Simulation of Space-borne Radar Observations of Precipitation at Ku and Ka Band

ABSTRACT

Global Precipitation Measurement (GPM) is poised to be the next generation precipitation monitoring mission from space following the success of TRMM mission. GPM core satellite will carry a Dual –frequency radar system called the DPR operating at two different frequencies, 13.6 GHz (Ku-band) and 35.6 GHz (Ka-band) in contrast to the single 13.8 GHz (Ku-band) precipitation radar (PR) on TRMM. The DPR will provide two independent measurements at every range bin. By combining the two measurements, various dual-wavelength algorithms will be used to retrieve two parameters of the drop size distribution (DSD). With the retrieval algorithms, precipitation rate can be estimated more accurately compared to the TRMM era, which is the key improvement expected from DPR on GPM.

There are numerous assumptions in the retrieval algorithms and they are based on our current understanding of the microphysical processes. One such example is the accuracy of retrieved DSD parameters using dual-wavelength algorithms relies strongly on correct identification of phase-height transition (PHT) from frozen to liquid hydrometeor along vertical profile of reflectivity (VPR). It is important to understand how these various features manifest themselves in dual-wavelength measurements.

Ground validation is an integral part of all satellite precipitation missions. During the GPM era there are going to be numerous single and dual-polarization radars around the globe that will be observing precipitation. If there is a way that one could use this ground based dual-polarization radar observation to simulate space borne dual wavelength measurement, then we have a globally diverse data set that can be used in the system development, algorithm development and validation.

This paper presents a technique to simulate DPR observation based on ground single and dual-polarization radar observations. In order to simulate more realistic VPR, PHT information along VPR, which is obtained through hydrometeor classification, was incorporated. The height where melting starts can be determined accurately using dual-polarization radar parameters, such as Zdr. The simulated VPRs were then used to test various dual-wavelength retrieval techniques. Classification and simulation results on convective storms show that PHT from partially-melted hydrometeors to rain drops can mostly be identified by simulated “apparent” reflectivity in higher frequency and difference of “apparent” reflectivity between the two channels. The DSD retrieved from simulated VPR using different retrieval techniques were compared and validated by those estimated from polarimetric parameters. Error of retrieved DSD in rain region and rainfall rate are evaluated.