Abstract
Endorheic lakes, lacking river outflows, are highly sensitive to environmental changes and human interventions. Central Asia (CA) has over 6000 lakes that have experienced substantial water level variability in the past century, yet causes of recent changes in many lakes remain unexplored. Modelling hydrological processes for CA lakes poses challenges in separating climatic change impacts from human management impacts due to limited data and long-term variability in hydrological regimes. This study developed a spatially lumped empirical model to investigate the effects of climate change and human water abstraction, using Shortandy Lake in Burabay National Nature Park (BNNP) as a case study. Modelling results show a significant water volume decline from 231.7x106m3 in 1986 to 172.5x106m3 in 2016, primarily driven by anthropogenic water abstraction, accounting for 92% of the total volume deficit. The highest rates of water abstraction (greater than 25% of annual outflow) occurred from 1989 to 1993, coinciding with the driest period. Since 2013, the water volume has increased due to increased precipitation and, more importantly, reduced water abstraction. Despite limited observational data with which to calibrate the model, it performs well. Our analysis underscores the challenges in modelling lakes in data-sparse regions such as CA, and highlights the importance and benefits of developing lake water balance models for the region.