The efficacy of multivariate regression models and GIS in Selecting SuitableSites for Rain Water Harvesting (Case Study: Tajareh Watershed)

Water scarcity in arid areas is a serious crisis. The most important step in using rainwater collection systems is to locate suitable areas. In this research, three methods of multivariate regression model and GIS have been used to locate the on-site and off-site rainwater collection method in Tejreh watershed. In this study, canopy, litter, rock and gravel, bare soil, CN, precipitation, slope and soil depth as independent variables and influence on in situ rainwater collection and maximum instantaneous discharge for non-in situ rainwater collection method The title of the dependent variable was considered. The multivariate regression model uses stepwise method, backward removal method, and forward method. And the standard step-by-step method, regression removal method, step-by-step method in collecting rainwater, non-in situ method have been used. The final results by matching the results of previous research show in step rainwater collection, stepwise method and between layers CN, soil, percentage of rock and gravel, and in non-in situ rainwater collection stepwise regression method Standard and among layers the percentage of litter, percentage of canopy, CN, slope, percentage of rocks and pebbles, amount of rainfall, percentage of bare soil and soil depth are known to be important in the equation. Finally, the importance of rain collection sites was divided into four classes: very good, good, medium and poor.