Towards Causal Representations of Climate Model Data
This work addresses the need for more interpretable and robust climate model emulation for climate scientists and policymakers, but it appears incremental as it builds on existing causal methods without claiming major breakthroughs.
The paper tackled the problem of improving the generalizability and interpretability of machine learning-based climate model emulators by exploring causal representation learning, specifically the CDSD method, and evaluated it on climate datasets to highlight challenges and promise.
Climate models, such as Earth system models (ESMs), are crucial for simulating future climate change based on projected Shared Socioeconomic Pathways (SSP) greenhouse gas emissions scenarios. While ESMs are sophisticated and invaluable, machine learning-based emulators trained on existing simulation data can project additional climate scenarios much faster and are computationally efficient. However, they often lack generalizability and interpretability. This work delves into the potential of causal representation learning, specifically the \emph{Causal Discovery with Single-parent Decoding} (CDSD) method, which could render climate model emulation efficient \textit{and} interpretable. We evaluate CDSD on multiple climate datasets, focusing on emissions, temperature, and precipitation. Our findings shed light on the challenges, limitations, and promise of using CDSD as a stepping stone towards more interpretable and robust climate model emulation.