A Footprint-Aware, High-Resolution Approach for Carbon Flux Prediction Across Diverse Ecosystems

TL;DR

This paper introduces FAR, a deep-learning framework that predicts high-resolution carbon flux across diverse ecosystems by integrating satellite data and tower measurements, improving monitoring of ecosystem carbon dynamics.

Contribution

We develop FAR, a novel footprint-aware deep learning model that jointly predicts spatial footprints and pixel-level carbon flux, enabling high-resolution ecosystem monitoring.

Findings

Achieves R2 = 0.78 in predicting monthly net ecosystem exchange.

Effectively combines tower data with satellite imagery for accurate flux estimation.

Demonstrates applicability across various ecosystem types.

Abstract

Natural climate solutions (NCS) offer an approach to mitigating carbon dioxide (CO2) emissions. However, monitoring the carbon drawdown of ecosystems over large geographic areas remains challenging. Eddy-flux covariance towers provide ground truth for predictive 'upscaling' models derived from satellite products, but many satellites now produce measurements on spatial scales smaller than a flux tower's footprint. We introduce Footprint-Aware Regression (FAR), a first-of-its-kind, deep-learning framework that simultaneously predicts spatial footprints and pixel-level (30 m scale) estimates of carbon flux. FAR is trained on our AMERI-FAR25 dataset which combines 439 site years of tower data with corresponding Landsat scenes. Our model produces high-resolution predictions and achieves R2 = 0.78 when predicting monthly net ecosystem exchange on test sites from a variety of ecosystems.