Spatial interpolation of meteorological fields using a multilevel parametric dynamic stochastic low-order model

Abstract

The paper focuses on a new method of spatial interpolation of air temperature and wind velocity fields in the troposphere. The method is based on Kalman filtering and a multilevel parametric dynamic stochastic low-order model. The key feature of the proposed model is that it has parameters, which are responsible for the altitude levels. Generally, models use so-called “shallow water” (shallow water approximation), and altitude correlation is not taken into account, or they may rely only on mandatory isobaric levels data, thus ignoring the data obtained for significant levels. Standard levels are located at considerable distances in altitude from each other and the altitude correlation there is not usually significant. By using parameters that are responsible for the altitude levels, this model allows us to estimate the effect that information coming from neighbouring altitude levels may have on the final estimate. The paper presents the results of a statistical estimation of the proposed spatial interpolation algorithm. A comparison of the results statistical estimation spatial interpolation of the proposed algorithm with a four-dimensional dynamic-stochastic model is given.