Central Asia (30&degN-60&degN, 50&degE-100&degE) is one of the typical arid and semi-arid regions in the world. Based on GIMMS ndvi3g v1 long-time series of NDVI, the threshold-based method (Jonsson and Eklundh, 2002) and inflexion-based method (Zhang, 2003) were used to extract the surface phenology data in Central Asia from 1982 to 2015. The spatial reference system is geographic coordinate system at a spatial resolution of 0.0833&deg (about 8 km). The dataset is consisted of two groups of data files: Group V1 and Group V2. The processing of the Group V1 development is as follows: first of all, the NDVI data were filtered and reconstructed by Double logistic function, and then the start of season (SOS), end of season (EOS), and length of season (LOS) were calculated by 20% dynamic threshold method. Meanwhile, the Group V2 was developed following the processing that the NDVI time series data was reconstructed using a series of piece wise logistic functions, and then the SOS, EOS, and LOS were calculated by the inflexion-based method. The dataset is archived in .hdr, .img and .tif data formats in 408 data files, with the data size of 126 MB (Compressed to 75.6 MB in one file).

Ma Yonggang, Liu Suhong. Yearly Data Set of Land Surface Phenology in Central Asia (1982-2015) [DB/J]. Global Change Data Warehouse, 2020. DOI: 10.3974/geodb.2020.01.05.V1