Work

Integrating multi-model frameworks to unravel the spatiotemporal dynamics of flash floods in the Tianshan Mountain, China

Year: 2025

Type: article

Abstract: • Integrated long - term flash flood and environmental data for spatial analysis and modeling. • Used multi - model coupling to identify change mechanisms across CTM subregions. • Found accelerated post - 1995 flood risks due to extreme precipitation under climate change. • Identified flood mechanisms: terrain - driven rainstorms (ETM), landscape diversity and dams (NSTM), intensified precipitation (SSTM). The Tianshan Mountain in China (CTM), a critical water resource and climate change hotspot in Central Asia, faces escalating flash flood events due to global climate change and intensified human activities. This study applied the Geodetector (GD) to select driving factors with spatiotemporal variability of flash floods in the CTM from 1975 to 2015. The filtered drivers were integrated with historical flash flood data using the Geographically and Temporally Weighted Regression model (GTWR) to explore the spatiotemporal heterogeneity of the driving effects. Subsequently, the Partial Least Squares Structural Equation (PLS-SEM) was used to explore the direct and indirect influence pathways among driving factors. The analysis revealed a fluctuating upward trend in flash floods, accelerating after 1995 and showing different trends in various subregions after 2010. By the early 21st century, a symmetrical north–south distribution pattern emerged, with extreme precipitation events as the key driver. Terrain rainstorms were the main trigger in the eastern Tianshan Mountains (ETM), while landscape diversity, reduced snowmelt, and artificial flood control mitigated floods in the northern Tianshan Mountains (NSTM). The southern Tianshan Mountains (SSTM) experienced significant flood changes due to abundant precipitation. This study constructs a comprehensive analytical framework for investigating flash flood changes in the CTM by integrating GD, GTWR, and PLS-SEM models and proposes flash flood management strategies based on the identified mechanisms in different subzones.

Source: Ecological Indicators

DOI: 10.1016/j.ecolind.2025.113259

Authors: Biao Zhang, Haiyan Fang, Guotao Zhang, Songqing Li, Shufang Wu, Kadambot H.M. Siddique