A comparison of observations in the tropical western Pacific from ground‐based and satellite millimeter‐wavelength cloud radars

Abstract

Millimeter‐wavelength cloud radar (MMCR) can provide information on the vertical structure of cloud fields and thereby improve our understanding of the spatial distribution of clouds and their role in the climate system. Here we consider the representativeness of ground‐based vertically pointing MMCR observations, which have been used in numerous climate studies. MMCR cloud statistics collected at Darwin, Australia, are compared against CloudSat (spaceborne) observations gathered in the near vicinity of the ground site. Overall, the total cloud occurrence vertical profiles observed by CloudSat and the ground‐based MMCR agree on a spatial scale of 4° × 4°, although CloudSat is found to observe more high reflectivity cloud than the ground‐based MMCR. Computed radar reflectivity using idealized atmospheric profiles suggests that rain (especially below the melting level) influences the observed reflectivities, and this appears to account for much of the differences in the observed distributions of radar reflectivity. After removal of precipitation profiles, CloudSat and ground‐based MMCR observations show reasonable agreement. Sampling uncertainty in the CloudSat observations makes comparison at smaller region spatial scales (e.g., 2 . 5°) difficult and unfeasible for analysis at the time scale of months. Comparison of CloudSat observations with the ground‐based data on scales of 4° and 7 . 5° works well. Comparison of total cloud occurrence and reflectivity distribution of nonprecipitating cloud from the MMCR and CloudSat at spatial scales from 4° to 7 . 5° show good agreement. This suggests that the properties of the nonprecipitating cloud are relatively homogeneous at this large scale.