A Vegetation Growth Pattern-Constrained Interpolation Method for High-Resolution Daily FPAR/LAI Reconstruction and NPP Spatial Disaggregation

High-spatial-resolution fractions of absorbed photosynthetically active radiation (FPAR) and leaf area index (LAI) are important for characterizing urban vegetation productivity, yet reconstructing their continuous annual dynamics remains challenging. In this study, a phenology-constrained interpolation method was developed to reconstruct annual daily 10 m FPAR and LAI, which are key inputs to MOD17-based net primary productivity (NPP) estimation. Taking Nanjing, China, as a case study, six Sentinel-2 images and 8-day 500 m MOD15A2H FPAR/LAI data in 2023 were integrated. A four-step logistic function was fitted for cropland, while two-step logistic functions were applied to the other six vegetation types. The fitted curves achieved R2 values of 0.69–0.95 for FPAR and 0.82–0.97 for LAI, indicating their capacity to characterize intra-annual vegetation dynamics. Cross-comparison with independent Sentinel-2 retrievals showed mean relative differences of −0.23% and −1.32% and mean absolute relative differences of 5.12% and 7.43% for FPAR and LAI, respectively. The reconstructed parameters were further used to illustrate NPP spatial patterns. The regional mean 10 m NPP differed by only 0.94% from the MOD17 500 m product, while showing stronger spatial heterogeneity. Maximum NPP was 1.31 versus 0.87 kg C·m−2·a−1; pixels below 0.2 kg C·m−2·a−1 accounted for 9.92% versus 0.84%, while those above 0.8 kg C·m−2·a−1 accounted for 12.2% versus 0.10%. Because meteorological forcing was not downscaled and independent ground-based carbon-flux validation was unavailable, absolute NPP estimates should be interpreted cautiously and require further local calibration, whereas the reconstructed high-spatial-resolution FPAR and LAI provide valuable inputs for characterizing fine-scale patterns of urban vegetation productivity.