新一代冰流模式乌阿及其在南极埃默里冰架的应用

A new-generation ice flow model úa and its application in Amery Ice Shelf,Antarctica

南极冰盖不仅是全球环境变化的指示器,其消融所产生成的淡水输入也是未来海平面上升的主要不确定性来源。数值模式是诊断冰流动力机制、评估冰盖物质损耗的重要手段。本文首先介绍了乌阿(冰岛语úa或英语Ua)冰流模式的基本原理,并利用该模式模拟东南极埃默里冰架的动态变化。乌阿冰流模式基于质量和动量守恒方程的垂直积分,在自适应不规则三角网格上求解微分方程,仅用少数参数规则即可构造适应冰流动力特征的网格结构,有效缩减运算时间。采用当前主流的模式边界数据集,针对埃默里冰架设计了两个试验。试验一为反演试验,试验中模式的代价函数在100次迭代后下降三个数量级,表明模拟的流速与遥感观测吻合(RMSE=13.35m·a^-1),但高频细节仍有待提高;试验二为预测试验,测试了模拟冰厚变化率的不确定性,以自由漂移量接近零为标准选出一组最优模型参数,最后假设埃默里冰架解体情景开展模拟,结果表明冰架解体会导致海平面上升(45.36±0.08)mm。随着资料更新迭代,基于最新发布的南极底部地形数据模拟效果是否提升还有待未来检验。

Antarctic Ice Sheet is not only the sensitive indicator of global environmental changes,but its melting is also the largest uncertainty of the future sea-level rise.The numerical model is an important methodology for diagnostic ice flow mechanism and assessing ice mass loss.This paper introduces the basic principles of a new generation ice flow dynamic model,úa,based on which the two experiments are implemented to simulate the dynamics of Amery Ice Shelf(AIS)in East Antarctic.Leveraging the adaptive triangular finite-element meshes,such model solves vertically integrated formulations of the mass and momentum conservation equations.The generated mesh accurately captures the glacial dynamics from only a few parametric criteria,which reduces the computation time.On the basis of mainstream datasets,two experiments are designed in the AIS area.In the first part of the inverse run,the value of cost function in the model reduces for three orders of magnitude.Although the fine-scale structures remain to be improved,the output flow pattern agrees well with the satellite observation(RMSE=13.35 m·a^-1).In the second part of the prognostic run,we test the sensitivity of ice thickness change rate,from which the optimal one with the minimum bias is selected to simulate the shelf’s collapse scenario.The results demonstrate the upstream ice system would contribute about(45.36±0.08)mm to sealevel rise.Since Antarctic topography data were updated,whether our model can be improved from the new data deserves further research.