8—
Conclusions

WEATHER FORECAST AND STORM WARNING PERFORMANCE

The National Weather Service uses five data sources to prepare local weather forecasts and storm warnings for Oahu:

  • Numerical guidance and analyses from the National Meteorological Center and the Fleet Numerical Oceanographic Center.

  • Locally produced analyses.

  • Remote weather stations, including telemetered raingauges.

  • Radar from the Federal Aviation Administration (FAA) and the Air Weather Service (Hickam Air Force Base).

  • Satellite imagery.

With respect to these five data sources, used to forecast Oahu's New Year's flood, the following comments can be made:

  • The numerical guidance products adequately described the evolution of synoptic events.

  • The local analyses captured the surface conditions.

  • The remote weather stations performed reasonably well. Telemetered stations are interrogated every 6 hours unless conditions warrant higher-frequency sampling. The two telemetered raingauges are automatically updated every 15 minutes. However, neither of the two raingauges was located near the center of the storm. Several raingauges eventually overflowed or malfunctioned, although the breakdowns occurred after flood warnings had been issued.

  • The FAA's radar display blocks out a 50-mile radius. Cells approaching the island are depicted, but cells developing over Oahu are undetected. The radar operators at Hickam did not report significant cells over the island. This is not surprising since 12,000-foot cells would not be cause for concern in many parts of the world. (Remember, the forecast was for thunderstorms.)



The National Academies | 500 Fifth St. N.W. | Washington, D.C. 20001
Copyright © National Academy of Sciences. All rights reserved.
Terms of Use and Privacy Statement



Below are the first 10 and last 10 pages of uncorrected machine-read text (when available) of this chapter, followed by the top 30 algorithmically extracted key phrases from the chapter as a whole.
Intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text on the opening pages of each chapter. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

Do not use for reproduction, copying, pasting, or reading; exclusively for search engines.

OCR for page 62
The New Year's Eve Flood on Oahu, Hawaii December 31, 1987–January 1, 1988 8— Conclusions WEATHER FORECAST AND STORM WARNING PERFORMANCE The National Weather Service uses five data sources to prepare local weather forecasts and storm warnings for Oahu: Numerical guidance and analyses from the National Meteorological Center and the Fleet Numerical Oceanographic Center. Locally produced analyses. Remote weather stations, including telemetered raingauges. Radar from the Federal Aviation Administration (FAA) and the Air Weather Service (Hickam Air Force Base). Satellite imagery. With respect to these five data sources, used to forecast Oahu's New Year's flood, the following comments can be made: The numerical guidance products adequately described the evolution of synoptic events. The local analyses captured the surface conditions. The remote weather stations performed reasonably well. Telemetered stations are interrogated every 6 hours unless conditions warrant higher-frequency sampling. The two telemetered raingauges are automatically updated every 15 minutes. However, neither of the two raingauges was located near the center of the storm. Several raingauges eventually overflowed or malfunctioned, although the breakdowns occurred after flood warnings had been issued. The FAA's radar display blocks out a 50-mile radius. Cells approaching the island are depicted, but cells developing over Oahu are undetected. The radar operators at Hickam did not report significant cells over the island. This is not surprising since 12,000-foot cells would not be cause for concern in many parts of the world. (Remember, the forecast was for thunderstorms.)

OCR for page 62
The New Year's Eve Flood on Oahu, Hawaii December 31, 1987–January 1, 1988 The small vertical extent of the precipitating clouds limited the effectiveness of satellite imagery in depicting local weather. The rain clouds were masked by overlying cirrus clouds. This is often the case with orographic rains in Hawaii. However, the satellite imagery did accurately depict the large-scale weather pattern. Overall, the forecast performance was what could be expected given the available technology. The synoptic-scale forecast was correct. Heavy showers were forecasted, and the Oahu Civil Defense Agency had been briefed. There was no apparent justification for changing the telemetered raingauges to a higher frequency of sampling. The unavailability of adequate radar information, coupled with high clouds masking the actual rain clouds and with no useful public reports until 7:00 p.m., December 31, left forecasters with no evidence to justify more frequent raingauge telemetry. CAUSE OF THE FLOOD EVENT The Oahu New Year's Eve flood event and the resulting damage were caused by a combination of four factors. The first was the extreme antecedent moisture conditions caused by the heavy rains that occurred during December 1987. The December rainfall was almost five times the normal monthly amount. The second factor was the extreme amount of rain that fell on New Year's Eve. Recurrence interval studies of rainfall frequency indicated that the rainfall was in excess of that expected for a 100-year event and probably as much as would occur in a 200-year event. These first two factors combined to generate large amounts of sediment and debris that quickly filled existing debris basins, blocked drainage channels, and diverted streams from their natural and man-made channels and was the major cause of damage to residences and infrastructure. This third factor, the sediment and debris, was the real culprit in the flood. The fourth factor was the failure of existing flood control facilities and structures. The Kawainui Marsh was originally designed as a flood control reservoir (U.S. Army Corps of Engineers, 1956). However, due to the deposition of sediment and debris since its initial construction and the lack of any systematic dredging, the reservoir's capacity had been significantly reduced. Furthermore, the top of the levee had settled approximately 1 foot (U.S. Army Corps of Engineers, 1988f). Therefore, the flood wave generated by the New Year's Eve storm probably traveled through the reservoir as overland sheetflow and overtopped the downstream levee. Furthermore, the drainage of the Kawainui Marsh via the Oneawa Canal is influenced by the backwater effects of tidal action from the ocean. However, the U.S. Army Corps of Engineers discharge rating curve for the Oneawa Canal ignores backwater tidal effects (see U.S. Army Corps of Engineers, 1956, Plate A-9). Finally, the debris basins, the concrete-lined channels, and the roadway crossing were all designed using clear water flow data. No design criteria for sediment and debris flows have been established by the city and county of Honolulu.

OCR for page 62
The New Year's Eve Flood on Oahu, Hawaii December 31, 1987–January 1, 1988 EMERGENCY RESPONSE AND INITIAL RECOVERY A unique set of meteorological and geographical circumstances created two different types of flood incidents to which emergency managers needed to respond. One was the rapid flash flood that occurred in the Hawaii Kai area of Oahu, and the other was the relatively slow flood that occurred in the Coconut Grove area of Kailua. In general, the following conclusions about the adequacy of the response to the flash flood episode can be drawn: The inability of the National Weather Service to issue a flood watch announcement prior to the onset of flooding resulted in a lack of predisaster mobilization efforts by emergency response agencies that delayed their ability to respond quickly to flood problems, especially in the Hawaii Kai area. The holiday created exceptional problems for the mobilization, coordination, and provision of emergency response personnel and services during the early flash flood period. As a result of these first two conditions, requests for assistance to stem flood problems on leeward Oahu during the early disaster period were not satisfactorily met. Similar problems occurred in the Coconut Grove area of Kailua, where floodwaters overtopped the levee of the Kawainui Marsh. The potential levee failure had been ignored as a possible flood threat. The following conclusions regarding the levee failure flood can be drawn: No monitoring of this particularly vulnerable area was conducted on New Year's Eve primarily because of expectations by officials and residents that the levee provided adequate flood protection to Coconut Grove residents. In essence, the levee provided a false sense of security. Because of a lack of attention to this area, no flood warning was issued to residents there that would have allowed them to take protective actions or to evacuate the area. Despite the lack of warnings and the difficulties associated with mobilization efforts on a holiday, no lives were lost due to the flooding. Emergency management personnel were able to provide adequate rescue assistance during the night, but their efforts were always in response to an event in progress. Early relief and recovery efforts were, by and large, successful, primarily because of predisaster planning. Adequate relief and sheltering services were available, beginning on New Year's Eve, due to the preidentification of possible shelters and personnel. The early recovery period was well coordinated, particularly with respect to efforts to clear roadways and drainage canals, again owing to predisaster planning arrangements.

OCR for page 62
The New Year's Eve Flood on Oahu, Hawaii December 31, 1987–January 1, 1988 In general, when emergency management agencies were able to rely on the implementation of predisaster plans, their efforts were largely successful. However, the lack of an early identification that a hazard actually existed resulted in delays in resource mobilization, coordination difficulties, and exposure of residents to potential harm. FLOODPLAIN MANAGEMENT Three general conclusions can be drawn from a review of the extent of protection offered by the primary nonstructural flood-mitigation effort, that is, the floodplain management program. Although a full assessment of the management program was not undertaken, the appropriateness of the flood vulnerability zones was reviewed. Approximately one-third of all damages [those in Flood Insurance Rate Map (FIRM) Zones A and AH] that occurred should have been anticipated; that is, for the characteristics of this storm, some damage and loss could have been expected. Given the characteristics of the storm, over half of the damage (that in Zones B and C) occurred in areas where such damage would not have been anticipated. Fifteen percent of the damage occurred in areas that had not yet been evaluated for flood hazard potential, even though the FIRMs had been updated as recently as 3 months prior to the disaster. These conclusions clearly indicate the need to extend flood insurance rate mapping to currently unmapped areas as well as the need to review and update the present maps. However, the difficulty of improving FIRMs should not be minimized. Until longer time series of streamflow and rainfall data become available, a fair degree of uncertainty is likely to persist in the maps.