Abstract
The classical rain attenuation correction scheme of Hitschfeld and Bordan (HIBO) and the newer iterative approach by Hildebrand (HL) are reconsidered. Although the motivation for the HL algorithm was an extension into ranges, where HIBO tends to be unstable, it is shown here that the contrary is the case. The finite-range resolution causes an intrinsic instability of HL already at moderate attenuation, where HIBO would still deliver stable results. Therefore, the authors concentrate the further analysis on HIBO, and confirm that the usual implementation of HIBO does not account correctly for finite-range resolution. They suggest a modified scheme that produces exact retrievals in the ideal case of perfect measurements.
For vertically pointing Doppler radars a new element is explored in the attenuation correction—namely, calculating rain attenuation κ and rainfall R from Doppler spectra via the raindrop size distributions (RSDs). Although this spectral scheme (SIBO) avoids the uncertainty of Z–R and Z–κ relations, the superiority of this approach is not a priori obvious because of its sensitivity to vertical wind. Therefore, radar rain rates, based on a Z–R relation and on RSDs, respectively, are compared with in situ measurements. The results indicate better agreement for RSD-based retrievals. Because κ is closely correlated with R, the authors assert the advantage of RSD-based retrievals of κ.
The application of HIBO and SIBO to real data shows that the uncertainty of standard Z–R relations is the main source of deviation between the two versions. In addition, the comparison of profiles suggests that the parameters of Z–R relations aloft can deviate considerably from near-surface values. Although artifacts cannot be excluded with certainty, there is some evidence that this observation actually reflects microphysical processes.
Corresponding author address: Gerhard Peters, Meteorological Institute, University Hamburg, Bundesstrasse 55, D 20146 Hamburg, Germany. Email: gerhard.peters@zmaw.de
