Study On Large Marine Oil Spills By Remote Sensing

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.