Spatial Analysis of Soil Moisture after Excessive Normal Precipitation of 1997-98 with Linear Modeling of Environmental Variables and Satellite Images

Understanding the behavior of phenomena requires attention to all its dimensions, and one way to understand their complexities is modeling. Soil surface moisture is a key variable for describing drought, water, and energy exchanges between Korea and the air, as well as assessing crop conditions. Soil moisture is affected by both environmental variables and many environmental variables such as runoff, soil erosion, and crop production, but is highly variable due to unstable spatial and temporal conditions. The purpose of this paper is to investigate, extract and evaluate the spatial model of soil moisture dispersal after more than normal rainfall in 1979-98 in Kurdistan province. In this regard, after analyzing soil moisture dispersion as dependent variable and precipitation variables, snow water equivalent, topographic moisture index and vegetation index were selected as independent variables. Then, using a general regression model (OLS) and geographically weighted regression (GWR), spatial modeling was performed. Based on the evaluation criteria, the results showed that the GWR model with R2 = 0.74 has better explanatory power and better estimation than the general regression model with R2 = 0.68. According to the results of the GWR model, snow water equivalent variable in the northern mountainous regions had the highest effect on evapotranspiration and the least effect on soil moisture. The obtained spatial model can identify low or moist soil areas in order to identify environmental potentials and improve decision making, allocation and spatial distribution of agricultural services.