积雪
可预测性
雪
环境科学
降水
气候学
地表径流
仰角(弹道)
气候变化
水资源
融雪
地理
气象学
地质学
生态学
海洋学
物理
几何学
数学
量子力学
生物
作者
Ben Livneh,Andrew M. Badger
标识
DOI:10.1038/s41558-020-0754-8
摘要
Mountain snowpack serves as an immense natural water reservoir, and knowledge of snow conditions helps predict seasonal water availability and offers critical early warning of hydrologic drought. This paradigm faces an impending challenge given consensus that a smaller fraction of future precipitation will fall as snow. Here, we apply downscaled hydrologic simulations from 28 climate model projections to show that by mid-century (2036–2065), 69% of historically snowmelt-dominated areas of the western United States see a decline in the ability of snow to predict seasonal drought, increasing to 83% by late century (2070–2099). Reduced predictability arises when peak snowpack approaches zero or because of decreased warm-season runoff efficiency. Changes in drought prediction skill show significant (P < 0.01) elevation dependence, with lower-elevation coastal areas most impacted by warming. Ancillary predictive information can only partially mitigate snow-based predictability losses to 65% of areas, underscoring the importance of declining future snowpack. Climate warming causes less mountain precipitation to fall as snow. Hydrologic simulations predict that in a high-end emissions scenario, this decreases the predictability of seasonal water resources across the western United States, with low-elevation coastal areas impacted most strongly.
科研通智能强力驱动
Strongly Powered by AbleSci AI