Vig-itant, this results in a factor of 3 overestlmatlon. Using this as a 

 calibration factor, the wave estimates made from GEOS may possibly be of use 

 as inputs for future modeling efforts. 



As part of the National Oil Pollution Contingency Plan Response, the 

 National Weather Service (NWS), NOAA, supplied the Federal On-Scene Coordin- 

 ator with special marine forecasts for the site of the wreck of the Avgo Mev- 

 ohant since the day of the grounding. Routine scheduling of the special 

 forecasts began late on December 15. The schedule has varied from as many as 

 six per day to as few as two per day, dependent upon need. Since the initial 

 request for forecast assistance, 252 forecasts have been provided up to 

 February 1, including 147 special forecasts and 103 special wind forecasts. 



Some operational problems were encountered by NWS, the major problem 

 being lack of feedback of weather conditions at the site. This was resolved 

 recently with the assistance of the Federal On-Scene Coordinator. 



Table 2-4 was prepared by NWS to give some insight into the accuracy of 

 the wind forecasts issued by NWS for the oil spill site. On-site winds were 

 not relayed to NWS in Boston on a real-time basis until mid-January. Thus, 

 operational forecasts were issued with the Nantucket Shoals Light Vessel 

 (L/V) as closest observational site, located 32 miles from the spill site on 

 Fishing Rip. Table 2-4A shows the wind verification based on Nantucket L/V 

 data. Table 2-4B shows the wind verification for the Fishing Rip site. The 

 tables are basically self-explanatory (MAE = mean average error). However, 

 it should be pointed out that the selection of categories was subjective and 

 based upon NWS operational experience, by which a forecast of wind direction 

 that verifies within 30° is considered excellent. Forty percent of the fore- 

 casts (Table 2-4B) fell within this range; 72% of the forecasts fell within 

 the range of 60°, considered a good-to-excellent forecast; and only 10% were 

 in the 90° category, which is considered poor and of no ooperational use. 



A wind speed in error by less than 5 knots is considered excellent (36% 

 were in this category), less than 10 knots good to excellent (72% in this 

 category) , and greater than 15 knots unsatisfactory (12%) . Even the last 

 category can be of limited use depending on the strength of the wind. 



While the table provides a statistical appraisal of the forecasts, it 

 does not necessarily directly reflect their utility. That utility should be 

 judged where possible on the overall effectiveness in assisting the DCS in 

 making proper operational judgments. 



One of the most serious problems arising from the spill was, and is, the 

 potential contamination of the beaches and shorelines along the New England 

 coast and possibly even the mid-Atlantic coast. Therefore, it was particu- 

 larly important to accurately forecast offshore winds (260-360°) and onshore 

 winds (90-160°). These were considered the two most important categories. 



In evaluating the utility of the forecasts of the other wind directions, 

 forecasts were definitely useful 84% of the time up to 12 hours and 81% of 

 the time up to 30 hours. 



36 



