Modeling Space–Time Dynamics of Aerosols Using Satellite Data and Atmospheric Transport Model Output

Kernel-based models for space–time data offer a flexible and descriptive framework for studying atmospheric processes. Nonstationary and anisotropic covariance structures can be readily accommodated by allowing kernel parameters to vary over space and time. In addition, dimension reduction strategies make model fitting computationally feasible for large datasets. Fitting these models to data derived from instruments onboard satellites, which often contain significant amounts of missingness due to cloud cover and retrieval errors, can be difficult. In this paper, we propose to overcome the challenges of missing satellite-derived data by supplementing an analysis with output from a computer model, which contains valuable information about the space–time dependence structure of the process of interest. We illustrate our approach through a case study of aerosol optical depth across mainland Southeast Asia. We include a cross-validation study to assess the strengths and weaknesses of our approach.