降水
环境科学
气候学
大气科学
生长季节
季节性
旱季
高原(数学)
气候变化
生态系统
雨季
水循环
极端气候
厄尔尼诺南方涛动
气候模式
热带
作者
Hongzhou Wang,Fangyue Zhang,Zhaofei Wu,Zunchi Liu,Guizeng Qi,Rongqi Tang,Zehong Huang,Yongshuo H. Fu
摘要
Abstract China's drylands have experienced a pronounced warming‐wetting trend in recent decades, yet its seasonal expression and hydrological implications remain unclear. Using high‐resolution gridded observations spanning 1982–2022, we show that recent wetting is driven not by changes in precipitation totals alone, but by a concurrent intensification of precipitation regimes, including extreme event frequency, intensity, and dry\wet spell structure. This intensification exhibits strong seasonal asymmetry: extreme precipitation trends during the growing season are generally weak or insignificant, whereas the non‐growing season shows robust increases in extreme precipitation frequency and intensity, accompanied by widespread shortening of dry spells. When normalized by seasonal baselines, extreme precipitation intensified more coherently during the non‐growing season than during the growing season, with stronger relative increases and significant absolute increases in R95p (a time‐invariant 95th percentile threshold) frequency and intensity. This seasonal contrast is spatially coherent across major dryland eco‐geographical subregions, with the Tibetan Plateau exhibiting a distinct dual response, characterized by increased consecutive wet days during the growing season and decreased consecutive dry days during the non‐growing season. Importantly, this hydroclimatic asymmetry between seasons is evident in temperature, atmospheric moisture, and precipitation, but is not reflected in soil moisture. This discrepancy may increase the risk of seasonal mismatches between water availability and ecosystem demand. These results highlight the importance of resolving precipitation changes at seasonal scales for understanding dryland hydroclimatic dynamics and informing water resource management under ongoing climate change.
科研通智能强力驱动
Strongly Powered by AbleSci AI