Constraining Microwave Brightness Temperatures by Radar Brightband Observations

A. Battaglia

Department of Physics, University of Ferrara, Ferrara, Italy

C. Kummerow

Department of Atmospheric Science, Colorado State University, Fort Collins, Colorado

Dong-Bin Shin

School of Computational Sciences, George Mason University, Fairfax, Virginia

C. Williams

CIRES, NOAA/Aeronomy Laboratory, Boulder, Colorado


Multichannel microwave sensors make it possible to construct physically based rainfall retrieval algorithms. In these schemes, errors arising from the inaccuracy of the physical modeling of the cloud system under observation have to be accounted for. The melting layer has recently been identified as a possible source of bias when stratiform events are considered. In fact, Tropical Rainfall Measuring Mission (TRMM) Microwave Imager (TMI) observations reveal systematic differences in the observed brightness temperatures between similar rain profiles that often differ only by the presence or absence of a bright band.

A sensitivity study of the scattering properties of the melting layer with different one-dimensional steady-state microphysical and electromagnetic models is performed. The electromagnetic modeling of the ice particle density and assumption of the ventilation coefficient parameterization is found to have the greatest impact on the extinction profiles. Data taken from a 0.915-GHz National Oceanic and Atmospheric Administration (NOAA) profiler during the Kwajalein Experiment (KWAJEX) field campaign are used to reduce the uncertainties in the modeling of the bright band. The profiler data reduce the number of viable parameterizations, which in turn leads to a reduction in the variability of the upwelling radiances (simulated at TMI angle) for different cloud simulations.

Using the parameterizations that best match the profiler data, the brightness temperatures TB generally increase if mixed-phase precipitation is included in the model atmosphere. The effect is most pronounced for systems with low freezing levels, such as a midlatitude cold front simulation. For TMI footprints at 10.65 GHz, the increase in the TB from the bright band generally increases with rain rate and changes by as much as ~15–20 K. At 19.35 GHz the maximum effect is found around 3–5 mm h−1 (~15 K), and at 37 GHz the maximum effect is around 1 mm h−1 (~10 K), while at 85.5 GHz the effect is always lower than 3 K.

Despite the reduction of uncertainties achieved by using 915-MHz profiler data, differences between parameterizations are still significant, especially for the higher TMI frequencies. A validation experiment is proposed to solve this issue and to further reduce the uncertainties in brightband modeling.