30m land-cover products (single-date) over China in 2013

Since the traditional large-area land-cover classification takes a lot of manpower for collecting training samples, we proposed a novel and automatic land-cover classification method to solve this difficulty. Specifically, we combined the time-series MOD09A1 reflectance and GlobCover2009 land-cover products to develop a priori spectral library (SPECLib). Then, the single-date land-cover results in 2013 were produced by using the SPECLib, Landsat reflectance, fusion classification method (MLC and SAM) and knowledged-based rules.