Shallow Lakes Water Level Estimation using Satellite Optical Imagery and Digital Elevation Models Over the Persian Plateau, Iran

Lakes' water levels have a dynamic behavior, and their variations are an essential subject for water resources research and management. These variations have a wide range of time scales, from short-term (daily) to long-term (yearly) scales. However, access to hydrological data is limited due to scarce observation stations, fragmented data holdings, and low data quality in developing countries. Satellite altimeters are considered the main source of water level estimation among remote sensing data. Although many seas and oceans are covered by altimetry satellites, currently, they have a huge gap in covering inland lakes. Accordingly, we proposed an alternative approach to estimate shallow lakes' water levels using typical optical imageries and digital elevation models. The water level is estimated based on the Area Elevation Model (AEM) approach, using MODIS surface reflectance product, ALOS DSM and Landsat JRC product as inputs to the model. The AEM helps extract the water level time series based on the information about water area obtained from satellite products using various spectral indices (NDWI_GNIR, NDVI, NDWI_RSWIR and MNDWI). The methodology was applied to eight shallow lakes in Iran using Google Earth Engine (GEE) platform. These lakes are located across the arid and semi-arid regions of the Persian Plateau, Iran. The lakes' water level in these regions is declining, and there is a great need for taking important measures by regional authorities for sustainable water management. Spectral indices and the effect of satellite resolutions were evaluated. Overall, this methodology can be the alternative approach for water level estimation for lakes with minimum or no ground observation and altimetry coverage.