Investigation of a dual-frequency surface reference technique for estimates of path-integrated attenuation

The surface reference technique (SRT) has been used to estimate path-integrated attenuation from airborne and spaceborne platforms and constitutes a key element in some of the rain retrieval methods that have been proposed. The SRT rests on the idea that the decrease in the apparent surface return in the presence of precipitation can be attributed to the attenuation by the precipitating medium along the radar beam. The method has shown itself to be useful when the path attenuation is much larger than the inherent fluctuations in the surface return. For the TRMM precipitation radar (PR) that operates at 13.8 GHz (Ku-band), this means that the method is typically applicable at moderate and heavy rain rates over certain ranges of incidence angles over ocean and land.

The Dual-Frequency Precipitation Radar (DPR), built by the Japanese Aerospace and Exploration Agency (JAXA), is scheduled to be launched on the Global Precipitation Mission (GPM) satellite in 2013. The frequencies of operation are Ku- (13.6 GHz) and Ka-band (35.5 GHz) and a question that arises is the performance of the SRT when applied to dual-frequency data from the surface using a matched-beam configuration that scans cross-track from nadir to approximately 90 on either side.

Data from the JPL APR 2 dual-frequency airborne radar show moderate to high correlations in the normalized surface radar cross-sections at Ku- and Ka-band over incidence angles from nadir out to approximately 250. Preliminary results over ocean indicate that the correlations in the surface returns at the two frequencies can lead to more accurate estimates of the differential path attenuation than an estimate of path attenuation at either frequency. However, deriving the path attenuations at either frequency from the differential attenuation leads to errors that tend to partially offset the benefits of the approach. The nature and magnitude of this error will be investigated. Other issues to be addressed are the performance of the method over land, the extension of the method to other frequency pairs, and the role of path attenuation in the retrieval of parameters of the particle size distribution.