High-resolution passive microwave observations of convective-systems over the Tropical Pacific Ocean

This paper presents high-resolution passive microwave measurements obtained in the western Pacific warm pool region. These measurements represent the most comprehensive such observations of convection over the tropical oceans to date, and were obtained from the Advanced Microwave Precipitation Radiometer (AMPR) aboard the NASA ER-2 during the Tropical Ocean and Global Atmosphere Coupled Ocean-Atmosphere Re sponse Experiment. The AMPR measures linearly polarized radiation at 10.7, 19.35, 37.1, and 85.5 GHz. Nadir brightness temperature scatterplots suggest that the three lower frequencies respond primarily to emission/absorption processes. Strong ice scattering is relatively rare, as absolute magnitudes of the ice-scattering signature do not approach those measured in strong convection over land. This is apparently related to the reported weaker updraft velocities over tropical oceans, which would create and suspend relatively smaller graupel or hail particles in the upper cloud. Observations within stratiform regions suggest that approximately 220 K is the minimum 85.5-GHz brightness temperature associated with ice scattering in regions of stratiform precipitation. In agreement with Other studies using high-resolution data, the relationships between data at the lower frequencies and the 85.5-GHz data exhibit considerable scatter. Traces through a hurricane eyewall and a squall line reveal the tilt of these convective systems away from the vertical. It is suggested that this observed tilt of convective lines is responsible, in part, for the finding that warm 10.7-GHz brightness temperatures (showing heavy rain at low levers) and cold 85.5-GHz brightness temperatures (showing large optical depth of ice particles aloft) are not consistently collocated. Observations of heavily raining clouds with little ice above or nearby are also presented, but it is shown that the heaviest rain rates are associated with ice scattering aloft. The AMPR data are averaged to a 24-km resolution, in order to simulate a satellite footprint of that scale. Brightness temperature relationships become more linear, though the scatter is not significantly reduced. The effects of nonhomogeneous beamfilling are obvious. A description of brightness temperature variability within the simulated satellite footprint is also presented. Similar descriptions could be used to develop a beamfilling correction to increase the accuracy of microwave rain-rate retrievals over the tropical oceans.