Progressive Refinement: A Strategy for the Calibration by Collateral Data of Short-Period Satellite Rainfall Estimates

Abstract

The progressive refinement approach is a strategy for generating new shorter-term (less than 10 days) satellite rainfall estimation algorithms from structural combinations of an easy-to-calibrate algorithm generating longer- term rainfall totals only, and a shorter-term algorithm that efficiently evaluates individual precipitation events. It is designed to address the problem of local calibration of the latter type of algorithm without recourse to high-quality ground data, which are frequently not available. A PR daily rainfall estimation method based upon the Bristol polar-orbiter effective rainfall monitoring integrative technique algorithm is described and evaluated.