Prediction method and system for vegetation wide-range regionalization negative anomaly

The invention discloses a vegetation large-range regionalization negative anomaly prediction method and system, and the method comprises the steps: carrying out the meshing of a research region, and extracting a time sequence of a parameter W, reflecting a vegetation state, of each grid point; preprocessing each lattice point sequence; calculating the number of synchronism negative anomalies among the grid points to obtain synchronism data of negative anomalies among the grid points; judging whether the occurrence of the negative abnormal event among the grid points has synchronism or not according to the synchronism data; and screening a grid point set having synchronism with the to-be-predicted grid point in the region, and predicting a future parameter W negative abnormal event of the to-be-predicted grid point according to the time sequence of the parameter W negative abnormal event of each grid point in the grid point set. According to the method, synchronism consideration is introduced into prediction o