关于广义朗之万动力学与全球平均温度建模
统计力学
2020-12-07 v2
摘要
气候科学采用层次化模型,在简化能量平衡模型(EBM)的可处理性与全球环流模型的细节之间做权衡。自Hasselmann的开创性工作以来,随机EBM已能处理气候涨落与噪声。然而,最近有观点认为观测表明全球平均温度对扰动的时域响应函数具有重尾特征。我们的互补方法利用了Hasselmann的EBM与物理学中原始均值回复随机模型——1908年朗之万方程——之间的对应关系。我们提出将统计力学中著名的Mori-Kubo广义朗之万方程(GLE)映射以推广Hasselmann的EBM。若存在长程记忆,则GLE可简化为分数朗之万方程(FLE)。我们描述了对应于GLE和FLE的EBM,简要讨论其解,并将其与Lovejoy的新分数能量平衡模型相联系。
引用
@article{arxiv.2007.06464,
title = {On Generalized Langevin Dynamics and the Modelling of Global Mean Temperature},
author = {Nicholas Watkins and Sandra Chapman and Aleksei Chechkin and Ian Ford and Rainer Klages and David Stainforth},
journal= {arXiv preprint arXiv:2007.06464},
year = {2020}
}
备注
First version was preprint of paper submitted on 15th May 2020 to Tenth International Conference on Complex Systems, Nashua, New Hampshire, now in press. Revised version adds a reference to previous work on the GLE in climate by Kondrashov et al, and corrects an error I had made in the relationship of Lovejoy et al's exponent H (not identical to Hurst's) and the memory paramter d