Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.
New Structure of the National Weather Service A major purpose of the NWS modernization is to improve dramatically the short-term forecasts of significant weather events and warnings of severe weather. The new observation and information systems will not merely replace antiquated equipment but, rather, will provide new insights into the evolution of small-scale atmospheric systems, as well as provide longer lead times and precision in forecasting small-scale, short-duration weather events. Details about the weather that have been lost between stations on the conven- tional weather map will now be as visible as the large-scale weather systems observed since the days of Benjamin Franklin. Moreover a 12- to 24-hour forecast of convective weather somewhere in a region, can be updated with observations and forecasts of precise locations, intensities, and life cycles of specific weather phenomena with lead times of 30 minutes to six hours. To achieve this aim, meteorologists and hydrologists must be able to observe their service domains continuously and must have a workload commensurate with the area covered, the short response time necessary for effective warning, and the effective range of available observations (e.g., Next Generation Weather Radar). These human factors must be paramount in evaluating field service structures proposed for the modernized NWS. WEATHER FORECAST OFFICES The Committee has examined the various configurations of the Weather Forecast Office (WFO) network that have been considered and endorses the network of 115 WFOs proposed in the strategic plan (DOC, 1989). The area of forecasting and warning responsibility for each WFO within a 115-station network appears to be a reasonable compromise. This network configuration will be validated by the Modernization and Associated Restructuring Demon- stration (MARD) to be conducted for one year in the midwestern United 35
States around 1993, a schedule that is in jeopardy because of continued delays in implementation of the Advanced Weather Interactive Processing System as discussed in Chapter 3. However, the Committee is very concerned about a report that the Department of Commerce has decided to modify the MARD to test the efficacy of a "two-tiered" network having about one-half as many WFOs as now planned while maintaining the current proposed network of 115 Next Generation Weather Radar (NEXRAD) units. Attempting to double the area covered by each WFO without a propor- tional increase in staff on shift could seriously jeopardize the ability of each WFO to deal effectively with small-scale weather events in the issuance of forecasts and warnings over such a large area. The Committee believes that a significant reduction in the total number of shift meteorologists will not be feasible until there are major advances in the quality and accuracy of small- scale numerical prediction models. Moreover, coordination of warnings with state and local government would also be degraded by doubling the area of responsibility for each WFO.1 Furthermore, the 115-station WFO network configuration coincides with the expected effective coverage of the new NEXRADs, which has a radius of around 200 km from each unit. Thus, each WFO can be located at or very near its associated NEXRAD system to take maximum advantage of high-resolution Doppler radar data for severe storm forecasts and warnings without the cost and complexity of relaying and remotely processing all of the data produced by each NEXRAD. The planned NEXRAD network (Figure 1) will provide nearly total coverage of the coter- minous United States, except for some gaps in the western United States where mountains block the radar signal. (Additional units will be installed in Alaska, Hawaii, and Puerto Rico.) In the two-tiered network alternative (about 50 WFOs and 115 NEXRADs) proposed by the Department of Commerce for testing in the MARD, the Committee understands that those NEXRAD locations that are not also WFOs would perform radar observation functions and issue warnings. If the staff at the NEXRAD-only offices does not include meteorologists on each shift to utilize fully the new technology being introduced in moderniza- tion, there is a danger that these offices will not be able to produce warnings and local forecasts of the requisite quality. The certification process (see Chapter 6) requires that the quality of the forecasts and warnings for all areas of the United States, regardless of distance from a weather office, be at least as high as today even though the number of offices will be cut in half. To achieve this level of performance, the quality of the output of each office after the restructuring must be substantially increased to compensate for the reduc- See also discussion and recommendation in the section on public institutions in Chapter 5. 36
37
tion in the number of offices. Limited staffing places this requirement in jeopardy. Conduct of the proposed test of the two-tier concept clearly would delay the MARD and add major complexities to an already difficult demonstration. For example, proper design of the demonstration would require isolating the personnel operating one kind of network being tested from those operating the other network so that the output of one would not influence the output of the other. Also, erroneous conclusions may be drawn from extrapolation of the MARD results to other geographic regions and time periods. Moreover, a two-tier test would surely increase significantly the difficulties involved in using the MARD results in the certification process required by Congress. Finally, the need to transmit the full-resolution data from two or three remote NEXRADs to a WFO and to merge these data in "real time" for use by meteorologists, although technically feasible, would add significantly to the complexity, cost, and the time required to implement both the MARD and, subsequently, the entire modernization. Recommendation: The Department of Commerce should carefully reconsider its decision to have the National Oceanic and Atmospheric Administration/National Weather Service conduct a two-tiered Mod- ernization and Associated Restructuring Demonstration because a configuration of significantly fewer than 115 Weather Forecast Offices will lead to serious degradation of weather services. Moreover, such an experiment would be much more complex and expensive, and would probably lead to a serious delay in the National Weather Service modernization. HYDROLOGY IN THE NATIONAL WEATHER SERVICE MODERNIZATION The nation's need for improved management of water resources and more accurate flood forecasting will increase during the 1990s. Growth in population in the arid western United States and increasing sensitivity in all parts of the country to precipitation anomalies will result in demands on the NWS for more detailed and more timely hyclrological forecasts. Modernization of the NWS presents two opportunities for improving hydrological services: (1) The detailed quantitative precipitation measure- ments and forecasts that will become available through new observational technology and forecasting capabilities will significantly improve both flash flood prediction and regional runoff estimates that will also impact forecasts for larger basins. (2) The development of new forecast techniques and more 38
powerful communications systems will promote better cooperation between the meteorologists producing forecasts and hydrologists than now exists, a deficiency highlighted by the executive summary of Hydrometeorological Service Operations for the 1990's (NWS Office of Hydrology, 1989). Each of the 13 River Forecast Centers (RFCs) will be collocated with a WFO after the NWS modernization is completed. Thus the Committee anticipates that proper integration of the new developments in hydrological science and practice (NRC, 1991) and the capabilities being created by the modernization of NWS operations could provide greatly improved hydrological services on all time scales. Interaction of Weather Forecast Offices and River Forecast Centers Forecasting the effects of extensive and persistent rainfall associated with large-scale weather systems requires strong collaboration between Weather Forecast Office (WFO) meteorologists and RFC hydrologists. Currently, RFC hydrologists use the temperature and Quantitative Precipitation Forecasts produced by meteorologists in the Weather Service Forecast Offices to prepare hydrological forecasts. The improved numerical weather prediction guidance expected in the 1990s should result in better utilization of Quantita- tive Precipitation Forecasts and temperature forecasts by hydrologists, thereby improving the quality of hydrological forecast services. Recommendation: Incorporation of improved Quantitative Precipita- tion Forecasts and associated uncertainties into the hydrologic models for short-range and long-term stream-flow forecasts is essential and requires collaborative scientific investigation by the National Weather Service and the academic community. Cross-training of both meteorologists and hydrologists will help to ensure optimum collaboration between RFC and WFO personnel during the prepara- tion of hydrological forecasts. The present lack of training in hydrology for meteorologists and the equivalent lack of meteorological training for hydrolo- gists have impeded collaboration. The planned assignment of cross-trained Hydrometeorological Analysis and Support personnel to RFCs and of hydrometeorologists to many WFOs should promote this much needed interaction. Recommendation: Training programs in meteorological practices for Hydrometeorological Analysis and Support hydrologists and in 39
hydrology for meteorologists should be established to promote maxi- mum interaction between Weather Forecast Office and River Forecast Center operational personnel. Under the existing infrastructure of academic meteorology and hydrology pro- grams, the development of a hydrometeorology track seems difficult. Perhaps the NWS can request that a group of universities investigate this issue further and recommend ways to implement the academic training of hydrometeorolo- gists. The interaction in the 1990s of WFO meteorologists and RFC personnel during rapidly developing situations, such as flash flood events, is less clear. Meteorologists are responsible for issuing flash flood warnings; this is usually done without hydrological input. Yet the hydrologist has important knowledge of river basins and the effect that given rainfall intensities have on basin runoff. Modernization of the NWS should facilitate interactions between RFC and WFO personnel during these rapidly developing situations. The result will be improved weather and hydrological flash flood forecasts. Recommendation: Hydrometeorological Analysis and Support func- tions at River Forecast Centers and the interaction of Hydrometeor- ological Analysis and Support personnel with Weather Forecast Office meteorologists require clarification and better definition, especially as they relate to flash flood situations. New techniques are emerging that can improve flood forecasting in small basins. However, NWS professionals must have the knowledge and tools to take advantage of this capability. For example, in addition to developing expert systems to select proper algorithms for converting Next Generation Weather Radar information to rainfall amounts, the opportunity now exists to use first principles and actually calculate the rainfall intensity based on diver- gence measurements. Such calculations, using radar and other data, can validate the use of a particular algorithm. Current efforts to develop a hydrological computer work station are com- mendable and should be continued. This work station will be helpful to the WFO meteorologists as well as hydrologists. The software being developed by hydrologists will permit more efficient integration of hydrological observa- tions (e.g., river and stream gauge data) and the meteorological data needed to produce better and more timely flash flood forecasts. However, develop- ment of the hydrological work station does not now seem to take into account the planned relationship of Hydrometeorological Analysis and Support person-
nel and WFO meteorologists. Meteorologists as well as hydrologists should be involved in this development work. Recommendation: Consultation with meteorologists should be included in the current and future development of software to be used at hydrological computer work stations. This software should be installed in all of the Advanced Weather Interactive Processing Sys- tem work stations at the River Forecast Centers and Weather Forecast Offices so that it is accessible to all of the meteorologists, hydrometeorologists, and hydrologists. The improvements in both small-scale weather forecasts and hydrological basin models expected in the 1990s should result in improved anticipation of when and where flash floods will occur. This, in turn, should result in more time being available for interaction between shift meteorologist and hydrolo- gist and longer warning lead times for the public. Problem Areas The anticipated needs for hydrological services in the 1990s mandate a major increase in hydrological observations. At present there are about 3000 stream gauge sites in the United States. According to members of the NWS Office of Hydrology, this is an order of magnitude lower than necessary. Meteorologists and hydrologists can only speculate what is occurring in areas devoid of gauges. The NWS modernization may mitigate inadequacies in gauge data to some extent by use of Next Generation Weather Radar (NEXRAD) and satellite observations. Current plans call for River Forecast Center (RFC) personnel to combine radar observations from multiple NEXRADs, using Advanced Weather Interactive Processing System facilities, to infer precipitation rates and accumulations. However, there is concern that the same general algorithms for estimating rainfall may not be applicable at all NEXRAD locations. These plans also raise the question of whether RFC staffing patterns will accommodate this increase in workload. Recommendation: The validity of using the same general Next Generation Weather Radar algorithms for determination of rainfall estimates in all seasons, in all weather conditions, and at all Next Generation Weather Radar locations should be tested. 41
Because of the large number of Weather Forecast Offices (WFOs) in each RFC area of responsibility, coordination will be required to ensure consistency in the forecasts of precipitation and temperature that the RFC uses. Careful plans must be developed to ensure that improved numerical weather prediction guidance, observations, and interaction of WFO meteorolo- gists and RFC personnel will address this problem as reorganization of the NWS proceeds. The anticipated workloads of Hydrometeorological Analysis and Support personnel at the RFCs may be more than they can accommodate. The interaction and shared responsibilities with WFO meteorologists require clearer definition. Cross-training and full mutual appreciation of the functions and responsibilities of Hydrometeorological Analysis and Support and WFO meteorological personnel are needed to ensure optimum collaboration. Adequate training of meteorologists in hydrology and RFC personnel in meteorology is thus a major prerequisite to improved hydrological-related watches, warnings, and services in the 1990s. Recommendation: In light of the National Weather Service modern- ization and restructuring, the workloads, responsibilities, interactions, and cross-training of meteorological, hydrometeorological, and hydro- logical personnel planned for Weather Forecast Offices and River Forecast Centers should be examined carefully and redefined. 42
