An Iterative Algorithm for Objective Wind Field Analysis

View More View Less
  • 1 Department of Mechanics and Structures, University of California, Los Angeles 90024
© Get Permissions
Full access

Abstract

A wind field interpolated from a limited supply of observed data often implies a specific field of divergence, or a violation of the conservation of mass. To minimize the divergence, various objective analyses have been proposed. Three different methods of analysis are investigated and compared with respect to the degree of minimization of wind divergence and the accuracy of wind data at a measured station. The methods are a variational formulation in which the residual error in the continuity equation is minimized, a fixed-vorticity technique, and the presently proposed algorithm in which the measured winds are held fixed while winds at adjacent points are adjusted in order to reduce the divergence. The reduction of wind divergence and the convergence of the iterative scheme are examined. Results of the objective wind field analysis at times depend on subjective judgment in selecting the appropriate initialization algorithm especially for regions with scattered stations.

Abstract

A wind field interpolated from a limited supply of observed data often implies a specific field of divergence, or a violation of the conservation of mass. To minimize the divergence, various objective analyses have been proposed. Three different methods of analysis are investigated and compared with respect to the degree of minimization of wind divergence and the accuracy of wind data at a measured station. The methods are a variational formulation in which the residual error in the continuity equation is minimized, a fixed-vorticity technique, and the presently proposed algorithm in which the measured winds are held fixed while winds at adjacent points are adjusted in order to reduce the divergence. The reduction of wind divergence and the convergence of the iterative scheme are examined. Results of the objective wind field analysis at times depend on subjective judgment in selecting the appropriate initialization algorithm especially for regions with scattered stations.

Save