Detection of Anthropogenic Changes in Climate Extremes

拟议的工作将使用统计最佳检测方法来提高人为信号与内部气候变率的比率，并允许更早地检测变化。预测极端事件时空演变的“指纹”将从气候模式模拟中得到。一项检测研究将首先应用模型数据，以探索哪些极值指数应该显示最早可检测的变化。项目还将审查哪种检测程序最有希望早期发现极端情况的变化，以及可以和应该应用哪种变量或变量组合。最有希望的方法将应用于观察。重点是极端温度，干湿期的长度和极端降水，特别是在几天内。将使用主要来自美国、加拿大、欧洲、俄罗斯、中国和日本的地面观测站的日常观测数据。这项工作很重要，因为它将增进人们对气候变化探测技术的了解。

The proposed work will use statistical optimal detection methods to improve the ratio of anthropogenic signal to internal climate variability and allow for an earlier detection of changes. 'Fingerprints' for the expected spatial and temporal evolution of changes in extremes will be derived from climate model simulations. A detection study first applying model data only, will be performed to explore which extreme value indices should show the earliest detectable changes. The PI will also examine which procedure of detection is most promising for an early detection of changes in extremes, as well as which variable or combination of variables can and should be applied. The most promising procedure will be applied to observations. The focus is on temperature extremes, length of wet and dry periods and precipitation extremes, particularly over several days. Daily observations from land stations, mainly from the US, Canada, Europe, Russia, China and Japan will be used. The work is important because it will lead to increased understanding on climate change detection techniques.