5:15 PM - 7:15 PM
[ACG54-P01] Analysis of Mirror Images of Airborne Radar for Development of Satellite-Borne Doppler Radar Algorithms
Keywords:Precipitation observation, Doppler radar, airborne radar
Precipitation radar onboard satellites has been used for global precipitation observations since the Tropical Rainfall Measuring Mission (TRMM), which carried the world's first Precipitation Radar (PR), and its succeeding mission, the Global Precipitation Measurement (GPM) core observatory, which carried the Dual-frequency Precipitation Radar (DPR). Currently, JAXA has begun development of the Ku-band Doppler radar (KuDPR) in cooperation with NASA's AOS. Although Doppler velocity observations from space are limited to the nadir direction, they represent the vertical motion of precipitation particles and can estimate the vertical air motion by providing an appropriate falling velocity. The KuDPR also needs to establish an evaluation method for Doppler velocity measurements. It is known that mirror image that is defined as the echo that transmit signal from radar reflected at the Earth surface scattered by the precipitation downward to the Earth and reflected again at the Earth surface then return to the radar, is effective tool for estimating Doppler velocity and attenuation.
In this study, a method for estimating the radar reflectivity factor (Z) without attenuation and a method for estimating the attenuation coefficient were developed based on observations by an airborne radar to confirm the effectiveness of this method, and a method for evaluating the relationship between Z and Doppler velocity was examined. The airborne radar used in this study was the APR2 of NASA/JPL, and data from a 2003 mission in and around Japan were used. The sum of Z profiles of direct and mirror echoes is the sum of the unattenuated Z profile plus the loss due to path integrated attenuation (PIA) and scattering efficiency at the ground surface, and the profile shape is the same as the unattenuated Z profile. Therefore, if mirror echoes can be observed up to the profile of very weak precipitation or snowfall, where the attenuation is negligible, the Z profile without attenuation can be reproduced. The profile of the difference in Z between the direct and mirror echoes represents the integral of the attenuation from the ground surface to the target altitude, so the slope in the direction of height represents the attenuation coefficient (dB/Km). Since the Doppler velocities of the direct and mirror echoes show opposite sign under ideal mirror image conditions, the sum of the two is considered to be zero. If this value is not zero in the actual observed data, some bias (e.g., vertical motion of the aircraft itself) must be considered.
We applied this method to the stratiform precipitation observed over the Pacific Ocean on January 19, 2003; although APR2 operated cross-track scans, we used only data near the nadir. As a result, mirror echoes were obtained up to altitudes with very small attenuation, allowing us to estimate a profile of Z without attenuation. Using this data, we examined the relationship between Z and attenuation coefficient, and found that Ka showed a very good relationship between Z and attenuation coefficient, but Ku did not show such a clear relationship. Possible reasons for this are that the method of obtaining the attenuation coefficient from the slope of the difference in Z between the direct and mirror echoes and/or the assumption that it is uniform in the height. On the other hand, the possibility that the drop size distribution changed with respect to Z should be considered. A good correlation between Doppler velocity and Z (without attenuation) was obtained. In the future, we will evaluate the method through comparison with the relational equation in previous studies.
In this study, a method for estimating the radar reflectivity factor (Z) without attenuation and a method for estimating the attenuation coefficient were developed based on observations by an airborne radar to confirm the effectiveness of this method, and a method for evaluating the relationship between Z and Doppler velocity was examined. The airborne radar used in this study was the APR2 of NASA/JPL, and data from a 2003 mission in and around Japan were used. The sum of Z profiles of direct and mirror echoes is the sum of the unattenuated Z profile plus the loss due to path integrated attenuation (PIA) and scattering efficiency at the ground surface, and the profile shape is the same as the unattenuated Z profile. Therefore, if mirror echoes can be observed up to the profile of very weak precipitation or snowfall, where the attenuation is negligible, the Z profile without attenuation can be reproduced. The profile of the difference in Z between the direct and mirror echoes represents the integral of the attenuation from the ground surface to the target altitude, so the slope in the direction of height represents the attenuation coefficient (dB/Km). Since the Doppler velocities of the direct and mirror echoes show opposite sign under ideal mirror image conditions, the sum of the two is considered to be zero. If this value is not zero in the actual observed data, some bias (e.g., vertical motion of the aircraft itself) must be considered.
We applied this method to the stratiform precipitation observed over the Pacific Ocean on January 19, 2003; although APR2 operated cross-track scans, we used only data near the nadir. As a result, mirror echoes were obtained up to altitudes with very small attenuation, allowing us to estimate a profile of Z without attenuation. Using this data, we examined the relationship between Z and attenuation coefficient, and found that Ka showed a very good relationship between Z and attenuation coefficient, but Ku did not show such a clear relationship. Possible reasons for this are that the method of obtaining the attenuation coefficient from the slope of the difference in Z between the direct and mirror echoes and/or the assumption that it is uniform in the height. On the other hand, the possibility that the drop size distribution changed with respect to Z should be considered. A good correlation between Doppler velocity and Z (without attenuation) was obtained. In the future, we will evaluate the method through comparison with the relational equation in previous studies.