Best Estimation of Hydrological Data in Subsurface Layers Based on Spatiotemporal Models

Abstract

This paper proposes a newly extended method for estimation of hydrological distributions of subsurface events in space and time based on observed data. To extend spatial estimates into spatiotemporal ones, a conversion parameter of temporal-distance is newly defined and introduced into the kriging equations using trend and covariance functions. This parameter equivalently converts a temporal gap into a spatial distance in time-related terms of those functions. It can be shown that the proposed method with the conversion parameter is useful and applicable to a spatiotemporal estimation via the numerical tests and be concluded that the conjunctive correlation of the ramdom component in space and time is better in this method.