Exploring ENSO-Induced Anomalies over North America in Historical and Future Climate Simulations That Use HadGEM2-ESM Output to Drive WRF

• Published in: Climate (Basel) Vol. 10; no. 8; p. 117
• Main Authors: Tristan Shepherd, Jacob J. Coburn, Rebecca J. Barthelmie, Sara C. Pryor
• Format: Journal Article
• Language: English
• Published: MDPI AG 01-08-2022
• Subjects: climate change; convection-permitting; El Niño Southern Oscillation; global climate modelling; regional climate modelling; weather research and forecasting model
• ISSN: 2225-1154
• DOI: 10.3390/cli10080117

Projected changes to the El Niño Southern Oscillation (ENSO) climate mode have been explored using global Earth system models (ESMs). Regional expressions of such changes have yet to be fully advanced and may require the use of regional downscaling. Here, we employ regional climate modeling (RCM) using the Weather Research and Forecasting (WRF) model at convection-permitting resolution and nested in output from the HadGEM2 ESM. We quantify ENSO teleconnections to temperature and precipitation anomalies in historical and future climate scenarios over eastern North America. Two paired simulations are run, a strong El Niño (positive ENSO phase) and a weak La Niña (negative ENSO phase), for the historical and future years. The HadGEM2 direct output and HadGEM2-WRF simulation output are compared to the anomalies derived from the NOAA ENSO Climate Normals dataset. The near-surface temperature and precipitation differences by ENSO phase, as represented by the HadGEM2-WRF historical simulations, show a poor degree of association with the NOAA ENSO Climate Normals, in part because of the large biases in the HadGEM2 model. Downscaling with the WRF model does improve the agreement with the observations, and large discrepancies remain. The model chain HadGEM2-WRF reverses the sign of the ENSO phase response over eastern North America under simulations of the future climate with high greenhouse gas forcing, but due to the poor agreement with the NOAA ENSO Climate Normals it is difficult to assign confidence to this prediction.