Research on vegetation dynamic change simulation based on spatial data mining of ANN-CA model using time series of remote sensing images

Abstract

Dynamic change of vegetation has become a very sensitive problem in China due to climate variability and human's disturbances in the Yellow river basin. Dynamic simulation and forecast of vegetation are regarded as an effective measure to decision support for local government. This paper presents a new method to support the local government's effort in ecological protection. In integrates cellular automata (CA) -artificial neural network (ANN) model with Geographical information system (GIS) and remote sensing. The proposed method includes three major steps: (1) to extract control factors; (2) to integrate CA and ANN models; (3) to simulate the selected area using CA-ANN model. The results indicted that the integrated approach can rapidly find condition of future vegetation cover that satisfy requirement of local relative department. It has demonstrated that the proposed method can provide valuable decision support for local government. the result indicts that NDVI of the vegetation has an increasing trend and characteristics of distribution concentration trend, but the change rate is become lower from the year 2007 to 2014 compared with the changes from the year 2000 to 2007. © 2010 IFIP International Federation for Information Processing.