Thermodynamic response of precipitation extremes to climate change and its impacts on floods over China

Abstract

<p indent=0mm>As the planet warms, shifts in the thermodynamic and dynamic environments of the climate system exert substantial impacts on the energy budget and water cycle. Understanding the response of precipitation extremes to climate change is essential for mitigating weather-related hazards and facilitating adaptive water resources management. The moisture holding capacity of the atmosphere, governed by the Clausius-Clapeyron (C-C) relationship, increases with temperature at a rate of about 7%/°C. Thermodynamically, precipitation intensity is expected to increase exponentially and worsen flood conditions with atmospheric warming. However, both observational and satellite data have revealed a nonmonotonic (hook) scaling of precipitation and storm runoff, in which extremes strengthen with warming temperature up to a peak point (<italic>T</italic>_pp) and decline thereafter. Few studies have yet considered whether and why the hook structure might change in a future warming climate. Moreover, existing studies have largely failed to detect and attribute changes of extreme precipitation to atmospheric thermodynamics versus dynamics, limiting reliable projection of future hydrological hazards and climate simulations. Here, we build upon existing studies to understand how precipitation and storm runoff scaling might change over Chinese mainland. We use gridded observational data to examine the temperature scaling of precipitation and storm runoff extremes, extending the C-C relationship. Using the ERA5 reanalysis dataset, we diagnose the impacts of total column water vapor (TCWV), convective available potential energy (CAPE) and relative humidity on precipitation extremes. After partitioning the thermodynamic and dynamic contributions of precipitation extremes based on the energy budget and water vapor balance, we quantify the underlying physical mechanisms of hook structures. Future climate scenarios are projected by using a large ensemble from 21 global climate models (GCMs) within Coupled Model Inter-comparison Project Phase-6 (CMIP6). Future streamflow series of 151 basins are projected by forcing hydrological model with bias-corrected climate model outputs. We quantify changes in the hook structures and their <italic>T</italic>_pp, and investigate implications of these shifts for future storms and flooding. The results show that relative humidity usually decreases with warming near-surface temperatures over most areas of China, while the TCWV and CAPE exhibit positive scaling rates with temperatures. The ERA5 reanalysis data over China reveals that atmospheric dynamics are responsible for the intensification of precipitation extremes and even flooding under very warm temperatures. The daily bias correction method reduces the systematical biases of GCMs outputs adequately, and captures extreme precipitation and temperatures well. By using four different hydrological models, we find that the Xinanjiang model typically exhibits the best performance in simulating daily streamflow over Chinese mainland. Projections of future hydro-climatological scenarios indicate that the hook structures do not set a firm upper-bound constraint on the intensification of extremes; rather, both the hook structure and the <italic>T</italic>_pp shift toward warmer temperatures in a warming climate. Alongside the shifts in <italic>T</italic>_pp, future storm and flooding extremes are projected to intensify by 20%−30%. These findings highlight the urgency in enhancing societal resilience to climate change, and understanding how the intensification of weather-related hazards might affect existing infrastructure and ecosystems.