Simulation of the capability of Ku, Ka and W tri-frequency satellite-borne radar measuring the three-dimensional structure of cloud and precipitation

Abstract

The capability of satellite radar working at Ku, Ka and W bands measuring the three dimensional structure of cloud and precipitation was studied by model simulations for three typical weather cases including a mid-latitude cyclone over land (MCL), a tropical typhoon (TT), and a tropical cyclone over ocean (TCO). First, cloud resolving model of Weather Research and Forecasting (WRF) was utilized to simulate the distribution of all types of hydrometeors in these cases. The cloud water path and cloud top temperature from the MCL simulation were validated with real satellite observations from Moderate Resolution Imaging Spectrometer (MODIS) on Aqua satellite. The WRF simulation correctly captured the main features of the storm including the spatial pattern, the geolocation, and the cloud top temperature, etc. Then, based on a satellite radar simulator developed by this study, the radar reflectivities at the Ku, Ka and W bands of those cases were calculated, and the radar reflectivity simulation of the MCL was also evaluated with real satellite measurements from W-band Cloud Profiling Radar (CPR) on CloudSat satellite. The vertical structure of radar reflectivity of the storm in the simulation was very close to the real measurements from CPR. Next, the characteristics of radar reflectivities at the three bands for different cloud and precipitation hydrometeors in the three cases were investigated. It was confirmed that the radar reflectivity factor decreases with the increasing frequencies (i.e. Ku, Ka and W bands). The radar echoes from non-precipitating hydrometers in both liquid and ice phase were much weaker than those from precipitating sized hydrometers, and normally cannot be detected by Ku and Ka band space-born radar. Regarding the capability of penetrating deep storm, the signal of W band radar can penetrate most clouds in the case of MCL and reach the surface very well. However, it started to be saturated from the height 6−9 km in areas with heavy liquid water content in the two cases of TT and TCO. The Ka band radar can detect the vertical distribution of hydrometeors in most areas of the three cases, but was saturated below 4−5 km in the eyewall area and the convection core areas of the cases of TT and TCO. The Ku band radar can penetrate all storms in this study. Moreover, the detecting errors of cloud top height and cloud bottom height for the three bands were quantified, assuming the detecting thresholds of Ku, Ka and W radars are 15, 5 and -35 dBZ. The results showed that W band radar had the smallest error ( 30 m) of detecting cloud top height, while for Ku and Ka radars, the errors were as high as several thousand meters particularly in the area with thin clouds due to their misdetection of small particles. In the convection core area with large and dense hydrometeors at the top layers, the errors for Ku and Ka significantly decreased to less than 100 m. Regarding cloud bottom detection, the penetrability of all radars not only depends on cloud geometrical thickness, but also on liquid water content (LWC). In MCL with low LWC the detecting errors for cloud bottom height from W band radar were only a few hundred meters, while reaching several thousand meters for TT and TCO with high LWC. The detecting error for cloud bottom for Ka radar was generally low (less than 300 m) but also increased greatly in strong precipitation areas. For Ku radar, the detecting errors of cloud bottom height were small in all cases except small area near the eyewall for TT. The results of this study provided some unique information to improve the understanding of the detecting capability of tri-frequencies satellite borne cloud and precipitation radar, particularly the differences among different types of storms. And the results can be used as references for designing associated parameters in radar system development.