Advancements of Space-Time Modeling in Environmental Statistics

Abstract: The project will review some essential drawbacks in the previous research and propose the necessary advancements in space-time modeling of environmental events. In this scenario, environmental statistics has become the most crucial tool to deal with this. In this project we choose two important types of events one is air pollution and the other is rainfall. To introduce the proposed advancements, we illustrate the space-time pattern of an environmental event for example air pollution, and its essential covariates in detail. We measure the impact of a social event on that air pollution statistically. We demonstrate the way of temporal  pattern analysis of an environmental event and describe a new probabilistic time-series prediction approach. We describe a new vine copula-based data-driven  spatial interpolation algorithm at different temporal replicates along the gaussian spatial random field. At last, we overview the extension of this spatial interpolation to the hybrid spatio-temporal interpolation algorithm. We demonstrate the possibility of the future extension of this project such as stationary Gaussian-Weibull process, complex blended spatial random field, etc.