Climate Informatics 2022

Clouds play a vital role both in modulating Earth's radiation budget and shaping the coupled circulation of the atmosphere and ocean, driving regional changes in temperature and precipitation. They are, therefore, opportune for solar radiation management techniques such as Marine Cloud Brightening, and Stratospheric Aerosol Injections. However, these techniques may have unintended consequences on regional climate patterns, which needs to be studied thoroughly. AIBEDO is a hybrid AI model framework developed to resolve the weaknesses of Earth System Models by generating rapid and robust multi-decadal climate projections. We will demonstrate its utility using marine cloud brightening scenarios—to avoid climate tipping points, and consequences of ill-planned interventions. Our hybrid model consists of a Spherical U-Net (spatial), and Multi-timescale LSTM (temporal) module bolstered by physics constraints. Our model will be trained on 19 CMIP6 model ensemble data, and validated against observational reanalysis data. The interactive nature of AIBEDO with parameter controls provides a unique tool for policymakers and modelers alike to perform ‘what-if’ investigations, run climate related scenarios, or identify consequences of ill-planned, or rogue MCB experiments.   Event link: https://ncics.org/news/events/ci2022/