Regional disparities in extreme precipitation trends across East Asia: Observation-constrained projection and attribution

Bao, Zhong-Rui , Xie, Yong-Kun , Shi, Jin-Sen , Zhao, Min , Mi, Jia-Qin

2026-02-01 ADVANCES IN CLIMATE CHANGE RESEARCH 2026   17(卷), 1(期), (35-47页)

查看原文

Extreme precipitation events pose growing risks across East Asia. Although various studies have examined the historical and future changes of extreme precipitation, substantial uncertainties in magnitudes and signs remain regarding future trends at the regional scale. Combining CMIP6 simulations with observations, we analyze historical changes and project more robust future extreme precipitation trends across four subregions. During 1979-2014, heavy precipitation cumulative intensity (HPCI) increased by 10.7 mm per decade (p < 0.05) in humid region and 12.2 mm per decade in the southern Tibetan Plateau, while it slightly increased (+0.1 mm per decade) in Northeast Asia and decreased (-0.1 mm per decade) in dryland. Although model simulations project increased HPCI across all subregions during 2015-2100, observation-constrained projections refine these projections, revealing more realistic and regionally divergent futures. Under SSP2-4.5, HPCI increases by 64% in the humid region and 119% in the southern Tibetan Plateau, respectively, by the end of the 21st century, with the magnitudes doubled under SSP5-8.5. Dryland shows moderate increases (35% under SSP2-4.5 and 26% under SSP5-8.5), whereas Northeast Asia declines slightly (-13% and -2%, respectively). Greenhouse gas emissions fuel the intensification of heavy rainfall in the humid region by elevating atmospheric humidity and altering large-scale circulation. Meanwhile, on the southern Tibetan Plateau, the increase in extreme precipitation is driven by greenhouse gases-through enhanced moisture delivery from the Bay of Bengal and upward motion-as well as by aerosol reduction, which modifies humidity and circulation. Our observation-constrained projections, coupled with mechanistic insights, yield more robust results than unconstrained ones-providing critical scientific support for climate-adaptive flood control and water management across East Asia's diverse regions.