There are extensive and fragile mountain-oasis-desert ecosystems in Central Asia. Lakes are not only the most important water resources in arid Central Asia, but also important resources for maintaining the unique ecosystem of the region. As most lakes in Central Asia lack in situ monitoring data, and satellite altimetry records are relatively short-term, the continuous long-term monitoring of lake water storage change (LWSC) in Central Asia is inadequate. To this end, Shuttle Radar Topography Mission digital elevation model data were used in conjunction with all available Landsat images to construct an empirical model of area–elevation–storage changes and estimate the maximum LWSC (maxLWSC) and minimum LWSC (minLWSC) of 9,619 lakes (watershed area > 10 km2) in Central Asia from 1990 to 2020. Subsequently, the random forest (RF) algorithm was used to quantitatively evaluate the associations among climate change, human activities, and LWSC. It was revealed that maxLWSC and minLWSC decreased by 27.94 and 144.21 Gt, respectively. The results also revealed that LWSC varied by elevation, whereby lakes below 3,500 m shrunk and those above 3,500 m expanded. The RF findings demonstrated that human activities are more important than climate change for LWSC in Central Asia. Up to 96 % of LWSC can be explained by human activities. This study comprehensively evaluates the characteristics of LWSC in Central Asia, providing the latest data to help alleviate the contradiction between water supply and demand in Central Asia and resolve cross-border water resource disputes, so as to realize the protection of fragile regional ecosystems.
Read full abstract