Future changes and uncertainty in decision-relevant measures of East African climate
Future changes and uncertainty in decision-relevant measures of East African climate
复制标题
DOI：
10.1007/s10584-019-02499-2
复制
发表时间：
2019-10-01
期刊：
影响因子：
4.8
通讯作者：
Way, Celia
中科院分区：
文献类型：
--
作者：
Bornemann, F. Jorge;Rowell, David P.;Way, Celia
制定非洲气候变化适应战略的必要性正变得越来越重要。例如，现在需要设计或调整使用寿命长的基础设施，以适应与过去或现在不同的未来气候。决策中使用的气候信息越来越有必要从传统的科学驱动的指标转变为决策驱动的指标。这一点在东非尤其重要，那里的适应能力和社会经济能力有限，使该地区极易受到气候变化的影响。在这里，我们采用跨学科咨询过程来定义和分析许多此类面向决策的指标。这些指标采取整体方法，涉及东非的农业、供水、渔业、洪水管理、城市基础设施和城市卫生等关键部门。对多模式气候预测的多方面分析提供了一个关于气候变化及其不确定性的以用户为中心的信息库，涵盖所有指标、季节和两个未来时间范围。描述了温度和降雨指标变化的空间特征和较大的模型间不确定性，以及蒸发和太阳辐射等其他相关指标的空间特征和不确定性。模型之间的关系指标也进行了探讨，形成两个明确的集群周围的降雨量和温度指标。后一种分析确定了模型权重可以或不可以在多个气候指标中应用的程度。进一步的工作现在必须集中在最大限度地提高模型预测的效用，并制定量身定制的基于风险的沟通战略。
The need for the development of adaptation strategies for climate change in Africa is becoming critical. For example, infrastructure with a long lifespan now needs to be designed or adapted to account for a future climate that will be different from the past or present. There is a growing necessity for the climate information used in decision making to change from traditional science-driven metrics to decision-driven metrics. This is particularly relevant in East Africa, where limited adaptation and socio-economic capacity make this region acutely vulnerable to climate change. Here, we employ an interdisciplinary consultation process to define and analyse a number of such decision-oriented metrics. These metrics take a holistic approach, addressing the key East African sectors of agriculture, water supply, fisheries, flood management, urban infrastructure and urban health. A multifaceted analysis of multimodel climate projections then provides a repository of user-focused information on climate change and its uncertainties, for all metrics and seasons and two future time horizons. The spatial character and large intermodel uncertainty of changes in temperature and rainfall metrics are described, as well as those of other relevant metrics such as evaporation and solar radiation. Intermodel relationships amongst metrics are also explored, with two clear clusters forming around rainfall and temperature metrics. This latter analysis determines the extent to which model weights could, or could not, be applied across multiple climate metrics. Further work must now focus on maximising the utility of model projections, and developing tailored risk-based communication strategies.