奥里油海上溢油扩散的数值模拟及信息熵预测

利用CFD商业软件PHOENICS3.6.1对奥里油在恒定流情况下在海面和海下的稳态扩散进行数值模拟,通过对扩散的浓度场分布规律的分析,并与经验数据对比,表明奥里油在海下的扩散与其它类型油的扩散有所不同,在扩散的初期油的浓度与水深呈正比例关系衰减,在后期浓度则随水深呈指数型衰减.并应用信息熵理论对污染带的浓度分布进行预测,结果和数值模拟的结果吻合良好,表明用信息熵预测奥里油海上溢油扩散的浓度分布是可行的.     

基于改进Faster R-CNN模型的SAR图像溢油检测方法

SAR（synthetic aperture radar）图像溢油暗斑准确识别对海上溢油应急工作具有重要的意义。为减少SAR图像特征提取、特征选择过程中人为因素对溢油检测精度的影响,本文将Faster R-CNN卷积神经网络模型引入SAR图像溢油检测并进行了改进。针对溢油暗斑形状多样及SAR图像背景复杂的特点,选用结构一致且实用性强的VGG16卷积网络获取图像特征,并使用软化非极大值抑制算法（Soft-NMS）进行优化。同时基于相同的数据集,提取常用的SAR图像几何特征、灰度特征和纹理特征,构建反向传播（backpropagation,BP）人工神经网络溢油检测方法并与Faster R-CNN方法进行对比。实验结果表明,基于改进Faster-RCNN模型的溢油检测方法溢油检测率达到0.78,且溢油检测虚警率低于0.25,相比BP人工神经网络溢油检测方法样本识别率、溢油检测率分别提高了4%和5%,溢油虚警率降低了5%。     

基于合成孔径雷达回波信号的海洋溢油监测方法研究

海洋油污染是各类海洋污染中最常见、分布面积最广且危害程度最大的污染之一。近年来,海洋特别是近海人类活动频繁,且随着海上运输和石油加工业的发展,油田井喷、钻井平台爆炸、船舶碰撞等所造成的溢油事故增多,因而,监测海洋溢油具有重要的经济和社会现实意义。研究采用MatLAB工具,通过图像预处理(图像校正和增强)、特征提取和神经网络识别等方法,对合成孔径雷达(SAR)海洋溢油图像进行处理,最终期望实现半自动区分SAR图像上各类目标,并进行多种神经网络方法效果比较。研究首先对SAR海洋溢油图像进行初步人工识别;然后进行图像预处理(几何校正、滤波处理等)和基于灰度共生矩阵的特征值计算;最后,借助神经网络方法对溢油区域和疑似溢油区域进行分类,输出分类处理后的图像。通过输出图像分析发现,神经网络能对SAR海洋溢油图像中溢油、海水、土地3类目标进行明确分类,且RBF神经网络模型精度高于BP神经网络。本文提出的半自动分类方法不仅能提高SAR图像处理效率,将分类目标扩充有溢油和非溢油扩充到溢油、海水、土地3类,提高图像处理的全面性,同时通过比较RBF和BP神经网络在SAR溢油图像分类上的具体优劣,有着较好实际意义。     

江苏滨海北区风电场溢油风险数值模拟

利用MIKE21_FM HD模块搭建基于非结构网格形式的滨海北区二维潮流动力模型,并根据工程区实测水文数据对模型进行验证。结果显示,模拟结果与实测值吻合度良好,工程海域的水动力特性能够得到准确的反映,且可为溢油模块提供可靠的水动力基础数据。采用MIKE 21/3SA溢油分析模块搭建溢油风险预测模型,针对8种特定溢油情景下的油膜扫海面积、典型油粒子运动轨迹以及残油量变化等进行了预测分析。预测结果显示,不同情景下的油膜扫海范围和油粒子运动轨迹具有较明显的差异性;风场对油膜运动轨迹和扫海范围起决定性作用,潮流场次之;油膜扫海面积随溢油时间逐渐增大,在夏季主导风落潮时刻达到最大的626.89km2;盐城自然保护区北区最早受溢油影响的时间为第29.5h,南区最早受溢油影响的时间为第54h。     

长江口深水航道内溢油轨迹的数值模拟

耦合海洋和溢油模型,建立起1个适用于长江口深水航道内溢油轨迹预报模型。海洋模型考虑了深水航道中导堤丁坝的影响,能够较好地模拟深水航道内流场,使物理场更加可信;溢油模型采用前国际上常用的随机游走和拉格朗日油粒子追踪法,预测油粒子的漂移扩散轨迹和扫海面积。研究表明:在深水航道中段发生的溢油事故,油粒子的漂移分布和扫海面积受导堤丁坝和流场的共同影响,涨急时刻溢油24h后油粒子的分布和扫海主要分布在导堤丁坝附近,落急时刻溢油的油粒子则大部分分布于导堤丁坝外,扫海面积也比涨急时刻大,对九段沙自然保护敏感区域产生一定程度的潜在生态影响。本文用数值实验的方法验证了海洋模型中考虑导堤丁坝与不考虑导堤丁坝相比,溢油轨迹预测是有差别的,考虑了导堤丁坝会对油粒子在导堤丁坝附近的漂移和扩散起阻挡约束和聚集的作用,没有考虑导堤丁坝的溢油扫海面积增大。     

基于BP神经网络的长江口北支河槽容积分析

根据实测水文及泥沙等资料,采用现在较成熟的且应用广泛的BP人工神经网络建立了北支0m以下河槽容积与大通流量、大通输沙量及北支分流比3个因子问的神经网络模型,网络结构为3.1-7-1,通过选择合适的参数,模型训练较好,预测结果与线性回归模型预测结果相近,说明BP神经网络模型能够广泛应用于河口水文等方面的预报.     

海上溢油油膜运动不同数值模拟方法的比较研究

为能够迅速响应突发的海上溢油事故,本文利用油粒子法建立了溢油应急模型,以预测油膜漂移轨迹和形状。为弥补油粒子法无法模拟溢油初期油膜扩展过程和无法响应溢油量变化的缺陷,本文设计实验,使用将油膜扩展与油粒子法结合的两阶段法,发现能够较好地模拟溢油扩展过程;并用分数布朗运动代替传统布朗运动模拟油粒子扩散过程,模拟结果能够响应溢油量变化,更好地模拟油膜分散破碎的过程。模型中还考虑了蒸发过程使油粒子数量减少。     

基于元胞自动机的海洋溢油模型

本文采用三维元胞自动机模型,对海洋溢油过程进行了模拟,对风流、水流等水平方向的影响系数进行了改进。考虑了蒸发、垂直扩散、岸边附着、溶解、乳化等因子的影响,并且引入乳化含水率将乳化因子进行实际量化,构建了海洋溢油模型。并以"Deep Spill"的溢油实验为基础进行了模拟,实验结果 Kappa系数达到0.902,与实际相比具有较好的一致性。本文对海洋溢油事故进行预测模拟,为提出合理有效的应急预案提供科学依据。     

西北太平洋柔鱼中长期预测方法研究

为了能更好预测西北太平洋柔鱼的资源量, 选择合适的预测方法及开发相应的预测系统颇为重要。利用相关性分析, 筛选出在产卵区显著影响西北太平洋柔鱼资源量的关键网格点, 并采用这些网格点的海表温度、产卵区适宜温度所占面积的比例和单位努力捕获量等数据组织样本, 然后利用线性回归、BP 神经网络、RBF 神经网络和支持向量机等预测方法进行实验。结果表明: 在西北太平洋柔鱼中长期预测中, BP 神经网络要优于其他方法。以相关性分析和BP 神经网络为基础建立的西北太平洋柔鱼资源量预测系统是有效可行的。     

无主漂油源集合预测方法研究

采用基于"拉格朗日"粒子追踪方法的油粒子数值溯源模式,引入天气预报中的"集合预报"方法,考虑气象、海洋环境背景场误差因素,研究无主溢油源的集合预测方法。对某次渤海中西部溢油溯源事件的数值模拟结果表明,溢油源集合预测分析结果与溢油源实际调查情况相符,表明无主漂油源集合预测方法科学、结论可信。     

多环芳烃油指纹应用于船舶溢油鉴别研究

溢油种类主要包括船舶燃料油和原油,二者性质的差异决定了鉴别方法也相应不同,寻求适合于船舶溢油的鉴别方法具有重要意义。在使用柱色谱层析方法对样品进行分离前处理的基础上,以气相色谱/质谱方法(GC-MS)为主要分析手段,对溢油样品和可疑船舶溢油源样品的多环芳烃油指纹特征进行对比,并在多环芳烃油指纹参数的基础上进一步进行多环芳烃内组成三角图分布特征与聚类分析研究,成功为珠江口水域某船舶溢油事故追踪到肇事溢油源。结果表明:取自丁船的油样和现场溢油样芳烃油指纹特征最为相近,是此次溢油事故的溢油源。受风化作用后的船舶燃料油中饱和烃类化合物数量稀少,且含量极低,不适合用于溢油鉴别,而多环芳烃类化合物较饱和烃类化合物而言具有含量高、种类丰富的特点,是该类溢油鉴别的主要油指纹依据。使用油指纹参数进行可疑溢油源识别时,充分考虑油品中有机分子所受风化影响程度的不同是风化条件下溢油鉴定的关键。因此,多环芳烃油指纹可以有效应用于船舶燃料油溢油的鉴别。     

福宁湾海域溢油扩散影响的数值模拟研究

应用非结构化网格形式的MIKE 21水动力模型及溢油模块建立福宁湾海域海工工程施工期溢油扩散的数学模型,对在静风、全年主导风及不利风等3种不同风况条件下,一个潮周期内包括涨憩、落急、落憩、涨急等4个典型潮时分别发生100 t燃料油持续泄露30 min的事故溢油工况进行溢油扩散数值模拟.分析显示,事故溢油在福宁湾海域的扩散过程主要受潮流和风的影响.计算结果给出了事故溢油发生后24 h内的影响范围及0. 05 mg/dm3油浓度(一类或二类水质标准)溢油最快到达福宁湾内主要敏感目标的时间,为海洋环境影响评价和溢油事故应急措施的制定提供了科学依据.     

高光谱特征波段在海洋典型溢油油种识别中的精度评估

溢油种类精准识别对快速有效地治理污染具有重要的意义,高光谱遥感在海面溢油油种识别中至关重要。为探讨海洋典型溢油油种识别的高光谱特征波段范围,通过设计室外模拟溢油实验,在获取原油、燃料油、柴油、汽油和棕榈油等5种油种的实测高光谱数据基础上,运用因子分析和光谱标准偏差分析法遴选溢油油种的光谱特征波段,并利用支持向量机模型开展基于光谱特征波段的油种识别精度评价。结果表明,基于光谱标准偏差分析和因子分析获得的特征波段的油种识别精度分别是83.33%和90.74%,与基于全波段的识别精度相比,整体精度分别提高了3.7%和11.11%。选取的特征波段(360～540 nm, 560～600 nm, 610～630 nm, 640～660 nm)可作为5种油种相互区分的最佳高光谱波段。     

基于MIKE SA模型的厦门西港海域溢油影响的数值模拟研究

基于MIKE SA溢油模块,以燃料油为油种,建立了厦门西港海域溢油模型,模拟静风、主导风向（东北东风）和不利风向（西南风）3种风场条件下,一个潮周期内涨急、高潮、落急和低潮4个时段发生10 t溢油后油膜的漂移路径和影响范围.结果显示,发生在厦门西港海域的溢油在海面的漂移过程主要受潮流和风的影响,其中潮流起着主导作用.不同风向条件下,24 h内油膜的影响范围不同,静风条件下溢油浓度超一类（或二类,≥0.05 mg/dm3）、超三类（≥0.30 mg/dm3）和超四类（≥0.50 mg/dm3）的总影响面积分别为31.33、19.63和11.74 km2;主导风向条件下溢油浓度超一类（或二类）、超三类和超四类的总影响面积分别为99.62、69.01和8.99 km2;不利风向溢油浓度超一类（或二类）、超三类和超四类的总影响面积分别为8.38、5.05和2.10 km2.该预测结果可给出溢油事故发生后的影响范围、影响程度和影响敏感目标的时间,可为溢油事故应急决策的制定及溢油损害评估提供科学决策和支持,提升厦门海域环境风险管理应急能力建设.     

Accumulation and depuration of petroleum hydrocarbons by black mussels. 2. Depuration of field-exposed mussels

Results of depuration studies of mussels collected from chronically polluted sites and those collected after a fuel oil spill tend to confirm laboratory-deduced indications that the rate of depuration decreases as the exposure time increases. Chronically polluted mussels lost their accumulated hydrocarbons with half-lives for depuration that varied between 25 and 45 days whereas, after the oil spill, mussels lost 90 per cent of their accumulated burden within two weeks with a depuration half-life of 3,5 days. These results indicate that information concerning the type of exposure could be obtained from depuration studies of mussels collected during monitoring programmes.     

Histopathological alterations in Mya arenaria following a # 2 fuel oil spill in the arthur kill, Elizabeth, New Jersey

Following a 2270000-liter #2 fuel oil spill on 1 January 1990 studies were carried out to evaluate histopathological alterations in a population of Mya arenaria inhabiting an impacted tidal mud flat in the Arthur Kill, Elizabeth, New Jersey. Approximately 30 clams were collected 1 week after the spill and on a monthly basis thereafter. The lesion incidence found in M. arenaria following the spill was compared to a base-line incidence established at the same location during 1988 and 1989. There was an increased incidence of moderate and severe lesions of the digestive gland, gills, and mantle 1 week after the spill, returning to pre-spill levels by 1 month. During the first 3 months post-spill an increase in lesion incidence of the intestines was observed which returned to previous levels by 6 months post-spill. The heart and kidney showed a latent increase in lesion incidence 2 and 3 months following the spill, which by 4 months returned to previous levels. Examination of the gonadal development indicated that ova and sperm development were not adversely affected and spawning occurred both in late spring and late summer. The siphon, gill, stomach, foot muscle, and body epithelium were not found to be adversely affected.     

Effects of a tanker accident and an oil blowout in Bristol Bay,Alaska, on returning adult sockeye salmon (Oncorhynchus nerka)—A simulation study

The effects of a tanker accident releasing 34000 tons of diesel fuel and a blowout releasing 3000 t/day of crude oil on adult sockeye salmon returning through Bristol Bay, Alaska, were simulated. Parameters in the simulation were chosen to maximize possible effects of the oil. Mortalities from the tanker accident were predicted to range from 2% to 18% of the adults passing through the spill area or 1% to 5% of the total returning population. From 3% to 7% of the adults surviving migration through the spill area, or 1% to 2% of the total population, could be tainted at or above 0·6 ppm of hydrocarbons in the flesh. As many as 30% of the adults returning to fishing grounds closest to the spill area could be tainted. Effects of the blowout on returning salmon were less severe than those of the tanker accident, with mortalities reaching a maximum of 0·2% of the adults passing through the area of the blowout, and no tainting predicted above 0·6 ppm.     

Some dynamics of spilled oil in a fractured ice field in Buzzards Bay, Massachusetts

A spill of #2 fuel oil occurred 28 January 1977 in a dynamic ice field in BBuzzards Bay, Massachusetts. An oil budget study revealed that within four days after the spill occurred, approximately 60% of the oil was incorporated into the fractured ice zones and onto the ice surface. The movement of the oil (initially spilled under the ice) into the fracture zones appears to be the result of water current below the ice and buoyancy spreading mechanisms. Areas of fracture zones and leads provided a surface outlet for the less dense oil as it moved under the ice. When the leads closed, the oil was trapped into tools on top of the ice at the floe edges. The further spread of the oil onto the surface of the ice floes appears to have been caused by the wind.     

应用极化合成孔径雷达检测海上溢油研究进展

海上溢油给海洋生态环境带来严重的影响,快速准确地探测溢油对于防灾减灾具有重要的意义。利用卫星遥感探测溢油已成为目前主要的检测手段,大多采用合成孔径雷达(SAR)数据,运用图像处理的方法,开展了多种溢油提取算法的研究,取得了较好的结果,但由于海洋的类溢油现象存在,造成提取信息的精度达不到要求。近年来,国内外运用极化SAR数据开展溢油信息提取研究,从极化分解与相位差等角度对溢油特性分析,能有效地区分一些类溢油现象,得到了较理想的结果。分析了应用SAR数据开展溢油信息提取的研究状况,总结了溢油极化SAR探测的研究,指出了目前研究中存在的不足,并提出了今后溢油极化SAR遥感监测的方向。     

The North Cape oil spill: hydrocarbons in Rhode Island coastal waters and Point Judith Pond.

On 19 January 1996, the North Cape oil barge ran aground near Moonstone Beach, RI, and spilled over 2700 metric tons of No. 2 fuel oil during a severe winter storm. High winds and rough seas drove the oil into the water column, and the oil spread throughout Block Island Sound and into several coastal salt ponds. Over 50 water samples were collected from Point Judith Pond (PJP) and the southern coast of Rhode Island for four months after the spill and analyzed for polycyclic aromatic hydrocarbons (PAHs) and total petroleum hydrocarbons (TPHs). These analyses revealed that at least 60 km2 of coastal waters were impacted from the spill. Maximum concentrations of sigmaPAHs and TPHs were 115 and 3940 microg l(-1), respectively. The percentage of sigmaPAHs relative to the TPHs for all samples varied from 0.2 to 43%, showing that there was no clear relationship between sigmaPAHs and TPHs for the whole dataset and likely resulting from spatial and temporal partitioning over the course of the spill. However, within the dataset, there were stronger correlations for distinct samples collected at similar locations and times. In PJP, water column concentrations of individual PAHs decreased at rates of 0.08-0.24 day(-1) and lower-molecular weight PAHs were removed faster than higher-molecular weight PAHs.