A probabilistic description of the wind over Liverpool Bay with application to oil spill simulations

被引:25
作者
Elliott, AJ [1 ]
机构
[1] Univ Wales, Ctr Appl Marine Sci, Marine Sci Labs, Menai Bridge LL59 5AB, Anglesey, Wales
关键词
synoptic winds; probability; transition; oil spill; contingency; Liverpool Bay; Irish Sea;
D O I
10.1016/j.ecss.2004.06.022
中图分类号
Q17 [水生生物学];
学科分类号
071004 ;
摘要
Surface winds from the UK Meteorological Office mesoscale (12 km grid) atmospheric model have been used to define the wind at a location in Liverpool Bay during 1997-2001. Winds from the SW (centred on 240degrees) with a speed of about 10 m/s (20 knots) were the most frequent, although weaker winds from the SE were also common. The wind spectra were red in character and showed no evidence for a peak at the synoptic (2-5 day) time scale; however, a zero-up-crossing analysis suggested a dominant periodicity at 3.1 days, and at this time scale the winds were spatially coherent over a distance of 300 km. A wind direction transition matrix was derived to quantify the probability with which the wind changed between two specified directions. This information was then used with an estimate of the mean duration of a wind event to compute a stochastic wind time series that contained a similar energy level, periodicity, and direction variability to the archived wind data. The archived and stochastic winds were then used in 1000 oil spill contingency simulations during which estimates of the mean and minimum times taken for oil to reach the coastline, and the percentage of the oil impacting selected sites were computed. The stochastic winds provided more realistic results, when compared against those derived using the wind archive, than those obtained using a wind rose representation of the winds. The derivation and use of a stochastic wind time series has application to a range of modelling studies. (C) 2004 Elsevier Ltd. All rights reserved.
引用
收藏
页码:569 / 581
页数:13
相关论文
共 17 条