Observational constraints on climate projections

拟议研究的目标是利用过去世纪的气候变化观测，以改善气候模型对未来气候变化的预测。气候模型是预测未来气候变化的最佳工具，为减少温室气体排放和适应未来气候变化的决策提供必要的关键信息。然而，由于气候系统的许多复杂性及其性质的不确定性，气候模型对气候许多方面未来变化的预测在模型之间存在很大差异。例如，联合国政府间气候变化专门委员会的最新评估报告得出结论，即使给定21世纪世纪排放的具体轨迹，21世纪世纪的全球变暖可能在2.4°C到6.4°C之间。虽然升温2.4 ℃对社会和环境的影响可能是可控的，但升温6.4 ℃的影响可能非常严重。因此，如果气候预测对决策者和整个社会更有用，就必须努力减少这些不确定性。这项研究的重点是利用过去气候变化的观测结果来改善我们对未来气候变化的预测。目前，世界各地正在进行大量的气候模型模拟，为联合国IPCC报告的下一份报告做准备。通过分析历史时期（1850-2010年）和未来的模拟气候变化，应该可以确定模拟过去气候变化和模拟未来气候变化之间的可靠关系。举一个简单的例子，我们可能会发现，过去变暖更多的模型在未来也会变暖更多。然后，我们可以使用过去气候变化的观测结果，要么排除一些对温室气体增加过于敏感或不够敏感的模型，要么我们可以重新调整模型预测，以便改善与观测结果的匹配。迄今为止进行的这类研究在限制未来预测方面取得的成功有限。但是，使用更好的模型进行新的模拟（许多将在2010年完成）的可用性，为以这种方式改进我们的气候预测提供了一个充满希望的新机会。

The objective of the proposed research is to use observations of climate change over the past century to improve climate model predictions of future climate change. Climate models are the best available tools for predicting future climate change, providing key information necessary to inform decisions on reducing greenhouse gas emissions, and adaptation to future climate change. However, due to the many complexities of the climate system and uncertainties in its properties, climate model projections of future change in many aspects of the climate differ strongly between models. For example the last United Nations IPCC assessment report concluded that, even given a specific trajectory of 21st century emissions, 21st century global warming could be anywhere from 2.4°C to 6.4°C. While the impacts on society and the environment from a warming of 2.4°C might be manageable, the impacts of a warming of 6.4°C would likely be very severe. Thus it is important to try to reduce these uncertainties, if climate predictions are to be more useful to policymakers and society at large. The proposed research focuses on using observations of past climate change to improve our predictions of future climate change. A large number of climate model simulations are currently being carried out around the world in preparation for the next report of the UN IPCC report. By analysing simulated climate change over both the historical period (1850-2010) and the future, it should be possible to identify robust relationships between simulated past climate change and simulated future climate change. To take a simple example, we might find that models which warm more in the past tend also to warm more in the future. We can then use observations of past climate change, either to rule out some models as being either too sensitive to greenhouse gas increases or not sensitive enough, or we can rescale model projections so that the match to observations is improved. Such research carried out to date has had limited success in constraining future projections. But the availability of new simulations with better models, many finishing in 2010, presents a promising new opportunity to improve our climate projections in this way.