Monitoring Changes in Soil Surface Moisture by Analyzing the Time Series of Landsat 8 Data in Gavdare Watershed, Kurdistan Province

Accurate information about the amount of moisture and its fluctuations can provide a suitable solution for preparing soil surface moisture maps, predicting the occurrence of soil storms and dust, forecasting floods, droughts, and other climatic phenomena, determining the irrigation time table and the grazing season. In this study, the Perpendicular Soil Moisture Index (PSMI) extracted from Landsat 8 images was used to estimate and determine surface soil moisture. Using field data, the relationship between this index and soil moisture was determined. After radiometric and geometric corrections of images and normalization of field data, processing of satellite images and extraction of PSMI index were performed. Simple regression analysis between field data and index values ​​was performed at different times. Then, the accuracy of produced maps was determined with statistical indicators including coefficient of correlation (R2), root mean square error (RMSE), mean absolute error (MAE) and Nash Sutcliffe efficiency coefficient. The results showed that estimated values with the PSMI index are completely dependent on the season and vegetation status. In the growing season and with high greenness coefficient, the accuracy of estimation was high. In dormant seasons, vegetation had a weak and moderate correlation with field data. The average coefficient of R2 in all sampling times was about 0.65. From the findings of this study, it can be concluded that this index can be used to monitor soil moisture in areas with suitable vegetation, irrigation planning in agricultural areas, and the beginning and end of the livestock grazing season.