Annual NDVI and FVC product with 30 m Spatial Resolution Over Qinghai-Tibetan Plateau

The dataset was produced by the following steps. The time-series Landsat surface reflectance was firstly generated as original time-series NDVI, and then the reconstruction method considering changing trends and seasonality and S-G filtering method were used to reconstruct the time-series NDVI to obtain a new and complete time-series NDVI. All the NDVI data during the annual growing season were subsequently averaged to obtain the annual average NDVI in the growing season, and the dimidiate pixel model was applied to estimate the annual average FVC in the growing season. Finally annual NDVI and FVC composite products with 30m spatial resolution were generated over the Qinghai -Tibetan Plateau in 1990, 2000, 2010, 2020, 2023. The accuracy validated results compared with the published products and actual values shows that the annual NDVI product is reliable，reasonable, and line with the actual situation; the temporal spatial distribution of the FVC product is consistent with that of GLASS FVC product and the 30m/15day FVC product over China; the R-square between estimated FVC and the actual FVC is about 0.93, the RMSE is less than 10%. The dataset solves the underestimation phenomenon caused by large numbers of cloudy contaminated values when the original Landsat data is directly used to genenerat annual composite products. The dataset can be applied to analyze and evaluate changes in vegetation ecosystem status and the impacts of climate changes in the Qinghai-Tibetan Plateau. The results were stored in GeoTiff format.