A dataset of Global MuSyQ 500 m and 4-day spatiotemporally continuous gross primary productivity product (2000–2020)

Gross Primary Productivity (GPP) of vegetation refers to the total amount of organic matter produced by photosynthesis per unit area per unit time in green plants, and is an important ecological indicator for evaluating the sustainable development of terrestrial ecosystems. At present, the temporal resolution of the global GPP products have a temporal resolution of 8 days. However, satellite observations cannot fully reflect vegetation changes at this temporal resolution. Based on MuSyQ LAI products, NIRv data, MCD12Q1 land cover classification data, ERA5 meteorological data and Koppen climate classification map, this study developed GPP retrieva algorithms for natural vegetation and farmland, respectively, to generate a vegetation GPP product (MuSyQ GPP) with a spatiotemporal resolution of 500 m at a 4-day interval from 2000 to 2022. For natural vegetation, 42 light use efficiency-based GPP models incorporating vegetation functional traits were constructed. For farmland, the estimation method of distinguishing stress impact factors during the growing season was used. Validation results showed good consistency between the MuSyQ GPP product and the FLUXNET tower observations (R²=0.57, RMSE=2.53 gC/m²/day). The product shows strong agreement with flux tower data in deciduous broadleaf forests and mixed forests (R² > 0.69) and captures the spatiotemporal variability of GPP well. For low vegetation sites such as grasslands, open shrublands, and woody savannas, the R² of this product is greater than 0.45, with the lowest RMSE observed in open shrublands (RMSE=1.32 gC/m²/day). For croplands types, the R² of the MuSyQ GPP product is 0.48. This product reduces estimation uncertainties in terrestrial vegetation GPP caused by spatial scaling effects and eco-physiological heterogeneity, and provides important data support for monitoring the carbon cycle of terrestrial ecosystems.