ACTIVE AND PASSIVE MICROWAVE REMOTE-SENSING OF PRECIPITATING STORMS DURING CAPE .1. ADVANCED MICROWAVE PRECIPITATION RADIOMETER AND POLARIMETRIC RADAR MEASUREMENTS AND MODELS

The Advanced Microwave Precipitation Radiometer (AMPR), an across-trac k scanning, four-channel (10.7, 19.35, 37.1, 85.5 GHz) total-power rad iometer system, was instrumented aboard a NASA ER-2 aircraft during th e 1991 CaPE (Convection and Precipitation/Electrification) project in central Florida. At a 20 km flight altitude, the AMPR provides fine-sc ale microwave imagery of Earth surfaces and its atmosphere, and is wel l-suited for diverse hydrological applications. During overflights of precipitation, coincident ground-based radar measurements were taken w ith the NCAR CP-2 dual-frequency, dual-polarization radar system. Afte r remapping the radar data into a format compatible with the AMPR scan ning geometry, the radar-derived profiles of rain, melting, and frozen hydrometeors are compared against the AMPR equivalent blackbody brigh tness temperature (T(B)) imagery. Microwave radiative transfer modelin g procedures incorporating the radar-derived hydrometeor profiles were used to simulate the multifrequency AMPR imagery over both land and o cean background ER-2 flights. Within storm cores over land, columnar i ce water paths up to 20 kg m-2 gradually depressed the 85 GHz T(B) as low as 100 K. The presence of tall vertical reflectivity columns encom passing > 20 kg m-2 columnar ice water path often produced 37 GHz T(B) < 85 GHz T(B) directly over the core. Over ocean, the 10 GHz channel provided the clearest correlation with the rainfall amounts, whereas t he 19 GHz channel saturated near 260 K past 10-15 mm hr-1 rain rate as determined by radar. Scattering by ice and melting ice at 37 GHz prod uced T(B) ambiguities over both raining and clear-ocean regions. Sensi tivity to the columnar mixed phase region via the intermediate frequen cies (19 and 37 GHz) is demonstrated and explained with the radar-deri ved T(B) modeling. By superimposing vertical profiles of cloud liquid water (which this radar cannot measure) upon the radar-inferred hydrom eteor structure, additional information on the location of the peak cl oud water and its amount relative to the vertical ice structure can be noted, along with a possible inference of the dominant ice particle s ize within the upper storm core. These results suggest that as the res olution of passive radiometric measurements approaches dimensions wher e the antenna beams become increasingly filled by the cloud, precipita tion retrieval via multifrequency T(B) input is well-suited to a verti cal profiling-type algorithm. This is further examined in Part II of t his manuscript, where the radar-derived vertical hydrometeor profiles are used to test the applicability of a multispectral cloud model-base d approach to passive microwave precipitation retrieval from space.