Upload
others
View
5
Download
0
Embed Size (px)
Citation preview
DSpace Institution
DSpace Repository http://dspace.org
Hydraulic and Water Resources Engineering thesis
2020-03
Identifying Soil Erosion Hotspot Area
Using GIS and MCDA Techniques,
Case Study of Dengora and Meno
Watersheds in Belesa Woredas,
Amhara Region, Ethiopia
Munye, Kefale
http://hdl.handle.net/123456789/11014
Downloaded from DSpace Repository, DSpace Institution's institutional repository
BAHIR DAR UNIVERSITY
BAHIR DAR UNIVERSTITY INSTITUTE OF TECHNOLOGY
SCHOOL OF RESEARCH AND POST GRADUTE STUDIES
FACULTY OF CIVIL AND WATER RESOURCES ENGINEERING
HYDRAULICS ENGINEERING PROGRAM
Identifying Soil Erosion Hotspot Area Using GIS and MCDA Techniques, Case Study of
Dengora and Meno Watersheds in Belesa Woredas, Amhara Region, Ethiopia
By
Kefale Munye Ejigu
March, 2020
BBaahhiirr DDaarr,, EEtthhiiooppiiaa
Identifying Soil Erosion Hotspot Area Using GIS and MCDA Techniques, Case Study of
Dengora and Meno Watersheds in Belesa Woredas, Amhara Region, Ethiopia
Kefale Munye Ejigu
A thesis submitted to the school of Research and Graduate Studies of Bahir Dar Institute of
Technology, Bahir Dar University in partial fulfillment of the requirements for the degree of
Master of Science in Hydraulics Engineering in the Faculty of Civil and Water Resource
Engineering
Advisor Name: Mamaru Ayalew Moges (Ph.D)
Co-Advisor Name: Seifu Admassu Tilahun (Ph.D)
March, 2020
Bahir Dar, Ethiopia
i
ii
2020
Kefale Munye Ejigu
ALL RIGHTS RESERVED
iii
iv
v
“This work is dedicated to my family, friends and who loved me. Special dedication goes to my
mother Mosiet Tefera and my father Munye Ejigu”.
BIOGRAPHICAL SKETCH
I was born in 1992 in the Amhara Regional State, East Gojjam province Mota district, Ethiopia. I
went to Keranio Elementary school at the age of seven and studied for 7 years up to grade seven
and then transfer to Sedie Elementary school for study grade eight. Then, I went to Sedie
vi
secondary high school since 2007/8 and attended grade 9 to 10. I took grade 10 General
Secondary Leaving Examination in 2008 and passed to Mota preparatory school. After studied
natural science discipline at Motta preparatory school, I have joined Hawasa University in 2012.
I studied Bachelor degree of Water Resource and Irrigation Engineering and graduated in July,
2016. After graduation in 2016, I have been employed at Woldia University as Assistance
lecturer position and worked for one year in 2016/2017 season. Then, providentially, I have been
sponsored by Woldia University in 2017 to study master‟s degree in Hydraulics Engineering at
Bahir Dar University Institute of Technology. I‟m interested to study Hydraulics structure,
hydrological modeling, GIS and remote sensing and Hydrology Therefore, my dream is to
specialize in water resources engineering, Hydraulics Engineering, Hydropower in the future.
ACKNOWLEDGEMENTS
Any accomplishment needs co-operation and efforts of others; this thesis could not have been
carried out without the intensive assistance and guidance of others. First of all, I would like to
vii
thank the LORD GOD with his mother St. Marry for his sympathy, kindness and grace up on me
in all my life.
I would like to thank Bahir Dar University and CARE Project for the financial support during
this work and to give this chance. I would like to extend my gratitude to Woldia University for
sponsorship. My sincere thanks also Bahir Dar University, Institute of technology for providing
and gaining educational knowledge.
I am very much indebted to my advisors Dr. Mamaru Ayalew and Dr.Seifu Admassu for their
unreserved guidance, encouragement and critical comments starting from proposal write up to
the research undertaking. My deepest, maximum respect and special thanks go to Dr. Adugnaw
Tadesse for their initiation and providing an excellent advice and constructive comments over
the course of this research.
I would also like to thank my colleagues Fasikaw Fentie and Berhanu Geremew for giving
valuable comment on this work. Besides, Hydraulics students Haile and Hydrology student
Temesgen were highly acknowledged for their contribution during land use land cover
classification data and GPS point collection.
I would like to express my sincere appreciation and special thanks are given to my deepest friend
Daniel Anmaw and their beloved friends for their support, encouragement and motivation to
undertake the research and in addition to that to give any technical support and building idea to my
life.
Finally, I would like to express my genuine thanks to all my families, my sisters Tarik Munye, my
two brothers (Lakachew Munye and Belachew Munye) and my best friends for giving me care and
love during my research work. Especially thanks given to my Mom Mosit Tefera and my father
Munye Ejigu who support and encourage me throughout my life to be a better person. I also would
like to extend my thanks to all my best friends and classmates for their knowledge sharing,
encouragement and constructive comments.
ABSTRACT
In Ethiopia soil erosion and land degradation has become a key issue, because of roughed and
steep slope topography that soil erosion become accelerating. The main objective of this study
viii
was to identify soil erosion hotspot areas in Dengora and Meno watersheds using Revised
Universal Soil Loss Equation (RUSLE) and Multi Criteria decision Analysis (MCDA)
techniques. Based on RUSLE model, the average annual soil loss of Dengora and Meno
watersheds were reaches up to 223.97 and 256.09 ton ha-1
yr-1
respectively. In the Dengora
watershed 70.4%, 18.7%, 10.74% and 0.14% of the total watershed area predicted soil loss
ranges between 0 to 15, 15 to 50, and 50 to 200 and above 200 ton ha-1
respectively. In Meno
watershed 76%, 16.54%, 7.3% and 0.14% of the total watershed area predicted soil loss ranges
between 0 to 15, 15 to 50, 50 to 200 and above 200 ton ha-1
respectively. On the other hand, the
GIS based MCDA technique considered five major factors land use, soil type, topographic
wetness index, stream power index and potential location of gullies. The factors were weighted
using pair-wise comparison matrix and weights were combined using Weighted Overlay Tool of
ArcGIS Spatial Analyst Toolbox to obtain the final erosion hotspot map. In Dengora watershed
9.7%, 64.5%, 18% and 7.8% of the total watershed area was highly, moderately, slightly and
currently not sensitive to soil erosion respectively. In Meno watershed 6.1%, 71.3%, 23.23% and
0.375% of the total watershed area was highly, moderately, slightly and currently not sensitive to
soil erosion respectively. Based on validation, field level observation, MCDA model prediction
was more accurate than RUSLE. Both of the watersheds were at moderate risk. Thus,
bottomlands of the watersheds under highly sensitive areas for erosion therefore immediate
attention for soil and water conservation practice. Therefore, both tools should be applied for
planning and targeting of watershed intervention.
Keywords: Dangora watershed, Erosion hotspot, GIS, Meno watershed, MCDA, RUSLE
ACRONYMS AND ABBREVIATIONS
AHP Analytical Hierarchy Process
Arc SWAT ArcGIS Integrated SWAT Hydrological Model
DEM Digital Elevation Model
ix
EHP Erosion Hazard Parameter
FAO Food and Agricultural Organization of the United Nations
GIS Geographic information system
GPS Global Positioning System
LULC Land Use / Land Cover
m.a.s.l mean above sea level
MCDA Multi-criteria Decision Analysis
MoWIE Ministry of Water, Irrigation and Electricity
Mt Million ton
NMA National Meteorological Agency
RUSLE Revised Universal Soil Loss Equation
SCRP Soil Conservation Research Program
SPI Stream Power Index
SWAT Soil and Water Assessment Tool
SWC Soil and Water Conservation
TWI Topographic Wetness Index
USGS United States Geological Survey
UTM Universal Transverse Mercator
TABLE OF CONTENTS
DECLARATION ........................................................................... Error! Bookmark not defined.
BIOGRAPHICAL SKETCH .......................................................................................................... v
ACKNOWLEDGEMENTS ........................................................................................................... vi
ABSTRACT .................................................................................................................................. vii
x
ACRONYMS AND ABBREVIATIONS .................................................................................... viii
LIST OF FIGURE......................................................................................................................... xii
LIST OF TABLE ......................................................................................................................... xiv
LIST OF APPENDIX .................................................................................................................. xvi
1. INTRODUCTION .................................................................................................................. 1
1.1. Background ............................................................................................................................. 1
1.2. Problem of Statement .............................................................................................................. 3
1.3. Objective of the Study ............................................................................................................ 4
1.3.1. Main Objective................................................................................................................. 4
1.3.2. Specific Objective ............................................................................................................ 4
1.4. Research Question .................................................................................................................. 4
1.5. Significance of the study ......................................................................................................... 4
1.6. Scope of the study ................................................................................................................... 5
1.7. Paper Organization.................................................................................................................. 5
2. LITERATURE REVIEW ....................................................................................................... 6
2.1. Soil and land degradation in Ethiopia ..................................................................................... 6
2.2. Forms of soil erosion .............................................................................................................. 7
2.3. Factors affecting soil erosion .................................................................................................. 8
2.4. Consequences of soil erosion ................................................................................................ 10
2.4.1. On-site Effects ............................................................................................................... 10
2.4.2. Off-site Effects ............................................................................................................... 11
2.5. Soil Erosion Modeling .......................................................................................................... 11
2.6. Application of GIS and Multi criteria Decision Analysis (MCDA) techniques ................... 12
2.7. Related previous studies on soil erosion ............................................................................... 13
3. MATERIALS AND METHODS .............................................................................................. 14
3.1. Description of the study area ................................................................................................ 14
xi
3.2. Topography ........................................................................................................................... 15
3.3. Climate .................................................................................................................................. 16
3.4. Soil Type ............................................................................................................................... 17
3.5. Land use land cover map ...................................................................................................... 18
3.6. Satellite Data and DEM ........................................................................................................ 20
3.7. Materials and Data Used ....................................................................................................... 21
3.8. Methodology ......................................................................................................................... 22
3.8.1. Data Collection .............................................................................................................. 22
3.9. Data Analysis ........................................................................................................................ 24
3.9.1. Modelling of soil erosion ............................................................................................... 25
3.9.2. RUSLE Factors Generation ........................................................................................... 25
3.9.3. Multi- Criteria Decision Analysis (MCDA) .................................................................. 30
3.9.4. Pair wise comparison ..................................................................................................... 37
3.9.5. The Analytic Hierarchy Process (AHP) and fundamental scale .................................... 37
3.9.6. Weighted overlay ........................................................................................................... 40
4. RESULTS AND DISCUSSIONS ............................................................................................. 41
4.1. Soil Loss factors .................................................................................................................... 41
4.1.1. Rainfall Erosivity Factor (R) ......................................................................................... 41
4.1.2. Soil Erodibility Factor (K) ............................................................................................. 41
4.1.3. Topographic (LS) factor ................................................................................................ 43
4.1.4. Land Cover Factor (C) ................................................................................................... 43
4.1.5. Land Management Practice Factor (P) .......................................................................... 45
4.2. Soil Loss Estimation ............................................................................................................. 46
4.3. Multi- Criteria Decision Analysis (MCDA) ......................................................................... 49
4.3.1. Land use land cover map ............................................................................................... 49
xii
4.3.2. Soil map ......................................................................................................................... 50
4.3.3. Topographic Wetness Index Factor ............................................................................. 52
4.3.4. Stream power index (STI) Factor .................................................................................. 54
4.3.5. Gully potential location map .......................................................................................... 56
4.4. Pairwise comparison for parameters ..................................................................................... 58
4.5. Identification of Soil Erosion Hotspot Areas ........................................................................ 60
5.CONCLUSIONS AND RECOMMENDATIONS .................................................................... 63
5.1. Conclusions ........................................................................................................................... 63
5.2. Recommendations ................................................................................................................. 64
6. REFERENCES ........................................................................................................................ .65
7. APPENDIX .............................................................................................................................. .72
LIST OF FIGURE
Figure 3-1 Location of Study area map…………………………………………………………14
Figure 3-2 Slope map and Digital elevation model of the watersheds………………………….15
Figure 3-3 Average monthly rainfall of Dengora (top) and Meno (bottom) watersheds……....16
Figure 3-4 Soil map of Dengora and Meno watersheds………………………………………...18
Figure 3-5 Land use land cover map for Dengora (top) and Meno (bottom) watershed……….20
Figure 3-6 Meno watershed transect walk……………………………………………………….23
xiii
Figure 3-7 Dengora watershed transect walk……………………………………………………24
Figure 3-8 General flow charts for RUSLE Generation………………………………………...30
Figure 3-9 Topographic wetness index map……………………………………………..............35
Figure 3-10 Stream power index (STI) map ……………………………………………………36
Figure 3-11 Workflow charts of the criteria weighting using MCDA in Arc GIS 10.1………....42
Figure 4-1 K-Factor map of Dengora (top) and Meno (bottom) watershed…………………….44
Figure 4-2 Topographic (LS) factor of Dengora (top) and Meno (bottom) watershed…………45
Figure 4-3 Land Cover (C) Factor map for Dengora (top) and Meno (bottom) watershed ……..46
Figure 4-4 Land management factor Dengora and Meno watershed…………………………...45
Figure 4-5 Estimated annual soil loss for Dengora (top) and Meno (bottom) watershed……….50
Figure 4-6 LULC sensitivity map……………………………………………………………….50
Figure 4-7 Soil type sensitivity map of Dengora and Meno watershed…………………….........54
Figure 4-8 TWI sensitivity class……………………………………………………………........56
Figure 4-9 SPI sensitivity class…………………………………………………………….…….57
Figure 4-10 Potential Location of Gully ………………………………………………………..59
Figure 4-11 Sample Gully of Dengora watershed……………………………………………….60
Figure 4-12 Photo of sample gully on Meno watershed………………………………………...61
Figure 4-13 Overall contributions of parameters for soil erosion…………………………….....63
Figure 4-14 Overall soil erosion risk map……………………………………………………….66
xiv
LIST OF TABLE
Table 3-1 Slope class category of watershed……………………………………………………15
Table 3-2 Major Soil type ............................................................................................................. 17
Table 3-3 Land use land covers type ............................................................................................ 18
Table 3-4 Type, purpose and sources of data/material ................................................................. 21
Table 3-5 Soil types with K values .............................................................................................. 26
Table 3-6 Land cover factor of study area .................................................................................... 28
Table 3-7 Land management factors (Wischmeier and Smith, 1978) .......................................... 28
xv
Table 3-8 Factor sensitivity classes .............................................................................................. 31
Table 3-9 Land use land covers accuracy assessment of study area..............................................33
Table 3-10 Saaty‟s 1977fundamental weighting scale of pair wise comparison .......................... 38
Table 3-11 RI on the basis of various sample size ........................................................................ 40
Table 4-1 RUSLE based soil loss severity class for Dengora watershed ..................................... 47
Table 4-2 RUSLE based soil loss severity class for Meno watershed ........................................ 47
Table 4-3 Dengora watershed LULC sensitivity class to soil erosion .......................................... 49
Table 4-4 Meno-watershed watershed LULC sensitivity class to soil erosion ............................. 49
Table 4-5 Dengora watershed soil type sensitivity class .............................................................. 51
Table 4-6 Meno-watershed soil type sensitivity classes ............................................................... 51
Table 4-7 Topographic wetness index sensitivity class………………………………………….55
Table 4-8 Stream power index sensitivity class………………………………………………….57
Table 4-9 Accuracy assessment of gully area ……………………………………………………61
Table 4-10 The influencing power of the factors………………………………………………...62
Table 4-11 Overall Dengora watershed Erosion sensitivity ....................................................... ..65
Table 4-12 Overall Meno- watershed Erosion sensitivity ............................................................ 65
LIST OF APPENDIX
Appendix A Error matrix accuracy totals for the classified image .............................................. 72
Appendix B AHP pair wise comparison matrix for Dengora watershed ..................................... 73
Appendix C AHP pair wise comparison matrix for Meno watershed ........................................ 73
Appendix D Actual Gully location for Dengora watershed ....................................................... 73
Appendix E Actual Gully location for Meno watershed ............................................................. 74
Appendix F Geographical Location of Meno and Dengora Watershed ...................................... 75
xvi
Appendix G Focal Group Discussion on Meno watershed……………………………………...76
Appendix H Focal Group Discussion on Dengora watershed………………………………….76
1
1. INTRODUCTION
1.1. Background
In Ethiopia soil erosion and land degradation has become a key issue, resulted on reducing the
dynamic capacity of the land which leads to loss of fertile top soil, decline in visual landscape
beauty, and decrease quality of water. This occurs through anthropogenic and natural activities.
Poor land use practice particularly inadequate soil and water conservation practice and
cultivation of steep slope significantly contributed to soil erosion and land degradation (Onyando
et al., 2005). Although many countries of the world suffer from erosion, due to lack of adaptive
capacity of their farming systems, lost top fertile soils and nutrients (Erenstein, 1999). Soil
erosion are increased due to deforestation, over grazing, poor farming practices and cultivating
marginal lands (Valentin et al., 2005). It has on-site and off-site impacts, the on-site impacts
includes loss of agricultural land leading to reduction in food production while off-site effects,
siltation of rivers and reservoirs leading to water quality deterioration (Asquith et al., 2005).
In the highlands of Ethiopia, soil erosion and land degradation is a serious and continues problem
that resulted in the loss of fertile top soil leading to low agricultural productivity(Hurni,
1993b).Lack of effective watershed management system and poor land use practices played
significant role in land degradation in the region (Setegn et al., 2009). About 1.3 billion ton of
fertile soil are lost each year and soil erosion and land degradation increase significantly due to
undulation and irregular topography in the area (Hurni, 1989b). According to various study in the
Ethiopia highlands, much of the lost land will economically insufficient in the near future.
According to the Ethiopian highland reclamation study (Yilma and Awulachew, 2009), in the
mid1980‟s, 27 million hectare or almost 50% of the high land area was significantly eroded, 14
million hectare seriously eroded and over 2 million hectare were beyond reclamation. Efforts
have been made to manage this loss. However, a number of previous studies have pointed out
that such arrangements were unsatisfactory and incompatible due to ineffective community
participation in planning stage, improper intervention selection, poor management after
construction, not integrated with biological conservation measures and others among the
smallholder farmers (Hurni, 1989b). This requires immediate action to estimate soil erosion rate
and identify high erosion source area in the watersheds.
2
Estimating soil erosion is the process of mathematically incorporating and describing soil
detachment, transport and deposition on land surfaces. Empirical methods are an inseparable part
of any erosion research to estimate the amount of sedimentation(Najm et al., 2013). At present, a
variety of erosion models exist focusing on different spatial scales (point to regional) and
temporal scales (event to continuous) with different degrees of complexity and precision to
address the practical implication of soil erosion at landscape level. However, researchers proved
that there is no single erosion model that can be universally accepted to apply in complex
watersheds. There is also no clear agreement in the scientific community on which kind of model
is more appropriate for simulation purposes in a specific ecological condition, as several
modeling alternatives exist all with potentials and limitations that need to be known. Therefore,
when using hydrological models as a tool for understanding erosion and deposition processes at
catchment scale, the model user should be aware of the possibilities and examinations of the state
of the art of the model and also understand the basic considerations when choosing a model.
The Revised Universal Soil Loss Equation (RUSLE) by (Renard et al., 1991)is the most widely
used soil erosion models (Salehi et al., 1991). The RUSLE predicts long-term average-annual
soil erosion for a range of sites where the mineral soil has been exposed to raindrop impacts and
surface runoff. The RUSLE is computer-based which replaces the tables, monographs, and
calculations with a keyboard entry (Dube et al., 2014, Desmet and Govers, 1996). The Multi-
Criteria Decision Analysis (MCDA), an instrument for improving GIS, could help users to
improve their decision-making processes. Multi-criteria Decision Analysis often compares
different alternatives based on specific criteria to identify sensitive areas of erosion. To explore a
range of alternatives in terms of goal conflicts and multiple criteria, the MCDA technique is used
(Voogd, 1983). In order to achieve this, a ranking of alternatives and compromise alternatives
according to their attractiveness must be produced (Janssen and Rietveld, 1990). Numerous
researchers have been study using MCDA techniques in particular areas to conserve natural
resources management (Tecle A, 1990, Malczewski, 1996). In this outcome, GIS based MCDA
technique helps to carry out the delineation of the most erosion prone area in study watersheds.
Generally this study explains that RUSLE model was used to estimate the soil loss and GIS
based Multi Criteria Decision Analysis (MCDA) was used for identification of erosion hotspot
source areas from the watersheds.
3
1.2. Problem of Statement
Soil erosion was a critical problem in Ethiopia. It causes agricultural productivity that leads to
food insecurity and rural poverty. According to the Ethiopian highland reclamation study, in the
mid 1980‟s, 27 million hectare or almost 50% of the high land area was significantly eroded, 14
million hectare seriously eroded and over 2 million hectare were beyond reclamation (Yilma and
Awulachew, 2009).
Lack of effective watershed management system and poor land use practices and natural causes
played significant effect on fertility of soil in the crop field and landscape. In highlands of
Ethiopia, in which Dengora and Meno watersheds was found, are facing severe problems arising
from excessive erosion that results decreasing agricultural productivity and decreasing storage
capacity of reservoir and increasing environment degradation. Due to decreasing crop production
community living in both watersheds categorized food insecure area since 1999 by the regional
government (BoFE, 1999). In most watersheds, gully formation and sheet erosion with exposure
of rock and stones on previously cultivated steep upper slopes are the most visible evidences to
show erosion problems in this area (Birru, 2007). Excessive soil erosion from Dengora
watershed Atilkanay artificial reservoir located at the outlet of the Dengora watershed reduced its
storage capacity (Atikayina Earthen rock fill dam) closes its outlet.
Soil and water conservation measure had been implemented in the watershed for the last
decades. However, community as well as government blindly recommended different measures
without any evidence on targeting as a result they aggravated instead of reducing soil erosion in
both watersheds. To reduce soil erosion problems in the watershed, rate of soil loss could be
estimated and also soil erosion source areas should be identified for intervention. This helps to
apply different soil and water conservation measures in the watershed. Therefore, the purpose of
this study concerns on identifying soil erosion source areas of Dengora and Meno watersheds for
conservation priority.
4
1.3. Objective of the Study
1.3.1. Main Objective
The main objective of this study was to identify soil erosion hotspot areas using RUSLE model
and MCDA techniques in Dengora and Meno watersheds.
1.3.2. Specific Objective
The specific objectives of this study were:
To quantify the annual soil loss rate using GIS integrated with RUSLE model
To identify erosion hotspot areas using GIS based on Multi-Criteria Decision Analysis
Technique.
1.4. Research Question
To adders the objective of the study, the following research question were answered
1. What was the annual soil loss rate of the Dengora and Meno watersheds?
2. Which parts of the watershed area are more prone to soil erosion?
1.5. Significance of the study
In the high lands of Ethiopia nearly 1.3 billion tons of fertile soil are lost each year and soil
erosion and land degradation increases significantly due to the undulate and irregular topography
of the area(Hurni, 1989a, MoWIE, 1993). This amount is found to be equivalent to an average
soil loss of 130 t ha_1
yr_1
from cultivated lands (FAO, 1986). Dengora and Meno watersheds
were one parts of watershed in the highlands of Ethiopia; highly affected by soil erosion
problems. Therefore, estimating soil loss rate and identifying erosion sensitive parts of the
watershed is very important task to give priority of conservation practice in the watersheds.
Hence, soil and water conservation method should be based on the identified soil erosion source
areas. Identifying and studying soil erosion hot spot areas were useful as an input data for
integrated watershed management, for decision makers and other researchers.
5
In general, these specific studies were significant contribution in mitigating erosion problem on
time and in a cost effective way. The planning and interventions were in line with the identified
erosion source areas of a watershed. Thus, the present study will attempt to identify conservation
priority areas in Dengora and Meno Watersheds on the basis of erosion risk using Revised Universal
Soil loss Equation (RUSLE) model and GIS based MCDA approach.
1.6. Scope of the study
The scope of the study was limited to identify soil erosion hotspot areas in Dengora and Meno
watersheds by using RUSLE and GIS based MCDA techniques. Soil erosion factors were
identified for the analysis and RUSLE model were used for the estimation of potential soil loss in
the watershed. For soil erosion by water, rainfall is the major agent to take place in a catchment.
In this specific study, soil erosion factors were identified for the analysis of RUSLE model which
is used for the estimation of potential soil loss in the watershed and to identify or map the erosion
hotspot area by using GIS based MCDA techniques.
1.7. Paper Organization
This paper is organized in to seven topics to achieve the designed objectives. The first chapter
deals with the background status of soil erosion and modeling of soil erosion technique with
RUSLE model and satellite image. Research questions, objectives, scope and significance of the
paper were stated clearly. Chapter two narrates the overall historical back ground of soil erosion
and land degradation in Ethiopia, consequences of soil erosion, soil erosion modeling,
application of GIS on this study and related previous study on soil erosion will clearly stated.
Chapter 3 is the main part of this paper that briefly describe about the study area, methodologies
and data analysis on soil erosion modeling and erosion hotspot areas identification using GIS
based RUSLE model and MCDA techniques. In Chapter 4 the result and discussion obtained
from the designed methodologies. This were described in the form of figures and tables and
discussed in scientific way. The sensitivity analysis and validation process of result were
addressed with in the previous study. Chapter five summarized the overall research finding and
give scientific recommendation for future researches. Chapter six comprises the reference part
where the paper reviewed for identifying the gaps and methods. The final chapter contains the
appendix session.
6
2. LITERATURE REVIEW
2.1. Soil and land degradation in Ethiopia
In Ethiopia, deforestation, rapid rate of soil erosion and degradation of land are a serious
environmental problem resulting food in insecurity and reducing agricultural productivity. Soil
erosion is one part of land degradation that affects the physical and chemical properties of soil
and resulting in on-site nutrient loss and off-site sedimentation of water resources in Ethiopia
(Hurni, 1993b). Other degradation processes include intensified runoff from grasslands and
related gulling, as well as high soil erosion rates from heavily degraded lands. The practices of
the small scale farmers are the main cause of these processes. Due to this Soil erosion and land
degradation is a great concern which constitutes to global environmental and economic problems
(Hurni, 1993b).
Studies suggested that high rates of soil erosion in Ethiopia is mainly caused by extensive
deforestation due to the prevalence of high demand for fuel wood collection and grazing into
steep land areas (Amsalu et al., 2007). On the other hand, soil erosion by water is the dominant
degradation process and occurs particularly on cropland, with annual soil loss rates on average of
42 tons ha-1
for croplands, and up to 300 tons ha-1
in extreme cases(Hurni, 1993b, Hurni,
1993a).Hawando (1995) also estimated that the amount of annual soil movement (loss) by
erosion ranges from 1,248 to 23,400 Mtyr-1
from 78 million ha of pasture and rangelands and
cultivated fields in Ethiopia. Using conventional soil loss measuring method, the six SCRP sites
of Ethiopia found a soil loss ranging from 18 to 214.8 tons/ha/year (Berhe, 1996).This by far
exceeds the natural rate of regeneration. FAO (1986) estimated that 50 % of the highlands are
significantly eroded of which 25 % are seriously eroded and 4 % have reached a point of no
return. Based on (MoWIE, 1993, Hurni, 1989a) estimation, the trans-boundary Rivers that
originate from the highlands of Ethiopia carry about 1.3 billion ton/year of sediment to the
neighboring countries.
It implies the phenomenon is very intense in Ethiopia, even though all parts of the country are
not suffering uniformly. The extent and severity of the problems different in spatial variations in
altitude, ecology, settlement, topography and land use system (Shiferaw, 2015). As of natural
7
resource degradation is the major environmental problems resulting for decline of agricultural
productivity (Tesfa and Mekuriaw, 2014).
The average rate of soil erosion in the country wide was estimated at 12ton ha-1
yr-1
, giving a total
annual soil loss of 1,493Mt(Sinore et al., 2017). The severity is much higher in agriculture land,
in which 85% of the total population depends on it to get their survival (Authority, 2012).
2.2. Forms of soil erosion
The rate and magnitude of soil erosion by water is controlled by the following factors: Rainfall
and Runoff, Soil Erodibility, Slope Gradient and Length, Cropping and Vegetation, Tillage
Practices (Philor, 2011).Depending on the stage of progress in the erosion cycle and the position
in the landscape, there are various forms of soil erosion by water. Splash, sheet, rill and gullies
are the most important ones(Mitiku et al., 2006).
Rain splash erosion: occurs when water falling directly on to the ground during rainstorms or
intercepted by the canopy and make contact with the ground(Morgan, 1995, Morgan, 2005).
Sheet erosion: water that cannot infiltrate in the soil will be changed in to runoff or overland
flow. Sheet erosion is occurring when the runoff does not concentrate. Thus, it uniformly moves
the productive topsoil particles fortified by rain splash down slope(Mitiku et al., 2006).
Rill erosion: is a concentrated runoff resulted from intensive rainstorms, which produces more
observable features of erosion often on steep slopes and in depressions and forms channels up to
50 cm deep (Descheemaeker et al., 2006).
Gully erosion: Gully is a deep channel created as a result of severe soil erosion, usually caused
by running water. Gully erosion occurs when concentrated flows of water scouring along flow
routes cause channels deeper than 0.5 m(Nyamai et al., 2012). It is an advanced stage of rill
erosion whereby surface channels have been eroded to the point where they cannot be smoothed
over by normal tillage operations. Gullies are caused by land husbandry activities that result in
increased surface runoff. These include improper land use, such as slash and burn or shifting
cultivation, failure to terrace sloping land, reduced vegetation cover as a result of burning of
vegetation and bush fires and poorly managed croplands. Livestock management, particularly
overstocking of livestock leads to overgrazing further reducing soil cover which results in
8
excessive runoff (Grissinger and Murphy, 1989). Gully erosion is formed when runoff water
accumulates and often recurs in narrow channels and removes the soil from narrow area to
deeper than 50 cm.
On the other hand, gully erosion can be formed from rill erosion (Nyssen et al., 2006). Moreover,
gullies are efficient. Gully erosion represents an important sediment source and sources and
pathways of runoff from hill slopes to sediment sinks located in stream channels on a catchment
scale, contributing on average 50–80% of sediment production by water erosion (Verstraeten and
Poesen, 2002). However, gully erosion rates are difficult to assess, particularly at the catchment
scale. The major contribution of remote sensing to gully erosion assessment has been the visual
interpretation of aerial photography.
2.3. Factors affecting soil erosion
The major factors that influence the extent and rate of soil erosion from any area are: climate,
soil properties, topography of the area, vegetation cover and land use.
Climate: -Climatic factors which affect the magnitude and rate of soil erosion are; precipitation,
humidity, temperature, evapotranspiration, solar radiation and wind velocity(Blanco-Canqui and
Lal, 2008). The effect of precipitation on soil loss is partly through the detaching power of
raindrops striking the soil surface and partly through the contribution of runoff. The raindrops
which pound on the soil surface either infiltrate into the soil or leave the field as surface runoff.
Runoff occurs when the precipitation rate exceeds the infiltration capacity of the soil, and then it
collects and flows across the land surface (Toy et al., 2002). In general, the rainfall erosivity is
the function of its intensity and duration, and the raindrops‟ mass, diameter and velocity
(Morgan, 1995, R. P. C, 2005). As the rainfall intensity and the mass, diameter and velocity of
raindrops increases, the soil would be ready to be washed away from the ground through storm
runoff.
Soil Properties: -The susceptibility of soil is dependent on the soil‟s texture, content of organic
matter, surface roughness, moisture and depth to be eroded by erosion agents (Mitiku et al.,
2006). Soil texture refers to the relative proportion of clay, silt and sand. Fine particles have
cohesive property, as a result, they can resist detachment but easy to be transported, whereas,
9
large particles are resistant to transport because they need greater energy to be transported (R. P.
C, 2005). Silts and sands are the least detachment resistant particles.
Organic materials stabilize soil structure and coagulate soil colloids so; it is possible to decrease
soil erosion (Blanco-Canqui and Lal, 2008).
Roughness of the soil surface provides storage of rainwater, that helps the water to soaks into the
soil slowly and if the depth and porosity of the soil is high, runoff has decrease through the
increment of infiltration volume.
Topography: - The slope steepness and slope length of an area has greater impact on soil
erosion rate; as slope steepness and length increases, the velocity and volume of surface runoff
increases (R. P. C, 2005). Sloping watersheds are known by rill, gully, and stream channel
erosion and steeper surfaces of the earth are prone to mudflow erosion and landslides(Blanco-
Canqui and Lal, 2008).
Vegetation Cover: -Vegetation determines the soil erosion in so many different ways; leaves
and stems which are called the above ground components, absorb some of the energy of falling
raindrops, running water and wind, so there would be less contact with the soil, while the below-
ground components which contain the root system help the soil to get mechanical strength (R. P.
C, 2005). Vegetation decreases the volume of run-off by increasing transpiration and evaporation
and therefore reduces soil moisture and increases soil organic content, which also increases soil's
water absorptive capacity (Joint, 1986). The effectiveness of vegetative cover to protect soil
erosion depends on plant species, density, age, and root patterns (Blanco-Canqui and Lal,
2008).Dense and short growing vegetation is more effective to decrease soil erosion by
detachment and runoff than tall and sparsely growing plants. The soil particles detached by
raindrops under the forest canopies without litter layer can be between 1.2 and 3.1 times those in
open ground (R. P. C, 2005), but, forests can protect the land from mass movement or land slide.
Deforestation and overgrazing are the cause for 43% and 29% of water erosion, respectively
(Oldeman, 1992).
Land Management Practices: -Land conservation practices like contouring, strip-cropping,
terraces, crop rotations, reduced tillage and leaving crop residue on the land helps to reduce soil
erosion directly or indirectly. Crop residues, like straw, stubble and maize stalks can reduce soil
losses by one halve or more depending on other factors (FAO, 1981)Terraces reduce slope length
and velocity of running water. Agroforesty or intercropping is also another method for the
10
reduction of soil erosion; the system evolve into more complex production systems that can
provide different benefits than annual crop production system (Winterbottom et al., 2013).
Integrated woody perennial plants protect the soil from erosion after the crops being harvested.
In general, improving land management practices can reduce soil loss by erosion agents through
increasing soil organic matter and moisture content and through different tillage operations
especially on sloppy areas. About 24% of soil erosion by water is caused by agricultural
mismanagement (Oldeman, 1992). As population increases, marginal lands are used, fallow
periods will be shorten or even no fallowing in some areas.
2.4. Consequences of soil erosion
The consequences of soil erosion can be seen on both where the soil is worn and deposited;
earth‟s surface can either being degraded or aggraded. The problem is threatening ecosystems
and human wellbeing throughout the world (Toy et al., 2002), because it results in significant
reduction in economic, social and ecological benefits of land for crop and other environmental
services. Soil erosion affects about one billion people globally; around 50% of them found in
Africa, but, more attention is given to other agricultural topics than to soil erosion and its
consequences (Blanco-Canqui and Lal, 2008).
2.4.1. On-site Effects
Some of on-site effects of soil erosion are loss of soil, formation of rills and gullies, reduction of
soil moisture and organic matter, and decrease surface soil depth. The soil lost through water
erosion particularly by sheet erosion is usually the most productive top soil containing plant
nutrients and humus; it can be lost forever if it is washed in to the sea. Soil formation is a very
slow process; it may takes 100 up to 400 years to form 1cm soil depth (Mirsal, 2008). Cropland
soils are often left bare after harvesting, as a result, the soil will be more susceptible to erosion
(Blanco-Canqui and Lal, 2008). The reduction of soil productivity over extended period is the
main onsite effect of soil erosion. In Ethiopia, the active soil erosion is turning many of the once
fertile and surplus production areas in to badlands (Emama et al., 2015). Highlands of the
country are considered as the most seriously degraded parts of the world and in general, it is
estimated that the country looses 1.9 to 7.8 billion tons a year and this cost the country close to 1
Billion ETB (Emama et al., 2015). Generally, soil erosion results in a decline of soil quality
11
leading to a decrease in crop and other agricultural productivities. particularly high in the major
crop production areas under intensive tillage and mono-cropping(Blanco-Canqui and Lal, 2008).
2.4.2. Off-site Effects
Erosion not only damages the immediate agricultural area where it occurs but also negatively
affects the surrounding environment. Sedimentation and water pollution are the main off-site
effects of soil erosion by water. For the conservation, development and utilization of our soil and
water resources, sedimentation should be the main concern (Julien, 2010). Sediment is the
product of erosion and it decreases the storage capacity and life expectancy of reservoirs,
increases flood damage and water treatment cost (Toy et al., 2002). The sediment delivered at the
outlet of a watershed /watershed sediment yield/ should be estimated before the designing of
reservoirs to analyze sedimentation and water quality problems. In Ethiopia, most of the
reservoirs that are built for different purpose are filled with sediment with in less than 50% of
their projected service lives (Braimoh and Vlek, 2008). On the other hand, the running water can
wash away fertilizers, pesticides and other chemicals that are supplied by farmers on the land. As
a result, streams‟, rivers‟, reservoirs‟ and other water bodies‟ pollution occurs and living
organisms in the system would be at risk.
2.5. Soil Erosion Modeling
Soil erosion modeling is the process of describing soil particle and processes of detachment,
transport and deposition mathematically on land surfaces (Judson, 1965 and Merritt et al., 2003).
Soil erosion modeling is used to: 1) predict and assess soil loss for conservation planning, project
planning, soil erosion inventories, and regulation. 2) Predict where and when erosion is occurring
and hence helping the conservation planner target to reduce erosion. 3) Understanding erosion
processes and their interaction for setting research priorities (Lal, 1994).
Empirical models are simplified representation of a system or phenomenon which is based on
experience or experimentation. RUSLE is one of such type of models. The computational and
data requirements for such models are usually less than for conceptual and physically based
models (Zhang et al., 1996). Empirical models are easy to implementation, reliance on easily
accessible data and produce relatively accurate results than others.
12
RUSLE has the ability to estimate the long term average annual rate of soil erosion on a field
caused by slope, rainfall pattern, soil type, topography, crop system and management practices
(Renard et al., 1997).
The model predicts erosion potential on a cell-by-cell basis in GIS environment. It is successful
in attempting to identify the spatial pattern of soil loss present within a large watershed area (Shi
et al., 2004). The introduction of GIS in this model is to isolate and query these locations to
recognize the role of each variable in contributing to the observed erosion potential value
(Saavedra, 2005). RUSLE estimates the average annual soil loss using the in the below equation
in section three.
2.6. Application of GIS and Multi criteria Decision Analysis (MCDA) techniques
GIS are powerful tools when applied to earth sciences and land use study. GIS procedures
involve managing, editing, and analyzing huge volumes of spatial data and their related thematic
attributes. However, available GIS software lack in relation to spatial analysis and cartographic
modeling, because they just offer deterministic analysis and overlay of maps (Openshaw, 1991);
(Fischer and Nijkamp, 1993). To overcome these deficiencies, GIS packages such as IDRISI and
SPANS currently include MCDA modules. This study has used MCDA techniques to improve
managing of thematic data. MCDA is a set of procedures designed to facilitate decision making.
The basic purpose is "to investigate a number of choice possibilities in the light of multiple
criteria and conflicting objectives" (Lillie et al., 1983).
Integration of GIS and MCDA could provide a powerful tool for studying allocation of activities
and spatial modeling. GIS provides an appropriate framework for the application of multicriteria
decision analysis methods, which are not capable of managing spatial data, the multi criteria
evaluation procedures add to GIS for the means of performing compromises on conflicting
objectives, while taking into account multiple criteria and the knowledge of the decision maker
(Carver, 1991). In the last years, several procedures of MCDA are included in GIS for urban and
regional planning for allocation of agricultural land use (Janssen and Rietveld, 1990), residential
quality assessment (Can, 1993) and land suitability (Joerin and Musy, 2000) reviewed several
MCDA procedures and the possibility of integrating in GIS.
13
Therefore, the MCDA is an effective tool for multiple criteria decision making issues.
Integration of the MCDA and GIS (GIS-MCDA) can help land use and environment protection
agents and managers to improve decision making processes. GIS enables the computation of
assessment factors, while MCDA aggregates them in to sensitivity index.
2.7. Related previous studies on soil erosion
There is no any study about identification soil erosion hotspot areas directly relates on Dengora
and Meno watersheds with similar methodology. But, related studies on other watershed as
shown below:
Assefa et al. (2015) studied identification of erosion hotspot area using GIS and MCE Technique
for Koga watershed in the upper Blue Nile basin. The result of the study indicates that 2% (440
ha) to be highly sensitive, 43% (9,460 ha) to be moderately sensitive, 16% (3,520 ha) to be
marginally sensitive and 32% (7,040 ha) currently not sensitive. The remaining 7% of the
watershed area (22,000 ha) is constraint to erosion. The lowland area near the dam is found to be
found most sensitive for erosion and sedimentation. The overall research result indicated that
most erosion hotspots areas were found in the lowland (more than 75% of erosion hotspot area of
the catchment. Finally the study recommended that it is extremely important to consider the
saturated areas during design of watershed management strategies. (Birru, 2007, Adugna et al.,
2015) studied the soil erosion assessment and control in north east Wollega using RUSLE. The
result of the study indicates that the annual rate of soil loss is in the range of 4.5 Mg ha-1
yr-1
in
forestland and 65.9 Mg ha-1
yr-1
in cropland. The study recommended that it needs to address
issues of farmers‟ education, secure land rights and access to credit in order to control soil loss
from cultivated land.
This specific study tried to identify soil erosion hotspot areas using GIS based multi criteria
decision analysis techniques and estimate soil erosion rate using RUSLE model in Degora and
Meno watersheds. The uniqueness of this study is identifying soil erosion rate and hotspot areas
by using RUSLE and GIS based MCDA technique with in considering all the factors that case
soil erosion.
14
3. MATERIALS AND METHODS
3.1. Description of the study area
The study area Dengora and Meno watersheds were located in the Tekeze basin of Amhara
National Regional State in East and West Belesa Woredas respectively (figure 3-1). The
Geographical coordinate of the Dengora watershed is ranges from 38o2‟35” to 38
o3‟15” E and
12o23‟0” to 12
o23‟30” N and Meno watershed is ranges from 37
o46‟2”to 37
o46‟45” E and
12o26‟50”to 12
o27‟25” N. Both watersheds area classified under semi aired agro climatic zone.
Figure 3-1 Location of Study area map
15
3.2. Topography
Both watersheds were characterized by highly rugged and undulating topography on the upper
part. Dengora watershed large portion of the watershed falls in to gently flat to undulating
terrain 41.29% of the land and 29.19% hilly terrain and Meno watershed 47.45% gently flat to
undulating terrain and 32.22% hilly terrain slope classes according to the FAO slope class
category (Table 3-1). The elevation difference of Dengora and Meno watersheds varies from
1898 to 2172 m and 1888 to 2091 m above mean sea level respectively (Figure 3-2).
Table 3-1 Slope class category of watershed
S/N Slope Class (%) Class Name
1 0-2 Flat to almost flat terrain
2 2-10 Gently flat to undulating terrain
3 10-15 Rolling terrain
4 15-30 Hilly terrain
5 >30 Steep dissented to mountainous
Source: (FAO, 2001)
Figure 3-2 Slope map and Digital elevation model of the watersheds
16
3.3. Climate
The climate data available from National Meteorological Agency (NMA) two rainfall stations,
Arbaya (2004 to 2013year) and Guala stations (2008 to 2015year) for Dengora and Meno
watersheds respectively used. The climate of watersheds can be characterized as semi-arid agro
climate with a mean annual rainfall of Dengora and Meno watersheds 949.5 and 841.85 mm/year
and average temperature of above 31.5OC and 30OC based on the data available in Dengora and
Meno watersheds respectively. Most rainfall occurs between July and August season (Fig 3-3).
Figure 3-3 Monthly rainfall of Dengora (top) and Meno(bottom)watersheds
0
50
100
150
200
250
300
350
400
Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec
2009
2010
2011
2012
2013
2014
2015
2008
0
50
100
150
200
250
300
350
Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec
2004
2005
2006
2007
2008
2009
2010
2011
2012
2013
17
3.4. Soil Type
According to FAO soil group, which obtained from Ministry of Water Irrigation and
Electricity(MoWIE), major soil types were distinguished in both of the study area. From these
soils Eutric leptosols were dominant in Dengora and Meno watersheds as presented in the Table
3-2 below.
Table 3-2 Major Soil type
Watersheds
Soil type
Area coverage
Area (ha) Percentage (%)
Dengora watershed Eutric leptosols 22.6 46.6
Leptic Luvisols 12.9 26.6
Heplic Luvisols 6.54 13.15
Exposed Rock 1.84 3.8
Total 48.5 100
Meno watershed
Eutric leptosols 73.6 76.7
HyperskeleticLeptosol 17.8 18.57
Lithic Leptosol 0.8 0.8
Total 96 100
18
Figure 3-4 Soil map of Dengora and Meno watersheds
3.5. Land use land cover map
Land use land cover in an area significantly influences the pattern and rate of erosion. The land
use land cover types identified from the land sat 8 satellite imagery using Maximum Likelihood
supervised classification was made in Arc GIS 10.1. Dengora and Meno watersheds were
categorized under five classes (Table 3-3). The dominant land use land cover was cultivated land
for both of the watershed.
19
Table 3-3 Land use land covers type
Watersheds
Land use/cover
Area coverage
Area(ha) Area (%)
Dengora watershed Forest Area 1.65 3.4
Cultivated land 10.53 21.7
Cultivated with terrace 19 39.2
Bush land 4.5 9.27
Shrub land 12.8 26.4
Total 48.5 100
Meno watershed
Forest Area 0.5 0.5
Cultivated land 46 48
Grazing land 0.44 0.46
Bush land 15.4 16
Shrub land 33.6 35
Total 96 100
20
Figure 3-5 Land use land cover map for Dengora (top) and Meno(bottom) watersheds
3.6. Satellite Data and DEM
Satellite images having 169/052 and169/051 path/row were downloaded from USGS archive for
Dengora and Meno watersheds respectively. The data were used to classify and map the land
use/ land cover map both of the watersheds area for the years of 2019 which helped to notice the
rate of soil loss within the study area.
21
The Digital Elevation Model (DEM) of study area, having 12.5 meter spatial resolution was
downloaded from the Websitehttp://gdex.cr.usgs.gov/gdex/ . This data had multipurpose in this
thesis; it has been used to delineate and map the watershed including the streams and slope map
to get values of topographic factor, and potential gully location in the study area.
3.7. Materials and Data Used
In this study, different data types and software were used. Data required for the GIS based
MCDA technique are spatial in nature. Land use and land cover data, DEM, Soil type, gully
location and Rainfall were collected for both of the watersheds. Those data types, software used
includes, ArcGIS 10.1 for GIS based DEM processing image classification and overlay analysis.
MS office packages for chart making, tabulation, and word processing. GPS and Digital Camera
were used for ground control field data collection. In the Table 3-4 software, data types and
sources for each category were presented below.
Table 3-4 Type, purpose and sources of data/material
S.N Data type Purposes Sources
1 DEM To delineate watershed and
generate slope
Website: http://gdex.cr.usgs.gov/gdex/
2 Land use To generate land use land cover
map(LULC)
Website: http://landsat.usgs.gov/landsat8.php
3 Soil type To produce soil map Ministry of Water and Energy (MoWE),
Addis Ababa
4 Rainfall To generate Rainfall map National Meteorological Agency (NMA),
Arbay metrological branch
5 GPS For ground control points (LULC
and gully location)
-
6 ArcGIS 10.1 For analyzing, Displaying,
overlaying the criteria and spatial
data
-
22
3.8. Methodology
To attain the objective of this study different methods were used to collect data, both primary
and secondary data from the field, institutions and organizations and download different websites
and also group discussion and participatory transact walk. After the data collection, data analysis
has undergone the second step that is to execute satellite image processing, land use land cover
classification and estimate the annual soil loss using RUSLE. Thirdly the MCDA was used to
identify the erosion hotspot areas including gully erosion. At the end the estimated soil erosion
loss and hotspot areas in Dengora and Meno watersheds by using RUSLE and MCDA techniques
have been overlaid to get the erosion hotspot areas.
3.8.1. Data Collection
This study results were achieved with the utilization of both primary and secondary data
collected from the study area and respective institutions and organizations.
3.8.1.1. Primary data collection
The primary data collection involves sample survey of geographic location of gully and ground
control point for land use land cover classification by using GPS and group discussion and
participatory transect walk for validation of gully location .Satellite data, cloud free Landsat8
image (LULC) of March 2019 and location of the study area was downloaded from USGS
website for land use land cover classification. From Shuttle Radar Topographic Mission
(SRTM), 12.5 x 12.5 m resolution Digital Elevation Model (DEM) was used to watershed
delineation, slope and flow accumulation generation of the study area. In addition to this, random
sample of gully area polygons were generated from Google earth for verification of gully
potential location map.
3.8.1.2. Secondary data Collection
Secondary data, such as; Rainfall data was collected from National Meteorological Agency
(NMA). Since, there was meteorological station found within the catchment nearby stations
around 3-5 Km from the study area namely; Arbay and Guala meteorological station was used to
determine the mean annual rainfall which is input factor for RUSLE. Soil data was collected
from Ministry of Water, Irrigation and Electricity (MoWIE).
23
3.8.1.3. Group Discussion and participatory Transect Walk
Focus group discussion is a qualitative research method and data collection technique in which a
selected group of people discusses on a given topic or issue in depth, facilitated by professional,
external moderator (Twinn, 1998). For the study, a total of four focus group discussion was made
with a group member of women and men independently in both watersheds. A total of 8 women
and 10 men per the focus group discussion were attended. Additionally, in-depth interview with
Kebele level development agents were made in each watershed. A total of four development
agents were interviewed and two development agents per watershed. After the discussion on
community and the development agents understanding soil erosion risk and resources status
(vegetation density, soil and water) and identified factors contributed to soil erosion in the
watersheds. After focus group discussion, members also participated in transect walk across the
watersheds, east to west and south to north direction as in figure 3-6 and 3-7. During the transect
walk land-use/land cover types, slope gradient, highly erodible area/gully location, slope length,
soil color, and drainage patterns were documented. For validation of hotspot areas, the transect
walk information was applied.
Poor land cover
active Gully
Figure 3-6 Meno watershed transect walk
24
Improper road
Poor land cover
Group Discussion
Gully
Collapsed stone bund
Participatory transect walk and observation
Figure 3-7 Dengora watershed transect walk
3.9. Data Analysis
After collecting the required data, spatial analysis was made to prepare MCDA criteria map and
RUSLE factors for the source area investigation and soil loss estimation respectively. Landsat 8
image together with intensive field point data collection was used to perform supervised land use
classification in ArcGIS environment. The output map is validated and used to produce land use
criteria map based on land use suitability classes. Soil map from Ministry of Water, Irrigation
and Energy (MoWIE) was directly used to produce soil criteria map based on soil type suitability
classes. DEM with 12.5 m resolution were used to produce three criteria maps: Topographic
wetness index, stream power index and potential location of gullies. Topographic wetness index
and stream power index of the catchment was predicted based on flow accumulation and slope of
the particular pixel.
Potential locations of gullies were predicted based on threshold concept of two criteria
topographic wetness index and stream power index. The parameters were identified and
classified in to sub classes to get the relative weights using pair wise comparison method and the
soil erosion factors were weighted overlay to produce final soil erosion source areas in terms
spatial representation in the watershed.
25
3.9.1. Modelling of soil erosion
Modeling of soil erosion and estimation of soil loss was done using revised Universal Soil Loss
Equation (RUSLE). This method is the newly emerged model from Universal Soil Loss Equation
(USLE) (Renard et al., 1991). RUSLE has the same formula with USLE but has several
improvements in determining factors. Hurni (1985) has modified the USLE to fit the Ethiopian
conditions. RUSLE model parameters such as Cover (C) factor, Rainfall Erosivity (R), Soil
Erodibility Factor (K), slope steepness length factor (LS) and Conservation practice (P) was
estimated and used for the estimation of mean annual soil erosion loss rate in Dengora and Meno
watersheds. All the five RUSLE parameters were overlaid by using the above equation in
ArcGIS with spatial analysis package of raster calculator.
where; A is the computed spatial average soil loss rate (t ha-1
yr-1
), R is the rainfall runoff
erosivity factor (MJ.mm.ha-1
yr-1
), K is the soil erodibility factor (Mgh MJ-1
mm-1
), L is the slope
length factor (dimensionless), S is the slope steepness factor (dimensionless), C is the cover
management factor (dimensionless), and P is the conservation support practice factor
3.9.2. RUSLE Factors Generation
3.9.2.1. Rainfall Erosivity (R) factor
The rainfall erosivity factor (R-factor) is based on kinetic energy considerations of falling rain
and represents a measure of the erosive force and intensity of rain in an ordinary year. The
potential ability of the rain to cause erosion is a function of physical characteristics of the
rainfall. To determine the erosivity factor, the maximum 30 minute intensity is required. But, for
the study area it is difficult to find the 30 minutes intensity. R can be calculated by (Hurni, 1985)
which is derived from a spatial regression analysis by (Helldén, 1987)for Ethiopian conditions.
R = -8.12 + 0.8562P-------------------------------------------------- (3.2)
Where, P is mean annual rainfall in mm
26
3.9.2.2. Soil Erodibility (K) factor
Soil erodibility is the resistance of soil to detachment from parent material and transport from its
original position, reflects the effect of soil properties: texture, infiltration, organic matter and
chemical content. Soils with a high percent content of silt and very fine sand particles, a low
organic matter content, poor structure and very low permeability will be most erodible, based on
soil characteristics (Zhang et al., 1996). Soils having high silt content are most erodible of all
soils. They are easily detached, tend to crust and produce high rates of runoff; erodibility values
for these types of soils are tend to be greater. Organic matter content reduces erodibility,
decreases susceptibility of the soil to detachment, and increases infiltration rates, which in turn
reduces runoff and erosion. The erodibility factor for both of the watersheds was calculated by
using (FAO, 1989). The soil erodibility factor, K measures the resistance of the soil to
detachment and transportation by raindrop impact and surface runoff. Soil erodibility is a
function of the inherent soil properties, including texture, structure, organic matter content and
permeability (Wischmeier and Mannering, 1969). In this research, FAO standard classification of
soil type was obtained from MoWIE . Therefore based on (FAO, 1989) the K-factor of the study
area was developed using this suggestion(Table 3-5).
Table 3-5 Soil types with K values
S/N Dengora watershed K-factor Meno watershed K-factor
1 Eutric leptosols 0.15 Hyperskeletic Leptosols 0.2
2 Leptic Luvisols 0.2 Lithic Leptosols 0.1
3 Heplic Luvisols 0.2 Eutric leptosols 0.15
4 Exposed Rock constraint
3.9.2.3. Topographic (LS) factor
The effect of soil erosion on topography on soil erosion is accounted by slope length (L) and
slope steepness (S). Digital Elevation Model (DEM) was calculated with multiple flow
algorithms. Multiple flow algorithms can divide flow between several output cells (Desmet and
Govers, 1996). The LS factor has been derived from slope and flow accumulation. Slope was
generated from 12.5m x 12.5m resolution DEM using ArcGIS. Flow accumulation was an input
27
for generating LS. To generate flow accumulation which is the unit contributing area first, any
spurious single cell sinks within the DEM is filled to produce a depression less DEM. In this
process, individual sink elevations are flattened. Then by using filled DEM the flow directions of
each DEM cell is calculated. From flow directions flow accumulation will be determined by
ArcGIS. Then the LS factor is estimated by using in the equation3.3using raster calculator
proposed by Wischmeier and Smith (1978).
(
) (
) 3.3)
Where LS is slope steepness length factor, the cell value is the resolution of DEM which is 12.5
m resolution and S is slope in percent generated from DEM.
3.9.2.4. Land Cover Factor (C)
Cover factor represents the ratio of soil loss under a given cover to that of bare soil. Surface
cover affects erosion by reducing transport capacity of runoff water and by decreasing the
surface area susceptible to raindrop impact (McCool, 1995). Increasing surface roughness
decreases transport capacity and detachment of runoff by reducing flow velocity. The typical
values of C factor range from 0 to 1. The value of 0 indicates wetlands and water bodies and 1 is
for bare land and urban area, as obtained from different literatures. Previous studies of the
RUSLE modeling, in our country, was limited to account for additional soil losses, than sheet
and rill forms of erosion that might occur from drainage channels. However channel erosion can
be accounted as the work of Rojas-González (2008). In this watershed, a new data field of land
cover type was available enabling to account the cover factors in soil loss estimation from the
drainage channels.
28
Table 3-6 land cover factor of study area
S/N Land use land cover C-factor value Source
1 Forest 0.01 Hurni,1985
2 Cultivated land 0.2 FAO,1984
3 Cultivated with terrace 0.15 Hurni,1985
4 Bush land 0.1 Hurni,1985
5 Close shrub 0.05 Hurni,1985
5 Grazing land 0.01 Hurni,1985
6 Road 0 Hurni,1985
3.9.2.5. Land Management Practice Factor (P)
The management practice factor is the ratio of soil loss with specific practice to the
corresponding loss of up and down slope tillage and describes the effectiveness of erosion
control practices. P factor is reflecting the positive impacts of management through the control of
runoff. On special emphasis how the management changes the direction and speed of runoff, but
also reflecting to some degree management practices that control the amount of runoff. The P
factor is commonly calculated by the method developed by Wischmeier and Smith (1978) as
indicated in Table 3-7.
Table 3-7 Land management factors (Wischmeier and Smith, 1978)
Land use type Slope (%) P-Factor
Cultivated Land
0-5
5-10
0.1
0.12
10-20 0.14
20-30 0.19
30-50 0.25
50-100 0.33
Other land use All 1
29
Wischmeier and Smith (1978) derived six unique P values for agricultural land units and single P
value for other land uses. This valuation doesn‟t account the P values of cultivated lands on grid
scale but only limited values in to six classes. Moreover, this valuation totally ignored cultivated
lands located on slopes above 100%. In Dengora watershed, as well in Ethiopian highlands,
many sloppy areas are cultivated without any slope limitation. In this farming practice, many
land units present to be cultivated to more than 100% slopes. To account these problems, a
regression analysis between P factor and slope was developed, equation 3.4 below, from the
method available by Wischmeier and Smith (1978). This development has a correlation
coefficient of 99% and solves the problem in estimating management factors of cultivated lands
on grid and watershed scale, especially for slopes cultivation above 100%.
SP 003.0099.0 ------------------------------------------------------------ (3.4)
Where P is management factor and S is the slope for cultivated lands (%). The conservation
practice factor (P) indicates the effect of conservation practices on soil erosion; where the land
has adequate conservation interventions it reduces soil erosion problems. Specific cultivation
practices affect erosion by modifying the flow pattern and direction of runoff and by reducing
the amount of runoff (Renard, 1983). Values for this factor were assigned with considerations of
the local management practices and based on values suggested by Hurni (1985). For this specific
study, P factor were generated from land use land cover map corresponding to slope by using
raster calculator in Arc GIS.
30
Figure 3-8 General flow charts for RUSLE Generation
3.9.3. Multi- Criteria Decision Analysis (MCDA)
Multi-criteria analysis often compares various alternatives with the help of certain criteria. These
criteria are often a translation of the study objectives. MCDA helps for watershed prioritization for
the process of identification of soil erosion risk areas or pockets for taking up soil conservation
practice on the priority basis.
Input Data
Rain fall
data
Landsat8
image
Soil map
DEM
Ground
truth data
Image pre-
processing
Image
classification
Land use/land
cover Map
Slope
Flow-Ac
LS-Factor
K -Factor C-Factor
R -Factor
P-Factor
Intersect
A=R*C*P*K*LS
Soil Loss
map
31
The outcomes are not in the form of a valuation but more often in the form of selection,
classification or ranking of alternatives. It often compares various alternatives with the help of
certain criteria to identify erosion hotspot areas.
In this study, used MCDA technique within GIS environment to identify the actual source of
erosion and map sensitive areas based on spatial dataset analysis. Weight of decision factors are
assigned based on their relative effect to erosion process. To perform MCDA by using GIS for
hotspot area identification in Dengora and Meno watersheds land use/cover, Soil type,
topographic wetness index ,Stream power index (SPI) and potential gully location were as
factors and GIS aided analysis has been done to obtain a map for each criterion. For multi criteria
evaluation of factor generation, three main types of data inputs were used. Those includes land
cover, DEM and soil type which used to generate soil erosion factor maps such as land use land
cover map, soil map, slope and potential gully location map of Dengora and Meno watersheds.
Finally those factors were reclassified and sensitivity analysis undergo as presented in Table 3-8.
Table 3-8 Factor sensitivity classes
Sensitivity classes Notation Explanation
Highly sensitive S1 Factors significantly accelerate erosion
Moderately sensitive S2 Factors clearly sensitive but has opportunity to reduce
Marginally sensitive S3 Factors significantly reduce erosion
Currently not sensitive S4 Factors that cannot support erosion
Source: (FAO, 1981)
3.9.3.1. Land use land cover map
Land use land cover is one of the most important factors that affect surface runoff and erosion in
a watershed. It enables to assess the resistance of terrain unit to erosion because of surface
protection. High erosion and quick response to rainfall are resulted from poor surface cover.
Land use land cover types identified from the Landsat 8 satellite imagery using GIS tool. In
remote sensing, there are various image classification methods such as supervised, unsupervised
and hybrid. Supervised classification can be used to cluster pixels in data set in to classes
corresponding to user defined training classes.
32
This classification type requires selecting training areas for the bases of classification. Various
comparison methods are then used to determine if specific pixel qualifies a class member.
Maximum Likelihood is a classification method in ArcGIS environment. Maximum likelihood
classification assumes that the statistics for each class in each band are normally distributed and
calculates the probability that a given pixel belongs to a specific class. Unless a probability
threshold is selected, all pixels are classified. Each pixel is assigned to the class that has the
highest probability.
For this specific study, the land use land cover classifications were classified using a supervised
classification algorithm. The supervised classification involved the selection of a number of
known sites for each class throughout each image and 30 ground control point were taken from
each land use land cover type. Once these sites were identified Maximum Likelihood supervised
classification was made in Arc GIS 10.1. By using supervised image classification of Dengora
and Meno watersheds were categorized under seven classes (Table 3.3 above). The dominant
land use land covers for both of the watersheds were cultivated land.
Accuracy assessment
Accuracy assessment is necessary for validation of image classification process by evaluating
how effectively pixels were correctly grouped. Error matrix is the basic for accuracy assessment.
The matrix give a cross tabulation of the class label predicted against the ground truth GPS data.
The error matrixes give very important information on image classification to both map user and
producer‟s community as shown appendix B.
Kappa is used to measure the accuracy between the remote sensing derived classification map
and the reference data indicated by the major diagonals and the chance agreement, which is
indicated by the row and column totals (Janssen and Rietveld, 1990). The Kappa coefficient was
computed by equation 3.6.
(3.6)
The overall accuracy is often the only accuracy statistic reported with predictive landscape
models (Congalton, 1991), but the error matrix provides a means to calculate numerous
additional metrics describing model performance.
33
The overall accuracy (Table 3.9) of the model is simply total number of correct classifications
divided by the total number of sample points.
Over all accuracy = (Number of pixels correctly classified) / (total number of pixel)
Table 3-9 land use land cover accuracy assessment of study area
Land Use/Dengora
watershed
Producer
Accuracy(%)
Omission
Error(%)
User
Accuracy (%)
Commission
Error (%)
Overall
Accuracy(%)
Kappa
Coefficient (%)
Forest Area 1 0 0.67 0.33 0.83 0.82
Cultivated land 0.72 0.28 0.83 0.17
Cultivated terrace 0.72 0.29 0.83 0.17
Bush land 0.86
1 0
Shrub land 0.83 0.27 0.83 0.17
Land Use/Meno
watershed
Producer
Accuracy (%)
Omission
Error (%)
User
Accuracy (%)
Commission
Error (%)
Overall
Accuracy (%)
Kappa
Coefficient (%)
Forest Area 0.8 0.2 0.67 0.33 0.9 0.86
Cultivated land 0.83 0.17 0.83 0.17
Grazing land 0.67 0.33 0.67 0.33
Bush land 0.72 0.28 0.83 0.17
Shrub land 0.67 0.33 0.67 0.33
Road 1 0 1 0
3.9.3.2. Soil type
Soil is one of the major factors for soil erosion. Resistance of soil to erosion depends on soil
properties such as soil texture, structure, soil moisture, roughness, organic matter content and
chemical and biological characteristics(Vrieling et al., 2007). Generally, soils having quicker
infiltration rates, high levels of organic matter and improved soil structure have a greater
resistance to erosion (Saavedra, 2005).
34
For this specific study, the soil data from Ministry of Water Irrigation and Electricity (MoWIE)
was used to produce soil criteria map by extracting with the boundary of Dengora and Meno
watersheds. Finally soil sensitivity map was developed based on soil erodibility.
3.9.3.3. Topographic wetness index map
Another important element considered for identification of erosion hotspot area was TWI. The
effect of topography on soil erosion is a multifarious, because the local slope gradient influences
flow velocity and rate of soil erosion. Erosion would normally be expected to raise with slope
steepness and slope length increments as a result of respective increases in velocity and volume
of surface runoff (R. P. C, 2005). At gentle and steep slopes the action of rain is enough to soil
erosion (Fauck, 1956). Topographic Wetness Index was used to define the effect of topography
based on saturated excess runoff mechanism. It characterizes spatial distribution of surface
saturation and surface runoff that were very important parameter for soil erosion analysis.
Topographic wetness index and soil moisture increases as contributing area increases and slope
gradient decreases, this implies that TWI has high correlation with saturation.The wetness of the
catchment or topographic wetness index was predicted based on flow accumulation and slope of
the particular pixel of a watershed.
Topographic wetness index (
)
Where; is local upslope contributing area from flow accumulation raster and is local slope
angle (degree).
35
Figure 3-9 Topographic power index map
3.9.3.4. Stream power Index (SPI) map
To locate potential gully formation areas, stream power index (SPI) has been used. SPI is very
useful for determining potential critical source area locations (Minnesota Leg/ Ref, 2014). SPI is
calculated as the product of the natural log of both slope and flow accumulation. High SPI values
areas on the landscape where high slopes and flow accumulations exist and thus areas where
flows can concentrate with erosive potential.
Stream power index
Where; is local upslope contributing area from flow accumulation raster and is local slope
angle (degree).
36
Figure 3-10 Stream power index (SPI) map
3.9.3.5. Potential Gully location map
For the prediction of potential gully location, the wetness of the catchment or topographic
wetness index and stream power index have been predicted based on flow accumulation and
slope of the particular pixel on the Dengora and Meno watersheds boundary to drive gully
locations (equation 3.7and 3.8) respectively. The potential locations of gullies was predict where
the two thresholds have satisfied that is Stream Power Index >16.8 and Topographic Wetness
Index > 6.8 (Lulseged and Vlek, 2005). TWI and SPI map were overlaid to generate combined
map of gully potential locations.
To compare with the actual gully site sample of gullies digitized from Google earth and its
location were collected by GPS (Appendix F) to validate the mapping by SPI and TWI. To
compare potential and actual gully erosion areas, gully areas identified by SPI were overlaid on
37
map from Google earth. High SPI values are the characteristics of hilly and upper parts of the
area and it shows areas of high erosion. Thus, reclassification was done to indicate low to high
gully potential.
3.9.4. Pair wise comparison
After generating the factor/criteria maps of soil erosion, transforming the factors into a standard
scale of measurement was obligatory. This is because multi criteria decision analysis (MCDA)
technique requires the evaluation criteria to be standardized to corresponding units since each
criterion map contains raw values. Therefore all criteria maps should be transformed into a
standard scale. For data standardization there are a number of methods in GIS environment. The
score range method is the most used procedure since it is a special case of the single value
function method which integrates the decision makers‟ preferences in mathematical function
(MalczewskiandRinner, 2015).After criterion standardization in the spatial multi criteria
evaluation techniques weight was assigned for each factor which indicates the importance of
each factor with respect to the other factor under consideration.
3.9.5. The Analytic Hierarchy Process (AHP) and fundamental scale
The Analytic Hierarchy Process (AHP) was a multi-criteria decision-making approach which
constructs a matrix of pair-wise comparisons (ratios) between the factors responsible for erosion.
If these erosion hazard parameters are scaled as 1 to 9, 1 indicates that the two factors equally
important and 9 indicated that the one factor is more important than other. Reciprocal of 1 to 9
(1/1 and 1/9) show that one is less important than other. The Table 3-10 explained Saaty‟sRating
Scale and the allocation of the weights. The identical AHP depends on the relative
importance of factors and participatory group of decision makers. To fill the comparison
matrix a comparison of each erosion hazards parameters (EHP) with other parameters are made
and in this way the total number of comparison comes out to be comparison matrix. The
diagonals elements of the matrix in that way if the judgment value is left side of 1,
then for filling the upper matrix actual judgment value has been used. If the judgment
value is right side of 1 than reciprocal have been used. The lower triangular matrix is
filled by taking reciprocal of upper triangular matrix. From the comparison matrix
priority vector is computed which is the normalized eigen vector of the matrix that can
38
be used to assign the weight for different factors. In the present study nine different
parameter factors may be termed as erosion hazards parameters (EHP) have been selected for
construction of AHP matrix.
Table 3-10 Saaty‟s 1977 Fundamental weighting scale of pair wise comparison
Description of
preference
Rating
scale
Reciprocal
values
Explanations of scales
Equally 1 1 two activities contribute equally to the objective
Equally to
moderately
2 1/2 Intermediate value
Moderately 3 1/3 Experience and judgment slightly favor one activity
over another
Moderately to
strongly
4 1/4 Intermediate value
Strongly 5 1/5 Experience and judgment strongly favor one activity
over another
Strongly to very
strongly
6 1/6 Intermediate value
Very strongly 7 1/7 An activity is favored very strongly over another
Very strongly to
extremely
8 1/8 Intermediate value
Extremely 9 1/9 The evidence favoring one activity over another highest
possible order of proof
Consistency check
The consistency of subjective judgment can be checked by estimating consistency ratio which is
the comparison between consistency index and random consistency index. The consistency
index (CR) can be computed by using equation 3.9:
CR=CI/RI --------------------------------------------------------- (3.9)
Where, CI is the consistency index and RI is the random consistency index.
39
The consistency index is a measure of consistency can be estimated using equation
1
max
n
nCI
-------------------------------------------------------- (3.10)
Where, is the principal Eigen value obtained from priority matrix and n is size of
comparison matrix.
After consistency is checked and pair wise comparison was done, then final weight could
overlaid to produce map of soil erosion source areas and identify soil erosion hotspot areas of the
Dengora and Meno watersheds.
Table 3-11 RI on the basis of various sample size
N 1 2 3 4 5 6 7 8 9 10
RI 0 0 0.58 0.9 1.12 1.24 1.32 1.41 1.45 1.49
Source: (Saaty, 1977)
This study was used a pairwise comparison technique to assign the weights of the decision
factors since; it is less bias than other techniques like ranking technique. In pairwise comparison
technique, each factor was matched head-to-head (one-to-one) with each other and a comparison
matrix was prepared to express the relative importance.
2.9.5.1. Standarding and Assigning Criteria Weights
AHP involves analyzing a series of alternatives or objectives with a view to ranking them from
the most preferable to the least preferable using a structured approach procedure proceeding
from the pair wise comparison of criteria to evaluate the weights that assign relative importance
to selected factors. The priority scales were deriving by calculating the eigenvector associated
with the principal eigenvalue of each comparison matrix (Saaty, 1980).
The numerical values then normalized by dividing each entry in the column by the sum of all the
entries in that column, so that they sum up to one. Following normalization, the values was
averaged across the rows to give the relative importance weight for each sub factor. After
estimation of final weight, the final priority could be determined using normalization and their
corresponding weights obtained from Saaty‟s AHP based multi-criteria decision analysis
40
(MCDA) and classified in different categories from very high, high, moderate, low and very low
priorities. In this study pairwise comparison method were introduced and applied to assign
weights to the criteria. Different decision makers may apply different criterion and assign
different weights for each criterion according to their preferences.
3.9.6. Weighted overlay
After soil erosion factor maps were generated for each factor maps (land use land cover map, soil
map, TWI, STI and potential locations of gullies) were reclassified based on sensitivity classes.
Relative weights were assigned to each factor depending on the relevance of each factor. Values
were assigned to each factors based on pair wise comparison criteria. Pair wise comparison
method was used to acquire the final weight of each factor. Based on factors final weight, the
reclassified map was overlaid identify erosion sensitive areas(Yesuph and Dagnew, 2019) to
obtain the combined effect of all factors and produce the final soil erosion source area map of
Dengora and Meno watersheds.
Figure 3-10 Workflow charts of the criteria weighting using MCDA in Arc GIS 10.1.
41
4. RESULTS AND DISCUSSIONS
4.1. Soil Loss factors
4.1.1. Rainfall Erosivity Factor (R)
The contribution of the erosive agent water (precipitation) is represented by the rainfall erosivity
factor R. The R as RUSLE factor is estimated originally from both rainfall and rainfall intensity.
However, as these data are usually unavailable in developing countries unless there are standard
meteorological stations, a common solution was to use correlations between the R-factor and
annual rainfall to estimate erosivity factor as derived by (Hurni, 1985) for Ethiopian condition
(Equation 3.2) with in a mean annual rainfall of 949.5 and 841.85mm/year for Dengora and
Meno watersheds which gives 804.84 and 712.67 mm respectively.
4.1.2. Soil Erodibility Factor (K)
Soil erodibility represents the effect of soil properties on soil erosion. The high k factor value
indicates the more vulnerable soil types to soil erosion and the smaller value shows less
vulnerable soil to soil erosion. In this study, FAO standard classification of soil type was
obtained from Ministry of Water, Irrigation and Electric (MoWIE). The soil type map was
presented in figure 2-3. Therefore based on (FAO, 1989) the K-factor of the study area was
developed using this suggestion (Table 3-5), the soils were converted to vector format in to grid;
and then analysis of K factor.
42
Figure 4-1 K-Factor map of Dengora (top) and Meno(bottom)watersheds
Based on soil erodibility map, Leptic Luvisols, Heplic Luvisols and Hyperskeletic Leptosols
were higher K values (0.2) indicated more sensitive to soil erosion, The other two soil types
(Eutric leptosols and Eutric leptosols) have medium k values (0.15) and Lithic Leptosols have
small k values( 0.1)and Exposed Rock has constraint to soil erosion. Generally based on soil
erodibility capacity (K) values the upper and lower parts of Dengora watershed was relatively
highly sensitive, while the lower part of the Meno watershed was highly sensitive to soil erosion.
The soil physical properties such as soil wetness, water holding capacity and infiltration rate
plays a great influence for the erodibility of soil.
43
4.1.3. Topographic (LS) factor
Topographic factors (LS) were one of the major contributors to soil erosion rate. The
topographic factor map which was developed from slope length and slope steepness as shown in
the above equation (3.3). The values of LS factor ranges from 0 to 41.19 and 0 to 46.69 for
Dengora and Meno watersheds respectively as indicated in Figure 4-2.
Figure 4-2 Topographic (LS) factor of Dengora (top) and Meno(bottom)watersheds
4.1.4. Land Cover Factor (C)
A land-use and land-cover map of the study area was prepared from Landsat 8 satellite image
acquired on 2019 and Maximum Likelihood supervised land use land cover classification type
was employed using ArcGIS software. In addition, ground truth data were used as a vital
reference for supervised classification, accuracy assessment and validation of the result. In
supervised image classifications technique, land use and land cover types were classified so as to
44
use the classified images as inputs for generating crop management (C) factor and support
practice (P) factor.
Based on the land cover classification map, a corresponding C value suggested by Hurni
(1985)as presented in Table3-6above was assigned in a GIS environment. The land cover factor
map was presented in figure 4-3.
Figure 4-3 Land Cover (C) Factor map for Dengora (top) and Meno(bottom) watersheds
45
4.1.5. Land Management Practice Factor (P)
The P-factor was assessed using major land cover and slope interaction adopted by Wischmeier
and Smith (1978) for Ethiopia condition. The data related to management practices of the study
watershed were collected during the field work. Therefore, values for this factor were assigned
considering local management practices based on slope ranges and taken the weighed value for
similar land use types. As indicated in Table 3-7 above the conservation practice (P) factor value
ranges from 0.1 to 0.33 for cultivated land and 1 for other land uses since no influence to soil
erosion. Based on conservation practice (P) values, P factor map was developed on GIS by
intersecting land use land cover type and slope for both watersheds (figure 4-4).
Figure 4-4 Land management factor Dengora (top) and Meno(bottom)watersheds
46
Soil erosion controlling measures predominantly are practiced in sloping and cultivated area. The
lower the P value, the more effective the conservation practice is supposed to be at reducing soil
erosion. Cultivated lands were the dominant land use scattered at lower part and slope classes of
the watershed. Since conservation (management) practice reflects the effect of the management
at a catchment that minimize rate of soil erosion, so the high value of conservation practice (P) at
cultivated lands were indicated more sensitivity while small values showed less effect to soil
erosion.
4.2. Soil Loss Estimation
The annual Soil loss rate was computed spatially by multiplying the RUSLE factors over the
watersheds by using raster calculator in ArcGIS. Based on the analysis, the magnitude of annual
soil loss of the watersheds was estimated to 0 – 223.97 and 0 – 256.09 ton ha-1
year-1
and their
mean annual soil loss 16.34 and 23.26 ton ha-1
year-1
for Dengora and Meno watersheds
respectively. The result of RUSLE model indicated that the upper part of the watersheds is the
most sensitive area with higher soil erosion rate (Table 4-1 and 4-2). The estimated soil loss both
of the watersheds is within the range of soil loss estimated for the Ethiopian highlands by the
Soil Conservation Research Program (SCRP), which was in the range of 0 to 300 ton ha_1
yr_1
(Hurni, 1985). In the highlands of Ethiopia and Eritrea soil losses are extremely high with an
estimated average of 20 metric tons ha-1
year-1
(Hurni, 1985) such that soil loss rate and the
spatial patterns are good argument and relatively conforms well compared to what can be
observed from literature. Therefore this model gives good estimate of soil loss at Dengora and
Meno watershed as it is parts of Ethiopian highlands.
RUSLE prediction has a limitation of considering gully erosion which is the main contributor to
soil loss in the Ethiopian highlands. A recent study by Tamene et al. (2017) in Tigray indicated
that the RUSLE model predicted higher soil loss rates at steep slopes and middle slope positions
as well as along gullies. The predicted soil loss in the Dengora and Meno watersheds was
relatively smaller on lower slope positions. This is due to the fact that, the slope of the watershed
is more dominating factor for the model. However, in bottom land of the watersheds indicated
that several active gully erosion, gully head, bank collapse and more cultivated practice were
found on the valley bottomlands. This observation is in line with what has been reported by
Tebebu et al. (2010) and Zegeye et al. (2016) in which the rise of perched groundwater table in
47
the saturated bottomland areas of the Debre Mawi watershed resulted in the formation and
expansion of gullies. Since as the observation of the watersheds the gullies were found in the
bottomlands of the watersheds. This study was to identify erosion hotspot areas from watersheds
so as to consider gully susceptible areas in the watersheds.
Based on (FAO, 1984) soil loss classification, the soil loss was classified into four major severity
classes. The severity class for both of the watersheds area indicated in Table 4-1/4-2 below
respectively.
Table 4-1 RUSLE based soil loss severity class for Dengora watershed
Class Soil loss (t ha-1yr-1) Area (ha) Percentage Description
I 0-15 34 70.4 Slight
II 15-50 9 18.7 Moderate
III 50-200 5.2 10.74 High
IV > 200 0.07 0.14 Very high
Total 48.3 100
Source: (FAO, 1984)
Table 4-2 RUSLE based soil loss severity class for Meno watershed
Class Soil loss (t ha-1yr-1) Area (ha) Percentage Description
I 0-15 72.6 76 Slight
II 15-50 15.8 16.54 Moderate
III 50-200 6.7 7.3 High
IV > 200 0.15 0.15 Very high
Total 96.2 100
Source: (FAO, 1984)
48
Figure 4-5 Estimated annual soil loss for Dengora (top) and Meno(bottom)watershed
Based on the results of this study (Table 4-1/4-2 and Figure 4-5), the soil loss were distributed to
all parts of the watershed and it is at slight to moderate risk. According to Hurni (1983) and
Morgan (2005) the tolerable soil loss is maximum soil loss which occurs from a land without
resulting land degradation (11 t ha-1
yr-1
). Analysis of the result show that; 70.4% and 72.6 % of
the Dengora and Meno watershed respectively soil loss rate within the range of the tolerable soil
loss rate and most parts of uplands of the watersheds area were exceeded the tolerable soil loss
rate.
49
4.3. Multi- Criteria Decision Analysis (MCDA)
4.3.1. Land use land cover map
From Dengora and Meno watersheds different land use land cover types were identified. The
image classification accuracy was conducted on classified images to determine how well the
classification process accomplished using error matrix. The overall accuracy for Dengora and
Meno watersheds land use land cover map were 83.33% and 90% and also kappa coefficient
82% and 86% respectively, which were under acceptable limit according to (Anderson, 1976).
Land use land cover map of the watersheds were weighted based on AHP comparison to evaluate
effects on soil erosion. Major land use land cover categories were reclassified and weighted
based on pairwise comparison to evaluate effects on soil erosion (Table 4-3/4-4).
Table 4-3 Dengora watershed LULC sensitivity class to soil erosion
S/N Land use/ land cover Area (%) Severity class
1 Cultivated land 21.72 S1
2 Cultivated land with trace 39.19 S2
3 Bush land 9.23 S3
4 Shrub land 26.3 S3
5 Forest land 3.41 S4
Table 4-4 Meno-watershed watershed LULC sensitivity class to soil erosion
S/N Land use/land cover Area (%) Severity class
1 Cultivated land 48 S1
2 Shrub land 35 S2
3 Bush land 16 S2
4 Forest land 0.5 S3
5 Grazing land 0.46 S3
50
The sensitivity classes were developed by regarding on land cover factors in Table 3-6 above,
this means a land which have high cover factor was assigned to highly sensitive (S1) and Vis
versa. As indicated in Table, cultivated land and bush land were more sensitive to soil erosion,
while forest lands were less sensitive to soil erosion. Roads were constraint for the soil erosion.
Figure 4-6 LULC sensitivity map
4.3.2. Soil map
The other criterion in MCDA to estimate potential soil erosion risk area is erodibility, which is
vulnerability of the soils to get eroded. The value of K-factor has been classified in the above
(Figure 4-1). The soil with low K-factor value is less susceptible to erosion agents and with high
K-factor value more susceptible. The major soil types which were obtained in both of the
51
watershed were reclassified and weighted based on their sensitivity classes to soil erosion using
pairwise comparison method in the Table below.
Table 4-5 Dengora watershed soil type sensitivity class
Soil type Area (%) Sensitivity class
Heplic Luvisols 18.1 S1
Leptic Luvisols 31.2 S1
Eutric Leptosols 48.7 S2
Exposed Rock 2 Constraint
Table 4-6 Meno-watershed soil type sensitivity classes
Soil type Area (%) Sensitivity class
Hyperskeletic Leptosols 19.34 S1
Lithic Leptosols 0.82 S3
Eutric Leptosols 79.91 S2
As presented in Table above, Leptic Luvisols, Heplic Luvisols and Hyperskeletic Leptosols were
highly sensitive (S1), Eutric Leptosols have the dominant soil in the watershed which were
moderately sensitive (S2) and Lithic Leptosols were slightly sensitive (S3) to soil erosion based
on soil erodibility and exposed rock was constraint in soil erosion due to the properties of each
soil.
52
Figure 4-7 soil type sensitivity maps
4.3.3. Topographic Wetness Index Factor
Another important factor considered for identification of erosion hotspot area was TWI which
can be used to quantitatively simulate upslope contributing area on soil erosion and soil moisture
conditions in a watershed and it is used as an indicator of static soil moisture content. It is also
useful for distributed hydrological modelling for describes the effect of topography, mapping
drainage, soil type, soil infiltration and crop or vegetation distribution on soil erosion. In this
study the TWI was extracted from Digital Elevation Model (DEM).Determining the saturated
excess runoff generation over the land represented with topographic wetness index. The re-
classified TWI map indicated (Figure 4-9 below and Table 4-7) below.
53
Table 4-7 Topographic wetness index sensitivity class
TWI Erosion sensitivity class
Up to 11.5 S3
11.5 to 16.5 S2
16.5 to high S1
Higher elevation areas have low WI values whereas lowest elevation areas high TWI. Hill slopes
in the watershed were characterized as low TWI values indicating dry areas whereas TWI values
increases at lower reaches of the watershed i.e., in piedmont and flood plains indicating as
possible source areas for saturated overland flow.
54
Figure 4-8 TWI sensitivity class
4.3.4. Stream power index (STI) Factor
The Stream Power Index (SPI) is a measure of the erosive power of flowing water. SPI is
calculated based upon slope and contributing area. SPI approximates locations where gullies
might be more likely to form on the landscape. As designated in the earlier methodological
sections of this study the Stream power index (SPI) factor was considered as the major factor
contributed to soil erosion in the study area. It is the rate of the energy of flowing water
expended on the bed and banks of a stream line. The re-classified SPI map indicated (Figure 4-9
below and Table 4-8) as shown below.
55
Higher SPI values has been also observed along both sides of streams in the watershed indicating
possible source of soil erosion due to concentrated flow of runoff.
Table 4-8 stream power index sensitivity class
SPI Erosion sensitivity class
Up to 5 S3
5 to 12 S2
12 to high S1
Figure 4-9 SPI sensitivity class
56
4.3.5. Gully potential location map
In this study gully locations were found water flow concentrated as small streams and at
waterway sides of the ground surfaces (stream routs). The potential locations of gullies both of
the watersheds were identified based on SPI and TWI threshold values by overlaying their maps
with “AND” Boolean operation in ArcGIS raster calculator using the expression SPI >12 and
TWI >6.8 to get best and realistic locations of gullies in the watershed. The resulted map (Figure
4-10) shows areas with no gully and with gully formation.
Figure 4-10 Potential Location of Gully
Gully formation follows almost along stream lines of the watersheds and the map clearly shows
that small gullies (plot level) were not captured by the threshold. Gully locations were high
sensitive class (S1) while no gully location less sensitive class (S3). The sensitivity classes were
used to reclassify gully map (Figure 4-10).
57
4.3.5.1. Validation of potential Gully location
Gully sites and density was identified from stream power index within the range of 12 to 16.47
and topographic wetness index of 6.8 to 15.38 value for Dengora watershed and Stream Power
Index values 12 to 16.84 and topographic wetness Index of 6.8 to 17.92 for Meno watershed
were overlaid to get potential gully location. The results of stream power index analysis for the
Dengora and Meno watershed indicated in the range of 0.72 to 16.48 and 2.88 to 16.84
respectively. According to Lulseged and Vlek (2005), areas in a watershed with a stream power
index of greater than 18 are susceptible to gully formation. However; the watershed gully
location indicated that up to the stream power index value of 12 which can be taken as a
threshold value of gully prone area in the study watershed. Similarly, the topographic wetness
index of the watershed can be used the threshold value of 6.8 and greater(Lulseged and Vlek,
2005) for susceptibility to gully erosion risk. This result can be taken as a good indicator of
threshold variability for gully susceptible area identification in the watershed. The overlaid map
of topographic wetness index and stream power index (Figure 4-10) indicated that gullies were
found. Based on the result, gullies were found along the natural streams lines within higher
stream power index (SPI) both of the watersheds. This can be validated during the transect walk
gullies are occur in the rout of the stream of the two watersheds as shown in figure below
Figure 4-11 Sample Gully of Dengora watersheds
58
active Gully
Figure 4-12 Photo of sample gully on Meno watershed
Randomly Sampled gullies were digitalized and its location was collected by GPS and cross
tabulated with gully location. Therefore, The validation were made to cross check gully potential
sites by overlaying on the watershed boundary and the validation were 78.6 % and 86.7% for
Dengora and Meno watersheds respectively accurate as indicated in Table 4-9.
Table 4-9 Accuracy assessment of gully area
Dengora watershed Meno-watershed
Potential areas No. ground points No. ground points
Potential to Gully 22 26
Not potential to gully 6 4
Total 28 30
Overall accuracy (%) 78.6 86.7
4.4. Pairwise comparison for parameters
Pairwise comparison matrix was prepared by comparing factors one to one based on pairwise
comparison scale which was broken down from 1 to 9 with the help of natural resource expert‟s
opinion. For assigning weights in this study, pairwise comparison method was used so as to
reduce the complexity of decision making since two components are considered at a time. The
highest value indicates absolute important and the reciprocal kept in the transpose position
59
indicating absolute insignificant (Appendix D). The weights of factors were computed after
normalizing the Eigen vector by its cumulative and multiplied by 100%. The reliabilities of
weights were checked by computing the consistency of comparison matrix which was 9.03 %
and4.6%for Dengora and Meno watersheds respectively which is under the accepted Consistency
Ratio (<10%). Accordingly, the pairwise weights were accepted and consistent, so, the process
was continued. Finally, the MCDA erosion intensity map of the area has been produced by
multiplying the four criterion layers by their weight derived from pair wise comparison (Table 4-
10) and then sum up the results by Weighted Linear Combination (WLC) equation in raster
calculator operation of ArcGIS.
Table 4-10 The influencing power of the factors
Criteria Weight (%)
Dengora watershed
SPI 34
Land Use 11
Soil Type 6.4
TWI 29
Gully 19.6
Meno watershed
SPI 17.5
Land Use 8.1
Soil Type 10.6
TWI 26.8
Gully 37
60
Figure 4-13 Overall contributions of parameters for soil erosion.
4.5. Identification of Soil Erosion Hotspot Areas
Based on the methodology designed for identification of soil erosion hotspot area all selected
factors were overlaid to identify the area sensitive to erosion as Highly, Moderate, slightly and
currently not sensitive (constraint). The sensitivity map (Figure 4-14) shows the relative ranking
of the erosion potential sites, generated by weighted overlay mapping, according to the weight of
concerned criteria, the most sensitive areas to erosion under the multi-criteria evaluation (S1)
spatially coincided with the actual gully locations. Therefore MCDA technique indicates more
accurate than RUSLE prediction as a result of RUSLE model were more sensitive with slope and
not considering gully but MCDA considering pairwise comparison of all factors that affect soil
erosion. This is consistent with the findings reported by Zegeye et al. (2016) and Poesen et al.
(2003) in that gullies are critical sediment source areas in the Ethiopian highlands and accurate
soil erosion prediction should properly address estimating gully erosion. Weighted overly of all
factors was important for soil erosion source area assessments to obtain the most sever sites at
Dengora and Meno watersheds. The combined overall results were presented in the Table below.
0
5
10
15
20
25
30
35
40
SPI Land Use Soil Type TWI Gully
Dengora watershed
Meno watershed
Soil erosion contributing factor
61
Table 4-11 Overall Dengora watershed Erosion sensitivity
S.N Area (ha) Area (%) Severity classes
1 4.7 9.7 Highly (S1)
2 19.64 64.5 Moderate (S2)
3 8.73 18 Slight (S3)
4 3.8 7.8 Currently not sensitive (S4)
Total 48.5 100
Table 4-12 Overall Meno- watershed Erosion sensitivity
S.N Area (ha) Area (%) Severity classes
1 5.9 6.1 Highly (S1)
2 68.5 71.3 Moderate (S2)
3 22.3 23.23 Slight (S3)
4 0.36 0.375 Currently not sensitive (S4)
Total 96 100
Figure 4-14 overall soil erosion risk map
62
The result showed that, for Dengora watershed 9.7% of the total watershed area was highly
sensitive, 64.5% moderate, 18 % slight and 7.8% were currently not sensitive (exposed rocky
areas) and for Meno watershed 6.1 % of the total watershed area was highly sensitive, 71.3 %
moderate, 23.23 % slightly sensitive and 0.375% of the total area were currently not sensitive to
soil erosion based on the combined effect of annual soil loss class and MCDA technique of soil
erosion hotspot area identification in the watershed. Generally the result indicated that the
Dengora and Meno watersheds were at moderate risk. Moreover, the MCDA revealed that the
upper part of the watersheds is slightly sensitive (S3) for soil erosion and this could be explained
by the fact that in the upland areas there are no gullies and also forest, shrub and bush land
covers.
63
5. CONCLUSIONS AND RECOMMENDATIONS
5.1. Conclusions
In this specific study, RUSLE model and MCDA technique were used to identify erosion hotspot
areas in the Dengora and Meno watersheds. RUSLE model prediction indicated that the upslope
portion of the watersheds is highly sensitive to erosion. Whereas the multi criteria decision
analysis (MCDA) technique indicated that the bottom slope or the saturated bottomland is highly
sensitive to erosion. Topographic wetness index and stream power index (SPI) were powerful
predictors for the potential gully formation, which coincided with our field gully mapping. The
results from model prediction in combination with a gully validation indicated that the Dengora
and Meno watersheds were overlay of TWI ≥ 6.8 and SPI ≥12 to be sensitive to gully erosion.
The overall result of this study, the bottomlands were the most important erosion- prone areas of
the watershed. Thus, these areas should be given priority in the intervention of integrated
watershed management practices focused on gully erosion areas.
The overall study indicated that most erosion hotspot areas were found in the valley bottomlands
(gully) of the watersheds, which was extremely important to consider valley bottom as
intervention areas during design of watershed intervention. Erosion prone area identification was
useful information for the evaluation and decision making about implementation of intervention
in the watersheds due to resource and manpower scarcity to treat the watershed at a time. From
the validation of results, MCDA technique is more powerful tool than RUSEL for planning and
targeting of intervention area.
64
5.2. Recommendations
The study was vital to reduce excessive soil erosion from lower part of Dengora and Meno
watersheds to an artificial reservoir Atikayina dam and Meno gravity dam respectively by
boosting sediment concentration close to the outlet. This study will provide detail information for
planners, decision makers and other concerned stakeholders to take effective soil and water
conservation practices in order to reduce soil loss by improving the water holding capacity of the
watershed. Based on the present research outlook, the following recommendations were drawn for
the sustainable watershed monitoring practice in the area;
Valley bottomlands of the watersheds characterized by high soil erosion prone area need
immediate attention to soil and water conservation practice. The local people should be
aware about the loss and encourage them to apply the effective intervention mechanisms to
tackle the problem.
In planning of watershed development programs, identifying the targeting area and
technology options is very important.
Better to use models that used compressive watershed characteristics (MCDA better than
RUSLE.
65
6. REFERENCES
adugna, A., Abegaz, A. & Cerdà, A. 2015. Soil erosion assessment and control in Northeast
Wollega, Ethiopia. Solid Earth Discussions, 7.
Amsalu, A., Stroosnijder, L. & De Graaff, J. 2007. Long-term dynamics in land resource use and
the driving forces in the Beressa watershed, highlands of Ethiopia. Journal of
Environmental management, 83, 448-459.
Anderson, J. R., E.E. Hardey, J.T. Roach And R.E. Witmer, 1976. A land use and land cover
classification system for use with remote sensor data. U.S. Geological Survey
professional paper 964, USGS, Reston, VA.
Asquith, P., Pathak, P. A. & Ritter, J. R. 2005. Short interest, institutional ownership, and stock
returns. Journal of Financial Economics, 78, 243-276.
Assefa, T. T., Jha, M. K., Tilahun, S. A., Yetbarek, E., Adem, A. A. & Wale, A. 2015.
Identification of erosion hotspot area using GIS and MCE technique for Koga watershed
in the Upper Blue Nile Basin, Ethiopia. American Journal of Environmental Sciences, 11,
245.
Authority, E. P. 2012. National Report of Ethiopia, The United Nations Conference on
Sustainable Development (Rio+ 20). Federal Democratic Republic of Ethiopia, Addis
Ababa.
Berhe, W. 1996. Twenty years of soil and water conservation in Ethiopia: A personal overview.
Regional soil conservation unit/SIDA, Nairobi, Kenya.
Birru, Y. 2007. Land degradation and options for sustainable land management in the Lake Tana
Basin (LTB), Amhara Region, Ethiopia. PhD Thesis, Centre for Development and
Environment (CDE), Geographic Institute, University Bern, Bern, Switzerland. Solid
Earth Discussions.
Blanco-Canqui, H. & Lal, R. 2008. No-tillage and soil-profile carbon sequestration: An on-farm
assessment. Soil Science Society of America Journal, 72, 693-701.
Braimoh, A. K. & Vlek, P. L. 2008. Impact of land use on soil resources. Land Use and Soil
Resources. Springer.
Can, A. 1993. Residential quality assessment. Geographic information systems, spatial
modelling and policy evaluation. Springer.
66
Carver, S. J. 1991. Integrating multi-criteria evaluation with geographical information systems.
International Journal of Geographical Information System, 5, 321-339.
Congalton, R. G. 1991. A review of assessing the accuracy of classifications of remotely sensed
data. Remote sensing of environment, 37, 35-46.
Descheemaeker, K., Nyssen, J., Rossi, J., Poesen, J., Haile, M., Raes, D., Muys, B., Moeyersons,
J. & Deckers, S. 2006. Sediment deposition and pedogenesis in exclosures in the Tigray
Highlands, Ethiopia. Geoderma, 132, 291-314.
Desmet, P. & Govers, G. 1996. A GIS procedure for automatically calculating the USLE LS
factor on topographically complex landscape units. Journal of soil and water
conservation, 51, 427-433.
Dube, F., Nhapi, I., Murwira, A., Gumindoga, W., Goldin, J. & Mashauri, D. 2014. Potential of
weight of evidence modelling for gully erosion hazard assessment in Mbire District–
Zimbabwe. Physics and Chemistry of the Earth, Parts A/B/C, 67, 145-152.
Emama, B., Mohammed, H. & Mohammed, S. 2015. A situational analysis of agricultural
production and marketing, and natural resource management systems in the Ethiopian
highlands.
Erenstein, O. C. 1999. The economics of soil conservation in developing countries: the case of
crop residue mulching.
FAO 1981. Food and Agricultural Organization.(1981). Guidelines for designing and evaluating
surface irrigation systems.Irrigation and Drainage Paper 45.FAO (Food and Agricultural
Organization of the United Nation, Rome.
FAO 1984. Food and Agricultural Organization. (1984). the Ethiopian Highlands Reclamation
Study : Final Report. Rome. (Vol.1-2).
FAO 1986. Food and Agriculture Organization .Ethiopian highlands reclamation study, Final
Report (Volume II), Food and Agriculture Organization (FAO) Rome.
FAO 1989. Food and Agriculture Organization. The state of food and agriculture (Vol. 37) UN,
37, 1-132.
FAO 2001. Preparation of Land Cover Database of Bulgaria through remote sensing and GIS,
Version 6. Bulgaria, Europe.
Fauck, R. 1956. Evolution of soils under mechanized cultivation in tropical areas. Trans. 6th Int.
Congr. Soil Sci. E, 593-596.
67
Fischer, M. M. & Nijkamp, P. 1993. Geographic information systems, spatial modelling and
policy evaluation, Springer.
Grissinger, E. & Murphy, J. Ephemeral Gully erosion in the loss uplands, Gardwin watershed,
Northern Mississippi, USA. Proc. 4th Int. River Sedimentation Symp. Beijing, China,
IASH Pub, 1989. 2541-266.
Hawando, T. 1995. The survey of the soil and water resources of Ethiopia. UNU/Toko.
Helldén, U. 1987.An assessment of woody biomass, community forests, land use and soil erosion
in Ethiopia. A feasibility study on the use of remote sensing and GIS [geographical
information system]-analysis for planning purposes in developing countries, Lund
University Press.
Hurni 1993a. Land degredation, famine, and land resource scenarios in Ethiopia.
Hurni, H. 1983. Soil erosion and soil formation in agricultural ecosystems: Ethiopia and
Northern Thailand. Mountain research and development, 131-142.
Hurni, H. 1985. Erosion-productivity-conservation systems in Ethiopia.
Hurni, H. 1989a. Applied soil conservation research in Ethiopia.
Hurni, H. 1989b. Degradation and conservation of the resources in the Ethiopian highlands.
International Mountain Society; Mountain Research and Development. 8(2/3):123−130. ,
Hurni, H. 1993b. Land degredation, famine, and land resource scenarios in Ethiopia.
Janssen, R. & Rietveld, P. 1990. Multicriteria analysis and geographical information systems: an
application to agricultural land use in the Netherlands. Geographical information systems
for urban and regional planning. Springer.
Joerin, F. & Musy, A. 2000. Land management with GIS and multicriteria analysis. International
transactions in operational research, 7, 67-78.
Joint, F. 1986. WHO expert committee on brucellosis. World Health Organ Tech Rep Ser, 740,
1-132.
Julien, P. Y. 2010. Erosion and sedimentation, Cambridge University Press.
Lal, R. 1994. Soil erosion research methods, CRC Press.
Lillie, R. J., Nelson, K. D., De Voogd, B., Brewer, J. A., Oliver, J. E., Brown, L. D., Kaufman, S.
& Viele, G. W. 1983. Crustal structure of Ouachita Mountains, Arkansas: A model based
on integration of COCORP reflection profiles and regional geophysical data. AAPG
Bulletin, 67, 907-931.
68
Lulseged, T. & Vlek, P. GIS-based landscape characterization to assess soil erosion and its
delivery potential in the highlands of northern Ethiopia. Proceedings of the 1st
International Conference on Remote Sensing and Geoinformation Processing in the
Assessment and Monitoring of Land Degradation and Desertification,(ICRS „05), 2005.
7-9.
Malczewski, J. 1996. GIS-based approach to multiple criteriagroup decision-making.
International Journal of GeographicalInformation Science 10 (8): . Water Resources
Research,, 321-339.
Mccool, D. 1995. Public policy theories, models, and concepts: An anthology, Prentice Hall
Englewood Cliffs, NJ.
Mirsal, I. A. 2008. Soil pollution, Springer.
Mitiku, H., Herweg, K. G. & Stillhardt, B. 2006. Sustainable land management: A new approach
to soil and water conservation in Ethiopia. Centre for Development and Environment
(CDE) and NCCR North-South.
Morgan, D. O. 1995. Principles of CDK regulation. Nature, 374, 131.
Morgan, R. P. C. 2005. Soil Erosion and Conservation, 3rd Edn. Blackwell Publishing, New
York, USA,.
MOWIE 1993. Ministry of Water, Irrigation and Electricity.(1993). Improvement of the
resource–population sustainability balance. Water Resources Development, MoWIE,
Addis Ababa, Ethiopia.
Najm, Z., Keyhani, N., Rezaei, K., Nezamabad, A. N. & Vaziri, S. H. 2013. Sediment yield and
soil erosion assessment by using an empirical model of MPSIAC for Afjeh & Lavarak
sub-watersheds, Iran. Earth, 2, 14-22.
Nyamai, M., Mati, B. M., Home, P. G., Odongo, B., Wanjogu, R. & Thuranira, E. 2012.
Improving crop productivity and water use efficiency in basin rice cultivation in Kenya
through SRI. Agricultural Engineering International: CIGR Journal, 14, 1-9.
Nyssen, J., Haregeweyn, N., Descheemaeker, K., Gebremichael, D., Vancampenhout, K.,
Poesen, J., Haile, M., Moeyersons, J., Buytaert, W. & Naudts, J. 2006. Modelling the
effect of soil and water conservation practices in Tigray, Ethiopia (vol 105, pg 29, 2005).
AGRICULTURE ECOSYSTEMS & ENVIRONMENT, 114, 407-411.
69
Oldeman, L. R. 1992. Global extent of soil degradation. Bi-Annual Report 1991-1992/ISRIC.
ISRIC.
Onyando, J., Kisoyan, P. & Chemelil, M. 2005. Estimation of potential soil erosion for river
perkerra catchment in Kenya. Water Resources Management, 19, 133-143.
Openshaw, S. 1991. A view on the GIS crisis in geography, or, using GIS to put Humpty-
Dumpty back together again. Environment and Planning A, 23, 621-628.
Philor, L. 2011. Erosion impacts on soil and environmental quality: Vertisols in the Highlands
Region of Ethiopia. Soil and Water Science Department, University of Florida.
Poesen, J., Nachtergaele, J., Verstraeten, G. & Valentin, C. 2003. Gully erosion and
environmental change: importance and research needs. Catena, 50, 91-133.
R. P. C, M. 2005. Soil Erosion and Conservation. 3rd Edn. Blackwell Publishing, New York,
USA,, .
Renard, A. F. 1983. Soil Conservation: Principles of erosion by water. In H. E. Dregne and W.
O. Willis, eds., Dry land Agriculture,. 155-176.
Renard, K. G., Foster, G. R., Weesies, G., Mccool, D. & Yoder, D. 1997. Predicting soil erosion
by water: a guide to conservation planning with the Revised Universal Soil Loss
Equation (RUSLE), United States Department of Agriculture Washington, DC.
Renard, K. G., Foster, G. R., Weesies, G. A. & Porter, J. P. 1991. RUSLE: Revised universal soil
loss equation. Journal of soil and Water Conservation, 46, 30-33.
Rojas-González, A. M. Soil erosion calculation using remote sensing and GIS in RÍO grande de
Arecibo Watershed, Puerto Rico. Proceedings ASPRS 2008 Annual Conference
Bridging the Horizons: New Frontiers in Geospatial Collaboration, Portland, Oregon,
April 28th–May 2nd, 2008.
Saaty, T. L. 1977. A scaling method for priorities in hierarchical structures. Journal of
mathematical psychology, 15, 234-281.
Saavedra, C. 2005. Estimating spatial patterns of soil erosion and deposition of the Andean
region using geo-information techniques: a case study in Cochabamba, Bolivia.
Salehi, F., Pesant, A. & Lagace, R. 1991. Validation of the universal Soil Loss Equation for
Three cropping systems under natural rainfall in Southeastern Quebec. Canadian
Agricultural Engineering, 33, 11-16.
70
Setegn, S. G., Srinivasan, R., Dargahi, B. & Melesse, A. M. 2009. Spatial delineation of soil
erosion vulnerability in the Lake Tana Basin, Ethiopia. Hydrological Processes: An
International Journal, 23, 3738-3750.
Shi, Z., Cai, C., Ding, S., Wang, T. & Chow, T. 2004. Soil conservation planning at the small
watershed level using RUSLE with GIS: a case study in the Three Gorge Area of China.
Catena, 55, 33-48.
Shiferaw, E. 2015. Awareness and views of farming households regarding land resource
degradation and conservationthe case of Bule Hora. Ethiopia.
Sinore, T., Adugna, O. & Melkamu, T. 2017. Community Perception on Land Degradation
Problems and Management Practices in Begi Woreda, Oromia Regional State, Ethiopia.
International Jornal Of Agriculture, Forestry and Fisheries, 47-54.
Tamene, L., Adimassu, Z., Ellison, J., Yaekob, T., Woldearegay, K., Mekonnen, K., Thorne, P.
& Le, Q. B. 2017. Mapping soil erosion hotspots and assessing the potential impacts of
land management practices in the highlands of Ethiopia. Geomorphology, 292, 153-163.
Tebebu, T., Abiy, A., Zegeye, A., Dahlke, H., Easton, Z., Tilahun, S., Collick, A., Kidnau, S.,
Moges, S. & Dadgari, F. 2010. Surface and subsurface flow effect on permanent gully
formation and upland erosion near Lake Tana in the northern highlands of Ethiopia.
Hydrology and Earth System Sciences, 14, 2207-2217.
Tecle A, Y. M. 1990. Preference ranking of alternativeirrigation technologies via a multi
criterion decision makingprocedure. Transactions of ASAE3 (5): . ,, 1509-1517.
Tesfa, A. & Mekuriaw, S. 2014. The effect of land degradation on farm size dynamics and crop-
livestock farming system in ethiopia: A Review. Open Journal of Soil Science, 4, 1.
Toy, T. J., Foster, G. R. & Renard, K. G. 2002. Soil erosion: processes, prediction,
measurement, and control, John Wiley & Sons.
Twinn, D. S. 1998. An analysis of the effectiveness of focus groups as a method of qualitative
data collection with Chinese populations in nursing research. Journal of advanced
nursing, 28, 654-661.
Valentin, C., Poesen, J. & Li, Y. 2005. Gully erosion: impacts, factors and control. Catena, 63,
132-153.
Verstraeten, G. & Poesen, J. 2002. Using sediment deposits in small ponds to quantify sediment
yield from small catchments: possibilities and limitations. Earth Surface Processes and
71
Landforms: The Journal of the British Geomorphological Research Group, 27, 1425-
1439.
Voogd, H. 1983. Multi-Criteria Evaluation for Urban andRegional Planning. Pion, Ltd.: London.
Vrieling, A., Rodrigues, S., Bartholomeus, H. & Sterk, G. 2007. Automatic identification of
erosion gullies with ASTER imagery in the Brazilian Cerrados. International Journal of
Remote Sensing, 28, 2723-2738.
Winterbottom, R., Reij, C., Garrity, D., Glover, J., Hellums, D., Mcgahuey, M. & Scherr, S.
2013. Improving land and water management. World Resources Institute Working
Paper). Accessed on April, 2, 2014.
Wischmeier, W. H. & Mannering, J. 1969. Relation of soil properties to its erodibility 1. Soil
Science Society of America Journal, 33, 131-137.
Wischmeier, W. H. & Smith, D. D. 1978. Predicting rainfall erosion losses-a guide to
conservation planning. Predicting rainfall erosion losses-a guide to conservation
planning.
Yesuph, A. Y. & Dagnew, A. B. 2019. Soil erosion mapping and severity analysis based on
RUSLE model and local perception in the Beshillo Catchment of the Blue Nile Basin,
Ethiopia. Environmental Systems Research, 8, 17.
Yilma, A. D. & Awulachew, S. B. 2009. Blue Nile Basin Characterization and Geospatial Atlas.
Improved Water and Land Management in the Ethiopian Highlands: Its Impact on
Downstream Stakeholders Dependent on the Blue Nile, 6.
Zegeye, A. D., Langendoen, E. J., Stoof, C. R., Tilahun, S. A., Dagnew, D. C., Zimale, F. A.,
Guzman, C. D., Yitaferu, B. & Steenhuis, T. S. 2016. Morphological dynamics of gully
systems in the subhumid Ethiopian Highlands: the Debre Mawi watershed. Soil, 2, 443-
458.
Zhang, L., O'neill, A. L. & Lacey, S. 1996. Modelling approaches to the prediction of soil
erosion in catchments. Environmental Software, 11, 123-133.
72
7. APPENDIX
Appendix A Error matrix accuracy totals for the classified image
Dengora watershed Reference
Cla
ssif
ied
Land use land
cover
Forest
Area
Cultivated
land
Cultivated
with terrace
Bush
land
Shrub
land
Total
Forest Area 4 0 0 1 1 6
Cultivated land 0 5 1 0 0 6
Cultivated terrace 0 1 5 0 0 6
Bush land 0 0 0 0 0 6
Shrub land 0 1 0 6 5 6
Total 4 7 7 7 6 30
Meno watershed Reference
Cla
ssif
ied
Land use land
cover
Forest
Area
Cultivated
land
Grazing
land
Bush
land
Shrub
land
Road Total
Forest Area 4 0 0 1 1 0 6
Cultivated land 0 5 1 0 0 0 6
Grazing land 0 0 4 1 0 0 5
Bush land 0 0 0 5 1 0 6
Shrub land 0 0 1 0 4 0 5
Road 0 0 0 0 0 2 2
Total 4 5 6 7 6 2 30
73
Appendix B AHP pair wise comparison matrix for Dengora watershed
Criteria STI Land use Soil type TWI Gully
STI 1 3 2 2 3
Land use 0.33 1 3 0.33 0.2
soil type 0.2 0.33 1 0.2 0.33
TWI 0.5 3 5 1 3
Gully 0.33 5 3 0.33 1
Cr=
=0.0903<0.1.Acceptable Cr value
Appendix C AHP pair wise comparison matrix for Meno watershed
Criteria STI Land use Soil type TWI Gully
STI 1 3 3 0.5 0.33
Land use 0.33 1 0.33 0.2 0.2
Soil type 0.33 3 1 0.33 0.33
TWI 2 5 3 1 0.5
Gully 3 5 3 2 1
Cr=
=0.0463<0.1. Acceptable Cr value
Appendix D Actual Gully location for Dengora watershed
longitude latitude longitude latitude
38.0441 12.3928 38.0419 12.3919
38.0441 12.3928 38.0375 12.3875
38.0444 12.3926 38.0372 12.3872
38.0451 12.3923 38.0383 12.3883
F1 1/9 1/7 1/5 1/3 1 3 5 7 9 F2
extreme very strong moderate equal moderate strong very extreme
Less important more important
74
38.0436 12.3907 38.0388 12.3888
38.0450 12.3903 38.0391 12.3891
38.0482 12.3892 38.0444 12.3944
38.0354 12.3870 38.0450 12.3950
38.0494 12.3879 38.0447 12.3947
38.0433 12.3888 38.0447 12.3947
38.0439 12.3928 38.0432 12.3932
38.0389 12.3889 38.0411 12.3911
38.0394 12.3894 38.0410 12.3910
38.0397 12.3897 38.0456 12.3956
38.0400 12.3900 38.0408 12.4622
Appendix E Actual Gully location for Meno watershed
Longitude(E) Latitude(N) Longitude(E) Latitude(N)
37.7781 12.4500 37.7782 12.4508
37.7778 12.4501 37.7783 12.4512
37.7774 12.4503 37.7786 12.4509
37.7771 12.4503 37.7783 12.4514
37.7769 12.4503 37.7782 12.4521
37.7769 12.4506 37.7707 12.4531
37.7769 12.4500 37.7695 12.4514
37.7769 12.4506 37.7715 12.4537
37.7767 12.4508 37.7710 12.4525
37.7761 12.4511 37.7707 12.4532
37.7753 12.4518 37.7701 12.4534
37.7739 12.4526 37.7717 12.4525
37.7747 12.4536 37.7714 12.4522
37.7781 12.4503 37.7701 12.4533
37.7724 12.4550 37.7698 12.4533
75
Appendix F Geographical Location of Dengora and Meno Watersheds
76
Appendix G Focal Group Discussion on Meno watershed
Appendix H Focal Group Discussion on Dengora watershed
77