Effects of Wildfires on Runoff and Erosionucanr.edu/sites/prepostwildfire/files/2671.pdf · Effects...

Effects of Wildfires on Runoff and Erosion

Lee H. MacDonald

Watershed Science Program

Colorado State University, Fort Collins, CO

Contributors

• Tedd Huffman (M.S., 2002);

• Juan Benavides-Solorio (Ph.D., 2003);

• Joe Wagenbrenner (M.S., 2003);

• Matt Kunze (M.S., 2003);

• Zamir Libohova (M.S., 2004);

• Jay Pietraszek (M.S., 2006);

• Daniella Rough (M.S., 2007);

• Duncan Eccleston (M.S., 2008);

• Keelin Schaffrath (M.S., 2009);

• Darren Hughes (M.S., 2010);

• Ethan Brown (M.S., 2009);

• Dr. John Stednick;

• Isaac Larsen (Research Assistant, 2004-2007);

• Sergio Alegre (visiting Ph.D. student, 2010).

Why the concern about wildfires?

Hayman Fire, Colorado: August 2004

Channel incision from a 20 mm/hr rain event after

the Cerro Grande Fire near Los Alamos, NM

Photo by John Moody, USGS

Alluvial fan from Saloon Gulch extending into

the South Platte River, Summer 2004(2 years after burning!)

Post-fire Hydrology

Objectives

1. Provide a process-based understanding of the effects of wild and prescribed fires on soils, runoff, and erosion;

2. Evaluate the relative importance of different controlling factors on post-fire erosion rates;

3. Determine the rate of recovery to pre-fire conditions;

4. Discuss how post-fire processes and recovery vary with increasing scale, and put the effects of wildfire in a broader context.

Post-fire Effects Vary with Burn Severity

1) High severity: complete consumption of organic horizon and alteration of the structure or color of the underlying mineral soil; loss of aggregates (“pulverization”):

2) Moderate severity: consumption of litter layer but no visible alteration of the surface of the mineral soil;

3) Low severity: only partial consumption of the surface litter.

Severity is not equal to intensity (heat loss per unit width per unit time), but severity and intensity often assumed to be closely correlated;

Why the sharp increase in runoff and

erosion after some high-severity wildfires?

1. Loss of canopy decreases interception and

evapotranspiration, increasing runoff;

2. Loss of litter decreases interception and exposes soil to

raindrop impacts (increased erodibility) and sealing;

3. Loss of soil organic matter disaggregates or pulverizes

the soil, and this increases soil erodibility;

4. Increase in soil water repellency can decrease

infiltration and increase surface runoff;

5. Loss of litter decreases surface roughness and

increases runoff velocities, increasing erosion;

Effects are synergistic, but which is most important?

Soil Water Repellency

Fire-induced soil water repellency

(DeBano, 1981)

Methods of Analysis

Water drop penetration time (WDPT):

• Apply drops at different depths, beginning at mineral soil surface;

• Indefinite waiting time;

• Assesses persistence of soil water repellency.

Critical surface tension test (CST):

• Apply 5 drops of de-ionized water;

• If 4 of 5 drops are not absorbed within 5 seconds, test

solutions with progressively higher ethanol concentrations

(increasing ethanol concentrations decrease surface tension);

• Critical surface tension (CST) is the tension of the first

solution that is readily absorbed into the soil (“strength”).

Critical surface tension in wild and

prescribed fires: High-severity sites(bottom two sites are prescribed fires)

30

40

50

60

70

80

0 3 6 9 12 15 18

Depth (cm)

Cri

tical su

rface t

en

sio

n

(dyn

es c

m-1)

Crosier M tn.

Hi M eadows

Bobcat

Lower Flowers

Dadd Bennett

Huffman et al., 2001

Critical surface tension in wild and prescribed

fires: Moderate-severity sites(bottom two sites are prescribed fires)

30

40

50

60

70

80

0 3 6 9 12 15 18

Depth (cm)

Cri

tical su

rface t

en

sio

n

(d

yn

es c

m-1)

Hi M eadows

Crosier M tn.

Bobcat

Lower Flowers

Dadd Bennett

Huffman et al., 2001

Critical surface tension in wild and prescribed

fires: Low severity sites(bottom two sites are prescribed fires)

30

40

50

60

70

80

0 3 6 9 12 15 18

Depth (cm)

Cri

tical su

rface t

en

sio

n

(d

yn

es c

m-1) Hi M eadows

Bobcat

Crosier M tn.

Lower Flowers

Dadd Bennett

Huffman et al., 2001

30

35

40

45

50

55

60

65

70

75

0 2.5 5 7.5 10 12.5

Depth (cm)

Me

dia

n c

riti

ca

l su

rfac

e t

en

sio

n (

dy

ne

s c

m-1

)

Unburned

Summer 2002 (10 months)

Summer 2003 (22 months)

Summer 2004 (34 months)

Median soil water repellency over time,

Star Fire, Tahoe National Forest

E. Chase, M.S.. thesis, Colorado State Univ., 2006

Mean soil water repellency by depth:

Unburned vs. burned sites, summer 2002

30

40

50

60

70

80

0 3 6 9 12Depth (cm)

Cri

tica

l su

rface

ten

sion

(d

yn

es c

m-1

)

Hayman fire (n=36)

Schoonover fire (n=18)

Trumbull: unburned

(n=39)

D. Rough, 2007

Soil water repellency from 2002-2004:

Upper Saloon Gulch, Hayman fire

30

35

40

45

50

55

60

65

70

75

80

0 3 6 9 12

Depth (cm)

Cri

tic

al s

urf

ac

e t

en

sio

n (

dy

ne

s c

m-1

)

2004

2003

2002

No water repellency

D. Rough, 2007

0 2 4 6 8 10 12 140

2

4

6

8

10

12

14

0

0.4

1

2

3

4

Meters

Mete

rs

Spatial variability in soil water repellency:

Plot H1, high severity, Hayman fire

Moles of

ethanol

per liter

Woods et al. 2007, Geomorphology

Summary: Soil Water Repellency

• Soils in unburned areas usually water repellent;

• Fire-induced water repellency is usually shallow

(maximum of 9 cm);

• May be stronger in prescribed fires due to higher fuel

loadings and slower rate of fire spread;

• Very high spatial variability;

• Relatively rapid recovery (≤ 2 years);

• Not present under wet conditions (~10-35 percent soil

moisture), depending on fire severity;

• CST faster and more consistent than WDPT.

Supporting Data

Three papers on my web site (type “Lee MacDonald” into google):

1. Huffman, E.L., L.H. MacDonald, and J.D. Stednick, 2001. “Strength and persistence of fire induced soil hydrophobicity under ponderosa and lodgepole pine, Colorado Front Range”, Hydro. Proc. 15: 2877-2892.

2. MacDonald, L.H., and E.L. Huffman, 2004. “Persistence and soil moisture thresholds”, Soil Sci. Soc. Am. J. 68: 1729-1724;

3. Doerr, S.H., R.H. Shakesby, and L.H. MacDonald, 2009. “Soil water repellency: a key factor in post-fire erosion?” In Restoration Strategies after Forest Fires, edited by A. Cerda and P.R. Robichaud, Science Publishers, Inc., Enfield, NH.

Sediment Production at the

Hillslope Scale

Untreated

High severity 319

Moderate severity 55

Low severity 34

Treated (all high severity)

Seeding and scarification with seeding 36

Straw mulch and straw mulch with seeding 60

Contour-felled logs 44

Ground-applied hydromulch 20

Aerially-applied hydromulch 20

Polyacrylamide 12

Total 600

Total plot years of data by treatment

0.0

2.0

4.0

6.0

8.0

10.0

High Moderate Low

Sed

imen

t yie

ld (

Mg

ha

-1)

Fire severity

Summer Rainfall (Jun-Oct)

Snowmelt (Nov-May)

Sediment yields by fire severity and season:

First two years after burning (Colorado)

Role of surface cover,

recovery over time,

and rainfall intensity

Sediment production: Summer 2001 (before Hayman fire)

0.00

0.20

0.40

0.60

0.80

1.00

1.20

1 2 3 4 5 6 7 8 9 10

Pairs of sediment fences (n = 20)

Sed

imen

t (k

g/m

2)

Mean percent ground cover in Upper Saloon Gulch in

2001 (prior to burning) and 2002 (after the Hayman fire)

0

10

2 0

3 0

4 0

5 0

6 0

7 0

8 0

9 0

10 0

1\ 8 1\ 9 1\ 12 1\ 13 1\ 14 1\ 15 1\ 16 1\ 17 1\ 18 1\ 19 1\ 2 0 1\ 2 1 1\ 2 2 1\ 2 3 1\ 2 4 1\ 2 5 1\ 2 6 1\ 2 7 1\ 2 8 1\ 2 9

Swale

Gro

un

d c

over (

%)

2001

2002

Sediment from 11 mm of precipitation in

45 minutes on 21 July 2002

Sediment production after Hayman fire:

21 July 2002 storm (11 mm in 45 minutes)

0.00

0.20

0.40

0.60

0.80

1.00

1.20

1.40

1 2 3 4 5 6 7 8 9 10

Pairs

Sed

imen

t (k

g/m

2)

Control

"Treated" (not

implemented)

Year 2001

69% Bare soil

Year 2002

17% Bare soil

Year 2000, 15 days after fire

96% Bare soil

Vegetation recovery over timeBobcat fire, sediment fence #9

Year 2003

12% Bare soil

y = -26.09Ln(x) + 70.03

R2 = 0.63

y = -21.86Ln(x) + 45.75

R2 = 0.60

y = -10.44Ln(x) + 26.21

R2 = 0.49

0

20

40

60

80

100

0 1 2 3 4 5 6 7 8 9 10

High severity

Moderate severity

Low severity

Time since burning (years)

Pe

rce

nt b

are

so

ilPercent bare soil vs. time since burning

y = 0.0029e0.095x

p < 0.0001

R2 = 0.61

n=345

0

10

20

30

40

50

0 20 40 60 80 100

Sediment yield vs. percent bare soil

Percent bare soil

Sedim

ent yie

ld (

Mg h

a-1

yr-

1)

Event-based sediment production vs. I30:

High-severity wildfires

R2 = 0.62

R2 = 0.50

0

2

4

6

8

10

0 10 20 30 40 50

30-minute maximum intensity (mm hr-1

)

Mean

sed

imen

t p

rod

ucti

on

(M

g h

a-1

)

1-3 years after burning

4-5 years after burning

Sediment production over time:

Pendola fire, Eldorado N.F.

Post-fire erosion vs. percent bare soil:

Pendola fire, Eldorado N.F.

Percent Bare Soil (%)

Sedim

ent p

roductio

n (t h

a-1

yr

-1)

0 10 20 30 40 50 60 70 80 90

0

2

4

6

8

10

12

14

16

181999-20002000-20012001-2002

y=0.12900.0577x

R2=0.77p=0.01

Is all this sediment coming from:

(a) rainsplash and sheetwash on

the hillslopes; or

(b) rill, gully, and channel erosion?

Upper Saloon Gulch: 10 July 2002

17 mm rain in 2 hours

Sediment yields from swales vs.

planar hillslopes in 2001: Bobcat fire

R2 = 0.57

p = 0.01

R2 = 0.53

p = 0.0070

0.2

0.4

0.6

0.8

1

1.2

1.4

0 50 100 150 200 250

Erosivity (MJ mm ha -1 h-1)

Sed

imen

t p

rod

ucti

on

(kg

m-2

)

Planar hillslopes 2001 Swales 2001

Measuring rill erosion, Hayman fire

Rill erosion in Swale 4: Storm on 21 August 2003

-40

-30

-20

-10

0

0 10 20 30 40 50 60 70 80 90 100

cm

cm

15-Aug-02

23-Aug-02

-25

-20

-15

-10

-5

0

0 10 20 30 40 50 60 70 80 90 100

cm

cm

15-Aug-02

23-Aug-02

-40

-35

-30

-25

-20

-15

-10

-5

0

0 10 20 30 40 50 60 70 80 90 100

cm

cm

15-Aug-02

23-Aug-02-40

-35

-30

-25

-20

-15

-10

-5

0

0 10 20 30 40 50 60 70 80 90 100

cm

cm

15-Aug-02

23-Aug-02

12

3 4

8 mm rainfall

I30 = 15.6 mm/hr

27.4 MJ mm/ha yr

Estimated sediment from rill erosion vs.

measured sediment: Hayman wildfire

2500

0

500

1000

1500

2000

3000

0 500 1000 1500 2000 2500 3000

26-Jun-03 14-Jun-03 20-Jul-03

12-Aug-03 6-Sep-03 16-Jun-04

28-Jun-04 28-Jul-04 8-Sep-04

1:1

Measured sediment in fence (kg)

Estim

ate

d s

edim

ent

from

rill

incis

ion (

kg)

Inferred sources of runoff and erosion

• About 80% of the sediment is coming from rilling on

the hillslopes;

• These and other data indicate that the post-fire runoff

is coming from the hillslopes, but most of the post-fire

sediment is coming from incision due to concentrated

flows (rill, gully, and channel erosion);

• See also Moody and Martin, 2001; 2009.

Controls on Post-fire Erosion

• Erosion rates most strongly related to percent bare

soil, which is primarily a function of fire severity and

time since burning;

• For a given percent cover and slope, rainfall intensity

is the dominant control, and erosion increases non-

linearly with rainfall intensity or erosivity;

• Soil water repellency can help reduce infiltration after

burning, but the rapid decay and spatial variability

suggests it is not the dominant control;

• Soil type is generally a third-order control, after cover

and rainfall intensity;

• Rainfall simulations and other work suggest that post-

fire soil sealing is limiting infiltration (SSSAJ, 2009).

Runoff and Water Quality at

Catchment Scale

Saloon Gulch and Brush Creek: A Paired Watershed

Study to Investigate the Effects of Thinning

Stream reaches: Summer 2001

Saloon Gulch Brush Creek

Saloon Gulch flume before Hayman fire

Saloon Gulch flume after first post-fire rainstorm

Saloon Gulch flume cleaned out after

first post-fire rainstorm

Saloon Gulch flume after second

post-fire rainstorm

Lower Brush Creek: Upstream of flume

Since runoff rates decline within 2-4 years after burning,

how long will it take to transport the excess sediment

out of this channel?

Bobcat fire, 8 years later: How long

until this becomes a forest again?

Hayman fire, 7 years later: How long

until this becomes a forest again?

Hayman fire,

seven years later:

How long until

this stops eroding

and degrading

water quality?

Buffalo Creek fire, 2009(13 years after burning)

Hypothetical erosion rates over

time from different sources

Ero

sio

n

Background

Wildfires

Thinning

.01

.1

1

10

Time

Roads

Conclusions: Part 2

• High-severity fires can dramatically increase runoff and erosion rates in headwater areas;

• Large sediment deposits in lower-gradient channels can result in long-term degradation of aquatic habitat;

• For more information, see my web site (type “Lee MacDonald” into google).

Questions?