GCAM-Gym offers common interface to test climate policy methods
GCAM-Gym: A Benchmark for Agent-Based Climate Policy Search
Computers and Society
Summary
Testing different methods to find the best climate policies is tricky because of many differences in how policies are controlled and evaluated. The authors created GCAM-Gym, a tool that allows many methods to be compared fairly using the same way to interact with a climate model. They ran tests on seven different methods across three types of challenges and found that one method called CMA-ES performed the best under limited testing attempts. They also provided the data and tools so others can see or repeat their comparisons without running the full climate model themselves.
What this means in practice
- •For climate policy analysts: Compare and optimize different climate strategies using a single consistent testing framework.
- •For ai algorithm developers: Test and benchmark reinforcement learning and search algorithms on realistic climate policy problems within GCAM-Gym.
Authors
Tianyu Zhang, Zhenghan Tai, Xinyu Wang, Irina Rish, Yoshua Bengio
Abstract
Comparing climate-policy search methods requires accounting for differences in policy control, available observations, feedback timing, and simulator budgets. We introduce GCAM-Gym, a Gymnasium-compatible environment that supports full-policy evaluation and stepwise control through a common interface and exposes the components of the policy objective. We compare seven search and control systems in three settings: target attainment and cost, cost-temperature trade-offs, and one-step delayed feedback. Each system is evaluated with ten seeds per setting, for a total of 3,210 attempted GCAM evaluations. Four systems share a budget of 24 attempts per seed; among these, CMA-ES achieves the highest interquartile mean (IQM) of the best returns from completed policy evaluations in all three settings. We release the evaluation archive and analysis tools to reconstruct the reported comparisons without running GCAM.