Towards an Integrated Capability to Explain and Predict Regional Climate Changes (EXPECT)

气候系统正在迅速变化，一些地区的极端情况有所增加，超出了气候模型模拟的预期。为了支持有针对性的气候适应战略，EXPECT将通过开发综合归因和预测气候的原型业务能力，对包括极端事件在内的区域气候变化进行可靠的评估和预测。这一雄心勃勃的目标与WCRP解释和预测地球系统变化的灯塔活动密切相关。EXPECT将确定和量化物理过程在年际至数十年时间尺度上控制区域气候变化，包括极端气候变化的机制。它将通过利用新的气候模拟和地球观测（EO），并将机器学习（ML）与物理方法相结合来实现这一目标。这项研究将针对与大气环流和陆地-大气相互作用有关的基本知识差距，这是目前气候预测和预测的主要限制，特别是在理解欧洲夏季极端气候变化方面。为了支持研究，并使更广泛的研究界受益，EXPECT将开发工具来有效地分析各种大型数据集，这些数据集托管在各机构的不同存储库中。这将有助于利用最近对高分辨率气候模式和地球观测数据的投资。EXPECT将进一步建立数据科学能力，对大量数据资产进行科学可靠、高效和可重复的分析，包括新颖的机器学习方法，并为气候科学界特别是下一代研究人员提供培训。因此，EXPECT将为社会和气候科学界带来重大的科学和技术进步，这些进步将持续到项目结束后，以支持WCRP的战略目标。

The climate system is changing rapidly and some regions have seen increases in extremes beyond what is expected from climate model simulations. To support targeted climate adaptation strategies, EXPECT will enable trustworthy assessments and predictions of regional climate change including extremes by developing a prototype operational capability for integrated attribution and prediction of climate. This ambitious goal is closely aligned with the WCRP Lighthouse Activity on Explaining and Predicting Earth System Change. EXPECT will identify and quantify the mechanisms by which physical processes govern regional climatic changes, including extremes, on interannual to multi-decadal time scales. It will do so by exploiting newly available climate simulations and Earth Observations (EOs), and by combining machine learning (ML) with physical methods. The research will target fundamental knowledge gaps related to atmospheric circulation and land-atmosphere interactions, which represent major limitations in current climate predictions and projections, and in particular in understanding changes in European summer extremes. To underpin the research, and benefitting the wider research community, EXPECT will develop tools to efficiently analyse a variety of large data sets in combination that are hosted in different repositories across institutions. This will facilitate the exploitation of recent investments into high-resolution climate models and EO data. EXPECT will further build data science capacity for the scientifically robust, efficient and reproducible analysis of the massive data assets, including novel ML approaches, and provide training for the climate science community and the next generation of researchers in particular. EXPECT will thus deliver significant scientific and technological advances for society and the climate science community that will last well beyond the project, in support of WCRP’s strategic objectives.