Global terrestrial gross primary productivity time series across satellite platforms and spatial resolutions: Supporting data for interannual variability analysis

This dataset contains globally aggregated gross primary productivity (GPP) time series supporting the interannual variability analysis in Figure 6 of the manuscript “Towards Consistent Carbon Monitoring: The Role of Satellite Resolution and the Magnifying Effects of Machine Learning in Terrestrial Carbon Cycle Modelling”.The archive includes 24 CSV files containing UFLUX-derived GPP estimates based on MODIS, Landsat-5, Landsat-7, Landsat-8, and Sentinel-2 inputs, together with benchmark time series from FluxSat, CARDAMOM, and the TRENDY v12 ensemble. The UFLUX experiments examine the influence of satellite platform and spatial resolution on the magnitude and temporal variability of global terrestrial GPP.Temporal coverage varies by dataset. MODIS and Landsat-7 time series cover 2001–2022, Landsat-5 covers 2001–2011, Landsat-8 covers 2014–2022, and Sentinel-2 covers 2019–2022. Benchmark records cover 2001–2019 for FluxSat, 2003–2022 for CARDAMOM, and 2000–2022 for TRENDY v12. The archive contains both monthly and annual records, with some missing months in individual time series.Each CSV includes a time column and GPP estimates. CARDAMOM files additionally provide the 2.5th and 97.5th percentile values, while the TRENDY v12 file provides the ensemble mean and the 25th and 75th percentile values.