Collaborative Research: EaSM3 Integration of Decision-Making with Predictive Capacity for Decadal Climate Impacts

在区域和年代际尺度上进行气候影响预测的必要性已得到广泛承认。工业、政府和社会越来越需要与决策有关的预测信息，以便能够对气候变率和变化引起的未来危害进行适当规划和适应，既减轻未来的成本，又使潜在效益最大化。气候科学家正在努力了解气候系统的年代际可预测性，并开发以5-30年为时间尺度的气候预测模式。然而，了解气候影响的预测能力和实现年代际预测科学的全部社会价值的平行努力却很少受到关注。该合作项目将利用现有的和正在开发的动力模型，结合先进的统计方法观测数据，了解我们对年代际气候影响的预测能力，并整合预测方法，使其更直接地满足社会对预测信息的需求。总体目标是改变来自多个学科的科学家和实践者对年代际气候预测的概念，使与决策相关的年代际预测能够改善气候变率和变化的规划和适应能力。该方法分为两个并发部分。第一部分将在多个高影响天气和气候现象以及多个利益相关者群体中发展对当前预测信息需求和使用的理解。核心研究将在第二部分进行，并将通过开发新的组合统计动态建模方法来确定我们对所需信息的预测能力，这些方法包含不确定性并且对不确定性具有鲁棒性。信息需要和信息提供之间的迭代过程将促进熟练的预测信息与决策的有效结合。一个关键的项目贡献是开发一个原型广义跨学科研究框架，将预测能力与决策相结合。

The need for climate impact predictions on regional and decadal scales is widely recognized. Industry, government and society increasingly require predictive information in decision-relevant terms to enable appropriate planning and adaptation to future hazards arising from climate variability and change, both to mitigate future costs and maximize potential benefits. Climate scientists are engaged in significant effort to understand the decadal predictability of the climate system and to develop models to predict the climate on 5-30 year timescales. However, parallel efforts to understand predictive capacity of climate impacts and realize the full societal value of decadal prediction science has received little attention.This collaborative project will utilize existing and developing dynamical models combined with observed data using advanced statistical methods to understand our predictive capacity for climate impacts on decadal timescales, and integrate predictive methods so that they more directly meet the societal needs for predictive information. The overarching goal is to transform how scientists from multiple disciplines and practitioners conceptualize decadal climate prediction, by enabling decadal predictions in decision relevant terms to improve planning and adaptive capacity for climate variability and change.The approach follows in two concurrent parts. Understanding of current predictive information needs and usage will be developed in Part I across multiple high-impact weather and climate phenomena and across multiple stakeholder groups. The core research will take place in Part II and will identify our predictive capacity for the needed information through development of new combined statistical-dynamical modeling approaches that incorporate and are robust to uncertainty. An iterative process between information needs and information provision will promote effective integration of skillful predictive information with decision-making. A key project contribution is the development of a prototype generalized interdisciplinary research framework to integrate predictive capacity with decision-making.