An icosahedron-based method for even binning of globally distributed remote sensing data

A new scheme is presented for binning globally distributed measurements. The scheme is based on a network of evenly distributed grid points, defined by repeated subdivision of a spherical icosahedron. Delanuany triangulation is then used to obtain bin perimeters for each grid point, which results in a network of bins that are evenly distributed across the entire globe and have uniform area. A modified winding rule is used to determine which datapoints are in which bin. This binning method is especially suited to remote sensing applications involving datasets covering polar regions, where conventional rectangular latitude/longitude bins introduce distortion and streaking into the binned data if noise is present. It also has the property that adjacent bins overlap, providing Nyquist sampling and preventing spatial aliasing. Tests on synthetic data show that this icosahedral binning scheme preserves underlying data trends and is robust to noise. (c) 2006 Elsevier Ltd. All rights reserved.