Centre de documentation
scientifique

A hybrid training approach for leaf area index estimation via Cubist and random forests machine-learning / Rasmus M. Houborg in ISPRS Journal of photogrammetry and remote sensing, vol 135 (January 2018)  

Public [article]  inISPRS Journal of photogrammetry and remote sensing > vol 135 (January 2018) . - pp 173 - 188 Titre : A hybrid training approach for leaf area index estimation via Cubist and random forests machine-learning Type de document : Article/Communication Auteurs : Rasmus M. Houborg, Auteur ; Matthew F. McCabe, Auteur Année de publication : 2018 Article en page(s) : pp 173 - 188 Note générale : Bibliographie Langues : Anglais (eng) Descripteur : [Vedettes matières IGN] Traitement d'image [Termes IGN] classification par forêts d'arbres décisionnels [Termes IGN] image RapidEye [Termes IGN] Leaf Area Index [Termes IGN] réflectance de surface [Termes IGN] régression Résumé : (Auteur) With an increasing volume and dimensionality of Earth observation data, enhanced integration of machine-learning methodologies is needed to effectively analyze and utilize these information rich datasets. In machine-learning, a training dataset is required to establish explicit associations between a suite of explanatory ‘predictor’ variables and the target property. The specifics of this learning process can significantly influence model validity and portability, with a higher generalization level expected with an increasing number of observable conditions being reflected in the training dataset. Here we propose a hybrid training approach for leaf area index (LAI) estimation, which harnesses synergistic attributes of scattered in-situ measurements and systematically distributed physically based model inversion results to enhance the information content and spatial representativeness of the training data. To do this, a complimentary training dataset of independent LAI was derived from a regularized model inversion of RapidEye surface reflectances and subsequently used to guide the development of LAI regression models via Cubist and random forests (RF) decision tree methods. The application of the hybrid training approach to a broad set of Landsat 8 vegetation index (VI) predictor variables resulted in significantly improved LAI prediction accuracies and spatial consistencies, relative to results relying on in-situ measurements alone for model training. In comparing the prediction capacity and portability of the two machine-learning algorithms, a pair of relatively simple multi-variate regression models established by Cubist performed best, with an overall relative mean absolute deviation (rMAD) of ∼11%, determined based on a stringent scene-specific cross-validation approach. In comparison, the portability of RF regression models was less effective (i.e., an overall rMAD of ∼15%), which was attributed partly to model saturation at high LAI in association with inherent extrapolation and transferability limitations. Explanatory VIs formed from bands in the near-infrared (NIR) and shortwave infrared domains (e.g., NDWI) were associated with the highest predictive ability, whereas Cubist models relying entirely on VIs based on NIR and red band combinations (e.g., NDVI) were associated with comparatively high uncertainties (i.e., rMAD ∼ 21%). The most transferable and best performing models were based on combinations of several predictor variables, which included both NDWI- and NDVI-like variables. In this process, prior screening of input VIs based on an assessment of variable relevance served as an effective mechanism for optimizing prediction accuracies from both Cubist and RF. While this study demonstrated benefit in combining data mining operations with physically based constraints via a hybrid training approach, the concept of transferability and portability warrants further investigations in order to realize the full potential of emerging machine-learning techniques for regression purposes. Numéro de notice : A2018-070 Affiliation des auteurs : non IGN Thématique : FORET/IMAGERIE Nature : Article nature-HAL : ArtAvecCL-RevueIntern DOI : 10.1016/j.isprsjprs.2017.10.004 En ligne : https://doi.org/10.1016/j.isprsjprs.2017.10.004 Format de la ressource électronique : URL article Permalink : https://documentation.ensg.eu/index.php?lvl=notice_display&id=89428 [article]

Exemplaires(3)

Code-barres	Cote	Support	Localisation	Section	Disponibilité
081-2018011	RAB	Revue	Centre de documentation	En réserve L003	Disponible
081-2018012	DEP-EAF	Revue	Nancy	Dépôt en unité	Exclu du prêt
081-2018013	DEP-EXM	Revue	Saint-Mandé	Dépôt en unité	Exclu du prêt

Regional simulation of ecosystem CO2 and water vapor exchange for agricultural land using NOAA AVHRR and Terra MODIS satellite data: Application to Zealand, Denmark / Rasmus M. Houborg in Remote sensing of environment, vol 93 n° 1 (30/10/2004)  

Public [article]  inRemote sensing of environment > vol 93 n° 1 (30/10/2004) . - pp 150 - 167 Titre : Regional simulation of ecosystem CO2 and water vapor exchange for agricultural land using NOAA AVHRR and Terra MODIS satellite data: Application to Zealand, Denmark Type de document : Article/Communication Auteurs : Rasmus M. Houborg, Auteur ; H. Soegaard, Auteur Année de publication : 2004 Article en page(s) : pp 150 - 167 Note générale : Bibliographie Langues : Anglais (eng) Descripteur : [Vedettes matières IGN] Applications de télédétection [Termes IGN] agriculture [Termes IGN] atmosphère terrestre [Termes IGN] couvert végétal [Termes IGN] covariance [Termes IGN] Danemark [Termes IGN] dioxyde de carbone [Termes IGN] flux [Termes IGN] Green Leaf Area Index [Termes IGN] image NOAA-AVHRR [Termes IGN] image Terra-MODIS [Termes IGN] indice de végétation [Termes IGN] nuage [Termes IGN] turbulence [Termes IGN] vapeur d'eau Résumé : (Auteur) While accurate information on ecosystem C02 and water vapor exchange is available at eddy covariance flux tower sites, method, methods to expand predictions of C02 and energy exchange to regional or global scales with high fidelity are lacking. The main objective of this study was to examine the applicability of land surface and atmospheric products from the Moderate Resolution Imaging Spectroradiometer (MODIS) and the Advanced Very High Resolution Radiometer (AVHRR) for assessing the spatial variation in C02 and water vapor fluxes for cloudless agricultural land pixels at the Island of Zealand, Denmark. The spatial distribution of green leaf area index, directbeam ark: diffuse solar radiation and air humidity was inferred on the basis of late morning MODIS data that was combined with afternoon AVHRR data to resolve the diurnal variation in air and surface temperature. These variables were used in a coupled "twoleaf' ecosystem model operating at an hourly time scale. The enhanced vegetation index (EVI) was strongly correlated with field measurements of green leaf area index (r2=0.91) and remained sensitive to variations in green biomass up to green leaf area indices of 45. Evaluation against standard meteorological data showed that instantaneous estimates of air temperature, actual vapor pressure and incoming solar radiation could be retrieved with overall root mean square errors of 2.5°C, 138.3 Pa and 47.7 Wm2, respectively. The combination of late morning and afternoon inferences made it possible to resolve the diurnal course in key model parameters, and predicted rates of ecosystem C02 and water vapor exchange were comparable to eddy covariance measurements at a single flux tower. A large spatial diversity in C02 and water vapor exchange was maintained throughout the study period due to significant regional variations in meteorological input variables and large spatial differences in canopy development. The results of this study stress the necessity of pixel based estimates for an accurate evaluation of regional budgets of C02 and water vapor exchange. Numéro de notice : A2004-426 Affiliation des auteurs : non IGN Thématique : IMAGERIE Nature : Article nature-HAL : ArtAvecCL-RevueIntern DOI : 10.1016/j.rse.2004.07.001 En ligne : https://doi.org/10.1016/j.rse.2004.07.001 Format de la ressource électronique : URL article Permalink : https://documentation.ensg.eu/index.php?lvl=notice_display&id=26953 [article]