2009-04-20: Towards a standard procedure for validation of satellite-derived cloud properties.

Validation of satellite-derived cloud properties with ground-based data is necessary to ensure meaningful use of satellite cloud observations. Validation is difficult because differences between satellite and ground values are not only due to errors in the data, but also due to validation issues.
Greuell and Roebeling [2009] distinguish two validation issues: (1) scale differences, i.e. satellite pixels tend to be much larger than the area viewed by the ground instruments, and (2) the parallax, i.e. satellite-observed clouds are shifted with respect to their vertical projection onto the ground. Comparing liquid water path data from the SEVIRI sensor and from ground-based microwave radiometers, Greuell and Roebeling optimized the validation procedure so that differences due to the validation issues were minimized. Their method incorporates a parallax correction, Gaussian averaging of the satellite and the ground data and a time scale for averaging the ground data that is much longer than the time that the clouds need to move across the considered satellite pixels. The recommendations from Greuell and Roebeling (2009) are probably also valid for the validation of other cloud-related properties.
Due to the parallax, a cloud observed by a ground-based MWR with its vertical view direction, is horizontally displaced in a satellite image.