January 31, 2009 Preface Tornado and coastal storm-induced injuries, fatalities and economic losses in the states of Wyoming and Washington were the motivation for this report on the feasibility of new technologies to fill coverage gaps in today’s national weather radar network. This report contains an assessment of current radar coverage in these regions from a meteorological, social, and economic point of view as well as an analysis of the improvements in coverage that would be achieved by installing additional radars in Wyoming and Washington. This  report was  funded by a congressional appropriation “to  determine the applicability to  northeastern Wyoming and other regions the feasibility of integrating a number of small-scale Doppler radar technologies into future National Weather Service observing systems.”  The small-scale Doppler radar technologies cited in this language are the R&D focus of the Center for Collaborative Adaptive Sensing of the Atmosphere (CASA) one of the nation’s 15 Engineering Research Centers chartered by the National Science Foundation. Under rigorous and continuous peer review from the science and engineering community, CASA is investigating a concept in which networks of small, low-cost radars could be deployed on rooftops, communication towers, and other infrastructure elements to address coverage gaps in today’s radar infrastructure.  The CASA Engineering Research Center is a partnership  among nearly 20 institutions, configured into academic, government, and industrial arms. This configuration enables the participants of the center to: develop the concepts behind small radar networks; translate these concepts into new technologies; and commercialize and implement the best concepts in practice. As of this writing, CASA is in its 6th year as a national Engineering Research Center; many of the  center’s  key  concepts  have  been  demonstrated  in  research  trials,  and  development  and  commercialization of various technologies is underway. This report was funded by a contract from NOAA to CASA through the University of Oklahoma with Dr. Jerry Brotzge as Principal Investigator. The University of Oklahoma, in turn, subcontracted part of the work to the College of Engineering at the University of Massachusetts - the organization that leads CASA. Researchers from both institutions contributed to this report. In addition to the meteorological, social, economic and radar coverage assessments in this report, the authors analyze the additional radar coverage that would be achieved by installing small “CASA-type” radars as well as additional larger  conventional weather radars in gap regions. Previous experience with weather radars, including gap-filling weather radars, shows that additional components of the overall “system solution” can be just as important as the radars themselves. These additional components include the operating software, the mechanisms for transporting data to users and integrating the new observations into the existing observational infrastructure, and the provision for ongoing operations and maintenance of the various components of the new radar system throughout its intended lifecycle. This report includes some benchmark cost information but stops short of predicting what it would actually cost for a manufactured solution to the radar gap problem. Such costing information is more appropriately obtained through Requests for Information and Requests for Proposals from radar technology developers and manufacturers who are set up for delivering solutions to problems such as those considered in this report. David J. McLaughlin, Ph.D. Professor of Electrical and Computer Engineering, UMass Amherst Director, NSF Engineering Research Center for CASA January 29, 2009 2 Acknowledgments During the writing of this report, a number of professionals freely gave of their time and effort, including National Weather Service forecasters, River Forecast Center, USGS, and U.S. Army Corps of Engineers hydrologists, emergency managers, industry and media representatives, airport managers, university professors, and U.S. Congressional staff. We deeply appreciate their valuable insight in making this report possible. Specifically, we wish to thank the following individuals for their time and efforts in the completion of this report: Tim Crum, Radar Operations Center, Norman, OK Kevin Kelleher, National Severe Storms Laboratory, Norman, OK Dr. David McLaughlin, Director of CASA, University of Massachusetts Bob Saffle, Noblis, Inc., NWS Office of Science and Technology, Washington D.C Wyoming: Kevin Lynott, MIC Riverton, NWS Mark Betsch, Central Region Headquarters, Kansas City, NWS Matthew Bunkers, SOO Rapid City, SD, NWS Dave Carpenter, MIC Rapid City, SD, NWS John Eise, MIC Cheyenne, NWS Jim Fahey, Service Hydrologist, Riverton, NWS Mike Hudson, Central Region Headquarters, NWS Chris Jones, WCM Riverton, NWS David Lipson, Riverton, NWS Brett McDonald, SOO Riverton, NWS Keith Meier, MIC Billings, MT, NWS Kirk Miller, Cheyenne, USGS Susan Sanders, WCM Rapid City, SD, NWS Melissa Smith, Service Hydrologist, Rapid City, SD, NWS Tony Amormino, , Casper, WY, K2 ABC Affiliate Wanda Burget, Peabody Energy Marilyn Connolly, Johnson County, EM Kimball Croft, Washakie County, EM Bill Dalton, Blasting Supervisor, Foundation Coal Kyle Gravlin, K2 ABC Affiliate, Casper, WY Bryan Hansen, Environmental Engineer, Peabody Energy Vernon Hill, Arapaho Tribe, EM Windy Hutchinson, Arch Coal Angelo Kallas, Green River/Sweetwater County, LEPC David King, Campbell County, EM Steve Laird, Foundation Coal Marion Loomis, Wyoming Mining Association Kathi Metzler, Fremont County, EM 3 Richard Ochs, Teton County, EM Michael J. Patritch, Research Manager, Wyoming Department of Transportation Ken Schultz, State Maintenance Engineer, Wyoming Department of Transportation Walt Stickels, City of Riverton, EM Gary Valentine, Sweetwater County, Airport manager Judy Valentine, Sweetwater County, EM Kevin Walker, Cokeville, (Lincoln County), EM Robin Bailey, Sen. Enzi staff Kelsey Campbell, Sen. Barrasso staff Jennifer Robinson, Cong. Barbara Cubin staff Sandy Tinsley, Sen. Enzi staff Washington: Brad Colman, MIC Seattle, NWS Brent Bower, Service Hydrologist Seattle, NWS Ted Buehner, WCM Seattle, NWS Kirby Cook, SOO Seattle, NWS Don Laurine, Portland, OR, RFC Steve Todd, MIC Portland, OR, NWS Lisa Anderson, Gray’s Harbor Public Utility District Don Carson, Northwest Cable News Eric Grimit, 3TIER, North America Dennis Lettenmeier, Civil Engineering, University of Washington Cliff Mass, Atmospheric Sciences, University of Washington Ross McDowell, Lewis County, EM Socorro Medina, Atmospheric Sciences, University of Washington Carrie Novick, Redman, OR, airport manager Larry Schick, U.S. Army Corps of Engineers Marilyn Smith, Puget Sound Energy Anne Sullivan, Grays Harbor County, EM Julie Palakovich, Sen. Cantwell staff Mike English, Sen. Cantwell staff Jeff Waters, Sen. Cantwell staff Mary McBride, Sen. Murray staff Sara Crumb, Cong. Norm Dicks staff Jason Park, Emergency management, Washington D.C. Other: V. Chandrasekar, Colorado State University Dave Pepyne, University of Massachusetts Jenniffer Santos, University of Delaware Baxter Vieux, University of Oklahoma Mike Zink, University of Massachusetts 4 Table of Contents Acknowledgments 3 Table of Contents 5 Executive Summary 7 Introduction 1. Introduction 13 System Performance 2. Socioeconomic and Climatological Risks 19 2.1 Wyoming 2.2 Washington 3. Current Weather Radar Coverage 47 3.1 Importance of low-level radar coverage 3.2 Low-level WSR-88D weather radar coverage 3.3 Ongoing enhancements 4. Warning System Performance 55 4.1 Wyoming 4.2 Washington Feasibility of Additional Weather Radar 5. Requisite Radar Attributes 65 5.1 Analysis and detection 5.2 Numerical weather prediction 5.3 Summary of radar needs 6. Potential Radar Solutions 71 6.1 Long-range vs. short range radar deployments 6.2 Potential radar deployments 6.3 Short-range weather radar (e.g., CASA) 7. Infrastructure and Integration 83 7.1 Radar integration 7.2 Cost issues Summary 8. Summary and Recommendations 89 8.1 Wyoming 8.2 Western Washington 8.3 Final recommendations References 97 Abbreviations 105 5 Appendixes A Radar Network and Surface Observing Systems 107 B Presidential Federal Disaster Declarations 123 C National Climatic Data Center Storm Event Archives 125 D Methodology for Determining Unblocked Radar Beam Height 129 6 Executive Summary I. Introduction Is it feasible to deploy additional weather radar in Wyoming and western Washington? This study assesses the meteorological need for and the feasibility of deploying additional radars to augment the current radar observing system in northeastern and southwestern Wyoming and coastal Washington. The current NEXRAD WSR-88D radar network is operated by the National Oceanic and Atmospheric Association (NOAA) through the National Weather Service (NWS). The catalysts for this study1 were two severe weather events, an F-2 tornado in northeastern Wyoming and a three-day coastal storm system in Washington that raised awareness of the radar coverage gaps that exist in the current observing system. These coverage gaps have several causes: the distance between radars (345 km in the western United States); radar beam blockage due to mountains and other terrain; and the “horizon problem” which prevents long  range radars from observing the lowest parts of the atmosphere [< 3 km above ground level (AGL)] at distances greater than 175 kilometers. Radar beams travel generally in a straight path, while the earth curves away from the beam. Therefore, at greater distances, radar can only observe the mid- and upper atmosphere as shown in the figure above (NRC 2005). Furthermore, weather climatology research demonstrates that many hazardous weather phenomena occur in the lowest 3 kilometers (~10,000 feet) of the atmosphere (NRC 1995). The authors’ approach to this radar feasibility assessment incorporates NOAA’s vision which is  “to understand and predict changes in the earth’s environment …to meet our nation’s economic,  social, and environmental needs.”  In addition, we also take a holistic approach to evaluating the radars as part of an end-to-end warning system where socioeconomic, meteorological, geographic and technical factors all contribute to the performance of the system. We focus on specific regions – northeastern and southwestern Wyoming and coastal Washington – with the expectation that these results may also be applicable to other “radar gap” regions across the  country. 1 This study is funded through a Congressional appropriation sponsored by Senators Mike Enzi of Wyoming and Maria Cantwell of Washington and managed by in the National Oceanic and Atmospheric Administration (NOAA) through the National Severe Storms Laboratory. The National Severe Storms Laboratory, in turn, contracted this study to CASA, the Engineering Research Center for Collaborative Adaptive Sensing of the Atmosphere, a National Science Foundation research center focusing on the development of low-cost, short-range Doppler radar networks. 7 Part I of this study evaluates current system performance in the radar gap regions to determine if service deficiencies exist. This section looks at socioeconomic risk factors, climatology, historical impacts of hazardous weather, current radar coverage, NWS performance metrics, and feedback from stakeholders, such as emergency managers. Based on these findings, Part II evaluates which radar attributes, network solutions, and supporting infrastructure are needed. Both short-range and long-range radar solutions are considered. II. Key Conclusions The summary information compiled for this report points to three key conclusions: Service deficiencies exist across the radar gap regions identified in Wyoming and western Washington. Case studies, stakeholder interviews, and NWS warning statistics all indicate that severe storm warning lead times are below-average for those regions with limited radar coverage at low levels (< 3 km AGL). Detections of precipitation and wind shear at low levels are limited in radar gap regions. Additional radar coverage below 3 km (10,000 ft) likely could improve public safety and reduce negative economic consequences from hazardous weather through improved real-time analysis and prediction. In Wyoming, some towns (e.g., Gillette) and critical infrastructure (interstate highways, coal mines) have limited low-level radar coverage. Additional low-level radar data are needed in these areas for improving winter weather quantitative precipitation estimation (QPE) and quantitative precipitation forecasts (QPF) as well as monitoring low-level storm features. Dual-polarization would also aid quantifying winter precipitation and summertime convection, and higher spatial and temporal collection would improve monitoring of severe storms. For coastal Washington, high socially vulnerable areas and weather sensitive industries (e.g., fishing industry) have limited low- level radar coverage. Greater radar coverage in this region, particularly over the ocean, could improve the analysis and prediction of synoptic systems. Furthermore, dual-Doppler estimated winds could aid in identifying areas of strong winds, and higher temporal and spatial sampling could improve QPE and QPF. Deploying additional weather radars in Wyoming and western Washington will require a system engineering approach to achieve an effective solution to the gap problem. Hardware costs, siting, tower infrastructure, communication and electric power requirements, installation, communications interfacing, software integration and long-term maintenance and operations all will need to be carefully considered. For the deployment of short-range radars, consideration must be given to the availability of multiple sites, small power radars and communication availability, and long-term maintenance. For the deployment of long- range radars, consideration must be given to the “social footprint” (e.g., visual impact, land  use) and specialized power and other infrastructure needs that accompany the installation of a large high-power radar systems. 8 More specific discussion of the meteorological assessment and feasibility of additional weather radar in Wyoming and western Washington are discussed below. Wyoming Significant radar coverage gaps exist in northeastern and southwestern Wyoming. The gap regions in Wyoming lack radar coverage because of the distance between radars and mountain blockage. In Campbell County, no radar coverage is available below 3 km AGL, and only 24% of Johnson County and 39% of Sweetwater County have coverage below 3 km AGL. Radar gap areas are regionally active areas for severe weather. Based on historical data, Campbell County is three times more likely to report a tornado and twice as likely to experience flooding than other counties in Wyoming. Regional weather hazards during the warm season include tornadoes, hail, and straight-line winds. During the winter, storms often have snow, ice, and winds as their primary hazards. A range of low to moderate societal vulnerability exists within the radar coverage gap regions. Overall, Wyoming has a low population density with many isolated communities. The complex terrain contributes to highly variable weather risks across the state. However, Wyoming ranks 8th in the country in the percentage of homes that are mobile home units, thereby increasing resident vulnerability to tornadoes and high winds. Wyoming ranks 5th in the country in the percentage of the population born out-of-state. Non-native residents may be more susceptible than long-term residents to flooding and winter weather hazards common to Wyoming. The principal socioeconomic risks in these counties are to the mining industry and transportation. Wyoming provides 40% of the nation’s coal, and 70% of this coal is  mined in Campbell County. The coal industry uses forecasts and nowcasts of convective activity, snow, and winds to conduct its operations and protect its workforce. In Sweetwater County, Interstate 80 is a major east-west transit freight transportation route which closes from 3 to 45 days annually due to wind and snow, with an estimated negative impact to the national economy of approximately $84,000 to $333,000 per hour. Interviews with National Weather Service forecasters having jurisdictional responsibility over radar gap regions indicate that the lack of observations in gap areas impact their ability to provide expert knowledge and warnings to stakeholders. Findings from the interviews are confirmed by emergency managers and reinforced by the NWS warning statistics. For the counties sampled in the radar gap areas, lead times for severe storms, tornadoes, and flash floods are all below parent WFO averages. Most verification statistics for these three counties are below or much-below national averages. For example, in Sweetwater County only 14% of thunderstorm warnings are issued in advance. Additional low-level coverage is recommended for the gap areas in Wyoming. Low- level radar data are needed for improving winter weather QPE and QPF as well as 9 monitoring low-level storm features. Dual-polarization would also aid quantifying winter precipitation and summertime convection, and higher spatial and temporal collection would improve monitoring of severe storms. Forty-two small X-band radars (21 radars in NE Wyoming and 21 radars in SW Wyoming) or two large S-band radars (one in each region), could provide this coverage. The 42 short-range (X-band) radars, deployed strategically along critical infrastructure, would provide extensive multi-Doppler coverage of wind and rain at low- levels (below 2 km AGL), thereby enabling real-time monitoring and improved prediction of warm-season severe thunderstorms and low-level winter weather. Two long-range weather radars could provide equivalent coverage at and above 2 km AGL. Western Washington Radar coverage gaps in coastal Washington are primarily caused by beam blockage. Much of the region is blocked by the Olympic Mountains to the west, and so there is virtually no coverage over the ocean where the majority of western Washington's weather hazards originate. Large synoptic storms are the primary weather events faced by these regions. Powerful, mid-latitude cyclones often come ashore in western Washington and are the primary weather hazard to the region. These storms bring large areas of high wind and precipitation which interact with the mountainous terrain and create highly localized and intense wind and rainfall, events which are difficult to predict in numerical forecast models. Population along the west coast of Washington exhibits high social vulnerability based on a national index. This index measures a population’s ability to prepare for and  recover from natural disasters. Counties along the Washington coast share a high societal vulnerability and yet experience the greatest impacts from Pacific storms. Western Washington has significant industries and transportation routes that contribute to the national and international economy and that are sensitive to hazardous weather. For example, the lack of weather radar coverage over the ocean poses a safety hazard to the fishing industry. Interstate 5 is one of the busiest routes in the nation and intersects a radar gap area. NWS forecasters lack observations over the ocean and reliable data over land. Because of the lack of radar data over the oceans, analysis, tracking and prediction of these large synoptic storms is difficult. Only satellite data and a few buoy observations are available for assimilation into NWP models. Furthermore, because of the highly complex terrain, radar blockage, and a low melting layer (~ 2 km AGL), radar reflectivity over land also poses problems.  Radar is often considered “a secondary tool”, since the  estimates may not reflect ground truth. 10