Storm Trackers – Engineering Research Center for Collaborative Adaptive Sensing of the Atmosphere CASA
The cornerstone of today's weather observation and warning system is a nationwide network of physically large, high-powered Doppler radars. These long-range radars are effective in mapping the middle and upper regions of the atmosphere, but as illustrated in Figure 1, they are blocked from observing the weather near ground level due to the Earth's curvature. This inability to "look down low" significantly limits the accuracy of current weather forecasts and warnings. As evidenced by the series of deadly tornado outbreaks across the central and southeastern United States in the spring of 2011, action needs to be taken to reduce the number of fatalities associated with these hazardous weather events.
A Center for Atmospheric Sensors
At the request of the White House and the National Academy of Engineering, the Engineering Research Centers (ERC) program was established at the National Science Foundation (NSF) in 1984 as a national priority to strengthen the competitiveness of U.S. industry. The goal was to establish centers that would develop a new interdisciplinary culture in engineering research and education in partnership with industry. Together they would advance knowledge and technology and educate new generations of engineers who understand industrial practice and the process of advancing technology, design and manufacturing, preparing them to work productively in industry upon graduation.
The ERC for Collaborative Adaptive Sensing of the Atmosphere (CASA) was started in 2003. CASA was formed with the vision of a transformative radar network technology that introduces a new, more accurate dimension to weather forecasting and warning, providing capabilities that did not exist previously. CASA has its origins in a relationship between university and industry reaching back more than 20 years before the center's inception in 2003. In 2000, the University of Massachusetts (UMass) Amherst and Raytheon decided to extend an existing microwave design-based educational collaboration into a systems-level education and research partnership as a way of advancing the missions of both organizations as well as the public good. Three years of hard work culminated in winning a coveted NSF Engineering Research Center grant with four core academic partners: UMass-Amherst, UOklahoma, Colorado State and UPuerto Rico – Mayaguez. After eight years, CASA is a highly effective ERC that has grown to include 20 different industry, academic and government partners. Along with Raytheon, CASA's industry partners have included IBM, ITT, HP, EWR Weather Radar, Vaisala, ParoScientific Inc., Vieux and Associates, OneNet and Weather News International. Government partners have included NOAA's National Severe Storms Laboratory, the National Weather Service and Environment Canada.
The CASA project pursues innovation to supplement or replace the present network of 150 large radars with thousands of small radars that can be deployed on cell phone towers, rooftops and other existing infrastructure. Utilizing Raytheon's radar design and active circuit card assembly (CCA) array technology expertise, a network of low-power, lower-cost sensors will be connected in an intelligent network providing distributed collaborative adaptive sensing of atmospheric conditions near the Earth's surface. The closer spacing of these radars will avoid the obstruction caused by the Earth's curvature and allow forecasters to directly view the lower atmosphere with high-resolution observations. This new dimension to weather observation leads to improved characterization and better forecasting of storms, resulting in improved warning and response to tornadoes and other weather-related hazards.
The central goal is to design and deploy a system that can sample the atmosphere when and where the user need is greatest, and provide accurate, timely and useful information. It must reliably issue alerts that the public trusts and responds to appropriately. To accomplish this, the system design must also address the software and computing architecture required to maintain the system's many resources, the data volume, and the communications and user interface requirements.
CASA's Systems Approach
The solutions to the different problems posed by such an ambitious project are to be found through CASA's work on three planes of engineering research and development, as shown in the ERC strategic plan diagram in Figure 2. CASA's system focus (upper plane) is a real-time distributed system capable of focusing its resources onto particular volumes of the atmosphere and delivering forecasts and other data to enable forecasters, emergency managers and the public to generate accurate alerts and make effective decisions when extreme weather events occur. Realizing an effective and efficient system requires a number of new enabling technologies (middle plane) and fundamental research to create new knowledge (bottom plane).
Multi-disciplinary research in the lowest plane spans electrical engineering investigations of electromagnetic waves interacting with the atmosphere as well as sociological and decision-theoretic studies about individuals and organizations responding to severe weather hazards, utilizing warnings and taking protective action. Whereas the lowest plane tends to be the "comfort zone" of academic faculties, CASA's strategic concept is an interdisciplinary orchestration of work in all planes (integrating the resources of industry and academia), from the definition of the system in the upper left, through targeted technology and fundamental investigations, culminating in the creation of end-to-end system test beds. It is in these test beds (upper right) that the various elements are combined and one learns whether CASA's central concept will be successful assisting users in their difficult jobs of predicting, warning and responding to tornadoes and other hazards.
In current operational weather radar networks, radar coverage is non-overlapping (except at high altitudes) and the radars
are operated largely independent of one another, repeatedly searching the entire volume around the radar via mechanical scanning. In contrast, an essential feature of CASA's approach is to arrange the radars to have full overlapping coverage so that every location in the network is visible to multiple radars. This permits the use of a radar control architecture shown in Figure 3 that coordinates the beam scanning of the radars in the network both collaboratively — to obtain simultaneous views of a region for data fusion-based algorithms such as multiple-Doppler wind field retrievals — and to adaptively optimize where and how the space over the network is scanned based on 1) the type of weather occurring there and 2) the data product needs of the system's users. The result of this collaborative adaptive sensing approach is network-level performance that exceeds the capabilities of its component radars in terms of update rate on key weather features, spatial resolution, sensitivity, and the ability to support multiple users and multiple applications. Such network advantages decrease the design requirements on the individual radars that make up the network. Key radar size and cost drivers — such as the antenna size and the peak transmitter power needed to achieve a particular level of resolution or sensitivity — are lower than they would need to be if the radars were not part of a collaborative, adaptive network. The result is an average transmitted power requirement of only several tens of watts per radar and a required antenna aperture diameter of only one meter.
Active electronically steered antenna arrays are a key enabling technology since they do not require the maintenance of moving parts and their size allows installation on towers and other existing infrastructure. A particular challenge in realizing a cost-effective network composed of thousands of radars will be to achieve an AESA design that can be volume-manufactured at low unit cost. Several thousand transmit/receive (T/R) channels are needed in each array. The realization of such an antenna benefits from leveraging commodity silicon radio frequency semiconductors to achieve T/R functions, in combination with very low-cost packaging, fabrication and assembly techniques. A prototype CCA AESA, being developed by Raytheon, is shown in Figure 4, along with an artist's rendering of antennas as they might appear mounted on a cellular communications tower and on the side of a building.
CASA's first test bed was a four-radar system integrated in central Oklahoma, directly in "tornado alley." Since 2007, this system has tracked dozens of severe thunderstorms moving through the region. Figure 5 illustrates the improved storm morphology achieved with the CASA prototype compared with the more limited resolution achieved with the operational NEXRAD radar network deployed across the nation. The test bed demonstrates capabilities that are beyond today's operational state of the art, including the ability to resolve high-resolution "hook echoes," which are important indicators of tornado genesis that are poorly resolved in today's radar network. The life-saving implications of these capabilities were on dramatic display during the May 24, 2011, tornado outbreak in southwestern Oklahoma. After touching down in McClain County, a tornado progressed on a zigzag path, first traveling east, then north, and then northeast near the town of Newcastle. Since the directional changes happened too quickly to be resolved by the national Doppler radar network, emergency management officials relied on the CASA imagery to follow the twister and move people out of its direct path while they staged rescue and response assets. Officials noted that CASA information was critical for their decision-making during the event as they worked to shelter 1,200 people. Without the imagery, they would not have known the tornado track was changing direction and they would have directed people to the wrong location.
Motivated by these trials, the CASA team is forming partnerships with weather offices, universities and government agencies in Australia, the U.K., Canada and elsewhere to explore the applicability of the technology for improved forecasting and response to floods, wind storms, bush fires and other hazards to life and property.