also more realistic, presumably because the mean currents generally include a 

 surface drift resultant from the mean winds which was not separately removed 

 before the wind-induced currents were added. The sole contribution for 

 subsurface drift (Section 2.3.6) presents results which have not been veri- 

 fied to date. 



The following six sections contain the contributed modeling efforts. It 

 should be noted that the majority of the figures contributed are contained 

 and referenced in Appendix VII. 



2.3.1 U.S. Coast Guard Oceanographic Unit 



The primary technique used by the USCG Oceanographic Unit in forecasting 

 oil drift from the Argo Mevchant was to sum the vector effects of wind cur- 

 rents, tidal currents, sea currents (mean or residual currents), and oil 

 leeway. Based on 24-hour forecast winds for the spill site, provided by the 

 National Weather Service in Boston, the forecast was made at about local noon 

 each day for a verification time of 1300 the following day. The summation 

 was done by a program from the following inputs: 



o Initial oil positions at 1300 the previous day. 



o Observed winds up to the present. 



o Forecast winds up to 1300 the following day. 



o Tidal currents. 



o Wind currents. 



Program output was oil positions at 12-hour intervals from 1300 the previous 

 day to 1300 the following day. A new blob of oil at the wreck site was gen- 

 erated every 12 hours. The positions obtained were reported by telephone 

 each day about noon to Cmdr. Morgan, who would prepare forecast limits of all 

 oil from these positions. Often a correction would be applied to the posi- 

 tions based on the difference between the forecast and the oil observed the 

 preceding day. This correction represented the sea current, which was ini- 

 tially part of the computer program, but was later dropped. 



The wind current used was a time-dependent Ekman model based on equation 

 (41) in a paper by Jelesnianski (1970) . The tidal current was derived from 

 tidal vectors near the wreck site as given by Haight (1942). The oil leeway 

 was taken as 1.2% of wind speed, directly downwind, based on measurements 

 taken on the scene by the NOAA-USCG SOR Team. 



For the first few days of the Avgo Merchant oil spill, and after January 

 8, manual techniques were used for forecasting, which is continuing at the 

 present time. Plans are underway to perform a more thorough analysis in the 

 near future of the computer forecast technique based on observed winds. 



With few exceptions, all oil sightings fell within the forecast oil 

 limits. Examples of these forecasts are shown as box outlines in Figures VII- 



42 



