Analyzing the variations in the water surface area of Taleqan Dam of Iran using ground-based and satellite observations

Abstract

Accurate and long-term series surface water extraction from dams are of great significance for developing the agroecosystem and sustainable management of water resources. By expanding the remote sensing data with an excellent spatial-temporal resolution, it is possible to monitor various hydrological parameters such as dams’ water level and surface area. This study provides a quick and robust method for open-surface water detection based on the Google Earth Engine (GEE) platform in Taleqan Dam. For this purpose, 90 Sentinel-2 images from September 2015 to November 2019 were used. First, the NDWI index separated water pixels from non-water pixels. Then binary segmentation was conducted to extract water pixels from NDWI images to calculate the water surface area. Comparing two time series of water surface area results obtained from Sentinel-2 images with in-situ data showed RMSE values of 0.43 km$^{\mathbf{2}}$, a correlation of 93.64%, and an RRMSE value of 4.09%. Also, the results showed that the most significant increase in the water surface area with a value of 1.22 km}$^{\mathbf{2}}$ occurred from June 27, 2018, to August 1, 2018, and also the most significant decrease. It also happened with 1.75 km}$^{\mathbf{2}}$ from May 23, 2017, to June 12, 2017. On the other hand, the variations average of the water surface area with a value of -0.03 km}$^{\mathbf{2}}$ shows a slight decrease in the water surface area of Taleqan Dam in the study period.