Skip to Main Content

Improved Ice and Mixed-Phase Precipitation Models for Combined Radar-Radiometer Retrieval Algorithm Applications

William S. Olson, JCET/Department of Physics
Benjamin Johnson, JCET/Department of Physics
Lijun Diao, Department of Physics

In the GPM era, Bayesian algorithms for estimating precipitation distributions from spaceborne passive microwave radiometer observations are expected to provide estimates that are reasonably accurate and practical for many applications in climate analysis and hydrometeorological modeling/prediction. These algorithms rely heavily on large supporting databases of precipitation vertical profiles that serve as candidate solution profiles in the Bayesian estimation framework. In recent years, the supporting databases of precipitation
profiles have been derived “empirically”, based upon high-resolution spaceborne radar observations in combination with collocated passive microwave radiometer data. In this case, physically-based descriptions, or models, of cloud/precipitation vertical profiles are fit to the combined radar-radiometer observations, and these best-fit vertical profile models can then be used to simulate upwelling brightness temperatures at any particular radiometer channel frequencies and polarizations, thereby establishing the relationships between precipitation vertical profiles and microwave brightness temperatures. It follows that the accuracy of radiometer-only precipitation estimates is dependent upon the accuracy of the cloud/precipitation vertical profile models utilized in the combined radar-radiometer algorithms.  At higher latitudes in particular, it will be important to properly model the extinction and scattering properties of ice and mixed-phase precipitation, as these constituents typically occupy a greater proportion of the total vertical column of precipitation at higher latitudes.  The objectives of the proposed investigation are (a) to develop realistic descriptions of the microwave single-scattering properties of ice and mixed-phase precipitation that are consistent with radar/radiometer observations over a range of frequencies (6 – 200 GHz), (b) to incorporate the ice/mixed-phase precipitation descriptions into vertical profile models and test these models against airborne radar/radiometer and in situ observations from field campaigns, and (c) to provide tested, modular versions of the vertical profile models to the satellite algorithm developers for radar-only and radar-radiometer algorithm applications. Modeling of ice- and mixed-phase precipitation will be based on realistic descriptions of particle size distributions, particle densities/geometries, liquid fractions and the relative populations of different particle geometries, drawn from field campaign data.  A thermodynamic model for melting precipitation will describe the transition from ice to rain in stratiform precipitation regions.  Based upon these ice/mixed-phase particle descriptions, single-scattering properties will be calculated using discrete dipole approximation software.  The descriptions of the precipitation particles and their single-scattering properties will be incorporated into vertical profile models that will be tested using an ensemble filtering inversion method, which can be used to identify models that are consistent/inconsistent with field observations and quantify the impact of model uncertainties on estimated precipitation profiles.  Vertical profile models consistent with the field observations will be used to create computationally-efficient model versions that will be tested in a satellite combined radar-radiometer algorithm and distributed to the precipitation algorithm development teams. The impact of the proposed investigation will be improved combined radar-radiometer estimates of precipitation profiles, especially at higher latitudes, and greater accuracy of radiometer-only algorithms that are supported by these profile estimates. The proposed work is, in effect, a continuation and extension of studies by the principal investigator in support of TRMM/GPM facility algorithm development and should lead to a better quantification of critical components of the Earth’s water and energy cycles.