FLUXtrapolation: A benchmark on extrapolating ecosystem fluxes

TL;DR

FLUXtrapolation is a new benchmark designed to evaluate machine learning models' ability to extrapolate ecosystem fluxes across diverse and shifting environmental conditions, aiding scientific understanding of global biogeochemical cycles.

Contribution

It introduces a comprehensive benchmark with scenarios for temporal, spatial, and temperature-based extrapolation, addressing a key challenge in flux upscaling under distribution shifts.

Findings

Baseline models perform similarly on median RMSE but differ on tail errors.

FLUXtrapolation reveals the difficulty of flux upscaling under distribution shifts.

Progress on the benchmark can improve global ecosystem flux estimates.

Abstract

We introduce FLUXtrapolation, a benchmark for extrapolating ecosystem fluxes under progressively harder distribution shifts. Ecosystem fluxes are central to understanding the carbon, water, and energy cycles, yet they can only be measured directly at sparsely located measurement towers. Producing global flux estimates therefore requires training models on observed sites using globally available covariates and predicting in unobserved regions, that is, upscaling. Flux upscaling is a challenging domain generalization problem that is affected by a shift in covariate distribution across climates, ecosystem types, and environmental conditions, as well as by conditional shift: important drivers remain unobserved at global scale. We provide a quantitative analysis of both these shifts in $P_X$ and $P_{Y\mid X}$. FLUXtrapolation is designed based on domain expertise on flux upscaling: it defines temporal, spatial, and temperature-based extrapolation scenarios and evaluates performance across held-out domains, temporal aggregations, and tail errors. In a pilot study, we find that baselines perform similarly under median hourly RMSE, but separate under the proposed tail-focused and multi-scale evaluation. FLUXtrapolation therefore poses a realistic and thus relevant challenge for machine learning methods under distribution shift; at the same time, progress on this benchmark would directly support the scientific goal of improving flux upscaling.