An integrated new approach for optimizing rainwater harvesting system with dams site selection in the Dewana Watershed, Kurdistan Region, Iraq.

Water scarcity in Kurdistan-Iraq has become a crucial problem, particularly in semi-arid regions, as a result of severe droughts over the last decades. One potential solution to this water shortage is using rainwater harvesting (RWH) techniques. In this study, optimal sites of RWH in the Dewana watershed were identified using a combination of remote sensing (RS) and geographic information system (GIS), with multi-criteria decision analysis (MCDA) models, including analytical hierarchy process (AHP) and weighted sum method (WSM). Sixteen thematic layers are used. As a result of the AHP and WSM models, 236.89 km2 and 267.15 km2 were identified as highly suitable areas for RWH techniques in the suitability index map. They identified 13.06 km2 (5.55%) and 58 km2 (21.81%) as highly suitable for constructing dams in the dam site selection maps. The present study found that 11 proposed dam sites are suitable for dam construction. The weighted product model (WPM) was used to rank the proposed dam sites, with Dams #10 and #2 being the top-ranked sites. Accuracy assessment results indicated that the WSM model outperformed the AHP model with an overall accuracy rate of 50.5% and 52.78%, respectively. However, the AHP model demonstrated a higher receiver operating characteristic (ROC) and an area under the curve (AUC) score of 1.00, while the WSM model had an AUC of 0.78.