Graduate Exam Abstract
July 9, 2010, 09:30 am
ESTIMATION AND CORRECTION OF WET ICE ATTENUATION FOR X-BAND RADAR
Abstract: In the past, single polarized X-band radars were primarily used (along with S-band radars) for hail detection, first by the Russians and then later for the National Hail Research Experiment (NHRE). But X-band radars were not used alone because of the large attenuation at frequencies around 10 GHz and higher, until dual-polarized radars were developed. This fact has brought attention to development and evaluation of correction techniques for rain attenuation in order to exploit the advantages of dual-polarized data. Past developed methods make use of the close relation between the differential propagation phase ?DP and path attenuation PIA. Their use is known to be successful in rain events, but in the presence of wet ice, these methods are no longer useful because the differential propagation phase is not affected by the isotropic wet ice. This factor was the basis to develop herein two different techniques for estimating the attenuation due to rain and wet ice separately and correct for the wet ice induced attenuation.
In this dissertation, two methods are investigated and evaluated. The first method uses the Surface Reference Technique (SRT) ?-adjustment method to correct for the attenuation. This method was first developed for the Tropical Rainfall Measuring Mission (TRMM) precipitation radar. We assume that S-band data is un-attenuated and is used as a reference. The difference in reflectivities at the end of the beam (defined as the average of the last ten gates with good data) is attributed to the total attenuation (sum of rain and any wet ice) along the propagation path. The attenuation due to the rain component, if any, is corrected for using the differential propagation phase. Then the ? value in the Ah(X)wet ice-Zh(X) power law relationship (with fixed exponent ?) is adjusted such that the reflectivities at S-band and the rain-corrected reflectivity at X-band at the end of the beam are forced to match. This adjusted ? is used to apportion the reflectivity backwards, which assumes the ? parameter is constant along the beam. Using the adjusted value, the attenuation due to wet ice is estimated separately from that of rain. This method is termed here as the SRT-modified correction method.
This method has been applied to different datasets. It was evaluated in both simulated and measured radar data. Using the Regional Atmospheric Modeling System (RAMS) model a supercell was simulated by Professors Cottons group at Colorado State University (CSU). A radar emulator was used to simulate radar measurements from this supercell at both X-band and S-bands. Results showed good agreement of both corrected reflectivity profiles and wet ice specific attenuation estimation. A dataset from the International H2O Project (IHOP) that had rain mixed with wet ice particles (mixed phase region) was analyzed too. It showed good agreement also, when comparing profiles; moreover wet ice attenuation contours showed agreement with high values of reflectivity as expected in wet ice regions. Data collected by the Center for Adaptive Sensing of the Atmosphere (CASA) radar network was analyzed along with both Next Generation Radars (NEXRAD) KTLX and KOUN data. For the light rain event (CASA/KTLX), the dual wavelength ratio at the maximum range was close to unity as expected for Rayleigh scattering. When corrected for wet ice, the specific attenuation showed agreement with high values in reflectivity at both bands. Finally, this method was applied to two different Cloud Physics Radar (CP2) radar data sets. In the CP2 data analysis, Mie hail signals were eliminated for the purpose of this research. Results from both datasets showed that resulting corrected reflectivity was comparable to the un-attenuated S-band data.
Given that an un-attenuated reference signal, like the one described before, might not always be available, a second method was developed without this assumption. This second method estimates the wet ice specific attenuation using a Ah(X)wet ice-Zh(X) power law relation with fixed coefficients. These fixed coefficients were retrieved using the same CP2 datasets and compared with previous findings. Then, assuming that the reflectivity is already corrected for rain attenuation, the Hitschfeld-Bordan forward correction method is used. To determine the areas where the correction method will be applied the Hydrometeor Identification (HID) algorithm was used. The HID data is used here to locate the first good range gate of the mixed phase region containing the wet ice. This method is termed as the Piece-Wise Forward correction method (PWF).
Similar to the first method, this second method was applied to different datasets. First it was applied to one of the two CP2 datasets available, where the Mie hail signal was eliminated. The resulting corrected reflectivity showed good agreement compared with the S-band un-attenuated reflectivity. Also this method was applied to the same convective dataset (from CASA) as the one previously analyzed with the SRT-modified method. It presented higher reflectivity values in wet ice identified areas, but lower values than those presented by the SRT-modified method. The results were also compared with the Networked Based (NB) method.
Adviser: Dr. V.N. Bringi
Non-ECE Member: Dr. Steve Rutledge, Atmospheric Science
Member 3: Dr. V. Chandrasekar
Addional Members: N/A
Program of Study: