Identification of suitable sites for flood spreading by integrating fuzzy logic and GIS & RS techniques

Flood control and storage are the main factors in combating desertification especially in arid and semi arid regions. This study aims at recognizing suitable sites for flood spreading by coupling fuzzy inference systems and RS & GIS techniques. In this resarch, Davarzan watershed selected as a one of the most flood prone area in northeastern Iran because of its strategic location in upstream of Etka agricultural regions. Various thematic maps of the study area including slope map, surface infiltration, pollution generating potential of each formation type, runoff volume and aquifer depth on 1:50000 scales were selected as inputs and analyzed using ARC GIS 9.3 package. Land use map on 2002s ETM image after field measurement and identifying training sites in ENVI 4.8 software is also extracted. For weighting and integration of information layers, any techniques such as index overlay, Boolean logic and fuzzy logic model has been used. Results showed that fuzzy gamma 0.9 and index overlay with maximum overlaying surface with the control sites (5 and 4% respectively) are the most appropriate methods. Screening output maps, it is revealed that alluvial fan located in upstream of Etka agricultural regions was unsuitable. It is mainly due to accelerated erodibility of formations PLc and K1l, lying on the outlet section and resulted in aggregated marl sediments on the alluvial fan which made it unsuitable for flood spreading.