC092 38 NW





156

y


Witt Rt. 215 35.2900 82.9100 Transylvania Slump Zatm NC095 39 E

Cv NCDOT I-40, may have beenrockfall, not sure

35.7800 83.1000 Haywood 1989 Slump Zsp NC102 39 S

wr NCDOT NC 215 35.3500 82.9100 Haywood 1988 Slump Zatm NC095 39 NE

Rd Pomeroy debris torrent 35.4600 82.6700 Buncombe 1977 Debris Flow Zatm NC095 40 SE

Rd Pomeroy Ext. rock rubble at toe,multi-headed flow, No. 28

35.4600 82.6700 Buncombe 1977 Debris Flow Zatm NC095 40 SE

Pomeroy Number 29 35.4600 82.6700 Buncombe 1977 Debris Flow Zatm NC095 41 EPomeroy Number 27 35.4800 82.6700 Buncombe 1977 Debris Flow Zatm NC095 41 E

Pomeroy Number 9 35.4900 82.6800 Buncombe 1977 Debris Flow Zatm NC093 41 W


Pomeroy Debris Torrent 35.4900 82.6700 Buncombe 1977 Debris Flow Zatm NC093 41 W

Pomeroy 35.5000 82.6700 Buncombe 1977 Debris Flow Zatm NC095 41 E


Pomeroy Number 1 35.4700 82.7000 Buncombe 1977 Debris Flow Zatm NC095 42 NW


Pomeroy 35.4900 82.6600 Buncombe 1977 Debris Flow Zatm NC095 46 EPomeroy Number 22 35.4800 82.6600 Buncombe 1977 Debris Flow Zatm NC095 1977 SE

Appendix E: SINMAP Results

Default Parameters - 30m DEM

Stable ModeratelyStable Quasi-Stable LowerThreshold UpperThreshold Defended Total

Area (km²) 3796.6 949.2 1216.3 1114.8 15.3 0.1 7092.3

% of Region 53.5 13.4 17.1 15.7 0.2 0.0 100.0

# of Landslides 12 14 41 73 0 0 140

% of Slides 8.6 10.0 29.3 52.1 0.0 0.0 100.0

LS Density 0.003 0.015 0.034 0.065 0.000 0.000

Recharge 50mm – 30m DEM



157

Stable Moderately

Stable

Quasi-Stable Lower

Threshold

Upper

Threshold

Defended Total

Area (km²) 1735.2 358.0 544.1 3112.1 1315.9 3.8 7069.1

% of Region 24.5 5.1 7.7 44.0 18.6 0.1 100.0

# of Landslides 2 2 5 46 84 1 140

% of Slides 1.4 1.4 3.6 32.9 60.0 0.7 100.0

LS Density 0.001 0.006 0.009 0.015 0.064 0.263

Recharge 125mm – 30mDEMStable Moderately

Stable

Quasi-Stable Lower

Threshold

Upper

Threshold

Defended Total

Area (km²) 1735.2 358.0 544.1 3103.3 1286.7 28.1 7055.4

% of Region 24.5 5.1 7.7 44.0 18.2 0.4 99.9

# of Landslides 2 2 5 46 81 4 140

% of Slides 1.4 1.4 3.6 32.9 57.9 2.9 100.1

LS Density 0.001 0.006 0.009 0.015 0.063 0.142


Stable Moderately

Stable

Quasi-Stable Lower

Threshold

Upper

Threshold

Defended Total

Area (km²) 1735.2 358.0 544.1 3102.1 1278.7 53.6 7071.7

% of Region 24.5 5.1 7.7 43.9 18.1 0.8 100.1

# of Landslides 2 2 5 46 78 7 140

% of Slides 1.4 1.4 3.6 32.9 55.7 5.0 100.0

LS Density 0.001 0.006 0.009 0.015 0.061 0.131


Stable Moderately

Stable

Quasi-Stable Lower

Threshold

Upper

Threshold

Defended Total

Area (km²) 1735.2 358.0 544.1 3102.1 1278.7 61.4 7079.5

% of Region 24.5 5.1 7.7 43.8 18.1 0.9 100.1

# of Landslides 2 2 5 46 77 8 140

% of Slides 1.4 1.4 3.6 32.9 55.0 5.7 100.0

LS Density 0.001 0.006 0.009 0.015 0.060 0.130

R h 125 C 1 25 30 DEM



158

Recharge 125mm, C = .1-.25 – 30m DEMStable Moderately

Stable

Quasi-Stable Lower

Threshold

Upper

Threshold

Defended Total

Area (km²) 2524.7 535.7 844.5 2355.6 811.2 3.8 7075.5

% of Region 35.7 7.6 11.9 33.3 11.5 0.1 100.1

# of Landslides 8 4 4 63 60 1 140

% of Slides 5.7 2.9 2.9 45.0 42.9 0.7 100.1

LS Density 0.003 0.007 0.005 0.027 0.074 0.263

Default Parameters - 10m DEM

Stable Moderately

Stable

Quasi-Stable Lower

Threshold

Upper

Threshold

Defended Total

Area (km²) 3673.9 868.0 1120.2 1229.4 44.3 0.6 6936.4

% of Region 53.0 12.5 16.1 17.7 0.6 0.0 100.0

# of Landslides 14 11 24 69 4 0 122.0

% of Slides 11.5 9.0 19.7 56.6 3.3 0.0 100.0

LS Density 0.004 0.013 0.021 0.056 0.090 0.000

Recharge 50mm- 10m DEM

Stable Moderately

Stable

Quasi-Stable Lower

Threshold

Upper

Threshold

Defended Total

Area (km²) 1591.9 322.7 494.4 2944.9 1514.0 24.8 6892.7

% of Region 23.1 4.7 7.2 42.7 22.0 0.4 100.0

# of Landslides 0.0 2.0 5.0 33.0 82.0 0.0 122.0

% of Slides 0.0 1.6 4.1 27.0 67.2 0.0 100.0

LS Density 0.000 0.006 0.010 0.011 0.054 0.000

Recharge 125mm - 10m DEMStable Moderately

Stable

Quasi-Stable Lower

Threshold

Upper

Threshold

Defended Total

Area (km²) 1591.9 322.7 494.4 2932.5 1455.2 70.8 6867.5

% of Region 23.2 4.7 7.2 42.7 21.2 1.0 100.0

# of Landslides 0.0 2.0 5.0 33.0 74.0 8.0 122.0

% of Slides 0.0 1.6 4.1 27.0 60.7 6.6 100.0

LS Density 0.000 0.006 0.010 0.011 0.051 0.113



159

Recharge 250mm - 10m DEM

Stable Moderately

Stable

Quasi-Stable Lower

Threshold

Upper

Threshold

Defended Total

Area (km²) 1591.9 322.7 494.4 2930.9 1434.2 126.4 6900.5

% of Region 23.1 4.7 7.2 42.5 20.8 1.8 100.0

# of Landslides 0.0 2.0 5.0 33.0 70.0 12.0 122.0

% of Slides 0.0 1.6 4.1 27.0 57.4 9.8 100.0

LS Density 0.000 0.006 0.010 0.011 0.049 0.095

Recharge 375mm -10m DEM

Stable Moderately

Stable

Quasi-Stable Lower

Threshold

Upper

Threshold

Defended Total

Area (km²) 1591.9 322.7 494.4 2930.8 1432.4 138.8 6911.0

% of Region 23.0 4.7 7.2 42.4 20.7 2.0 100.0

# of Landslides 0.0 2.0 5.0 33.0 70.0 12.0 122.0

% of Slides 0.0 1.6 4.1 27.0 57.4 9.8 100.0

LS Density 0.000 0.006 0.010 0.011 0.049 0.086

Recharge 125mm, C = .1-.25 – 10m DEM

Stable Moderately

Stable

Quasi-Stable Lower

Threshold

Upper

Threshold

Defended Total

Area (km²) 2306.1 495.1 789.9 2269.2 959.3 50.2 6869.8

% of Region 33.6 7.2 11.5 33.0 14.0 0.7 100.0

# of Landslides 5.0 4.0 5.0 45.0 56.0 7.0 122.0

% of Slides 4.1 3.3 4.1 36.9 45.9 5.7 100.0

LS Density 0.002 0.008 0.006 0.020 0.058 0.139



160

Appendix F: SHALSTAB Results

Cohesion 2000, SFA 26 – 30m DEM

PERCENT CUMPERCENT INSTABILITY NUMSLIDES PERSLIDES CUMSLIDES Area km² CUMAREA LS DEN

14.87 14.87 Chronic Instability 10 45.45 45.45 243.8 243.8 0.041

20.50 35.37 < -3.1 5 22.73 68.18 336.2 580.0 0.015

11.40 46.76 -3.1 - -2.8 2 9.09 77.27 186.9 766.9 0.011

10.55 57.32 -2.8 - -2.5 2 9.09 86.36 173.1 940.0 0.012

6.40 63.72 -2.5 - -2.2 0 0.00 86.36 105.0 1044.9 0.000

0.81 64.53 > -2.2 1 4.55 90.91 13.3 1058.2 0.075

35.47 100.00 Stable 2 9.09 100.00 581.8 1639.9 0.003

Cohesion 2000 SFA 45 – 30m DEM



161

Cohesion 2000, SFA 45 30m DEMPERCENT CUMPERCENT INSTABILITY NUMSLIDES PERSLIDES CUMSLIDES Area km² CUMAREA LS DEN


2.00 2.04 < -3.1 3 13.64 13.64 32.8 33.5 0.092

4.19 6.23 -3.1 - -2.8 2 9.09 22.73 68.7 102.2 0.029

9.63 15.85 -2.8 - -2.5 5 22.73 45.45 157.8 260.0 0.032

9.20 25.05 -2.5 - -2.2 1 4.55 50.00 150.8 410.8 0.007

1.99 27.04 > -2.2 0 0.00 50.00 32.6 443.4 0.000

72.96 100.00 Stable 11 50.00 100.00 1196.6 1639.9 0.009

Cohesion 9427.41, SFA 45 (SINMAP Upper Bound) – 30m DEM



0.00 0.00 < -3.1 0 0.00 0.00 0.0 0.0 0.000

0.00 0.00 -3.1 - -2.8 0 0.00 0.00 0.0 0.0 0.000

0.01 0.01 -2.8 - -2.5 0 0.00 0.00 0.2 0.2 0.000

0.04 0.05 -2.5 - -2.2 0 0.00 0.00 0.6 0.8 0.0000.03 0.08 > -2.2 0 0.00 0.00 0.5 1.3 0.000

99.92 100.00 Stable 22 100.00 100.00 1638.7 1639.9 0.013

Default Results – 30m DEM



4.17 4.21 < -3.1 3 13.64 13.64 68.3 69.0 0.044

7.18 11.39 -3.1 - -2.8 4 18.18 31.82 117.7 186.7 0.034

14.33 25.72 -2.8 - -2.5 6 27.27 59.09 235.1 421.8 0.026

12.31 38.03 -2.5 - -2.2 2 9.09 68.18 201.8 623.6 0.0102.53 40.55 > -2.2 0 0.00 68.18 41.4 665.0 0.000

59.45 100.00 Stable 7 31.82 100.00 974.9 1639.9 0.007

Cohesion 0, SFA 26 (SINMAP Lower Bound) – 30m DEM



24.52 50.11 < -3.1 6 27.27 77.27 402.1 821.8 0.015

9 60 59 71 -3 1 - -2 8 3 13 64 90 91 157 5 979 3 0 019



162

9.60 59.71 3.1 2.8 3 13.64 90.91 157.5 979.3 0.019

8.72 68.43 -2.8 - -2.5 0 0.00 90.91 143.0 1122.2 0.000

5.20 73.63 -2.5 - -2.2 2 9.09 100.00 85.2 1207.5 0.023

0.38 74.01 > -2.2 0 0.00 100.00 6.3 1213.8 0.000

25.99 100.00 Stable 0 0.00 100.00 426.2 1639.9 0.000

Cohesion 0, SFA 45 – 30m DEM

PERCENT CUMPERCENT INSTABILIT NUMSLIDES PERSLIDES CUMSLIDES Area km² CUMAREA LS DEN0.04 0.04 Chronic Instability 0 0.00 0.00 0.7 0.7 0.000

2.00 2.04 < -3.1 3 13.64 13.64 32.8 33.5 0.092

4.19 6.23 -3.1 - -2.8 2 9.09 22.73 68.7 102.2 0.029

9.63 15.85 -2.8 - -2.5 5 22.73 45.45 157.8 260.0 0.032

9.20 25.05 -2.5 - -2.2 1 4.55 50.00 150.8 410.8 0.007

1.99 27.04 > -2.2 0 0.00 50.00 32.6 443.4 0.000

72.96 100.00 Stable 11 50.00 100.00 1196.6 1639.9 0.009

Cohesion 2000, STA 35 – 30m DEM

PERCENT CUMPERCENT INSTABILIT NUMSLIDES PERSLIDES CUMSLIDES Area km² CUMAREA LS DEN


7.55 8.62 < -3.1 4 18.18 31.82 123.8 141.4 0.032

9.66 18.28 -3.1 - -2.8 5 22.73 54.55 158.4 299.8 0.032

13.97 32.25 -2.8 - -2.5 2 9.09 63.64 229.1 528.9 0.009

9.61 41.86 -2.5 - -2.2 2 9.09 72.73 157.6 686.6 0.013

1.73 43.60 > -2.2 0 0.00 72.73 28.4 715.0 0.000

56.40 100.00 Stable 6 27.27 100.00 925.0 1639.9 0.006

Cohesion 0, STA 35 – 30m DEM

PERCENT CUMPERCENT INSTABILIT NUMSLIDES PERSLIDES CUMSLIDES Area km² CUMAREA LS DEN3.34 3.34 Chronic Instability 3 13.64 13.64 54.8 54.8 0.055

14.04 17.38 < -3.1 9 40.91 54.55 230.2 285.1 0.039

13.00 30.38 -3.1 - -2.8 2 9.09 63.64 213.1 498.2 0.009

14.94 45.31 -2.8 - -2.5 3 13.64 77.27 244.9 743.1 0.012

9.42 54.73 -2.5 - -2.2 1 4.55 81.82 154.5 897.6 0.006

1.49 56.23 > -2.2 0 0.00 81.82 24.5 922.1 0.000

43.77 100.00 Stable 4 18.18 100.00 717.8 1639.9 0.006



163

Total 10m DEM Haywood - 26 1922 0 2


32.57 32.57 Chronic Instability 20.0 86.96 86.96 534.8 534.8 0.037

17.09 49.66 < -3.1 0.0 0.00 86.96 280.7 815.5 0.000

8.52 58.18 -3.1 - -2.8 3.0 13.04 100.00 139.9 955.4 0.021

9.14 67.32 -2.8 - -2.5 0.0 0.00 100.00 150.1 1105.5 0.000

6.60 73.92 -2.5 - -2.2 0.0 0.00 100.00 108.3 1213.8 0.000

3.05 76.97 > -2.2 0.0 0.00 100.00 50.2 1264.0 0.000

23.04 100.01 Stable 0.0 0.00 100.00 378.3 1642.3 0.000




14.85 36.20 < -3.1 5.0 21.74 78.26 243.7 594.3 0.021

9.28 45.48 -3.1 - -2.8 2.0 8.70 86.96 152.4 746.7 0.013

10.92 56.40 -2.8 - -2.5 2.0 8.70 95.65 179.3 926.0 0.0118.33 64.73 -2.5 - -2.2 0.0 0.00 95.65 136.8 1062.8 0.000

4.02 68.75 > -2.2 0.0 0.00 95.65 66.0 1128.8 0.000

31.26 100.01 Stable 1.0 4.35 100.00 513.3 1642.1 0.002




6.85 10.01 < -3.1 7.0 30.43 56.52 112.6 164.5 0.062

7.26 17.27 -3.1 - -2.8 1.0 4.35 60.87 119.2 283.7 0.008

12.95 30.22 -2.8 - -2.5 4.0 17.39 78.26 212.7 496.4 0.019

12.78 43.00 -2.5 - -2.2 3.0 13.04 91.30 209.9 706.3 0.0146.83 49.83 > -2.2 0.0 0.00 91.30 112.1 818.4 0.000

50.17 100.00 Stable 2.0 8.70 100.00 823.9 1642.3 0.002




11.20 18.22 < -3.1 5.0 21.74 65.22 183.9 299.1 0.027

9.83 28.05 -3.1 - -2.8 1.0 4.35 69.57 161.4 460.5 0.006



164

14.55 42.60 -2.8 - -2.5 5.0 21.74 91.30 239.0 699.5 0.021

12.46 55.06 -2.5 - -2.2 1.0 4.35 95.65 204.7 904.2 0.005

6.27 61.33 > -2.2 0.0 0.00 95.65 103.0 1007.2 0.000

38.67 100.00 Stable 1.0 4.35 100.00 635.0 1642.2 0.002

Total 10m DEM Haywood - 35 1922 9427.41 2


0.00 0.00 Chronic Instability 0.0 0.00 0.00 0.0 0.0 0.0000.11 0.11 < -3.1 2.0 8.70 8.70 1.8 1.9 1.111

0.24 0.35 -3.1 - -2.8 1.0 4.35 13.04 3.9 5.8 0.256

0.92 1.27 -2.8 - -2.5 2.0 8.70 21.74 15.1 20.9 0.132

2.40 3.67 -2.5 - -2.2 3.0 13.04 34.78 39.4 60.3 0.076

2.59 6.26 > -2.2 2.0 8.70 43.48 42.6 102.9 0.047

93.74 100.00 Stable 13.0 56.52 100.00 1539.3 1642.2 0.008




2.29 2.63 < -3.1 8.0 34.78 34.78 37.6 43.2 0.213

3.37 6.00 -3.1 - -2.8 2.0 8.70 43.48 55.4 98.6 0.036

8.46 14.46 -2.8 - -2.5 3.0 13.04 56.52 139.0 237.5 0.022

11.85 26.31 -2.5 - -2.2 7.0 30.43 86.96 194.6 432.1 0.036

7.70 34.02 > -2.2 0.0 0.00 86.96 126.5 558.6 0.000

65.99 100.00 Stable 3.0 13.04 100.00 1083.6 1642.2 0.003


PERCENT CUMPERCENT INSTABILIT NUMSLIDES PERSLIDES CUMSLIDES Area km² CUMAREA LS DEN0.09 0.09 Chronic Instability 0.0 0.00 0.00 1.6 1.6 0.000

0.99 1.09 < -3.1 5.0 21.74 21.74 16.3 17.9 0.307

1.69 2.78 -3.1 - -2.8 2.0 8.70 30.43 27.8 45.7 0.072

4.92 7.70 -2.8 - -2.5 3.0 13.04 43.48 80.8 126.5 0.037

8.50 16.20 -2.5 - -2.2 4.0 17.39 60.87 139.5 266.0 0.029

6.39 22.59 > -2.2 2.0 8.70 69.57 104.9 370.9 0.019

77.41 100.00 Stable 7.0 30.43 100.00 1271.3 1642.2 0.006

T t l 10 DEM H d 45 1922 9427 41 2



165

Total 10m DEM Haywood - 45 1922 9427.41 2



0.00 0.00 < -3.1 0.0 0.00 0.00 0.1 0.1 0.000

0.01 0.01 -3.1 - -2.8 1.0 4.35 4.35 0.1 0.2 8.598

0.03 0.04 -2.8 - -2.5 0.0 0.00 4.35 0.5 0.7 0.000

0.15 0.19 -2.5 - -2.2 1.0 4.35 8.70 2.4 3.1 0.417

0.33 0.52 > -2.2 0.0 0.00 8.70 5.4 8.5 0.00099.48 100.00 Stable 21.0 91.30 100.00 1633.7 1642.2 0.013

Total Default Values - 10m DEM Haywood



3.96 4.30 < -3.1 8 34.78 34.78 65.0 70.6 0.123

5.21 9.51 -3.1 - -2.8 2 8.70 43.48 85.5 156.1 0.023

12.38 21.89 -2.8 - -2.5 5 21.74 65.22 203.2 359.3 0.025

15.61 37.50 -2.5 - -2.2 6 26.09 91.30 256.3 615.6 0.023

9.52 47.02 > -2.2 0 0.00 91.30 156.4 772.0 0.000

52.98 100.00 Stable 2 8.70 100.00 870.1 1642.1 0.002

Documents

Haywood County Landslide Data