Multi-variable Hard Physical Constraints for Climate Model Downscaling

Jose González-Abad, Álex Hernández-García, Paula Harder, David Rolnick, José Manuel Gutiérrez

arXiv preprint · 02 August 2023

In one sentence. Introduces multi-variable hard constraints so deep-learning downscaled climate variables keep physical relationships, demonstrated on temperature downscaling from GCM output.

Abstract

Global Climate Models (GCMs) are the primary tool to simulate climate evolution and assess the impacts of climate change. However, they often operate at a coarse spatial resolution that limits their accuracy in reproducing local-scale phenomena. Statistical downscaling methods leveraging deep learning offer a solution to this problem by approximating local-scale climate fields from coarse variables, thus enabling regional GCM projections. Typically, climate fields of different variables of interest are downscaled independently, resulting in violations of fundamental physical properties across interconnected variables. This study investigates the scope of this problem and, through an application on temperature, lays the foundation for a framework introducing multi-variable hard constraints that guarantees physical relationships between groups of downscaled climate variables.

Labels

Labels and summary are generated by Claude and reviewed weekly. Wrong? Edit this entry.