Global land surface 250 m 8 d fraction of absorbed photosynthetically active radiation (FAPAR) product from 2000 to 2021

[摘要] The fraction of absorbed photosynthetically activeradiation (FAPAR) is a critical land surface variable for carbon cyclemodeling and ecological monitoring. Several global FAPAR products have beenreleased and have become widely used; however, spatiotemporal inconsistencyremains a large issue for the current products, and their spatialresolutions and accuracies can hardly meet the user requirements. Aneffective solution to improve the spatiotemporal continuity and accuracy ofFAPAR products is to take better advantage of the temporal information inthe satellite data using deep learning approaches. In this study, the latestversion (V6) of the FAPAR product with a 250 m resolution was generated fromModerate Resolution Imaging Spectroradiometer (MODIS) surface reflectancedata and other information, as part of the Global LAnd Surface Satellite(GLASS) product suite. In addition, it was aggregated to multiple coarserresolutions (up to 0.25 ∘ and monthly). Three existing global FAPARproducts (MODIS Collection 6; GLASS V5; and PRoject for On-Board Autonomy–Vegetation, PROBA-V, V1) were used togenerate the time-series training samples, which were used to develop abidirectional long short-term memory (Bi-LSTM) model. Direct validationusing high-resolution FAPAR maps from the Validation of Land European Remotesensing Instrument (VALERI) and ImagineS networks revealed that the GLASS V6FAPAR product has a higher accuracy than PROBA-V, MODIS, and GLASS V5, withan R 2 value of 0.80 and root-mean-square errors (RMSEs) of 0.10–0.11at the 250 m, 500 m, and 3 km scales, and a higher percentage (72 %) ofretrievals for meeting the accuracy requirement of 0.1. Global spatialevaluation and temporal comparison at the AmeriFlux and National EcologicalObservatory Network (NEON) sites revealed that the GLASS V6 FAPAR has agreater spatiotemporal continuity and reflects the variations in thevegetation better than the GLASS V5 FAPAR. The higher quality of the GLASSV6 FAPAR is attributed to the ability of the Bi-LSTM model, which involveshigh-quality training samples and combines the strengths of the existingFAPAR products, as well as the temporal and spectral information from theMODIS surface reflectance data and other information. The 250 m 8 d GLASS V6 FAPAR product for 2020 is freely available at https://doi.org/10.5281/zenodo.6405564 and https://doi.org/10.5281/zenodo.6430925 (Ma, 2022a, b) as well as at the University of Maryland for 2000–2021( http://glass.umd.edu/FAPAR/MODIS/250m , last access 1 November 2022).

[学科分类] 眼科学