Comparison of SWAT and GeoWEPP model in predicting the impact of stone bunds on runoff and erosion processes in the Northern Ethiopian Highlands

Abstract

Soil degradation is a major issue in the Ethiopian highlands which are most suitable for agriculture and, therefore,
support a major part of human population and livestock. Heavy rainstorms during the rainy season in summer
create soil erosion and runoff processes which affect soil fertility and food security. In the last years programs for
soil conservation and afforestation were initiated by the Ethiopian government to reduce erosion risk, retain water
in the landscape and improve crop yields.
The study was done in two adjacent watersheds in the Northwestern highlands of Ethiopia. One of the watersheds
is developed by soil and water conservation structures (stone bunds) in 2011 and the other one is
without soil and water conservation structures. Spatial distribution of soil textures and other soil properties were
determined in the field and in the laboratory and a soil map was derived. A land use map was evaluated based
on satellite images and ground truth data. A Digital Elevation Model of the watershed was developed based on
conventional terrestrial surveying using a total station. At the outlet of the watersheds weirs with cameras were
installed to measure surface runoff. During each event runoff samples were collected and sediment concentration
was analyzed.
The objective of this study is 1) to assess the impact of stone bunds on runoff and erosion processes by
using simulation models, and 2) to compare the performance of two soil erosion models in predicting the
measurements. The selected erosion models were the Soil and Water Assessment Tool (SWAT) and the Geospatial
Interface to the Water Erosion Prediction Project (GeoWEPP). The simulation models were calibrated/verified for
the 2011-2013 periods and validated with 2014-2015 data. Results of this comparison will be presented.

Full text

University of Natural Resources and Life Sciences, Vienna
Department of Water, Atmosphere and Environment
.
Introduction
In the Ethiopian Highlands deforestation increased the vulnerability of the soils
due to rainfall driven soil erosion (Addis et at. 2016; Nyssen et al. 2000). To
tackle this soil erosion problem soil and water conservation strategies are
considered were constructed.
Two neighboring subwatersheds, Abakaloye and Ayaye, were selected to
investigate the impact of soil and water conservation structures on soil erosion
processes. Soil and water conservation structures were applied in the Ayaye
subwatershed while Abakaloye subwatershed was used as a reference without
soil and water conservation structures. Soil and Water Assessment Tool (SWAT)
(Arnold et al., 1998)and GeoWEPP (Renschler, 2003) models were used.
Materials and Methods
Abakaloye and Ayaye subwatersheds are located in Gumara-Maksegnit watershed in
the Lake Tana basin, Ethiopia.
located between 12°25’24’’ and 12°25’54’’ latitude and between 37°34’56’’ and
37°35’38’’ longitude (Figure 1). Altitude ranges from 1998 to 2150 m asl.
Mean annual rainfall is 1170 mm of which more than 90 %of the rainfall occurs
during the rainy season (June to August).
The average monthly maximum and minimum temperatures are 28.5 oC and 13.3
oC respectively.
Comparison of SWAT and GeoWEPP model in predicting the impact of stone
bunds on runoff and erosion processes in the Northern Ethiopian Highlands
N.D. Melaku1,2*, J.Flagler5, C.S. Renschler5, H. Holzmann1, C. Zucca3, F. Ziadat4, S. Strohmeier3 H. K. Addis2, W.
Bayu and A. Klik1
(1) University of Natural Resources and Life Sciences, Vienna, Austria (2) Amhara Agricultural Research Institute,
Gondar Agricultural Research center, Gondar, Ethiopia (nigus.melaku@students.boku.ac.at), (3) International
Center for Agricultural Research in the Dry Areas (ICARDA), Amman, Jordan, (4) Food and Agricultural
Organization of the United Nations (FAO), Rome, Italy, (5) University at Buffalo, NY, USA
For the Abakaloye and Ayaye subwatersheds, separate SWAT and GeoWEPP
project were set for daily runoff and sediment yield. Sequential Uncertainty
Fitting-2 (SUFI-2), a SWAT-CUP2012 program was used to optimize the
parameters of the SWAT using daily observed runoff and sediment yield data.
DEM, climate data, landuse map, soil map and management data were used as
input files. Runoff and sediment yield were monitored at the outlets of both
subwatersheds (Fig. 4).
Fig.i1 Map of the study area
Results
Table 1 Daily runoff
Fig. 6 Observed and simulated daily runoff for calibration (a) and validation (b) period at the
outlet of Abakaloye subwatershed
Fig. 3 Slope classes, soil map, land use classes and sub basins
Fig. 2 Treated Ayaye subwatershed with field experiments
Fig. 4 Runoff and sediment monitoring
Conclusions
ArcSWAT and GeoWEPP were used to simulate runoff and sediment yield from
Ayaye and Abakaloye subwatersheds. Both ArcSWAT and GeoWEPP provided
satisfactory prediction for runoff while sediment prediction was low in both models.
Acknowlegements
The Austrian Development Agency (ADA), International Center for Agricultural Research in Dry Areas
(ICARDA) and Organization for Economic Co-operation and Development (OECD)
Fig. 7 comparison of observed and SWAT simulated daily sediment yield
Table 1. Model comparison
Subwatersheds
Sediment yield (t/ha)
Model
Simulated
Observed
ArcSWAT
Ayaye
21.8
33.5
40.7
68.7
49.3
Abakaloye
62.6
GeoWEPP
Ayaye
49.3
Abakaloye
62.6
1
Fig. 8 offsite (left) and Onsite (right) soil loss at both Subwatershed
References
Addis H K, Strohmeier S, Ziadat F, Melaku ND, Klik A. 2016. Modeling streamflow and
sediment using SWAT in the Ethiopian Highlands. Int J Agric&BiolEng, 2016; 9(5): 51-66.
Arnold JG, Srinivasan R, Muttiah RS, Williams JR., 1998. Large area hydrologic modeling and
assessment part I:Model development. Journal of the American Water
ResourcesAssociation, 1998;34(1): 73－89.
Renschler, C. S. 2003. Designing geo-spatial interfaces to scale process models: The GeoWEPP
approach.Hydrol. Proc.17(5): 1005-1017.
Fig. 9 comparison of observed and WEPP simulated daily runoff and
daily sediment yield
Calibration
y = 0.687x + 9.990
R² = 0.57
NSE = 0.39
Validation
y = 0.871x + 8.402
R² = 0.698
NSE = 0.22
0
10
20
30
40
50
60
010 20 30 40 50 60
Observed sediment (t ha-1)
Simulated sediment (t ha-1)
Calibration
y = 0.794x + 10.70
R² = 0.63
NSE = 0.43
Validation
y = 0.703x + 7.135
R² = 0.542
NSE = 0.31
0
15
30
45
60
75
90
015 30 45 60 75 90
Observed sediment (t ha-1)
Simulated Sediment yield (t ha-1)
Subwatersheds
Calibration
Validation
R2
NSE
R2
NSE
Ayaye (Treated)
0.78
0.67
0.51
0.57
Abakaloye (Untreated)
0.71
0.58
0.59
0.52
1
Fig.5 Observed and simulated daily runoff for calibration (a) and validation (b) period at the outlet of
Ayaye subwatershed