Ploton Pierre auteur aut IRD Mortier F. auteur aut IRD Réjou-Méchain Maxime auteur aut IRD Barbier Nicolas auteur aut IRD Picard N. auteur aut IRD Rossi V. auteur aut IRD Dormann C. auteur aut IRD Cornu G. auteur aut IRD Viennois G. auteur aut IRD Bayol N. auteur aut IRD Lyapustin A. auteur aut IRD Gourlet-Fleury S. auteur aut IRD Pélissier Raphaël auteur aut IRD text journalArticle eng text/pdf reformatted digital access Mapping aboveground forest biomass is central for assessing the global carbon balance. However, current large-scale maps show strong disparities, despite good validation statistics of their underlying models. Here, we attribute this contradiction to a flaw in the validation methods, which ignore spatial autocorrelation (SAC) in data, leading to overoptimistic assessment of model predictive power. To illustrate this issue, we reproduce the approach of large-scale mapping studies using a massive forest inventory dataset of 11.8 million trees in central Africa to train and validate a random forest model based on multispectral and environmental variables. A standard nonspatial validation method suggests that the model predicts more than half of the forest biomass variation, while spatial validation methods accounting for SAC reveal quasi-null predictive power. This study underscores how a common practice in big data mapping studies shows an apparent high predictive power, even when predictors have poor relationships with the ecological variable of interest, thus possibly leading to erroneous maps and interpretations. specialized AFRIQUE CENTRALE 082 021 020 11 1 4540 [11 ] 2020 2041-1723 https://www.documentation.ird.fr/hor/fdi:010080646 10.1038/s41467-020-18321-y 2041-1723 [F B010080646] https://www.documentation.ird.fr/hor/fdi:010080646 https://horizon.documentation.ird.fr/exl-doc/pleins_textes/divers21-02/010080646.pdf IRD - Base Horizon / Pleins textes 2021-03-10 2025-02-24 fdi:010080646 fre