A meshless method using global radial basis functions for creating 3-D wind fields from sparse meteorological data

Abstract

An efficient, global meshless method has been developed for creating 3-D wind fields utilizing sparse meteorological tower data. Meshless methods do not require the need for a mesh in order to connect node points. In this study, node points are placed within the computational domain based on topological features. Wind speeds and directions are obtained from a set of instrumented meteorological towers. Inverse weighting is used to initially establish wind vectors at all nodal points. The Kansa technique, employing global basis functions, is then used to create a mass-consistent, 3-D wind field. The meshless method yields close approximations to results obtained with a high-order finite element technique. The method was implemented using MATLAB.