Mesoscale Variability of Tropical Precipitation: Validation of Satellite Estimates of Wave Forcing Using TOGA COARE Radar Data

Abstract

Satellite-derived brightness temperature has been used to estimate tropical precipitation. Ricciardulli and Garcia applied it to quantify forcing of atmospheric waves that are excited by tropical cumulus convection and propagate into the middle atmosphere. Because of the broad coverage of the satellite data, this method provides exclusively dense information on wave forcing and is especially valuable for middle-atmosphere modeling. However, the validity of the method has not been investigated, which is done in this study using radar-derived precipitation during the Tropical Ocean and Global Atmosphere Coupled Ocean?Atmosphere Response Experiment (TOGA COARE) field experiment. The method is shown to overestimate the variance of precipitation so that the wave forcing derived with it is too strong. The overestimation is most severe at coarse resolution, reaching nearly an order of magnitude at a grid scale of 2°, which is comparable to typical resolutions of current global climate models. Although the comparison was made using data from a limited region in the western Pacific, it is suggested that the method overestimates wave forcing globally. The probability distribution of the mean radar precipitation on the meso-? scale fits well to the gamma distribution, while that for the satellite-derived precipitation does not. The latter shows a high probability of extreme grid mean precipitation, and this contributes to the overestimation. The frequency spectra of radar and satellite precipitation showed some similarity in shape, but differences are evident at subdiurnal frequencies. In addition to the satellite method, the Geostationary Operational Environmental Satellite (GOES) Precipitation Index (GPI) is also investigated. GPI shows a similar, but better, performance to estimate wave forcing.