Study of Dual-Wavelength PIA Estimates: Ratio of Ka- and Ku-band Attenuations Obtained from Measured DSD Data

Accurate attenuation corrections to the measurements of the Ku- and Ka-band dual-wavelength precipitation radar (DPR) aboard the Global Precipitation Measurement (GPM) satellite is crucial for estimates of precipitation rate and microphysical properties of hydrometeors. Surface reference technique (SRT) provides a means to infer path-integrated attenuation (PIA) by comparing differences of normalized surface cross sections (σ0) between rain and rain-free areas. Although single-wavelength SRT has been widely used in attenuation correction for airborne/spaceborne radar applications, its accuracy relies on the variance of σ0 in rain-free region. Dual-wavelength surface reference technique (DSRT) has shown promising ways to improve accuracy in PIA estimates over single-wavelength as a result of that the variance of the difference of PIA between two wavelengths (δPIA) is much smaller than the variance of σ0 at single wavelength arising from high correlation of σ0 between Ku- and Ka-bands. However, derivation of PIA at either wavelength from DSRT requires an assumption of the ratio of Ka- and Ku-band PIAs (p). Inappropriate assumption of this ratio will lead to the bias of PIA estimates. In this study the ratio p will be investigated through measured DSD data. The attenuation coefficients at Ku and Ka bands are first computed directly from measured DSD spectra, and then regression analysis is performed to the data points (Ku- and Ka-band attenuation coefficients) in obtaining p values for rain. Taking an advantage of large collection of the DSD measurements from various GPM Ground Validation (GPM GV) programs, the results of the ratio p will be examined from different climatological regimes. Because PIA is affected by all types of hydrometeors contained in the columns of radar measurements, the synthetic profiles composed of different types of hydrometeors are constructed using measured DSD to look into impacts of different phase hydrometeors on the p values. To generate these profiles fully-, partially- and un-correlated DSD data are employed in an attempt to describe wide dynamic range of microphysical structures of hydrometeors. Bright-band model is employed to take into account of mixed-phase region, and additional attenuations due to cloud water are also included in the profiles.