Robust historical and future drying trends in Central Asia evidenced by the latest observation and modeling datasets

Abstract

As one of the largest arid and semiarid regions, Central Asia (CA) is prone to drought, which imposes significant impacts on human communities and ecosystems. Understanding the historical and future wetting/drying trend with the backdrop of climate change is paramount to sustainable development in CA. However, previous studies for the historical period yielded inconsistent results due to different data, study durations and methods used and those for the future period are rare. By analyzing the latest generated long-term (1894–2020) homogenized station observations, multiple global climate model (GCM) outputs and their dynamically and statistically downscaled results, we find robust historical and future drying trend in CA, especially in the growing season (April–September). Though there is an increasing tendency in regional precipitation during 1894–2020 in CA, the Standard Precipitation Evapotranspiration Index (SPEI) shows a decreasing trend due to the dominating influence of regional warming. Compared to the non-growing season (October–March), the decreasing trend of SPEI is more profound in the growing season. Moreover, the SPEI calculated based on the GCM outputs and their dynamically and statistically downscaled results consistently shows future drying trend in CA throughout the 21st century, which robustly holds against the approaches used to calculate potential evapotranspiration (i.e., Thornthwaite and Penman-Monteith equations). Besides SPEI, the simulated soil moisture of surface layer also exhibits a decreasing tendency. All these lines of evidence suggest robust historical and future drying trends in CA, which have important implications for climate change adaptation in this region.