Operational evaluation of the real-time attenuation correction system for CASA IP1 testbed

Collaborative Adaptive Sensing of the Atmosphere (CASA) Engineering Research Center pursues a network paradigm of sensing of the atmosphere with X-band radars. The first generation testbed of CASA, called Integrated Project 1 (IP1) is currently semi-operational. This testbed consists of four identical dual-polarization X-band radars which form a network topology of a diamond-like shape in Okalahoma, US. In order to achieve the fast response to hazardous weather events, the system is designed with real-time algorithms (e.g., clutter suppression, rain attenuation correction) that ensure good data quality for higher level detection and prediction algorithms. In this paper, we evaluate the real-time attenuation correction system from an operational aspect.

During the semi-operational phase of the CASA IP1, the attenuation correction algorithm is being monitored, maintained, and improved in a case-by-case basis. Despite many experimental and operational changes to the system during this phase, the attenuation algorithm is relatively robust and working continuously. Although there were a few times it produced not so pleasing results (artifacts), these cases were all analyzed and solutions were developed to avoid those scenarios. The analysis revealed that the artifacts were caused by problems encountered in the operational environment (e.g., the change of system phase, and the units interface, etc.) By solving problems encountered in this semi-operational environment, the attenuation correction algorithm is made more robust everyday and gradually adapts to a full-operational phase.

Many types of precipitation, e.g., strong squall line, stationary front, arctic front, scattered cells, and etc., were observed and recorded by CASA IP1 in the summer and winter of 2006. Results of attenuation corrected reflectivity and differential reflectivity for these events are presented here. About 10 months (since May, 2006) of specific attenuation and differential attenuation statistics are analyzed. Cross-platform evaluation, as well as with the NEXRAD radars in the proximity is also analyzed where there was coverage from more than one radar at the same time.