Machine Learning (ML) is a rapidly evolving interdisciplinary field with great potential to advance climate downscaling, providing innovative approaches to enhance the performance and scalability of downscaling techniques. Promising results have already emerged, demonstrating its capability to complement and extend standard methods. CORDEX regional activities and Flagship Pilot Studies (FPS) are playing an active role in advancing these developments, contributing to the exploration of new ML downscaling approaches and methods, and fostering collaboration across the climate science community. Continued efforts in this area are expected to drive further innovation and improve our understanding of regional climate processes. Effective coordination across key initiatives is crucial for addressing challenges such as establishing standards, conducting intercomparison and evaluation, identifying and advancing critical research gaps, and organizing collaborative ensemble production.
The primary objective of this CORDEX Task Force (TF) is to develop a strategic plan for coordinating machine learning-based downscaling activities. This will require collaboration and coordination with relevant external initiatives, which will be identified through a landscaping activity to assess the current status of the field. Within CORDEX, the Task Force will work closely with ongoing regional, FPS and TF activities, particularly the TF on Convection, to align km-scale downscaling activities. Additionally, it aims to support contributions to CORDEX-CMIP6/7 downscaling experiments, promote best practices, and facilitate knowledge exchange to address key challenges in the field.
