当前位置:科学网首页 > 小柯机器人 >详情
季风滞后揭示了大气记忆效应
作者:小柯机器人 发布时间:2025/5/7 15:21:24

近日,德国波茨坦大学Anders Levermann团队发现季风滞后揭示了大气记忆效应。相关论文发表在2025年5月6日出版的《美国科学院院刊》杂志上。

在地球的气候系统中,海洋、冰冻圈和植被表现出滞后行为,其状态取决于其过去,而不仅仅是其当前的边界条件。大气的快速混合时间尺度被认为抑制了这种多稳态的必要记忆效应。研究组表明,大气柱内的水分积累会在季风环流中产生滞后现象,与海洋储热无关,并在相同的太阳辐射量下产生两种稳定的大气状态。 

因此,季风降雨的动态是两种稳定状态之间的季节性过渡。观测数据显示了由此产生的滞后现象,并在大气环流模型中再现,在该模型中,滞后现象随着海洋记忆的减少而增加,并表现出两种不同的状态,这两种状态持续了60多年。它们通过大气柱内的水分积累而稳定下来,大气柱携带的信息跨越的时间尺度比混合的典型时间尺度要长得多。这两个州之间突然转变的可能性对全球季风降雨的未来演变产生了影响,这对目前养活20多亿人的农业生产力至关重要。

附:英文原文

Title: Monsoon hysteresis reveals atmospheric memory

Author: Katzenberger, Anja, Levermann, Anders

Issue&Volume: 2025-5-6

Abstract: Within Earth’s climate system, the ocean, cryosphere, and vegetation exhibit hysteresis behavior such that their state depends on their past and not merely on their current boundary conditions. The atmosphere’s fast mixing time scales were thought to inhibit the necessary memory effect for such multistability. Here, we show that moisture accumulation within the atmospheric column generates hysteresis in monsoon circulation independent of oceanic heat storage and yields two stable atmospheric states for the same solar insolation. The dynamics of monsoon rainfall is thus that of a seasonal transition between two stable states. The resulting hysteresis is shown in observational data and reproduced in a general circulation model where it increases with decreasing oceanic memory and exhibits the two distinct states that persist for more than 60 y. They are stabilized by moisture accumulation within the atmospheric column that carries information across time scales much longer than those typical for mixing. The possibility of abrupt shifts between these two states has implications for the future evolution of global monsoon rainfall that is crucial for the agricultural productivity currently feeding more than two billion people.

DOI: 10.1073/pnas.2418093122

Source: https://www.pnas.org/doi/abs/10.1073/pnas.2418093122

期刊信息
PNAS:《美国科学院院刊》,创刊于1914年。隶属于美国科学院,最新IF:12.779
官方网址:https://www.pnas.org