32nd Conference on Radar Meteorology

P8R.10

Automated sea clutter identification

PAPER WITHDRAWN

Dawn L. Harrison, UK Met Office, Exeter, United Kingdom

Sea clutter is notoriously difficult to identify and remove automatically from radar imagery. The reflectivity values are similar to precipitation and, unlike ground clutter, the frequency of occurrence is less than permanent and is highly dependent on sea state. Historically, processing of UK radar data has relied on fixed maps to identify areas where the lowest scan(s) are subject to frequent sea clutter contamination. The philosophy of the latest radar data processing system (Radarnet IV) is to eliminate the reliance on subjectively derived information which requires regular, or even occasional, manual updating.

A wide range of methods have been proposed to discriminate between precipitation and clutter. These methods include using the vertical reflectivity gradient, ray tracing techniques and using sets of fixed clutter maps. The method described in this paper uses the vertical profile of probability of detection (PoD), rather than reflectivity, to establish locations which will be prone to sea clutter. It is based on the observation that the frequency of sea clutter occurrence decreases with increasing height. The difference in PoD at successive scan elevation angles, together with restrictions on range and the application of a land/sea mask, has proved effective at identifying locations where sea clutter occurs, particularly where the clutter is the result of side-lobe echoes. At these locations, the reflectivity data are flagged as unusable, and data from higher elevation scans is used in preference for deriving surface rain rate estimates.

This paper describes and illustrates a sea clutter identification technique in detail, shows examples of the sea clutter problem from radars in the UK network and provides an assessment of the effectiveness of the technique.

Poster Session 8R, Operational radar applications
Thursday, 27 October 2005, 1:15 PM-3:00 PM, Alvarado F and Atria

Previous paper  Next paper

Browse or search entire meeting

AMS Home Page