Extreme events in the past and future - A comparative assessment for the Hai He river and the Poyang lake basins

未来气候变化对土地利用和水资源管理的影响在很大程度上取决于极端天气的相关变化。未来只能通过使用全球气候模式来评估，而这些模式目前缺乏表征此类事件所需的空间分辨率。此外，全球气候模式由于分辨率较低，无法纳入导致极端事件的所有过程。因此，此类运行的降尺度是必要的，并且只有采用高分辨率区域模型的动态降尺度才能捕获导致极端情况的必要非线性过程和过程链。本联合提案的目标是对21世纪两个气候非常不同的流域，即海河流域和鄱阳湖流域，提供极端天气事件影响土地利用和水资源管理行为的不确定性估计。为了实现这一目标，我们将首先在观测和使用最先进的极值统计重新分析数据集的基础上，分析这两个地区过去的极端事件。为了评估流域极端事件的未来，将进行全球气候运行的动态降尺度。这首先需要通过与统计基础上的观测结果进行比较，对当前气候及其极端情况的模拟进行评估。这将使我们能够使用最合适的区域气候模型，并选择最适合两个可能不同的区域的参数化设置。中国合作伙伴将提供观测数据集并执行全球气候运行的动力降尺度，德国合作伙伴将在中国合作伙伴研究所安装动力降尺度程序，并从观测和模拟中生成极端值的统计数据。对趋势和不确定性的评价将在密切合作下进行。

The impact of future climate change on land-use and water resource management is strongly dependent on the related changes in weather extremes. The future can only be assessed by the use of global climate models, which currently lack the necessary spatial resolution to represent such events. Moreover, global climate models are not able to incorporate all processes leading to extremes because of their low resolution. Thus downscaling of such runs is necessary, and only dynamical downscaling with high-resolution regional models is able to catch the necessary non-linear processes and process chains leading to extremes. The goal of this joint proposal is to provide estimates including their uncertainties of the behaviour of extreme weather events impacting land-use and water management for the 21st century for two climatically very different catchments, namely the Hai He river and the Poyang lake basins. To this goal we will first analyse the past of extreme events in both regions on the basis of observations and reanalysis data sets using state of the art extreme value statistics. Dynamical downscaling of global climate runs will be performed in order to evaluate the future of extreme events in the catchments. This necessitates first an evaluation of simulations of the current climate and its extremes by comparison with observations on a statistical basis. This will enable us to use the most appropriate regional climate model and to select the parametrisation setup most suitable for both regions, which might be different. While the Chinese partners will provide the observation data sets and perform the dynamical downscaling of global climate runs, the German partners will install the dynamical downscaling procedure at the Chinese partner institute, and generate the statistics of extremes both from observations and the simulations. The evaluation towards trends and uncertainties will be performed in close cooperation.