ACTIVE AND PASSIVE REMOTE-SENSING OF PRECIPITATING STORMS DURING CAPE.2. INTERCOMPARISON OF PRECIPITATION RETRIEVALS OVER LAND FROM AMPR RADIOMETER AND CP-2 RADAR
Fs. Marzano et al., ACTIVE AND PASSIVE REMOTE-SENSING OF PRECIPITATING STORMS DURING CAPE.2. INTERCOMPARISON OF PRECIPITATION RETRIEVALS OVER LAND FROM AMPR RADIOMETER AND CP-2 RADAR, Meteorology and atmospheric physics, 54(1-4), 1994, pp. 29-51
One of the recent campaigns devoted to precipitation studies using bot
h active and passive microwave remote sensing systems was the Convecti
on and Precipitation/Electrification Experiment (CaPE), which took pla
ce in central Florida during the summer of 1991. During CaPE, the airb
orne Advanced Microwave Precipitation Radiometer (AMPR), having four c
hannels at 10.7, 19.35, 37.1 and 85.5 GHz and the National Center for
Atmospheric Research CP-2 multiparameter radar at S-band (3 GHz) and X
-band (10 GHz) were operated simultaneously. In this paper, we compare
estimated hydrometeor liquid/ice water contents and surface rainrates
, both retrieved from the AMPR radiometer and CP-2 radar measurements,
for a case study consisting of a heavy precipitating storm over land
near Cape Canaveral on August 12, 1998. The multi-frequency radiometer
-based retrieval scheme uses a cloud-precipitation dataset generated f
rom a cloud model and extended by a physically-constrianed Monte Carlo
procedure, along with a discrete-ordinate radiative transfer model an
d a principal component statistical technique to help formulate non-li
near regression equations for the sought-after hydrometeor quantities.
By applying linear discriminant analysis, the algorithm is used to es
timate column integrated liquid/ice water contents, as well as the ver
tical profiles of these quantities to within a specified accuracy. Rai
nfall rates are estimated either by non-linear regression or by a suit
able fallout model. The analysis has confined itself to along-track na
dir-looking AMPR measurements to avoid complications with variable pol
arization mixing and geometric distortion for off-nadir observations.
Considering the different model assumptions used in the two types of r
etrieval algorithms and the diverse geophysical information content wi
thin the two types of measurements, substantial agreement between the
radar- and radiometer-derived retrievals has been achieved for the col
umnar liquid/ice water contents and rainrates.