Monitoring changes in natural ecosystems is considered essential to natural resource management. Despite the global importance of the lakes' quality monitoring, there is currently a research gap in the simultaneous predictive modeling of lakes' land-use changes and ecosystem measurements. In the present study for projecting the water bodies of lakes and their surrounding ecosystems, the land-use changes and the landscape analysis of different periods, i.e., 1987, 2002, 2018, and 2030, are studied using remote sensing data and various metrics. The trend of land-use and landscape changes is projected for 2030. The results indicate significant degradation of rangelands and forests due to the conversion to agriculture and construction and the declining trend of lakes' water body and their transformation to salt lake and salt lands. The increase of agricultural lands and the overuse of groundwater wells upstream of the lakes could be one of the reasons for this decline. Decreasing the lakes' water body and subsequently increasing salt lands are considered a severe threat to human health and the ecosystem services of the lakes. Besides, the dust generated by salt lands could also decrease crop yield in the study area.
Keywords: Artificial intelligence; Lake degradation; Land-use change; Landscape metrics; Machine learning; Remote sensing.