# The FLUXCOM ensemble of global land-atmosphere energy fluxes

arXiv: https://arxiv.org/abs/1812.04951 · HTML: https://mlsanity.com/paper/1812.04951/

## Key facts

- Cited by 8 papers in the map (data through 14 Oct 2026).
- First posted on arXiv 11 Dec 2018.

## Built on by (most cited first, 8 of 8)

- [OpenForest: A data catalogue for machine learning in forest monitoring](https://mlsanity.com/paper/2311.00277/) (2023)
- [Differentiable modeling to unify machine learning and physical models and advance Geosciences](https://mlsanity.com/paper/2301.04027/) (2023)
- [Recurrent Neural Networks for Modelling Gross Primary Production](https://mlsanity.com/paper/2404.12745/) (2024)
- [Machine Learning Research Trends in Africa: A 30 Years Overview with Bibliometric Analysis Review](https://mlsanity.com/paper/2304.07542/) (2023)
- [Deep Learning and Earth Observation to Support the Sustainable Development Goals](https://mlsanity.com/paper/2112.11367/) (2021)
- [Machine Learning Workflows in Climate Modeling: Design Patterns and Insights from Case Studies](https://mlsanity.com/paper/2510.03305/) (2025)
- [The Rise of AI in Weather and Climate Information and its Impact on Global Inequality](https://mlsanity.com/paper/2603.05710/) (2026)
- [Earth Observation Foundation Models for Terrestrial Ecohydrology: From Representation Learning to Process Inference](https://mlsanity.com/paper/2608.15282/) (2026)

## Abstract (arXiv)

Although a key driver of Earth's climate system, global land-atmosphere energy fluxes are poorly constrained. Here we use machine learning to merge energy flux measurements from FLUXNET eddy covariance towers with remote sensing and meteorological data to estimate net radiation, latent and sensible heat and their uncertainties. The resulting FLUXCOM database comprises 147 global gridded products in two setups: (1) 0.0833deg resolution using MODIS remote sensing data (RS) and (2) 0.5deg resolution using remote sensing and meteorological data (RS+METEO). Within each setup we use a full factorial design across machine learning methods, forcing datasets and energy balance closure corrections. For RS and RS+METEO setups respectively, we estimate 2001-2013 global (± 1 standard deviation) net radiation as 75.8±1.4 W m-2 and 77.6±2 W m-2, sensible heat as 33±4 W m-2 and 36±5 W m-2, and evapotranspiration as 75.6±10 × 103 km3 yr-1 and 76±6 × 103 km3 yr-1. FLUXCOM products are suitable to quantify global land-atmosphere interactions and benchmark land surface model simulations.