First echo development in trade-wind cumulus

The S-Pol radar (S and Ka band, dual-wavelength) operated on Barbuda in the Lesser Antilles from late November 2004 to late January 2005, scanning trade wind cumulus clouds over the open ocean. Barbuda was selected because of its low elevation (<100m) and the lack of islands N or E, so that the clouds in the NE trades are relatively undisturbed.

One main objective of the project (RICO) was to characterize the precipitation development in trade cumulus using radar and aircraft. This paper reports early results of an ongoing, systematic study of “first echoes” in the clouds, using the S-band data (10-centimeter wavelength; this radar operated 24 hours a day) with ZDR as extra evidence on drop size distributions. Bragg scattering may dominate the echo from the earliest stages of small, active cumulus up to 5 to 10 dBZ at this wavelength.

Using PPI scans at several elevation angles as a function of time, precipitation from the small cumulus is characterized by calculating a kind of total dBZ and total ZDR; not direct averages, but the values that would be found if the radar beam size and shape matched the cloud cross-section. These can be used to characterize the total precipitation size distribution considering the cross-sections as a whole, if Bragg scattering does not interfere. To remove that, the values are calculated with data thresholded at two values of dBZ: 9dBZ, which definitely excludes almost all echo dominated by Bragg scattering, but may omit some cloud for which the echo is dominated by hydrometeor scattering; and -1 dBZ, for which some Bragg scattering may be included.

Using either threshold, the ZDR values are very low in an example presented, and they appear by inspection to be low in the trade-wind cumulus in general, though that is to be confirmed. Particularly in the early echoes, the ZDR data show that the rain is dominated by small drops, unlike previous studies in Georgia and Florida that showed that the early radar echoes from small cumulus tended to be small quantities of several-millimeter drops, giving appreciable values of ZDR at low values of dBZ.

The other main conclusion, also awaiting testing by more extensive analysis, is the finding that many of the small cumulus that produce precipitation can be tracked backwards for very long times by the Bragg scattering: up to an hour, or more. If one defines this as the time scale for precipitation formation, it can be considerably longer than has been reported before in maritime cumulus; and if this result is generally valid, it and the other tentative conclusion above indicate that ultra-giant CCN probably have little importance for precipitation formation in these clouds.