3.3
1. INTRODUCTION
For many years, evidence suggested that local officials -- when confronted with making decisions that involved impending weather emergencies -- normally were ill-equipped to respond using current and localized information at their fingertips. The key missing ingredients were relatively easy and cost-effective access to timely radar information from the national network of WSR-88Ds and proper training in data acquisition and use. Information-based tools simply were not available at the community level. Consequently, dividends at the local level from the NWS modernization seemed destined to be limited. Yet, a favorable work environment existed in Oklahoma to create a program that would benefit the public safety community.
Unisys Corporation an authorized vendor of WSR-88D NEXRAD data through its NIDS Agreement with the NWS approached the Oklahoma Climatological Survey (OCS) in May of 1995. Unisys presented an opportunity to redistribute NIDS data to public safety officials who were state and/or locally supported. Today, 65 public safety officials across Oklahoma (Fig. 1) are well versed in the availability and use of information from 15 NEXRAD units that provide surveillance of Oklahoma (Fig. 2). More importantly, success stories continue to accumulate regarding the local application of timely NEXRAD information.
The success of OK-FIRST has resulted from partnerships at several levels. These partnerships include a federal-state-community relationship between the NWS, OCS, and many local agencies. Yet, perhaps the most critical linkage was the public-private relationship established between OCS and Unisys. It was clear that Unisys desired to break new ground to reach a market not fully tapped. This manuscript is about a NIDS success story we believe resulted from a convergence of technology, software, a favorable work environment, and a positive corporate business policy.
2. HISTORICAL PATH TOWARD NIDS
The NWS faced a number of operational challenges with its RRWDS arrangement; these included management of a great number of users, monitoring and control of services, and individual user support. Certain RRWDS sites were quite popular because they provided surveillance of major population centers (e.g., the radar at Pax River covered Baltimore, MD and Washington, D.C). Providing a data-sharing capability at Pax River bordered on the unmanageable when, at one point, about 100 users were individually attached to this single WSR-57. NWS resources were stretched beyond the breaking point. As a result, some users were not serviced during critical weather events, primarily because the NWS focused on providing watches and warnings during critical weather events. In the 1990s, the NWS devised a plan to correct inherent problems associated with the RRWDS era. Under a program called the NEXRAD Information Dissemination Service (NIDS), the NWS created a limited-access user agreement with four commercial weather data vendors. The NWS received immediate benefit from the NIDS program by off-loading management and user support to the NIDS vendors. As the NIDS program progressed, the NWS benefited by outsourcing individual client management to NIDS vendors such as Unisys Corporation. In the latter half of the 1990s, as all WSR-88D radars were commissioned, the benefit of the NIDS program became clearly evident. Competition between the NIDS vendors created an environment where pricing was controlled by market pressure, allowing widespread distribution of radar data to thousands of users. NIDS clients supported the creation of value-added radar products, while the vendors promoted rapid adoption of the new WSR-88D. Since the NIDS Agreement limited the number of vendors that interfaced with the NWS, the NWS could concentrate on its critical mission. The result was an alliance between the NWS and the commercial weather industry that benefited the nation as a whole. 3. WHY OK-FIRST?
In the 1970s, the NWS permitted commercial weather users to connect directly to NWS radars at various places across the country. Under this arrangement, commercial weather users provided their own communications equipment to connect to a Radar Remote Weather Display System (RRWDS) that provided digitized radar data. However, due to funding limitations, the NWS did not install RRWDS units at all locations. This fact led to another class of radar users (separate from RRWDS-enabled radars) whereby commercial weather vendors sup-plied equipment to access called gap filler radars. In addition, radars west of the Rocky Mountains simply were not available to outside users.
OK-FIRST was born out of a desire to strengthen and upgrade the public-safety component of Oklahomas emergency warning system. OCS built OK-FIRST upon its successes in designing, installing and operating the Oklahoma Mesonet (Brock et al., 1995) and upon an educational outreach program (known as EARTHSTORM) that moved near real-time Mesonet data into the K-12 schools of Oklahoma (McPherson and Crawford, 1996). This base of credibility and experience was essential to the launching of OK-FIRST. In addition, the widespread availability of the Internet by the early 1990s plus establishment of a state-operated, fiber-optic network (known as OneNet) delivered an essential ingredient for OK-FIRST (namely, a reliable, cost-effective system for two-way, high-speed telecommunications). A final reason to establish OK-FIRST centered around the fact that a centralized collection and distribution system for NEXRAD data did not exist within the NWS, and thus, was a methodology that the NWS simply could not address. On the other hand, the Unisys opportunity became a major benefit of OK-FIRST: namely, that analysis and use of WSR-88D data could occur through a network approach. For example, a public-safety official located midway between Oklahoma City and Tulsa could view their area of responsibility using two or more WSR-88Ds (illustrated in Fig. 3). The advantage is two-fold. First, the public safety official can cor-roborate image information between multiple radars and second, he can mitigate those rare but real-life situations when a given WSR-88D might require maintenance at the most inopportune moment.
While the NWS significantly upgraded and mod-ernized its operations during the past ten years, for the most part, local public safety agencies concerned with environmental emergencies still suffer from a lack of: (1) modern technology, especially in rural areas; (2) easy access to improved information from the modernized NWS; and (3) training on how best to use the dividends from a modernized NWS.
4. THE OK-FIRST/NIDS DISSEMINATION SYSTEM
Morris (1998) documented the design and implementation of the OK-FIRST decision-support system for public safety agencies in Oklahoma. The first building block is a low-end, UNIX-based web server networked to the NIDS ingest system purchased from Unisys. All NIDS products (Morris and Duvall 1999; their Table 1) generated during each volume scan from 15 WSR-88Ds (Fig. 2) are acquired via satellite from Unisys in Kennett Square, PA.
It is fortunate that the Internet rapidly became a reliable technology to mesh with OK-FIRST for the transfer of digital information to local officials in remote locations. A critical decision was made to share NIDS data using Internet browsers rather than pre-generating images at the primary file server. The OK-FIRST methodology simply stores and forwards data files in a native NIDS format. Thus, data processing and image building occurs on the user-end via browser plug-in software developed at OCS. Based upon extensive feedback from the 65 trained users and from our NWS colleagues, we believe the plug-in software and accompanying web pages are simple and intuitive. More importantly, the interactive plug-in software permits the data to be manipulated at the local level. For example, Figure 3 illustrates a two-screen display of reflectivity images as observed at the height of a major tornado outbreak in central Oklahoma on 4 October 1998.
5. NIDS USAGE IN OK-FIRST
Other products that found favor with OK-FIRST graduates include (Fig. 5): (1) reflectivity at the second-lowest elevation angle, (2) base reflectivity out to 248 nmi, (3) Doppler velocity at the lowest elevation angle, (4) composite reflectivity with 16 data-level resolution (also used to access algorithm output after 15 August 1998), (5) storm relative velocity, (6) vertically-integrated liquid (VIL), and (7) the storm-attribute table. Each of these products com-posed 5% or less of the number of files transferred (i.e., between 5,000 and 20,000 during this 16-month period). Emergency managers downloaded VIL products more often than did other user groups to help distinguish the most severe of storms and to move storm spotters away from damaging hail cores. Perhaps the best testimony on the application of OK-FIRST information to the public safety mission in Oklahoma occurred on 4 October 1998 a day when 20 tornadoes occurred, setting a new national record for most tornadoes in one state on any one day in October. During a 24-hour period centered on this outbreak, the OK-FIRST server processed 86,500 On this day, a significant tornado struck in the City of Moore (between Oklahoma City and Norman). Gayland Kitch, Emergency Management Director for the City of Moore, deployed his storm spotters prior to this tornado. One spotter was deployed about 5 miles due west of Moore. As the storms intensified and moved toward Moore, the OK-FIRST display was used to relocate this spotter a few miles south. Ultimately, this spotter provided the first knowledge of a large wall cloud that eventually spawned the tornado which, in turn, destroyed/severely damaged about a dozen homes and apartment buildings. This spotters call, along with a radar-based warning from the NWS in Norman, led to the activation of Moores warning system. In addition, this particular spotter report likely played an important role in the issuance of the NWS warning. The spotter who made the call later indicated that
had he not been relocated, he would not have been in the proper location to see the wall cloud. In the
words of Gayland Kitch: This scenario is exactly what OK-FIRST was designed to do!
The 65 OK-FIRST participants were trained in one of three different workshops. Each participant received a total of eight training days. Their use of OK-FIRST information (Figs. 4-5) reflects system use following their individual training. The number of NIDS files downloaded from the OK-FIRST server exceeded 428,000 files (Fig. 5) between June 1997 and September 1998 (a period that witnessed a significant drought and only limited outbreaks of severe weather). Of this number, 332,600 files were base reflectivity images from the radars lowest elevation angle (for 80% of the total NIDS files shared). This fact is not surprising since Oklahoma television stations have been broadcasting reflectivity images for nearly three decades.
requests and shared nearly 25,000 radar files.
6. SUMMARY
Yet, these accomplishments likely would not have occurred during the 1990s without assistance from Unisys Corporation. They made possible a network approach toward meeting the information needs of modernized public safety agencies. We consider this accomplishment to be a NIDS success story. Thus, it would appear that Unisys is poised to replicate an OK-FIRST-like concept beyond the borders of Oklahoma.
The goal of OK-FIRST is to implement an interactive, 21st century information and support system for public safety agencies to provide timely, detailed, and relevant environmental information and forecasts that, in turn, help public safety agencies harness the information age. We believe we have successfully demonstrated the feasibility, relevance, and utility for such a system.
7. REFERENCES
Brock, F.V., K.C. Crawford, R.L. Elliott, G.W. Cuperus, S.J. Stadler, H.L. Johnson, and M.D. Eilts, 1995: The Oklahoma Mesonet: a technical overview. J Atmos. Oceanic Tech., 12, 5-19. Crawford, K.C., D.A. Morris, R.A. McPherson, H.L. Johnson, M.A. Shafer, J.M. Wolfinbarger, and T.W. Hughes, 1998: OK-FIRST: A decision-support system for public safety agencies. Preprints, 14th Intl Conf. on IIPS for Meteor., Ocean., and Hydro. Amer. Meteor. Soc., Phoenix, AZ., 5-9. McPherson, R.A., and K.C. Crawford, 1996: The EARTHSTORM project: encouraging the use of real-time data from the Oklahoma Mesonet in K-12 classrooms. Bull. Amer. Meteor. Soc., 77, 749-761. Morris, D.A., 1998: The design and implementation of a web-based decision-support system for public safety agencies. Preprints, 14th Intl Conf. on IIPS for Meteor., Ocean., and Hydro. Amer. Meteor. Soc., Phoenix, AZ., 10-14. Morris, D.A. and C. Duvall, 1999: An evaluation of the use of real-time weather data by public-safety agencies. Preprints, 15th Intl Conf on IIPS for Meteor., Ocean., and Hydro. Amer. Meteor. Soc., Dallas, TX., in press. Wolfinbarger, J.M., R.A. Young, and T.B. Stanley, 1988: Interactive software for viewing NEXRAD Level 3 data on the World Wide Web. Preprints, 14th Intl Conf. on IIPS. Amer. Meteor. Soc., Phoenix, AZ., 208-213.