Scholarlink

<p>The densely populated East Asia is vulnerable to precipitation extremes. By utilizing a deep learning downscaled high-resolution (0.1°) dataset CLIMEA-BCUD (Climate Change for East Asia with Bias corrected UNet Dataset), changes in precipitation extremes over East Asia under different emission scenarios are investigated. Evaluation against observations from different sources (i.e., reanalysis-based, in situ-based and satellite-based) shows that CLIMEA-BCUD can reasonably reproduce the spatial patterns of the extreme precipitation indices, although it tends to overestimate CWD (consecutive wet days) in the Indo-China Peninsula. CLIMEA-BCUD exhibits good agreement with the magnitude of the observations and presents an obvious improvement in terms of bias, root mean square error, and correlation coefficient compared with the driving CMIP6 (Coupled Model Intercomparison Project Phase 6) models. The frequency and intensity of precipitation extremes are projected to increase in most parts of East Asia, especially over southern latitudes such as India and Indo-China. More pronounced increases in R10mm are also projected over the Tibetan Plateau. Record-breaking events, even those that break historical records by much higher magnitude, are becoming more frequent in a warmer climate. During 2071–2100, precipitation extremes that break the historical records by two or more standard deviations are three to five times more likely to occur somewhere in East Asia under SSP5-8.5 compared to those under SSP1-2.6. Potential hotspots of such record-breaking precipitation extremes are high-altitude areas such as the Tibetan Plateau, where the probability of experiencing record-breaking R95p, which breaks historical records by at least two standard deviations, is 40% under the high-emission scenario.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Future projections of precipitation extremes over East Asia based on a deep learning downscaled CMIP6 high-resolution (0.1°) dataset

  • 期刊论文

  • 发表日期：

    2025年1月13日

查看全文
  • Yi Yang,
  • Hai Lin,
  • Yi Xu,
  • Hang Pan,
  • Guangtao Dong,
  • Jianping Tang

摘要

The densely populated East Asia is vulnerable to precipitation extremes. By utilizing a deep learning downscaled high-resolution (0.1°) dataset CLIMEA-BCUD (Climate Change for East Asia with Bias corrected UNet Dataset), changes in precipitation extremes over East Asia under different emission scenarios are investigated. Evaluation against observations from different sources (i.e., reanalysis-based, in situ-based and satellite-based) shows that CLIMEA-BCUD can reasonably reproduce the spatial patterns of the extreme precipitation indices, although it tends to overestimate CWD (consecutive wet days) in the Indo-China Peninsula. CLIMEA-BCUD exhibits good agreement with the magnitude of the observations and presents an obvious improvement in terms of bias, root mean square error, and correlation coefficient compared with the driving CMIP6 (Coupled Model Intercomparison Project Phase 6) models. The frequency and intensity of precipitation extremes are projected to increase in most parts of East Asia, especially over southern latitudes such as India and Indo-China. More pronounced increases in R10mm are also projected over the Tibetan Plateau. Record-breaking events, even those that break historical records by much higher magnitude, are becoming more frequent in a warmer climate. During 2071–2100, precipitation extremes that break the historical records by two or more standard deviations are three to five times more likely to occur somewhere in East Asia under SSP5-8.5 compared to those under SSP1-2.6. Potential hotspots of such record-breaking precipitation extremes are high-altitude areas such as the Tibetan Plateau, where the probability of experiencing record-breaking R95p, which breaks historical records by at least two standard deviations, is 40% under the high-emission scenario.