Concerning the robustness of predicting the forced response of the El Niño—Southern Oscillation-Indian summer monsoon (ENSO-ISM) teleconnection based on 9 CMIP-class models, we come to distinguish three time periods. (1) In the late 20th century, the trend was insignificant in most models, which does not permit a statement on robustness in this period. Thus the weakening of the teleconnection reported from the observational data might not be truly forced. (2) Furthermore, we find it typical that under global warming in the first part of the 21st century, the teleconnection is strengthening or non-decreasing. This considerable inter-model robustness is owed to an increasing ENSO variability as well as coupling strength. (3) At the end of the 21st century, however, under strong forcing, the teleconnection change is not robustly modeled: the ENSO variability change is not projected robustly across models, either with respect to the start or the rate of the ENSO variance decline, competing this time with an increase of the coupling strength. The difference in the projected coupling coefficient between models is mainly attributed to the disagreement in the projected ENSO-induced changes in the regional Hadley cell and SST patterns over the ISM domain, despite the agreement on the projected eastward shift of the Walker cell over the Pacific Ocean.

关于基于9个CMIP级模型预测厄尔尼诺-南方涛动-印度夏季风（ENSO-ISM）强迫响应的鲁棒性，我们区分了三个时间段。(1) 在 20 世纪末，大多数模型的趋势并不显着，因此不允许对这一时期的稳健性进行陈述。因此，观测数据报告的遥相关性的减弱可能并不是真正被迫的。(2)此外，我们发现，在21世纪上半叶全球变暖的情况下，遥相关性正在加强或没有减少，这是典型的。这种相当大的模型间稳健性归因于 ENSO 变异性和耦合强度的增加。(3) 然而，到了 21 世纪末，在强强迫作用下，遥相关变化没有得到稳健的建模：无论是相对于 ENSO 方差下降的开始还是速率，ENSO 变率变化都没有在模型中稳健地预测，这次与耦合强度的增加进行竞争。模型之间预测耦合系数的差异主要归因于对 ENSO 引起的区域哈德来环流和 ISM 域海温模式的预测变化的分歧，尽管对太平洋上沃克环流的预测东移达成了一致。海洋。