Accurate　and　automatic　monitoring　and　assessment　of　vegetation　changes　are　important　to　support　food　security,　ecosystem　balance　and　global　climate　regulation.　Compared　with　urbanization　and　deforestation,　vegetation　changes　such　as　afforestation,　cropland　reclamation　and　variations　in　cropping　intensity　were　understudied.　This　study　aimed　to　propose　an　Automatic　Method　for　detecting　Multiple　vegetation　Changes　(AMMC)　through　a　knowledge-based　strategy.　Five　temporal　indices　were　proposed　in　order　to　fully　characterize　different　vegetation　types　from　four　aspects:　vegetation　abundance,　temporal　dispersion,　primary/minor　temporal　continuity,　and　growing　season　length.　The　AMMC　takes　advantage　of　the　knowledge　on　the　expected　temporal　trajectories　of　vegetation　dynamics,　which　could　be　tracked　based　on　their　corresponding　trends　in　these　temporal　indices.　The　efficiency　of　the　proposed　AMMC　method　was　verified　with　its　applications　in　central　east　China　using　500　m　8　day　composite　MODIS　datasets　from　2001　to　2016.　An　overall　accuracy　of　94.75%　was　achieved　when　evaluated　with　3,011　reference　sites.　Results　revealed　that　there　were　totally　7,180　km(2),　3,610　km(2)　and　3,280.5　km(2)　areas　of　afforestation,　cropland　reclamation　and　variations　in　cropping　intensity　in　central　east　China,　respectively.　This　study　verified　that　afforestation　efforts　were　succeeded,　but　＂Grain　for　Green＂　project　was　not　as　expected　since　more　cropland　was　reclamation　implemented　at　less　favorable　biophysical　conditions　than　cropland　retirement.