| Abstract: |
This paper presents passive microwave retrievals of ocean surface wind speed given coincident radar rain rate for hurricane Patricia in the Pacific Ocean. These observations are the result of near-simultaneous flights of a high-altitude NASA aircraft operating the Hurricane Imaging Radiometer (HIRAD), and a lower-altitude NOAA P-3 aircraft operating a Lower-Fuselage Precipitation Radar (LFR) and a Stepped Frequency Microwave Radiometer (SFMR). Using sensor fusion, the resulting wind speed and rain rate retrievals provide unprecedented high-resolution coverage of the wind and rain structure of this powerful hurricane.
On October 22, 2015, HIRAD obtained brightness temperature (Tb) images of hurricane Patricia, while a NOAA P-3 aircraft flew a near-simultaneous “Fig-4” pattern. This paper presents wind speed retrievals obtained using HIRAD multi-frequency Tb’s @ 5, 6 & 7 GHz, given the NOAA Lower-Fuselage Precipitation Radar (LFR) reflectivity.
SFMR measures both wind speed and rain rate very well, but the spatial coverage is severely lacking. On the other hand, the P-3 LFR provides excellent rain reflectivity spatial patterns, but the rain rate accuracy is poor. Finally, HIRAD provides excellent wide swath coverage that hurricane surveillance requires, but the wind speed retrievals are significantly degraded by heavy rain. So, taken alone, each sensor has serious limitations, but there is significant advantage in the sensor fusion of HIRAD, SFMR and LFR data to improve hurricane measurements.
In this paper, LFR was used to produce a composite radar reflectivity image of the hurricane. Using a simultaneous SFMR rain rate measurements, an empirical Z-R relationship was established between the reflectivity and the SFMR rain rate (R). Next, using this relationship, the LFR composite reflectivity was transformed into the hurricane rain field. Then, using this LFR rain rate as a priori information, the HIRAD Tb’s were used to retrieve wind speed (WS) using a constrained maximum likelihood estimation (C-MLE).
The combined active/passive hurricane retrieval algorithm will be described, and wide swath images of hurricane wind speed will be shown, which will display high-resolution features of the hurricane structure. Moreover, comparisons of collocated HIRAD and SFMR WS and LFR and SFMR rain rates will also be presented, which will validate the HIRAD and LFR retrievals. |
