Extended Abstract (1.2M)P3A.9
Comparisons of raindrop size distributions from the TRMM Precipitation Radar and the C-POL ground radar
Yuxiang Liu, Colorado State University, Fort Collins, CO; and V. N. Bringi and T. D. Keenan
The rain rate estimated from the 2A25 algorithm of the Tropical Rainfall Measuring Mission (TRMM) precipitation radar (PR) is strongly related to the a-adjustment method for attenuation correction of the measured reflectivity Z profile. This greatly depends on an appropriate initial specific attenuation (k) and reflectivity (Z) relation (k=aiZb), and a reliable retrieval of the multiplicative factor, e, used to adjust the initial a value in the k-Z relation. The normalized gamma drop size distribution (dsd), whose three parameters are Nw, D0 and µ, can be retrieved from e and corrected Z data available in the 2A25 files (note that µ is fixed at 3 in version 5 of 2A25). The dsd parameters can also be retrieved from the C-POL dual-polarized ground radar measurements of Zh, Zdr and Kdp. Comparison of the Nw and D0 parameters from the PR and the C-POL radar can lead to valuable insights about the performance of the 2A25 algorithm in different rain events, particularly convective, over land and ocean. Eight cases in different locations, e.g., Darwin, South China Sea and Sydney respectively, with TRMM overpass coincidences were studied. Five out of eight events were over ocean and the rest were over land. Data sets from the two different radar systems were gridded and aligned before analysis, taking into account different viewing aspects, different resolution volume sizes, and distortion due to space radar movements etc. Emphasis was on those PR beams with high path integrated attenuation (PIA), where reliable e values could be used to retrieve the Nw parameter; the D0 was retrieved using the corrected Z from PR, the Nw and fixed µ=3. Thus the a-adjustment method could be validated by comparing the dsd retrievals from the two radars. For PR beams with low PIA, the initial dsd parameters in 2A25 algorithm were compared with ground radar retreivals to see if they were appropriate. Our studies showed that the range of variability of the dsd parameters (Nw and D0) were similar for the two radars for convective rain events over the ocean. However, for convective events over land this was not the case with the PR-Nw values tending to be systematically on the lower side (as compared with C-POL retrievals), and D0 values tending to be systematically on the higher side. This indicates a possible bias in the PR retrieval of Nw and D0 for strong convective rain over land which may be related to the different statistics of e over land (as opposed to ocean).
Poster Session 3A, TRMM/GPM Poster
Friday, 8 August 2003, 1:30 PM-3:30 PM
Previous paper Next paper