Spatial validation reveals poor predictive performance of large-scale ecological mapping models
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.
Main Authors: | , , , , , , , , , , , , |
---|---|
Format: | article biblioteca |
Language: | eng |
Subjects: | P01 - Conservation de la nature et ressources foncières, K01 - Foresterie - Considérations générales, U30 - Méthodes de recherche, forêt tropicale, écologie, technique de prévision, cartographie, modèle mathématique, modèle de simulation, qualité, défaut, http://aims.fao.org/aos/agrovoc/c_24904, http://aims.fao.org/aos/agrovoc/c_2467, http://aims.fao.org/aos/agrovoc/c_3041, http://aims.fao.org/aos/agrovoc/c_1344, http://aims.fao.org/aos/agrovoc/c_24199, http://aims.fao.org/aos/agrovoc/c_24242, http://aims.fao.org/aos/agrovoc/c_6400, http://aims.fao.org/aos/agrovoc/c_24158, http://aims.fao.org/aos/agrovoc/c_1432, |
Online Access: | http://agritrop.cirad.fr/596514/ http://agritrop.cirad.fr/596514/1/ploton_NC20.pdf |
Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
id |
dig-cirad-fr-596514 |
---|---|
record_format |
koha |
spelling |
dig-cirad-fr-5965142024-01-29T03:02:02Z http://agritrop.cirad.fr/596514/ http://agritrop.cirad.fr/596514/ Spatial validation reveals poor predictive performance of large-scale ecological mapping models. Ploton Pierre, Mortier Frédéric, Rejou-Mechain Maxime, Barbier Nicolas, Picard Nicolas, Rossi Vivien, Dormann Carsten F., Cornu Guillaume, Viennois Gaëlle, Bayol Nicolas, Lyapustin Alexei I., Gourlet-Fleury Sylvie, Pélissier Raphaël. 2020. Nature Communications, 11:4540, 11 p.https://doi.org/10.1038/s41467-020-18321-y <https://doi.org/10.1038/s41467-020-18321-y> Spatial validation reveals poor predictive performance of large-scale ecological mapping models Ploton, Pierre Mortier, Frédéric Rejou-Mechain, Maxime Barbier, Nicolas Picard, Nicolas Rossi, Vivien Dormann, Carsten F. Cornu, Guillaume Viennois, Gaëlle Bayol, Nicolas Lyapustin, Alexei I. Gourlet-Fleury, Sylvie Pélissier, Raphaël eng 2020 Nature Communications P01 - Conservation de la nature et ressources foncières K01 - Foresterie - Considérations générales U30 - Méthodes de recherche forêt tropicale écologie technique de prévision cartographie modèle mathématique modèle de simulation qualité défaut http://aims.fao.org/aos/agrovoc/c_24904 http://aims.fao.org/aos/agrovoc/c_2467 http://aims.fao.org/aos/agrovoc/c_3041 http://aims.fao.org/aos/agrovoc/c_1344 http://aims.fao.org/aos/agrovoc/c_24199 http://aims.fao.org/aos/agrovoc/c_24242 http://aims.fao.org/aos/agrovoc/c_6400 http://aims.fao.org/aos/agrovoc/c_24158 Afrique centrale http://aims.fao.org/aos/agrovoc/c_1432 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. article info:eu-repo/semantics/article Journal Article info:eu-repo/semantics/publishedVersion http://agritrop.cirad.fr/596514/1/ploton_NC20.pdf text cc_by info:eu-repo/semantics/openAccess https://creativecommons.org/licenses/by/4.0/ https://doi.org/10.1038/s41467-020-18321-y 10.1038/s41467-020-18321-y info:eu-repo/semantics/altIdentifier/doi/10.1038/s41467-020-18321-y info:eu-repo/semantics/altIdentifier/purl/https://doi.org/10.1038/s41467-020-18321-y info:eu-repo/semantics/reference/purl/https://rdcu.be/b7ofW info:eu-repo/grantAgreement/EC/H2020/696356//(EU) ERA-NET for Monitoring and Mitigation of Greenhouse Gases from Agri- and Silvi-Culture/ERA-GAS |
institution |
CIRAD FR |
collection |
DSpace |
country |
Francia |
countrycode |
FR |
component |
Bibliográfico |
access |
En linea |
databasecode |
dig-cirad-fr |
tag |
biblioteca |
region |
Europa del Oeste |
libraryname |
Biblioteca del CIRAD Francia |
language |
eng |
topic |
P01 - Conservation de la nature et ressources foncières K01 - Foresterie - Considérations générales U30 - Méthodes de recherche forêt tropicale écologie technique de prévision cartographie modèle mathématique modèle de simulation qualité défaut http://aims.fao.org/aos/agrovoc/c_24904 http://aims.fao.org/aos/agrovoc/c_2467 http://aims.fao.org/aos/agrovoc/c_3041 http://aims.fao.org/aos/agrovoc/c_1344 http://aims.fao.org/aos/agrovoc/c_24199 http://aims.fao.org/aos/agrovoc/c_24242 http://aims.fao.org/aos/agrovoc/c_6400 http://aims.fao.org/aos/agrovoc/c_24158 http://aims.fao.org/aos/agrovoc/c_1432 P01 - Conservation de la nature et ressources foncières K01 - Foresterie - Considérations générales U30 - Méthodes de recherche forêt tropicale écologie technique de prévision cartographie modèle mathématique modèle de simulation qualité défaut http://aims.fao.org/aos/agrovoc/c_24904 http://aims.fao.org/aos/agrovoc/c_2467 http://aims.fao.org/aos/agrovoc/c_3041 http://aims.fao.org/aos/agrovoc/c_1344 http://aims.fao.org/aos/agrovoc/c_24199 http://aims.fao.org/aos/agrovoc/c_24242 http://aims.fao.org/aos/agrovoc/c_6400 http://aims.fao.org/aos/agrovoc/c_24158 http://aims.fao.org/aos/agrovoc/c_1432 |
spellingShingle |
P01 - Conservation de la nature et ressources foncières K01 - Foresterie - Considérations générales U30 - Méthodes de recherche forêt tropicale écologie technique de prévision cartographie modèle mathématique modèle de simulation qualité défaut http://aims.fao.org/aos/agrovoc/c_24904 http://aims.fao.org/aos/agrovoc/c_2467 http://aims.fao.org/aos/agrovoc/c_3041 http://aims.fao.org/aos/agrovoc/c_1344 http://aims.fao.org/aos/agrovoc/c_24199 http://aims.fao.org/aos/agrovoc/c_24242 http://aims.fao.org/aos/agrovoc/c_6400 http://aims.fao.org/aos/agrovoc/c_24158 http://aims.fao.org/aos/agrovoc/c_1432 P01 - Conservation de la nature et ressources foncières K01 - Foresterie - Considérations générales U30 - Méthodes de recherche forêt tropicale écologie technique de prévision cartographie modèle mathématique modèle de simulation qualité défaut http://aims.fao.org/aos/agrovoc/c_24904 http://aims.fao.org/aos/agrovoc/c_2467 http://aims.fao.org/aos/agrovoc/c_3041 http://aims.fao.org/aos/agrovoc/c_1344 http://aims.fao.org/aos/agrovoc/c_24199 http://aims.fao.org/aos/agrovoc/c_24242 http://aims.fao.org/aos/agrovoc/c_6400 http://aims.fao.org/aos/agrovoc/c_24158 http://aims.fao.org/aos/agrovoc/c_1432 Ploton, Pierre Mortier, Frédéric Rejou-Mechain, Maxime Barbier, Nicolas Picard, Nicolas Rossi, Vivien Dormann, Carsten F. Cornu, Guillaume Viennois, Gaëlle Bayol, Nicolas Lyapustin, Alexei I. Gourlet-Fleury, Sylvie Pélissier, Raphaël Spatial validation reveals poor predictive performance of large-scale ecological mapping models |
description |
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. |
format |
article |
topic_facet |
P01 - Conservation de la nature et ressources foncières K01 - Foresterie - Considérations générales U30 - Méthodes de recherche forêt tropicale écologie technique de prévision cartographie modèle mathématique modèle de simulation qualité défaut http://aims.fao.org/aos/agrovoc/c_24904 http://aims.fao.org/aos/agrovoc/c_2467 http://aims.fao.org/aos/agrovoc/c_3041 http://aims.fao.org/aos/agrovoc/c_1344 http://aims.fao.org/aos/agrovoc/c_24199 http://aims.fao.org/aos/agrovoc/c_24242 http://aims.fao.org/aos/agrovoc/c_6400 http://aims.fao.org/aos/agrovoc/c_24158 http://aims.fao.org/aos/agrovoc/c_1432 |
author |
Ploton, Pierre Mortier, Frédéric Rejou-Mechain, Maxime Barbier, Nicolas Picard, Nicolas Rossi, Vivien Dormann, Carsten F. Cornu, Guillaume Viennois, Gaëlle Bayol, Nicolas Lyapustin, Alexei I. Gourlet-Fleury, Sylvie Pélissier, Raphaël |
author_facet |
Ploton, Pierre Mortier, Frédéric Rejou-Mechain, Maxime Barbier, Nicolas Picard, Nicolas Rossi, Vivien Dormann, Carsten F. Cornu, Guillaume Viennois, Gaëlle Bayol, Nicolas Lyapustin, Alexei I. Gourlet-Fleury, Sylvie Pélissier, Raphaël |
author_sort |
Ploton, Pierre |
title |
Spatial validation reveals poor predictive performance of large-scale ecological mapping models |
title_short |
Spatial validation reveals poor predictive performance of large-scale ecological mapping models |
title_full |
Spatial validation reveals poor predictive performance of large-scale ecological mapping models |
title_fullStr |
Spatial validation reveals poor predictive performance of large-scale ecological mapping models |
title_full_unstemmed |
Spatial validation reveals poor predictive performance of large-scale ecological mapping models |
title_sort |
spatial validation reveals poor predictive performance of large-scale ecological mapping models |
url |
http://agritrop.cirad.fr/596514/ http://agritrop.cirad.fr/596514/1/ploton_NC20.pdf |
work_keys_str_mv |
AT plotonpierre spatialvalidationrevealspoorpredictiveperformanceoflargescaleecologicalmappingmodels AT mortierfrederic spatialvalidationrevealspoorpredictiveperformanceoflargescaleecologicalmappingmodels AT rejoumechainmaxime spatialvalidationrevealspoorpredictiveperformanceoflargescaleecologicalmappingmodels AT barbiernicolas spatialvalidationrevealspoorpredictiveperformanceoflargescaleecologicalmappingmodels AT picardnicolas spatialvalidationrevealspoorpredictiveperformanceoflargescaleecologicalmappingmodels AT rossivivien spatialvalidationrevealspoorpredictiveperformanceoflargescaleecologicalmappingmodels AT dormanncarstenf spatialvalidationrevealspoorpredictiveperformanceoflargescaleecologicalmappingmodels AT cornuguillaume spatialvalidationrevealspoorpredictiveperformanceoflargescaleecologicalmappingmodels AT viennoisgaelle spatialvalidationrevealspoorpredictiveperformanceoflargescaleecologicalmappingmodels AT bayolnicolas spatialvalidationrevealspoorpredictiveperformanceoflargescaleecologicalmappingmodels AT lyapustinalexeii spatialvalidationrevealspoorpredictiveperformanceoflargescaleecologicalmappingmodels AT gourletfleurysylvie spatialvalidationrevealspoorpredictiveperformanceoflargescaleecologicalmappingmodels AT pelissierraphael spatialvalidationrevealspoorpredictiveperformanceoflargescaleecologicalmappingmodels |
_version_ |
1792500020582285312 |