Quantifying future climate change
Quantifying future climate change
复制标题
DOI：
10.1038/nclimate1414
复制
发表时间：
2012-06-01
影响因子：
30.7
通讯作者：
Stephenson, David B.
中科院分区：
文献类型：
--
作者：
Collins, Matthew;Chandler, Richard E.;Stephenson, David B.
科学界、政策制定者和其他利益攸关方对未来气候的量化预测的需求越来越大。复杂程度各不相同的气候模型被用来进行预测，但模型中的近似和不足或“错误”意味着这些预测是不确定的，有时会探索非常广泛的可能的未来。这里用一个通用框架描述了量化不确定性的技术，通过这个框架，模型被用来探索过去的气候和气候变化与未来预测之间的关系。可以改变模型参数，以产生一系列对过去气候的不同模拟，然后将这些模拟与使用“指标”的观测值进行比较。如果通过观测比较可以将模型参数限制在较窄的范围内，则预测也可以限制在较窄的范围内。讨论了不同实现的优点和缺点。
Quantitative projections of future climate are in increasing demand from the scientific community, policymakers and other stakeholders. Climate models of varying complexity are used to make projections, but approximations and inadequacies or 'errors' in models mean that those projections are uncertain, sometimes exploring a very wide range of possible futures. Techniques for quantifying the uncertainties are described here in terms of a common framework whereby models are used to explore relationships between past climate and climate change and future projections. Model parameters may be varied to produce a range of different simulations of past climate that are then compared with observations using 'metrics'. If the model parameters can be constrained to a tighter range as a result of observational comparisons, projections can also be constrained to a tighter range. The strengths and weaknesses of different implementations are discussed.