Downturn in scaling of UK extreme rainfall with temperature for future hottest days
Downturn in scaling of UK extreme rainfall with temperature for future hottest days
复制标题
DOI：
10.1038/ngeo2596
复制
发表时间：
2016-01-01
影响因子：
18.3
通讯作者：
Blenkinsop, Stephen
中科院分区：
文献类型：
--
作者：
Chan, Steven C.;Kendon, Elizabeth J.;Blenkinsop, Stephen
根据温度和饱和蒸汽压之间的克劳修斯-克拉珀龙关系，极端日降水量被认为以每K 6.5%的速率随温度增加（1）。在全球范围内，观测到极端日和小时降水量的各种标度关系，有证据表明，在某些地区，亚日极端降水量的标度超过6.5%/K（2-4）。只有高分辨率气候模型才能模拟这种缩放关系（5）。在这里，我们使用英国南部的一个具有三分之一尺度分辨率的模型进行实验，研究未来气候中每小时极端降水强度的尺度（6）。我们的模型模拟了现今的标度关系，每K为6.5%，与观测结果一致。模拟的未来极端降水的总体增加遵循相同的关系。然而，英国极端降水强度下降，在温度高于约22摄氏度的温度范围内，是不是很好地采样，在现今的整合，作为一个更频繁发生的反气旋天气系统的结果。反气旋产生更多的日数，白天有较强的加热，但不利于深强对流风暴的发展。我们的结论是，未来的极端每小时降水强度不能简单地从现在的温度标度外推，并展示了使用区域表面温度作为标度变量的陷阱。
Extreme daily precipitation is thought to increase with temperature at a rate of 6.5% per K according to the Clausius-Clapeyron relationship between temperature and saturation vapour pressure(1). A wide range of scaling relationships has been observed globally for extreme daily and hourly precipitation, with evidence of scaling above 6.5% per K for sub-daily extreme precipitation in some regions(2-4). Only high-resolution climate models can simulate this scaling relationship(5). Here we examine the scaling of hourly extreme precipitation intensities in a future climate using experiments with a model for the southern UK with kilometre-scale resolution(6). Our model simulates the present-day scaling relationship at 6.5% per K, in agreement with observations. The simulated overall future increase in extreme precipitation follows the same relationship. However, UK extreme precipitation intensities decline at temperatures above about 22 degrees C-a temperature range that is not well sampled in the present-day integration-as a result of a more frequent occurrence of anticyclonic weather systems. Anticyclones produce more days with strong daytime heating, but are not favourable to the development of deep intense convective storms. We conclude that future extreme hourly precipitation intensities cannot simply be extrapolated from present-day temperature scaling, and demonstrate the pitfalls of using regional surface temperature as a scaling variable.