Precipitation estimates from satellite passive microwave radiometry have proven useful in a variety of applications, including numerical weather prediction, atmospheric general circulation modeling, and climate analysis. Since the mid-1990's, the Goddard Profiling (GPROF) algorithm has been used successfully to provide estimates of surface rain rate based upon satellite microwave radiance data; it is currently the TRMM facility algorithm. As major contributors to GPROF, the Principal Investigator and his colleagues helped to improve the physical basis of the algorithm and exploited methods to characterize the error in algorithm estimates of precipitation and latent heating.
The GPROF algorithm is currently supported by a set of cloud/radiative model simulations that establish the physical relationships between sensor-observed radiances and the precipitation/latent heating profiles that are estimated. Systematic errors in these modeled relationships, as well as other factors that make the relationships non-representative, have led to biases in GPROF estimates, particularly in precipitation vertical structure. As an alternative, the supporting data for passive microwave algorithms have recently been derived "empirically", based upon high-resolution spaceborne radar observations from the Precipitation Radar (PR) in combination with collocated microwave radiometer data from the TRMM Microwave Imager (TMI). In these studies, physical models of precipitation vertical profiles are fit to combined PR/TMI observations, and the best-fit vertical profile models can then be used to simulate upwelling microwave radiances at any specified frequencies/polarizations, thereby establishing the relationships between precipitation vertical profiles and upwelling radiances applicable to any satellite microwave radiometer.
The investigators of this proposal recently developed a combined PR/TMI method for estimating precipitation vertical profiles over ocean regions, calculated the upwelling microwave radiances associated with these profiles, and then incorporated the profiles/radiances in a TMI-only algorithm for estimating precipitation and latent heating. The new TMI-only method yielded precipitation estimates with less bias than the current GPROF algorithm; however, the combined PR/TMI method did not make full use of the higher-frequency TMI data due to uncertainties in the modeling of ice- and mixed-phase particles, and it was otherwise sub-optimal for calculating radiances at microwave frequencies other than those of the TMI. The objectives of the proposed study are, therefore, (1) to improve the physical description of ice- and mixed-phase precipitation in the vertical profile model of the combined method, evaluating the resulting model using available field campaign observations, (2) to generalize the combined method to estimate the environmental parameters needed to calculate radiances at other radiometer frequencies, and (3) to evaluate the resulting combined method using TMI data, alternative radiometer data, and ground validation radar observations.
The proposed investigation will contribute to the development of combined radar/radiometer methods that will yield estimated profiles of precipitation over ocean surfaces with reduced bias. Using these profiles as supporting data for radiometer-only algorithms, the resulting algorithms should also yield precipitation profiles with less bias. The reduction of biases in precipitation/latent heating estimates is critical for numerical weather prediction model and general circulation model data assimilation applications, as well as for climate analysis applications. The proposed study is therefore relevant to NASA's objectives of improving our predictive capability for weather, quantifying precipitation fluxes in the global water cycle, and understanding the role of precipitation/heating in the climate system and improving our predictive capability for climate.
