Analysis of ENSO's Response to Unforced Variability and Anthropogenic Forcing Using CESM
The El Niño-Southern Oscillation (ENSO) significantly influences global temperature and precipitation patterns. Understanding how ENSO may change with climate is a major challenge, given the internal (natural) variability of the system and relatively short observational record. Earth system modeling approaches that capture the effect of natural variability, or more precisely the variability within the coupled system in the absence of time-varying external forcing, can be used to characterize the importance of internal variability when considering potential anthropogenic climate trends. Here we use a low resolution CESM ensemble sampling the joint effect of coupled ocean and atmosphere internal variability to analyze potential ENSO responses to anthropogenic forcing.
We find the ENSO statistics within the ensemble are relatively stable for pre-industrial, historical, and future conditions, though we find large variability between different simulations due to effects of internal variability. The primary role of natural modulations in this ensemble highlights the importance of careful assessment of ocean-atmosphere internal variability in ENSO projections.
Here we analyze the effect of coupled internal variability on changes in ENSO under anthropogenic global warming using the Community Earth System Model (CESM). We present results from a ~5,000-year control run with constant pre-industrial conditions and a 50-member climate change ensemble experiment, consisting of historical hindcasts (1850–2005) and future projections to 2100 following representative concentration pathway 8.5 (RCP8.5). Given this large single-model ensemble, we are able to use simple statistical analyses to compare the effects of anthropogenic climate change with the effects of natural modulations in ENSO sea surface temperature (SST) metrics, as well as how internal variability may change with global warming. Changes in eastern Pacific ENSO SST metrics due to climate change are secondary to the model’s natural modulations; however, central Pacific ENSO amplitude significantly decreases, to an extent comparable with natural modulations. We also assess the sensitivity of internal variability estimates to ensemble size.