Journal of Hydro-environment Research

Flood Risk assessment using multi-criteria in the case of Upper Awash river basin,
Ethiopia
--Manuscript Draft--

Manuscript Number:

Article Type: Research paper

Keywords: Flood Risk; GIS; Flood hazard; Multi criteria Analysis

Corresponding Author: reta Birhanu endale, MSc.

Assosa University
Addis Ababa, Addis Ababa ETHIOPIA

First Author: reta Birhanu endale, MSc.

Order of Authors: reta Birhanu endale, MSc.

Abstract: Flood is among the most common and destructive natural hazards causing extensive
damage to infrastructure, public and private services, the Environment, the economy
and devastation to human settlements. The objective of this thesis is to develop a flood
hazard map, flood risk map in GIS environment. To produce the flood hazard seven
flood causative factors were considered, these are the reclassified, Elevation, Rainfall
intensity (precipitation), Drainage Density, Elevation, Slope, Distance to River, soil type
and Land use/cover map of the study area. Flood risk assessment was done by taking
the population and land use land cover as element at risk. Flood hazard analysis for
Upper Awash basin indicates that 1.245%, 20.93%, 76.6% and 1.23% of the area are
characterized by Very high, High, Moderate and Low flood hazard level respectively.
The result of flood risk analysis shows that 3.8, 34.04, 40.42, 9.43 and 2.31 percent of
the area considered in the upper awash basin were subjected to very low, low,
moderate, high and very high flood risk respectively.

Opposed Reviewers:

Powered by Editorial Manager® and ProduXion Manager® from Aries Systems Corporation
Cover Letter

Date:29/11/2022
Journal name: COMPUTERS, ENVIRONMENT AND URBAN
SYSTEMS
Article type: Research Article

I am enclosing herewith a manuscript entitled “Flood Risk assessment using multi-

criteria in the case of Upper Awash river basin, Ethiopia” for publication in
Computers, Environment and Urban system for possible evaluation. The
Corresponding author of this manuscript is Reta Birhanu Endale and contribution
of the authors as mentioned below with their responsibility in the research.

Sl Author full Email ID Contribution of Complete address

no name author
Phone : +251923593151
1 Reta Birhanu Retabirhanubr12@gmail.co planning, execution,
Endale m or analysis of this Fax: +251-057-775-2958
study, reading and P.O.Box : 18
approval of
submission Address: Ethiopia, Assosa
manuscript

1. UNDERTAKING

With the submission of this manuscript I would like to undertake that:

 Authors of this Article (paper) have directly participated in the planning,

execution, or analysis of this study;

 Authors of this paper have read and approved the final version submitted;

 The contents of this manuscript have not been copyrighted or published

previously;

 The contents of this manuscript are not now under consideration for
publication elsewhere;
  The contents of this manuscript will not be copyrighted, submitted, or
published elsewhere, while acceptance by the Journal is under consideration;

 There are no directly related manuscripts or abstracts, published or

unpublished, by any authors of this paper.

 My Institute’s Assosa University representative is fully aware of this

submission.

2. RESEARCH AND MANUSCRIPT RELATED DETAILS

Submitted manuscript is a (Research Article)

The research project was conducted under the supervision of: Ermias Teferi
(PHD.)

This research was conducted from 10/8/2021 to 29/11/2022

Manuscript File Click here to view linked References

1 Title

2 Flood Risk assessment using multi-criteria in the case of Upper Awash river
3 basin, Ethiopia
4

5 Author

6 Reta Birhanu Endale, MSc

7 Affiliation

8 Reta Endale1
1
9 Assosa University
1
10 Department of Surveying Engineering, Assosa University, Ethiopia

11

12 Corresponding author information

13 Name: Reta Birhanu Endale
14 Email: retabirhanubr12@gmail,com
15 Phone : +251923593151
16 Fax: +251-057-775-2958
17 P.O.Box : 18
18 Address: Ethiopia, Assosa
19

1
20 ABSTRACT
21 Flood is among the most common and destructive natural hazards causing extensive damage to
22 infrastructure, public and private services, the Environment, the economy and devastation to
23 human settlements. The objective of this thesis is to develop a flood hazard map, flood risk map
24 in GIS environment. To produce the flood hazard seven flood causative factors were considered,
25 these are the reclassified, Elevation, Rainfall intensity (precipitation), Drainage Density,
26 Elevation, Slope, Distance to River, soil type and Land use/cover map of the study area. Flood
27 risk assessment was done by taking the population and land use land cover as element at risk.
28 Flood hazard analysis for Upper Awash basin indicates that 1.245 %, 20.93 %, 76.6% and
29 1.23% of the area are characterized by Very high, High, Moderate and Low flood hazard level
30 respectively. The result of flood risk analysis shows that 3.8, 34.04, 40.42, 9.43 and 2.31
31 percent of the area considered in the upper awash basin were subjected to very low, low,
32 moderate, high and very high flood risk respectively.

33

34 Keywords

35 Flood Risk, GIS, Flood hazard, Multi criteria Analysis

36

37

38

39

40

41

42

43

44

45

46
47

2
48 1. INTRODUCTION
49 1.1 Background
50 Flood problem has been reported almost everywhere in the world with much more pronounced
51 effects in the developing countries due to their low incomes, poor housing facilities, inadequate
52 warning systems and preparedness generally grouped by (Ayala, 2002) as social, economic,
53 political and cultural vulnerabilities. Large and damaging floods occur every year. Heavy floods
54 visited Pakistan, India and China in the summer of 2010, Colombia from October to December
55 2010 and Australia during the austral summer 2010/11. The maximum estimated annual damage
56 caused by river floods in one country was recorded in China in 2010, where a total loss of US$51
57 billion was reported. In 2010, there were nearly 2000 immediate fatalities from monsoonal
58 flooding in Pakistan ( Syvitski & Brakenridge, 2013)

59 Flood is a natural phenomenon that results in the temporary submerging with water of a land that
60 does not occur under normal conditions European Union (2007). According to this study Floods
61 affect more than 70 million people all over the world and more 800 million people are living in
62 the areas exposed to floods and 3700 floods were recorded in the world from 1985 to 2014,
63 causing the loss of tens of billions USD (Omid, Hossein, & Mosa, 2016). Floods are caused due
64 to numerous factors among them rainfall, geomorphological features, land use/land cover, soil
65 type, drainage density, slope and urbanization (Ibrahim , Lionelle , Tiki , & Gabriel, 2018).
66 These factors could be exacerbated by the fair socio-economic aspect of the population.

67 The most recent cases of floods are those of August 2012, 2014, 2015 and 2017 with negative
68 consequences and damages as such as loss of human beings lives, destruction settlements and
69 plantations, displacement of populations, the resurgence of hydric diseases. In this context, the
70 management of floods integrates assessment of different factors of flooding and mapping of
71 flood risk areas is important.

72 According to the report of the report of United Nations Office for the Coordination of
73 Humanitarian Affairs (2006) indicates that Ethiopia was also experiencing flooding in many
74 parts the country. The figure shows that a total of 357,000 people were affected by flooding in
75 Ethiopia from which 136,528 people were homeless due to the flood. In Amhara region in 2006
76 extreme flooding affects and displaced 43,127 and 8,728 peoples respectively. On 5 August
77 2006, torrential rains caused the Dechatu River to burst its banks, causing serious flash flooding

3
 78 in the region of Dire Dawa. The United Nations Office for the Coordination of Humanitarian
79 Affairs (OCHA 2006), reported that ,000 people were reported as being displaced, 300 missing
80 and 200 dead and the report also indicates that 6000 people were displaced from their living area.
81 In addition, the Disaster Preparedness and Prevention Agency (DPPA) confirmed that floods in
82 South Omo Valley killed 364 people and displaced approximately 6000 – 10,000 more following
83 heavy rains which forced the Omo River to burst its banks and flood five villages in remote
84 southern Ethiopia.
85
86 In Ethiopia 80% of rain is received during the summer rainy season is concentrated from June to
87 September causing an extreme flooding in some part of the countries and the over flow of river
88 causes destruction and debris flood in some areas. Ethiopia embraces flood events annually in
89 different parts of the country and all the above flood related issue is the reason for doing this
90 thesis. The situation requires appropriate flood risk assessment. To do this Remote sensing (RS)
91 and geographic information system (GIS) techniques have made significant contributions.

92 1.2 Objectives
93 1.2.1 General Objectives
94 The main objective of this study is to assess the flood risk of Upper Awash River Basin.
95 1.2.2 Specific Objectives
96 The specific objectives of this study are:

97  To develop different thematic layer of flood factors

98  To develop flood hazard map of the Upper Awash River Basin
99  To develop flood risk map of the upper awash basin
100 .
101
102
103
104
105

4
106 2. MATERIAL AND METHODS
107 2.1 Study Area
108 The Awash River Basin on of the basin in Ethiopia with a geographic location from 8.5⁰ N and
109 12 ⁰ N. the total land area covered by this basin is of 110,000 km2,from this 64 000 km2 of the
110 land area is included under the Western Catchment, drains to the main river or its tributaries. The
111 remaining 46 000 km2, most of which comprises the so called Eastern Catchment, drains into a
112 desert area and does not contribute to the main river course.
113
114 The Awash Basin is traditionally divided into four distinct zones. These are; Upper Basin, Upper
115 Valley, Middle Valley and Lower Valley.
116
117 Upper Awash River Basin is situated in the central Ethiopia in oromia regional state. The Upper
118 Awash River Basins contains sub-basins like hombole sub-basin, Mojo sub-basin, Melka Kuntire
119 sub-basin, Kakaki Subbasin, Bello Sub-basin and Teji sub-basins.
120
121 2.2 data collection

122 2.2.1 Primary data

123 As a primary data ground coordinate has been collected using Global Positioning System in
124 order to identify the location of Bridges which is used as an input for HEC-RAS/HEC-GeoRAS,
125 flood Inundation area and validate the collected ground features of earth’s truth to the classified
126 satellite image of the study area (for image classification Accuracy Assessment).

127 2.2.2 Secondary data

128 The DEM data of 12.5m resolution has been downloaded from Alaska Satellite facility and
129 having downloaded the DEM the study area was extracted and missing data were filled in GIS
130 environment Using a GIS extension Arc hydro tools.
131 The study Area is extracted from two adjacent scene satellite imageries of Path 169 Row 54 of
132 2017 layer stacking has been carried out in order to combine different bands. The images were
133 the mosaicked, the Upper Awash Basin area was subsetted from the mosaicked scene and
134 different image enhancement techniques were applied to extract the land cover information for
135 image classification using ERDAS IMAGINE 2014 software.

5
136
137 The study area LandUse/Landcove has been extracted from LANDSAT imagery obtained from
138 USGS websites and Maximum likelihood supervised classification technique has been carried
139 out using ERDAS imagine software to identify different land use land cover classes. In addition
140 to the satellite image in this study flood discharge data for gauging station located in different
141 part of the study area has been collected from Ministry of Water Irrigation has been for
142 modelling flood inundation using HEC-RAS model.
143
144 2.3.3 Software and materials
145 Please See Error! Reference source not found.
146 2.3 Methods
147 Besides, flood risk occurrence results from natural and anthropogenic factors ( Danumah, Odai,
148 Saley, & Szarzynski, 2016), that means it is a complex phenomenon. For all that, Multi-Criteria
149 Analysis (MCA) approach has become widely used (Wang et al., 2011; Sowmya et al., 2015) to
150 solve complex problems and to assess flood risk. The aim of this study is to assess the efficiency
151 of Multi-Criteria Analysis (MCA) approach to identify and map the potential flood risk areas by
152 using remote sensing and GIS. Flood zonation mapping and hazard assessment implicate
153 multiple criteria and factors geographically related (Booij, 2005; Minea, 2013; Xu et al., 2013).
154 Remote sensing and GIS have shown significant contribution in natural disaster analysis (Omid,
155 Hossein, & Mosa, 2016). However, some studies have been done on flood risk mapping and
156 assessment using remote and GIS in this geographical locality (Leumbe et al., 2015), and had
157 allowed getting a strength database on the different factors causing flooding, but without
158 considering their pondered weight proposed by methods for multi-criteria making decision. The
159 best known and the most widely used is Analytic Hierarchy Process (AHP), developed by (Saaty,
160 2008).

161 The methodology of this study relies mainly on analysis of image acquired from different
162 sources and other data sets. The objective of this thesis is to develop a flood risk map in GIS
163 using Multi criteria Analysis. To determine the flood vulnerable areas, Multi Criteria Analysis
164 (MCA) was used.

6
165 2.3.1 Multi criteria Analysis
166 Multi-Criteria Analysis is a structured decision-making tool used to solve complex problems.
167 MCA describes any deliberate and organized approach used to determine overall preferences
168 among alternative options with a view to ranking them from the most important to the least
169 important (Md. Sadequr Rahman, 2014). The outcome of MCA is a set of weights for the various
170 alternatives.

171 The factors that were used as an input to for multi-criteria analysis was preprocessed in
172 accordance to the criteria set to develop flood hazard analysis. So Eigen vector for the taking
173 flood causative factors like Slope, Drainage Density, Distance to river Rainfall, Elevation, Soil
174 and Land use/Land cover was computed using Analytical hierarchy process (AHP) of Multi-
175 criteria Analysis.

176 In practice, individual decisions are never the same and do not agree perfectly therefore, the
177 degree of consistency achieved in the scaling of the matrix is measured by a Consistency Ratio
178 (CR) which shows the probability that the scaling were generated randomly. A CR less than or
179 equal to 0.1 indicates an acceptable reciprocal matrix and a ratio over 0.1 indicates that the
180 matrix should be reexamined (T. L. Saaty, 2008). Reexamining the matrix involves locating
181 inconsistent decisions with regards to the importance of criteria and revising these decisions by
182 comparing the pair of criteria again.

183 Calculating Consistency Ratio (CR)

184 Consistency Index (C.I) = (Λmax –n)/(n-1) where n is number of compared factors
185 Consistency ratio (C.R) = (C.I)/ Random index (R.I)
186
187 2.3.2 Flood Hazard Mapping
188 Main causative factors for the occurrence of flooding Hazard in the study area were identified
189 based on the existing realities and knowledge of the study area.
190
191 Flood hazard analysis of Upper Awash basin was carried out using weighted sum overlay
192 analysis of the seven reclassified flood causative factors (i.e. reclassified rainfall, reclassified
193 drainage density, reclassified slope, reclassified distance to river, reclassified elevation,

7
194 reclassified soil and reclassified landuse Layers using Multi-criteria Evaluation technique (MCE)
195 using model builder in ArcGIS environment.
196
197 2.3.3 Flood Risk Mapping
198 Flood risk analysis was carried out by applying the following general risk equation (Shook,
199 1997).
200 Risk = (Elements at risk)*(Hazard*Vulnerability)………..………….………Equation 1.1
201 Before flood risk analysis the flood hazard analysis was performed using multi criteria
202 evaluation (MCE). To carry out the M , seven flood disaster causative factors such as elevation,
203 slope, drainage density, Distance to river, Soil, Rainfall and land-use were developed and
204 weighted. Then weighted overlay technique was computed in ArcGIS10.7 Model Builder to
205 generate flood hazard map.
206
207 The vulnerability is assumed to be one by considering the degree of loss to be total for the study
208 area. Finally to generate flood risk map of the study area prepared by overlaying flood hazard
209 map with two elements at risk layer (land use and population density) using weighted sum
210 overlay analysis technique in ArcGIS 10.7 environment by assigning equal weight for each
211 inputs.
212

213 3. DATA ANALYSIS AND RESULT

214 3.1 Flood Hazard Assessment

215 At data analysis stage, all the factor layers are ready to be combined in order to assess the flood
216 hazard zones in the study area. If all datasets were equally important, it could be possible to
217 combine them simply. However, from the principal eigenvector calculation, the relative
218 importance of each parameter was determined. Therefore, the higher the weight, the more
219 influence a particular factor will have in the flood generation. Accordingly, the factor layers were
220 combined by applying the following formula in the raster calculator of spatial analyst extension
221 in ArcGIS environment. It was done systematically using ArcGIS model builder.
222

8
223 To produce the flood hazard map weighted sum overlay has been carried out in a GIS
224 environment. This process requires a spatial data for the criteria being used for the weighted
225 overlay process. The criteria that are considered as flood causative factors in this thesis are;
226 Elevation, Rainfall intensity (precipitation), Drainage Density, Elevation, Slope, distance to
227 River, soil type and Land use33/cover. Raster maps for elevation, slope, and drainage are derived
228 from ALOS PALSAR DEM which has a resolution of 12.5m. A raster map is created for
229 precipitation through surface interpolation Inverse distance weight (IDW). Soil Map has been
230 collected from Ethiopia Ministry of water irrigation and energy.

231 The following flood causative factors were combined for the development of flood hazard map
232 by using Model builder in ArcGIS environment.

233 3.1.1 Slope

234 The slope raster was derived from the ALOS PALSAR DEM using the ArcGIS Spatial Analyst
235 extension of surface module, which enabled to classify the area according to the steepness and
236 the gentleness of the terrain.
237
238 Then the slope raster was reclassified in to five sub group using the natural breaks (jenks)
239 classification and the frequency of slope degree in the study area on the histogram The
240 reclassified slope was given a value 1 to 5 with the higher value of 5 showing high influence,
241 resulting in very high flood rate, while the lower value of 1 showing very low influence,
242 resulting in very low flood rate.
243 The break values and the description of the new slope classes, See Error! Reference source not
244 found. below

245 3.1.2 Drainage Density

246 Drainage density is an inverse function of infiltration . Greater drainage density indicates high
247 runoff for basin area along with erodible geologic materials, and less prone to flood. Thus the
248 rating for drainage density decreases with increasing drainage density.
249 DEM was used to extract the drainage network, to calculate the drainage density of the streams.
250 Arc Hydro10.7 software, which works as an extension on ARC GIS 10.7 version software was
251 used to generate drainage network map of the sub-basin
252

9
253 3.1.3 Rainfall Data
254 The annual average of the rainfall data from 1988-2018 has been downloaded from the center of
255 Hydrometeorology and Remote Sensing (CHRS) Portal developed by University of California. It uses
256 neural network function classification/approximation procedure to compute an estimate of rainfall rate
257 at each 0.250 *250 pixel of the infrared brightness temperature image provided by the geostationary
258 satellites. The PERSIANN (Precipitation Estimation from Remotely Sensed Information using Artificial
259 Neural Networks) system was based on geostationary infrared and later extended to include the use of
260 both infrared and daytime visible imagery. The reclassified rainfall was given a value 1 to 5 with the
261 higher value 5, showing high influence resulting in very high flood rate, while the lower value 1,
262 showing very low influence resulting in very low flood rate.
263 3.1.4 Elevation Factor

264 The raster Elevation was derived from filled DEM which has a resolution of 12.5 in a GIS
265 environment using 3D analyst tool. The resulting raster had elevation values. Then the elevation
266 was reclassified into five groups using equal interval scheme. Based on their susceptibility to
267 flooding, the lower the elevation value the higher the flood hazard was and the higher the
268 elevation value the lower was the flood hazard. The rank and the level of hazard of the elevation
269 classes are given in Error! Reference source not found. below.

270 3.1.5 Distance to River

271 The distance to active channel is one of the contributing factor in flood hazard assessment.. The
272 distance to the active channel has been computed by buffering on at 500 m both sides of the
273 channels with spatial analyst and 3–D analyst. After buffering, the values were divided into four
274 classes with respect to hazard assessment namely, very high (0-1000m), high (1000-1500m),
275 moderate (1500-2000m), low (2000-300) and very low (>3000m) based on our field experience.

276 3.1.6 Land Use/ Land Cover Factor

277 The land use of the study area was classified from the subsetted image using two adjacent scene
278 of Landsat 8 acquired on 01/10/2017 and 01/17/2017, having path and row of 169/54. Then the
279 Land use/Land cover class has been identified from the subsetted image using one of the
280 classification techniques supervised classification with the aid of 250 sample points using
281 signature editor.

10
282 Classified pixels were clustered into the following five general categories: Built-up Area,
283 Agricultural lands, forest, barren-land and water bodies. As shown in figure below. Then each of
284 the land use types was further reclassified for the overlay analysis. Accordingly, Built-up area,
285 was given more weight, which is equal to 5, water bodies was given weight 4, Barren land was
286 given weight of 3, agricultural land was given weight 2 and Forest was given weight of 1.
287

288 3.1.7 Soil

289 The soil factor of the study area was derived from the FAO standard classification of Ethiopian
290 soil. The characteristics of each soil group are analyzed based on hydrologic soil grouping
291 system. Accordingly, the soil group of the study area was grouped into five general classes and
292 converted to raster format. Further, the soil raster layer group was reclassified into five groups.
293 And new values reassigned in order of their flood hazard rating. Soil type that has very high
294 capacity to generate very high flood rate is ranked to 5 and the one with very low capacity in
295 generating flood rate is ranked to 1; therefore, Pellic Vertisols are ranked to 5, Chromic Vertisols
296 are ranked to 4, Chromic Luvisols are ranked to 3, Euthric Nitosols are ranked to 2, and
297 Lithosols are ranked to 1
298 .
299 3.2.2 Pairwise Comparison Method
300 The criteria for analysis are the element which deemed to have causative influence over flooding
301 in the basin. Matrix is constructed for these flood causative factors based on Eigen vector
302 principles. The computed weight value of each flood causative factors is given below.

303 𝚲max = 7.49

304 Consistency Index (C.I) = ( 𝚲max –n)/(n-1) where n is number of compared factors
305 = ( 𝟕. 𝟒𝟗 –7)/(7-1)

306 = 0.0817
307 Consistency ratio (C.R) = Consistency Index (C.I)/ Random index(R.I)
308 Where R.I is obtained from Table
309 The factors being compared in this analysis are seven and the corresponding R.I value is 1.32
310 Consistency ratio (C.R) = 0.0817/1.32
311 = 0.061 which is acceptable

11
312 3.3 Weighted Overlay Analysis
313 The weighted overly tool in spatial analyst is applied for the analysis. The six reclassified raster
314 layers were weighted based on the weights of the pairwise comparison process and then
315 combined based on the assigned rankings used in the reclassification. If all datasets were equally
316 important, it could be possible to combine them simply. However, from the principal eigenvector
317 calculation, the relative importance of each parameter was determined. Therefore, the higher the
318 weight, the more influence a particular factor will have in the flood generation.
319 3.4 Flood Risk Assessment
320 Flood risk analysis is computed by Weighted Overlay setting equal importance to all factors as
321 shown in the following table. The developed factors for overlay are Flood hazard analysis layer
322 and the two elements at risk in the study area these are Population density and land use for flood
323 risk analysis.
324 3.4.1 Population Density
325 To quantify the assets under potentia threat, the woreda wise population density has been chosen
326 as an important variable. Woreda shape of upper Awash basin is updated with population data
327 from census of (2015) which is downloaded from the website of Ethiopian Central statistics
328 Agency. Gross population density calculation method is used to calculate the number of person
329 per square kilometers for each woredas. Then further the data layer was reclassified into fiver
330 sub-factors which are classified using equal interval method. And new values re-assigned in
331 order of increasing number of population that is more susceptible to flood hazard.

332 4. DISCUSION
333 4.1 Flood Hazard
334 In this study, the flood hazard zone of the study area was delineated by considering the seven
335 parameters. The final combined resulting map obtained after the weighted overlay analysis was
336 flood hazard map of the study area Fig below. From this map, the different levels of flood hazard
337 in the study area was calculated and given in Table below. The result has revealed that, 1.245 %
338 of the basin was characterized by very high flood hazard level and 20.93 % of the study area was
339 characterized by high flood hazard level. On the other hand 76.6% of the study area fall under
340 medium hazard level zone. The rest of the study area which is 1.23% falls under a level low
341 hazard. On the basis of historical information, the upper part of Upper Awash basin becho

12
342 Woereda was the most commonly inundated area due to its geomorphic settings as it comprises
343 of topographically flat land.
344 Areas of different Woredas in the Basin that are subjected to different flood hazard levels were
345 also calculated. The further analysis for flood hazard for different woredas of the upper awash
346 basin is shown inError! Reference source not found. below.
347 Flood hazard level for different land use land cover class revealed that agricultural land is more
348 subject to flood hazard. The further analysis for flood hazard in each land use land cover class is
349 given in the following table Error! Reference source not found..below

350 4.2 Flood Risk

351 Flood risk analysis is computed by Weighted Overlay setting equal importance to all factors
352 (Fig). The developed factors for overlay are Flood hazard analysis layer and the two elements at
353 risk of the study area which are Population density and land use for flood risk analysis
354 (table).According to the flood risk map (Fig ), it was estimated that 3.8, 34.04, 40.42, 9.43 and
355 2.31 percent of the area considered in the upper awash basin were subjected respectively to very
356 low, low, moderate, high and very high flood risk. Further analysis for element at risk and risk in
357 each woreda is given in the following tables
358 Further analysis for flood risk assessment revealed that 0.003 % of forest, 0.51% of Agriculture,
359 1.16% of Barren Land and 2.1% of Built-up area of the upper awash basin fall under very high flood
360 risk level, and it is also observed that 0.17% forest, 18.42% Agriculture, 2.82% barren land and
361 2.18% Built-up land use land cover of the study area are classified in to high flood risk levels.

13
362 5. CONCLUSION

363 Flooding in Upper Awash causes a considerable damage to life and property. The damage is
364 more pronounced in Agricultural Land which is a large coverage of the study area and people
365 living in the surrounding area of Awash River. The damage is mainly caused during the rainy
366 season of the country.

367 The decision factors identified are rainfall, population Density, distance to active channel,
368 landuse land cover, slope of the area, Elevation, drainage density and Soil of the study area.
369 Thematic maps for each factors were prepared using several image processing techniques and
370 GIS operation at different scale. Each of the thematic layers (Classified data sets) were brought
371 to same scale.

372 Analytical Hierarchy Process (AHP) is multi criteria decision making techniques which provide
373 a systematic approach for assessing and integrating the impacts of various parameters, involving
374 several level of dependent and independent, qualitative as well as quantitative information. The
375 weightage has been assigned taking into the consideration the flood hazard and vulnerability
376 potential of the area, which ultimately lead to the risk zonation. Using ARC-GIS model builder,
377 Overlay analysis of all the thematic layers was carried out to generate separate Flood hazard
378 Map, Flood Vulnerability map and their combination to produce final risk map. The point to be
379 considered during overlay analysis is that all the thematic layers should have discrete values
380 rather than continuous values for assigning weightage. The format of the thematic layers should
381 be in GRID to suit the current framework of AHP analysis..

382 The assessment of the flood Risk due to the flooding was made with regard to element at risk
383 which area the land use pattern in the flood areas and population density per Sqr Km with
384 integration of flood hazard analysis of the study area. The assessment of the flood risk
385 assessment indicates that a large percentage (more than 84 %) of vulnerable area lies in flood
386 plain area i.e. agricultural land followed by barren land and built up area 10.6% and 1.36%
387 respectively.

14
388 Flood hazard map were validated using GPS data collected in the field for places with Flood
389 inundation. The results are satisfactory and it validates the logic followed in the analysis and the
390 model developed. It has been observed that in Hazard map the dominant factor is slope followed
391 by distance to river, drainage density, Rainfall, Elevation, soil and LULC, which ultimately
392 contribute to the flooding of the study area. Population density and Land use Land cover is
393 considered as contributing factor to the vulnerability of the study area. Hence, the final risk map
394 generated by combining all the dominant factors as well as the contributing factors to flood risk.

395
396

397

398

399

400

401

402

403

404

405

406

407

15
Figures

16
17
18
19
20
21
Figures captions

Figure 1 Study Area

Figure 2 Reclassified Slope Map

Figure 3 Reclassified drainage density

Figure 4 Reclassified Rainfall Map
Figure 5 Reclassified Elevation value
Figure 6 Reclassified distance to River Map
Figure 7 Supervised classification LULC

Figure 8 Reclassified Soil type Map

Figure 9 Reclassifie population density map
Figure 10 Flood Hazard Map of Upper Awash basin
Figure 11 Flood Risk Map of Upper Awash

22
List of Tables
Table 1Data source and Material

NO DATA Source
1 Satellite Image LANDSAT
2 DEM aster website
3 Soil map ethio_soil map by extracting
4 Population Central statistics agency
5 rainfall data Ethiopian Meteorology

Software and hardware’s

6 -
ArcGIS version 10.7
7 Microsoft Excel -

Table 2 Slope classification and level of hazard

Slope (Degree) Class Rank Level of Hazard

0-3 Flat 5 Very High
3-8 Undulating plain 4 High
8-14 Undulating to rolling 3 Moderate
14-23 Rolling to hilly 2 Low
23-74 Mountainous 1 Very low

Table 3 Drainage density reclassification value

Drainage Density(KM/KM2) Rank Level of Hazard

0-0.071 1 Very Low

0.071-0.174 2 Low
0.174-0.261 3 Moderate
0.261-0.361 4 High
0.361-0.825D 5 Very High

23
Table 4 Reclassified rainfall data value

Rainfall in mm Rank Level of Hazard

877.02-972.48 1 Very Low

972.48-1067.94 2 Low

1067.94-1163.39 3 Moderate

1163.39-1258.86 4 High

1258.86-1354.32 5 Very High

Table 5 Elevation class hazard level

Elevation Rank Level of Hazard

1538-1944 5 Very High

1944-2350 4 High

2350-2756 3 Moderate

2756-3162 2 Low

3162-3568 1 Very Low

Table 6 Weight computed by Eigen vector principles

Factors Slope D/ce-River Drng dnsty Rainfall Elevation Soil LULC

Weight 0.27 0.21 0.17 0.13 0.11 0.06 0.05

24
Table 7 Factors for food risk analysis

Factors Weight Sub-factors Ranking

very low 1
low 2
Flood hazard 0.3333 moderate 3
high 4
very high 5
108 – 243.7 1
243.7 – 1454.4 2
Population density 0.3333 1454.4– 3722.6 3
3722.6 – 5261.3 4
5261.3 – 9903.0 5
Forest 1
Agriculture 2
Land use types 0.3333 Water Bodies 3
Barren Land 4
Built-up Area 5

Table 8 Flood hazard level of Upper Awash basin

Hazard level AREA_HA_ Area(%)

very low 73123.89 6.696534
Low 185584.4 16.99543
Moderate 296448.2 27.14811
High 318087.8 29.12982
Very High 218722 20.03011

Table 9 Cross tabulated area of land use class by flood hazard level

LEVEL FOREST AGRICULTURE WATERBODIES BARRENLAND BUITUP

very high 6188.76 5418.59 0.33 1030.67 1032.13

High 17543.46 147509.73 169.41 41346.98 23301.22

Medium 5825.78 670031.11 16979.45 87028 61567.41
Low 2.75 5012.98 125.36 2983.03 5405.27

25
Table 10 Flood risk level for different Land use type

FLOOD RISK LEVEL (Sq. Km)

LULC VERY LOW MEDIUM HIGH VERY

SN LOW HIGH
1 Forest 174.28 60.79 31.28 17.94 0.35

2 Agriculture 1465.64 2244.30 2415.83 1991.69 55.26

3 Water Bodies 3.16 38.9713 123.12 4.72 1.110

4 Bare land 130.07 331.93 397.47 305.27 125.0

5 Buitup 76.53 149.61 206.94 235.95 227.12

26
 Reference
Alcántara-Ayala, I. (2002). Geomorphology, natural hazards, vulnerability and prevention of
natural disasters in developing countries. Geomorphology, 47(2-4), 107–124.
https://doi.org/10.1016/s0169-555x(02)00083-1

Booij, M.J. 2005. Impact of climate change on river flooding assessed with different spatial
model resolutions. J Hydro., 303, pp.176-198.

Brakenridge, G. R., Syvitski, J. P., Overeem, I., Higgins, S. A., Kettner, A. J., Stewart-Moore, J.
A., & Westerhoff, R. (2012). Global mapping of storm surges and the assessment of
coastal vulnerability. Natural Hazards, 66(3), 1295–1312. https://doi.org/10.1007/s11069-
012-0317-z

Danumah, J.H., Odai, N.S., Saley, B.M. and Akpa, L. 2016. Flood risk assessment and mapping
in Abidjan district using multicriteria analysis (AHP) model and geoinformation
techniques, (Côte d’Ivoire). Geoenvironmental Disasters, 2016. Available from:
https://geoenvironmental-disasters.springeropen.com/articles/10.1186/s40677016-0044y

Ethiopia: Dire dawa floods - ocha situation report no. 1 - ethiopia. ReliefWeb. (2006, August 7).
Retrieved July 13, 2022, from https://reliefweb.int/report/ethiopia/ethiopia-dire-dawa-
floods-ocha-situation-report-no-1

Ethiopia: Dire dawa floods - ocha situation report no. 4 - ethiopia. ReliefWeb. (2006, August 7).
Retrieved July 13, 2022, from https://reliefweb.int/report/ethiopia/ethiopia-dire-dawa-
floods-ocha-situation-report-no-1

Ethiopia: South omo valley floods ocha situation report no. 1 - ethiopia. ReliefWeb. (2006,
August 17). Retrieved July 13, 2022, from https://reliefweb.int/report/ethiopia/ethiopia-
south-omo-valley-floods-ocha-situation-report-no-1

Gebre SL, G. Y. S. (2015). Flood hazard assessment and mapping of flood inundation area of the
Awash River basin in Ethiopia using GIS and Hec-GeoRAS/HEC-Ras Model. Journal of
Civil & Environmental Engineering, 05(04). https://doi.org/10.4172/2165-784x.1000179

Gebre SL, G. Y. S. (2015). Flood hazard assessment and mapping of flood inundation area of the
Awash River basin in Ethiopia using GIS and Hec-GeoRAS/HEC-Ras Model. Journal of
Civil & Environmental Engineering, 05(04). https://doi.org/10.4172/2165-784x.1000179

Leumbe O., Bitom L., Mamdem L., Tiki D. and Ibrahim A. 2015. Cartographie des zones à
risques d’inondation en zone soudano-sahélienne: cas de Maga et ses environs dans la
région de l’extrême-nord Cameroun. Afrique Science, 11(3), pp.45-61.

Minea, G. 2013. Assessment of the flash flood potential of Bâsca river catchment (Romania)
based on physiographic factors. Cent Eur J Geosci., 5, pp.344-353.

27
Omid, R., Hossein, Z. and Mosa, B. 2016. Flood hazard zoning inYasooj region, Iran, using GIS
and multi-criteria decision analysis. Natural Hazards and Risk, 7(3), pp.1000-1017

Saaty, T. L. 2008. Decision Making With the Analytic Hierarchy Process. International Journal of
Services Sciences, 1(1): 83-98.

Saaty, T. L. (1994). how to make a decision the analytic hierarchy process. interfaces, 24, 19-43.
- references - scientific research publishing. (n.d.). Retrieved July 13, 2022, from
https://www.scirp.org/(S(351jmbntvnsjt1aadkposzje))/reference/referencespapers.aspx?ref
erenceid=1951458

Shook, G. 1997. An Assessment of Disaster Risk and Its Management. Disasters, 21(1): 77-78.

Sowmya, K., John, C.M. and Shrivasthava, N.K. 2015. Urban flood vulnerability zoning of
Cochin City, southwest coast of India, using remote sensing and GIS. Nat Hazards, 75,
pp.1271-1286.

Wang, Y., Li, Z., Tangand, Z. and Zeng, G. 2011. A GIS-Based Spatial Multi-Criteria Approach
for Flood Risk Assessment in the Dongting Lake Region, Hunan, Central China. Water
Resources Management, 25: pp.3465-3484.

Xu, C., Chen, Y., Zhao, R. and Ding, H. 2013. Responses of surface runoff to climate change
and human activities in the arid region of Central Asia: a case study in the Tarim River
Basin China. Environ Manag., 51, pp.926-938.

https://www.researchgate.net/publication/333618356_Geospatial_Based_Flood_Risk_Assessme
nt_The_Case_of_Kewet_District_Amhara_Region_Ethiopia

https://www.researchgate.net/figure/Drainage-density-classes-and-weights_tbl1_299978233

https://www.statsethiopia.gov.et/census-2007-2/

https://www.researchgate.net/figure/Saatys-nine-point-pairwise-comparison-
scale_tbl1_306500145

28
Conflict of Interest

Conflict of Interest

There is No conflict of interest