Radar rainfall data quality control by the Influence Function Method

Abstract

The statistical concept of the influence function is applied to detection of outliers in radar rainfall fields. The method can identify observations which are inconsistent with the spatial correlation in the field. A Monte Carlo experiment has been performed to test the method for daily and hourly radar rainfall data and to compare it to other simple methods such as range and gradient checks. The results of that study indicate the usefulness of the method in the detection of outliers in real time. They also show the potential of the method to deal with outliers resulting from certain types of anomalous propagation. This paper is not subject to U.S. copyright. Published in 1987 by the American Geophysical Union.