Tuesday, 7 August 2007
Halls C & D (Cairns Convention Center)
Handout (359.8 kB)
For most weather radars, such as the WSR-88D, reflectivity, radial velocity and spectrum width are the only parameters estimated. Recently, a technique to retrieve near-surface refractivity has been developed by Fabry et. al. [J. Atmos. Oceanic Technol., 14, 978-987, 1997]. The technique relies on the returned phase from ground clutter which changes according to the refractivity of the atmosphere. Until recently, most refractivity measurements have been focused on S-band radars. These radars are usually designed, however, to observe long ranges and are therefore limited in range by the earth curvature effect. As part of the CASA NSF Engineering Research Center, higher-frequency X-band radars have been designed for observations of the lower atmosphere. The initial network (IP-1) consists of four radars and is located in south-west Oklahoma. Because of the closer spacing of these radars (approximately 30~km), in comparison to the WSR-88D network, the IP-1 network is less susceptible to the earth curvature effect and can provide more complete coverage of estimated refractivity. A significant challenge arises, however, with shorter wavelength radars in the implementation of refractivity retrieval. The refractivity retrieval technique relies on the phase change between two radar scans. One is referred to as the reference scan and the other as the measurement scan. A typical field of phase-change between these two scans exhibits a phase wrapping signature that depends on the refractivity change between the two scans and the radar wavelength. For X-band radars, the phase obviously wraps more frequently in comparison to S-band radars, which makes subsequent processing steps problematic. To mitigate this problem, we have proposed an algorithm called Differential Refractivity Retrieval (DRR), which accumulates phase differences from scan-to-scan rather than over a longer time period, as is currently the practice. As a result, typical atmospheric changes over such a short time (less than 5 min) do not cause a significant change in signal phase, minimizing phase wrapping. As a possible drawback, error accumulation caused by the DRR algorithm will be investigated as a limitation of the technique. A field experiment was conducted during REFRACTT-2006 using a mobile X-band radar (XPOL) developed by the University of Massachusetts. The XPOL was positioned next to the CHILL radar (S-band) for side-by-side comparisons. Results from the XPOL radar and the CASA IP-1 network will be presented to illustrate the feasibility of refractivity retrieval using an X-band radar.
- Indicates paper has been withdrawn from meeting
- Indicates an Award Winner