Spatiotemporal Evolution Simulation of Volcanic Ash Cloud from Remote Sensing Image

摘要

Aiming at the dynamic monitoring and diffusion forecasting of volcanic ash cloud, this paper presents a new spatiotemporal evolution simulation method of volcanic ash cloud from the satellite remote sensing image. It couples on cellular automata (CA) and artificial neural network (ANN) (i.e., CA-ANN) in terms of cell spatiotemporal database construction, neural network structure confirmation and model training and verifying, and finally realized the simulation and forecast of volcanic ash cloud in space and time. The simulation is performed with test data and the real volcanic ash cloud spatiotemporal evolution case is discussed and verified in our empirical study. Experimental results show that the proposed method has the advantages of high accuracy and good image quality compared with the traditional methods.