Comparative study of data assimilation methods in environmental models

摘要

In this paper two data assimilation methods are proposed for the estimation of air pollution. From the analysis and experimental results, it is concluded that if the isotropic and intrinsic hypothesis is satisfied, the Kriging approach is a simple data assimilation method with acceptable accuracy in case the number of sampling points is relatively large. A data assimilation method using Kalman filtering can provide a more accurate estimation, and it can also be implemented with only a few observation points. The computation burden using Kalman filtering is however heavier than that using Kriging. In our future research the data assimilation methods proposed in this paper are going to be put into a real life application.