Use of Multiquadric Interpolation for Meteorological Objective Analysis

Abstract

The method of multiquadric interpolation is described and compared to the Barnes and Cressman methods of meteorological objective analysis. The method of multiquadric interpolation uses hyperboloid radial basis functions to fit scattered data to a uniform grid. Results for an analytical function indicate that the method is more accurate than the Barnes or Cressman methods. Application to actual meteorological data indicates that multiquadric interpolation produces excellent analyses that retain small-scale features resolved by the observations in any subregion of the analysis.