CarbonBench: A Global Benchmark for Upscaling of Carbon Fluxes Using Zero-Shot Learning

ArXi:2603.09868v1 Announce Type: new Accurately quantifying terrestrial carbon exchange is essential for climate policy and carbon accounting, yet models must generalize to ecosystems underrepresented in sparse eddy covariance observations. Despite this challenge being a natural instance of zero-shot spatial transfer learning for time series regression, no standardized benchmark exists to rigorously evaluate model performance across geographically distinct locations with different climate regimes and vegetation types.