AI as a Bridge between ARM Observationsand E3SM for improving Clouds and Precipitation

Abstract

Science Challenge Improving simulations of clouds and precipitation is the central challenge in reducing present-day biases and future projection uncertainties in Earth System Prediction models such as the DOE Energy Exascale Earth System Model (E3SM). To “advance a robust predictive understanding of Earth’s climate and environmental systems”, the DOE Atmospheric Radiation Measurement (ARM) user facility was created three decades ago for high-quality ground-based observations. While our physical understanding of cloud processes has improved, models continue to suffer from long-standing biases. We believe that AI offers an unprecedented opportunity to improve the prediction of clouds and precipitation by building a bridge between ARM data and E3SM.