Reconstructed NDVI Dataset for the CASA (Carnegie-Ames-Stanford Approach) Model (2019–2023)

This dataset provides a reconstructed Normalized Difference Vegetation Index (NDVI) time series covering the period 2019–2023, developed to support ecological and carbon-cycle modeling applications. The dataset integrates multi-source satellite observations from Landsat 8 OLI and MODIS (MOD13Q1) products and applies a Gap-Filling and Savitzky–Golay (GF–SG) reconstruction method to correct for cloud contamination, sensor noise, and missing observations. It contains two main components: (1) model-ready NDVI composites standardized for direct input into the CASA model for NPP and NEP estimation, and (2) monthly reconstructed NDVI data for vegetation dynamics, phenological monitoring, and long-term trend analysis. All files are provided in GeoTIFF format with a spatial resolution of 30 meters and projected in WGS 84 / UTM Zone 50N. This dataset enables accurate assessment of vegetation productivity, ecological restoration effectiveness, and carbon source/sink variations in regions such as Beijing and similar urban–rural transition zones. The data are made freely available for academic and non-commercial purposes; users must provide proper attribution when using the dataset in research, teaching, or publications.