investigation of flooding and causative factors in balegli chay watershed by gis, rs, and ahp techniques

introduction floods are major disasters worldwide that causes serious damage to agriculture, fisheries, housing and infrastructure. they can also disrupt socio-economic activities. as the severity and frequency of flood events have considerably increased, there is a growing global concern about the need to decrease flood related fatalities and associated economic losses. just in 2010, 178 million people were affected by floods and the total financial losses in the exceptional years such as 1998 and 2010 exceeded $40 billion. this study has presented an exhaustive methodology of using geographical information system (gis), satellite imagery interpretation, rain gauge station data and analytical hierarchy process (ahp) method of flood hazard mapping for surveying flood hazard in balegli chay watershed and sub-basins. the study area is located in the northwest of iran, in the west of ardabil and east of azerbaijan sharghi provinces where the balegli chay is located. this river is one of the most important rivers in the azerbaijan, because its sub-watershed is one of the vulnerable areas to flood damages. also this river passes through the middle of the ardabil city where there are many communication lines, infrastructure, and recreation areas next to this river. the balegli chay watershed is one of garasu sub-basins, which is drained from alborz mountain range. this basin has an area about 1036 km2 in the upstream of polehalmas hydrometer station. high mountain ranges and harsh slope changes at the foot of mountains create special hydro-geomorphical characteristics for the watershed and due to these characteristics, risk of flooding is high. the savalan mountain with 4811 meters elevation is located in the northwest. a flat plain that called ardabil plain with 1432 meters elevation is located in the middle parts and downstream of the balegli chay basin. the average elevation of the study area is 2014 meters above mean sea level (m.s.l) and ranging from 1424 to 4811 meters above m.s.l. in this study, flood hazard map was created based on the integrated effects of physical and climatic factors. first, eight physical factors including size of sub-basin, slope, aspect, time of concentration, drainage density, bifurcation ratio, miller coefficient, curve number and two climatic factors including annual rainfall and the maximum daily rainfall were used to produce the flood hazard map. arcgis application was used to produce layers of the factors, along with the spatial conditions of the study area obtained by field investigation and satellite images. materials and methods in order to determine weights of causative factors, ahp method based on the data collected from expert responses by questionnaire has been adopted in this research to quantify the weights given to each component and subcomponent. the pairwise comparison matrix makes pairwise comparisons of the criteria. this method applies basic scale with 1 to 9 values to determine the extent of relative priorities of two criteria. weighted linear combination method (wlc) is the most common technique in analyzing the multi-scale evaluation. in this method, decision making principle calculated the value of each ai option by equation 1:                                                                                                                                                         (1) where, wj is weight of j criterion; xij is a value asigned in place i in relation to j criterion. in other words, this value can indicate appropriate degree of i in relation to j criterion; n is total number of criteria and aj is a value which finally will attach to i location.  saaty considered consistency tests for pair evaluation, including consistency index (c.i.) and consistency ratio (c.r.) (eq. 2 and 3).                                                                                                                                      (2)                                                                                                                                                   (3) where, λmax is the maximum eigenvalue of the matrix, n is the matrix rank, and r.i. (random index) is the consistency index of the random matrix. saaty regarded the comparison being randomly generated when the c.r. approached 1 and the consistency being higher when c.r. approached 0. in general, c.r. ≤ 0.1 was considered acceptable, while c.r. > 0.1 showed the inconsistency and they had to be re-compared. expert choice software was used to analyze the numerical values of inundation factors. with the input values in pairwise comparison and weights calculated, consistency ratio (cr) was found as 0.08. this indicated a reasonable level of consistency in the pairwise comparison of the factors. after calculating the weight among factors in the hierarchies, the weight of the overall hierarchy should also be calculated. the weights acquired by ahp were linked to the shacks as attribute data in the arcgis. the total weights of every contributing factor were calculated by ahp.  the maps of all factors were overlaid in arcgis to create raster maps of flood vulnerable areas with 10*10 meters cell size. the pixels with the highest weights in each raster indicate high flood hazard risk areas. finally, the combined map shows that the areas with higher pixel values are more susceptible to flooding. results and discussion the produced layers of flooding factors in arcgis and flooding weight map of balegli chay watershed are shown in figure1. the results show that in south and northeast of the study area, influence of these factors are very high whereas, in the middle and outlet parts of the watershed it is slight.   figure 1. produced layers in arcgis and flooding weight map of balegli chay watershed   conclusion according to the results it can be indicated that the sagezchi chay (h) sub-basins in the south of the watershed is in the first rank in term of flood occurrence. in contrast, sub-watershed of latran chay (a) in the east of the watershed is in the lowest rank of flooding. these results are greatly dependent on combined and integrated effectiveness of physical and climatic factors on flood production. the results of such studies are quite helpful in flood control projects and assessment of the causative factors on flooding in the watershed for more efficient management practices.