【作者】 刘丙新；

【导师】 李颖；

【作者基本信息】 大连海事大学 ， 环境科学与工程， 2010， 硕士

【摘要】 海上大型溢油事故具有污染面积大、持续时间长、清理难度大的特点,这就导致对环境造成的破坏更加严重。如果能够快速地发现溢油区域、确定溢油面积并对溢油漂移趋势进行预测和跟踪,将会对溢油的清理工作带来很大帮助。目前常用遥感监测方式对溢油事件进行监测,其中光学遥感因其良好的视觉效果而被广泛应用。对于溢油清理来说,能否以最快的速度获得油污信息将直接影响清理工作的速度与效果,因此作为油污信息获取常用手段的遥感监测,如何在溢油发生后确定适合的遥感数据,并快速获取所需的遥感数据,成为一个关键性问题。欧空局ENVISAT卫星的MERIS数据、美国NOAA卫星的AVHRR数据和NASA的Terra/Aqua卫星的MODIS数据是目前溢油监测中应用较多的光学传感器数据。在以往的溢油事件监测中,这三种数据都曾成功地对油污进行识别和监测,但由于数据特性的差异,每种数据又会有不同的适用环境。本文在研究利用三种数据对两起溢油事件进行监测的基础上,对三种数据在溢油监测方面的优劣势进行比较,并提出每种数据的最佳适用情况。另外,‘本文还综合利用AVHRR数据和MODIS数据对发生在墨西哥湾的原油泄漏事故进行了动态分析。通过比较分析发现,AVHRR数据更适合于海上大型非持久性油的溢油事件的动态分析,可对油污在较短时间内的动态情况进行监测,并且对大面积较厚油污的监测效果良好;MODIS数据因其较短的重访周期在海上持久性油污染事件的动态监测方面具有一定优势；MERIS数据在光照条件良好的情况下,能够发现较小斑块的油污,能够反映更多的油污染细节信息,并且其高光谱数据更有利于区别油污与假目标。另通过对溢油覆盖范围的动态分析,验证了利用遥感数据综合监测溢油方案的可行性。更多还原

【Abstract】 Large oil spills has the characteristics of large polluted area, long duration, difficult to clean up, which lead to more serious destroy to the environment. If the waters covered by oil spill could be found and the area of oil and the drift trends of oil spill could be known quickly, it will be a great help for us to clean-up work. At present, remote sensing monitoring methods is used to monitor the oil spill events, especially optical remote sensing, because of its good visual effects, is widely used. For oil spill clean-up, weather we can get the oil information as quickly as possible will affect the speed and efficiency of clean-up work directly. Therefore, how to determine and quickly obtain the needed remote sensing data when the oil spill happened become a key problem.At present, MERIS data of ESA’s ENVISAT satellite, AVHRR data of U.S.’s NOAA satellite and MODIS data of NASA’s Terra/Aqua satellite are widely applicated optical sensor data in oil spill monitoring. In the past, this three types data have been successfully used to identify and monitor the oil pollution. But because of differences in data characteristics, each data has different suitable environment for its application. Based on making use of this three type data to monitor two oil spill events, this paper compared the advantages and disadvantages of the three data in the aspect of oil spill monitoring and propose the best application of each data. In addition, the paper also makes use of Multi-temporal MODIS and AVHRR data to dynamically analyze the oil spill which happened in the Gulf of Mexico.By comparing and analyzing, it was found that AVHRR data is more suitable to dynamically analyze the oil spill with the characteristics of large and non-persistent oil at sea, especially the large thick oil. MODIS data have some advantages in dynamic monitoring for the persistent oil pollution which lasted a long time at sea. In the condition of good sun light, MERIS data can reflect more detailed information about oil pollution. At the same time, the hyperspectral data can help distinguishing the oil pollution and false target better. By the oil spill’s dynamic analysis, we found that the composed use of different remote sensors’data to monitor oil spills is feasible, what’s more, the drift trends of oil spill at sea depends largely on the local environment of surface currents.更多还原