Constraining uncertainty in glacial system evolution

冰和气候系统的快速、非线性变化可能为地球系统的变化提供了影响最大的途径，但也是地球系统过程中最难预测的方面。要量化这类事件在冰和气候等复杂和高度相互关联的系统中发生的概率是一项非常具有挑战性的任务。然而，优秀科学的一个定义特征是对不确定性的严格量化。可以说，这种量化的缺陷在当前关于气候变化的辩论中扮演着重要角色。 这项研究将采取多方面的方法来推断冰川(冰和气候)系统演变的不确定性估计，特别是在潜在的快速变化的背景下。我们将进一步开发一种统计上严格的校准方法，将观测数据和建模相结合，以限制冰和气候系统过去和未来的演变。将这一方法应用于过去，将对重大不稳定事件前后冰川系统的状态建立更清晰的图景。此外，通过在这个框架内部署越来越复杂的模型，我们将测试当前的模型和理解在多大程度上能够复制和解释过去的冰盖演化和不稳定事件。这些测试反过来将有助于估计与未来冰盖演变有关的不确定性以及与气候系统其他部分的相关相互作用，解决当前气候变化评估中的一个重大差距。 更广泛地说，这项工作将是改进以观测为基础的方式严格分析复杂非线性物理系统动力学的方法的个案研究。

Fast, non-linear changes in the ice and climate systems potentially offer the highest-impact routes to changes in the earth system, but are also the hardest aspects of earth system processes to predict. It is a very challenging task to quantify the probabilities of such events occurring in such complex and highly interconnected systems as ice and climate. Yet, a defining feature of good science is the rigorous quantification of uncertainty. Deficiencies in such quantification arguably plays a significant role in the current debate over climate change. This research will take a multifaceted approach to inferring uncertainty estimates for the evolution of the glacial (ice and climate) system, especially in the context of potential rapid changes. We will further develop a statistically rigorous calibration methodology to integrate observational data and modelling to constrain past and future evolution of the ice and climate system. Application of this methodology to the past will build a clearer picture of the state of the glacial system before and after major instability events. Furthermore, by deploying increasingly sophisticated models within this framework, we will test the extent to which current models and understanding can replicate and explain past ice sheet evolution and instability events. These tests will in turn facilitate the estimation of uncertainties relating to future ice-sheet evolution and associated interactions with the rest of the climate system, addressing a major gap in current climate change assessments. More generally, the work will be a case study in improving methodologies for rigorously analysing the dynamics of complex non-linear physical systems in an observation-grounded manner.