During the summer of 2005 the Center for Collaborative Adaptive Sensing of the Atmosphere (CASA) will deploy a four-node array of low-cost, low-power X-band radars as a proof-of-concept test bed. This test bed, named NetRad, is designed to demonstrate a radar network that responds adaptively and collaboratively to the data that is sensed. In other words, the radars respond to detected atmospheric features and adapt scanning strategies to optimize the measurement and tracking of the features. Such a system is needed to more quickly identify and recognize hazardous weather as well as to provide the spatial and temporal data necessary for improvement in model prediction. The ability of NetRad to collaboratively detect hazardous features depends on the spatial resolution of the data, the network spacing and configuration.

The general area for NetRad was chosen based on several factors, including geography, existing infrastructure, and proximity to end-users. One of the primary goals of CASA is to improve detection of severe thunderstorm winds, including tornadoes; thus, the test bed was placed in central Oklahoma due to the relatively frequent occurrence of severe thunderstorms in the area. Oklahoma is also preferred because of the extensive weather-observing networks already deployed within the state. In addition to the standard operational suite of observing systems, Oklahoma is home to the Oklahoma Mesonet, the Atmospheric Radiation Measurement (ARM) Program Southern Great Plains (SGP) Network and the Agricultural Research Service Rainguage Micronetwork. To supplement NEXRAD radar coverage, the NetRad network was placed about midway between two existing NEXRAD radars. Finally, NetRad was placed upstream of Oklahoma City, a major metropolitan area. This location satisfied the greatest need for the emergency managers and media outlets in the area.

The specific network spacing and configuration were chosen based on optimization experiments. The collaborative operation of the NetRad radars should allow for multi-Doppler analysis for improved detection resolution and wind retrieval. Calculations of intersecting viewing angles were first done for hypothetical radar arrays to determine the radar spacing that would maximize the dual-Doppler coverage area given the proposed radars’ operating specifications. Searching the existing OneNet communications infrastructure for towers that approximated the ideal radar spacing yielded three viable radar sites. The general region for the fourth site was chosen using the first three sites and considerations of the dual-Doppler viewing angle. Details of the terrain, power availability and human factors were considered to identify some specific candidates for the fourth site. Finally, detailed site surveys and engineering studies were done to select the actual location to be used. This process can be extended to configure a larger network, which will be necessary for second phase of the CASA program, for which nine or more radars are planned.