Knowledge Bank

Altimeter data gridding is a process of predicting the sea surface heights on a regular grid, based on data values given along the altimeter subsatellite tracks. The gridded form of altimeter data is useful in such applications as the contouring of the sea surface, and the comparison of various data. This report gives actual gridding examples and corresponding contour maps of the sea surface, to illustrate the effect of grid spacing, number and configuration of data points used to predict a grid value, and prediction method used. The prediction methods tested include least squares collocation, Bjerhammar's inverse distance weighting method, Hardy's multiquadric method using a hyperboloid kernel, a general harmonic kernel method using Krarup's kernel, and a piecewise biquadratic surface fitting method.