Logo do repositório
 

Scale matters: Spatial resolution impacts tropical leaf phenology characterized by multi-source satellite remote sensing with an ecological-constrained deep learning model

dc.contributor.authorSong, Guangqin
dc.contributor.authorWang, Jing
dc.contributor.authorZhao, Yingyi
dc.contributor.authorYang, Dedi
dc.contributor.authorLee, Calvin K.F.
dc.contributor.authorGuo, Zhengfei
dc.contributor.authorDetto, Matteo
dc.contributor.authorAlberton, Bruna [UNESP]
dc.contributor.authorMorellato, Patricia [UNESP]
dc.contributor.authorNelson, Bruce
dc.contributor.authorWu, Jin
dc.contributor.institutionThe University of Hong Kong
dc.contributor.institutionShenzhen Campus of Sun Yat-sen University
dc.contributor.institutionBrookhaven National Laboratory
dc.contributor.institutionPrinceton University
dc.contributor.institutionUniversidade Estadual Paulista (UNESP)
dc.contributor.institutionInstituto Tecnológico Vale
dc.contributor.institutionNational Institute for Amazon Research (INPA)
dc.date.accessioned2025-04-29T20:08:23Z
dc.date.issued2024-04-01
dc.description.abstractAccurate monitoring of tropical leaf phenology, such as the leaf-on/off status, at both individual and ecosystem scales is essential for understanding and modelling tropical forest carbon and water cycles, and their sensitivity to climate change. The discrepancy between tree-crown size and pixel size (i.e., spatial resolution) across orbital sensors can affect the capability of cross-scale phenology monitoring, an aspect that remains understudied. To examine the impact of spatial resolution on tropical leaf phenology monitoring, we applied a spectral index-guided, ecologically constrained autoencoder (IG-ECAE) to automatically generate a deciduousness metric (i.e., percentage of upper canopy area that is leaf-off status within an image pixel) from simulated VIS-NIR PlanetScope data at a range of resolutions from 3 m to 30 m, as well as from VIS-NIR data of three satellite platforms with the same range of spatial resolutions (3 m PlanetScope, 10 m Sentinel-2, and 30 m Landsat-8). We compared the deciduousness metrics derived from the simulated and satellite data to corresponding measurements derived from WorldView-2 (three sites) and local phenocams (four sites) at five tropical forest sites. Our results revealed that: (1) the IG-ECAE model captured the amount of deciduousness across spatial scales, with the highest accuracy obtained from PlanetScope, followed by Sentinel-2 and Landsat-8; (2) coarser spatial resolutions led to lower accuracies in tropical deciduousness monitoring, as demonstrated by both simulated PlanetScope data across various spatial resolutions and real satellite data; and (3) while not as accurate in capturing fine-scale tropical phenological diversity as PlanetScope, Sentinel-2 provided satisfactory monitoring of deciduousness seasonality at the ecosystem level consistently across all phenocam sites, whereas Landsat-8 failed to do so. Collectively, this study provides a robust assessment for advancing cross-scale tropical leaf phenology monitoring with potential for extension to pan-tropical regions and highlights the impact of spatial resolution on such monitoring efforts.en
dc.description.affiliationResearch Area of Ecology and Biodiversity School of Biological Sciences The University of Hong Kong
dc.description.affiliationSchool of Ecology Shenzhen Campus of Sun Yat-sen University, Guangdong
dc.description.affiliationDepartment of Environmental and Climate Sciences Brookhaven National Laboratory
dc.description.affiliationDepartment of Ecology and Evolutionary Biology Princeton University
dc.description.affiliationDepartment of Biodiversity Bioscience Institute Sao Paulo State University UNESP, Sao Paulo
dc.description.affiliationBiodiversity and Ecosystem Services Instituto Tecnológico Vale
dc.description.affiliationEnvironmental Dynamics Department National Institute for Amazon Research (INPA)
dc.description.affiliationUnespDepartment of Biodiversity Bioscience Institute Sao Paulo State University UNESP, Sao Paulo
dc.description.sponsorshipInnovation and Technology Fund
dc.description.sponsorshipNational Natural Science Foundation of China
dc.description.sponsorshipIdNational Natural Science Foundation of China: 31922090
dc.identifierhttp://dx.doi.org/10.1016/j.rse.2024.114027
dc.identifier.citationRemote Sensing of Environment, v. 304.
dc.identifier.doi10.1016/j.rse.2024.114027
dc.identifier.issn0034-4257
dc.identifier.scopus2-s2.0-85184149192
dc.identifier.urihttps://hdl.handle.net/11449/307089
dc.language.isoeng
dc.relation.ispartofRemote Sensing of Environment
dc.sourceScopus
dc.subjectDeep learning
dc.subjectEcosystem deciduousness
dc.subjectLeaf phenology
dc.subjectPhenological diversity
dc.subjectSatellite remote sensing
dc.subjectSpatial resolution
dc.subjectSpectral unmixing
dc.subjectTropical forest
dc.titleScale matters: Spatial resolution impacts tropical leaf phenology characterized by multi-source satellite remote sensing with an ecological-constrained deep learning modelen
dc.typeArtigopt
dspace.entity.typePublication

Arquivos

Coleções