基于非平稳极值理论的我国极端降水气候变化的机理研究

The observational analysis on climate changes of precipitation extremes in China are tending to be complete. However, more and more overestimated and record-breaking precipitation extremes in China are reported in recent years, which is indicating our knowledge on the mechanisms of precipitation extremes are far from complete. As a result, under the rapid changing climate background currently, the past precipitation extremes studies based on static climatology, simple qualitative and non-parametric methods have not meet the operational requirements any more. This project will adopt the newly-developed non-stationary extreme value theory, from the point of regional climate changes. Large-scale circulation indices will be statistically modeled with parameters of precipitation extremes PDF and tested rigidly, to study the parametric and quantitative mechanisms of the climate of precipitation extremes in China. The feasibility of attribution by precipitation extremes will be explored. The latest CMIP5 output of various future scenarios will be used to discuss the trend and probability of precipitation extremes in China during the late 21st century based on obtained statistical models. This project is aiming to explain how the low-changing large scale circulation profoundly affect the rare occurred and small regional precipitation extremes, which could be a new way to climate extremes study in China.

我国极端降水气候变化的事实研究已日趋完备。然而，近年我国多地发生超出预估的创纪录的极端降水事件，又表明我们对于极端降水气候变化的机理仍然缺乏充分的认识。这促使我们意识到，在目前气候快速变化的大背景下，过去那种静止气候态的、简单定性的、非参数化的极端降水机理研究已越来越不能适应现实的要求。本项目拟采用国际上最新发展的非平稳极值理论，从区域气候变化的角度出发，将大尺度环流因子与极端降水的分布参数进行详实可靠的统计建模，参数化地定量化地研究我国极端降水气候变化的机理，探索性地研究极端降水事件的归因可行性问题，并利用CMIP5耦合模式的情景输出，基于前述模型，探讨变暖背景下，未来我国极端降水发生的可能趋势。本项目主要拟深入解释缓慢变化的大尺度环流背景如何深刻影响到罕见的、局地小尺度变化的极端降水这一关键科学问题，为我国极端气候事件的机理和归因研究探索一条新的道路。