spill of 2010 in the Gulf of Mexico. With Landsat imagery, satellites 5 and 7, the spread and appearance of oil slicks will be identified and manipulated for best display. The usefulness of Landsat imagery in relation to oil spills will also be examined with other alternatives mentioned in the discussion. Gulf of Mexico Oil Spill 2010 using landsat imagery Allie Potter Remote Sensing GSC312, Dr. Haluk Cetin Introduction Oil spills are calamitous for extensive reasons: destruction of environment, waste of natural resources, loss of life, destroying the ways people make their living are but a few of the main reasons. This is why cleanup is essential and the faster the better. Using satellite imagery to measure the scope and the spread of the oil is exceedingly valuable for recovery and cleanup as well as prosecution of the organizations responsible. In order for identification of the spill to take place, knowledge of the anatomy of a spill must exist. Oil and water are not known to mix. In an ocean, where the spill originates below the surface of the water, oil may deviate from the norm of floating to and remaining at the surface. There are three layers where oil may settle: the surface, midwater, and the ocean floor. On the surface, oil floats along in what is known as an oil slick. Oil slicks are pushed around and spread by currents and wind. Beneath this lies midwater. The oil may mix with a different density of ocean water before floating to the surface and remain in an underwater cloud, an oil plume. An oil plume may contain a lot of water with clumps of oil mixed in. This is due to the dispersants sprayed on the underwater spill so different plumes may have different densities, thus settling higher or lower in the water rather than rising up to float on top. In the BP oil spill, a total amount of 1.8 billion gallons of dispersants where sprayed on the spill. The third option is oil sinking to the seafloor due to adherence to particles in the water (The Ocean Portal Team). Research is still being done to fully understand the dynamics of oil spills and the roles that dispersants, ocean currents, and the resulting impacts on regional life forms have. Here the focus will remain on oil slicks, oil slicks are much easier to observe with satellite imagery than the other two, ocean floor oil and oil plumes. It is observed that oil slicks rest on top of the ocean creating a very distinctive shape that may appear shiny or shimmery. Oil slicks can be seen from satellite fairly easily, barring cloud cover. There are still uncertainties about how best to locate and measure oil plumes, various techniques that are being researched include different meters and sensors, laser diffraction, and the use of sonar (Cooperative Institute for Ocean Exploration, Research and Technology). But with oil slicks, the distinctive shape that oil has combined with the slicks movement determined by wind and water result in a very visual type of deduction as seen below. Methods Using the USGS Global Visualization Viewer, the path and row of the location of the spills origins where located and inputted for download. The images and bands were downloaded in a TZ format. Double unzipping the file produced all of the products necessary to pyramid the images and open them with Erdas Imagine 2011. In order to ascertain the visual acuity of the specified event, I modified the downloaded data to better enhance the accident and disseminate the desired information. In my effort to disseminate this information more readily, I adjusted the visual data to an enlarged format using the method of subsetting. Subsetting is selecting a smaller portion of the image from the whole image so as to crop and enlarge for smaller file size and increased speed and efficiency of data analysis. I selected the image with the largest region of spill using an inquire box to maintain the exact parameters for replication to subset each of the data sets. Now each of the images of the spill, at their specified points in time, are enlarged and visually offered in an easily discernable fashion. Following this I modified the bands displayed to create an artificial color scheme to better represent the spill. There were three different band variations that I found to be of varying usefulness. The first, TM False Color Infrared produced an image where land was a reddish hue and the oil was a hazy/cloudy hue. The second proved the best, TM False Natural Color where the land turned a vivid green and the oil became a distinguishable purple. The Third was fairly distinctive, it was TM Thermal Infrared Composite. Many other variations were useful in creating a very easily identifiable difference so that the oil could be easily discerned from its surroundings. Results In hindsight, I realized that Landsat data, while extremely inexpensive (AKA free) and easily obtainable, isnt the best data to use in this scenario. Upon much research and analysis on different types of data and different ways to approach remote sensing of the Gulf of Mexico BP oil spill the conclusion was reached that Landsat is good for band layer and contrast analysis, and perhaps oil spill detection but not as good for detailed spread analysis. TerrSAR-X satellite imagery, however, is useful and is commonly used for detecting oil spills all over the world and has a much higher resolution. This satellite shoots pulses at the Earths oceans and the echoes show the presence of anomalies (Christensen) that with experience and deduction can be differentiated form one another. This German satellite, launched in 2007, has a revisit period of 11 days. It is a high resolution satellite that has many feasible practical applications, oil spill observations and change detection among them (Scheuchl, Koudogbo and Petrat). A problem I consistently ran into was the weather. Meteorological impacts on visual quality continually and consistently interfered with my ability to obtain quality images for oil spill analysis. This problem was shared by many at the time of the spill when time was limited and oil was continuing to spread, regardless of whether good pictures were taken of it or not. Due to this, I was able only to obtain consistent images of the original site of the spill, but not the space subsequent to it. This proved problematic as the oil spill in the images stayed true to their pattern until the image captured on July 12, 2012. This image showed the oil spill had drastically decreased to less than a quarter of its previous size, a result that while not impossible, seems exceedingly improbable. My peers and I hypothesized that rather than dissolving in so short a period of time, about a month, it must have moved out of the image, though no trail was apparent. Not a single image for the space of a year was available in usable form from Landsat 5 and 7 from those that I searched. Additionally, I was unable to obtain regular images. My images included May 1 st , May 9 th , May 17 th , and July 12 th . Upon further research I discovered other satellites had better images available but no more consistent than the ones I possessed, the weather being a vexing problem for all satellites. Discussion In the future there are great leaps to be made to expand this study to a more useful point. Foremost, what is the point of studying an oil spill if you can only see the uppermost piece? Research has shown that for underwater oil spills, oil slicks are just the tip of the figurative ice burg. More data and more information needs to be had on the specifics of oil plumes, their densities and where they settle in the water, how frequently do oil plumes rise and/or sink, how long it takes for dissipation due to spread and chemical breakdown, in essence, what are the dynamics of oil plumes? Of vital importance is the rate that oxygen is absorbed to break down these chemicals and the effects this has on life in those areas. Can whole sections of the ocean recover from being sucked dry of oxygen? One interesting thought for a future experiment: can tropical storms and hurricanes hasten the breakdown of oil spill pollutants or would it further the negative impact on coastal regions in the area or specifically would it further the harm to those vital coastal wetlands? Another aspect to consider would be how much of the oil from the spill sunk below to decompose at the bottom of the ocean? How does it affect the life there and how is its rate of dispersal compared to those above? What means would there be, if existing in an oil plume or oil slick were preferable for cleanup purposes, for the underwater oil to rise above and be rid of? Addressing all parts of the spill are equally important and all need to be addressed fully and until the problem is solved and future events are prevented. This study of the movements and presence of oil spills in the gulf has value in that those slicks can be fairly easily tracked and taken care of. Without this knowledge hundreds of miles of ocean would have to be traversed, analyzed, mapped and cleaned. Using satellite to determine the overall picture is preferable and convenient. It saves time, lives, landscape, and money in the fixing of an event that should have been better maintained in the first place.
References Christensen, Arnfinn. Detecting oil spills with satellite. Prod. ScienceNordic. Copenhagen, 24 February 2012. <http://sciencenordic.com/detecting-oil-spills-satellite>. Cooperative Institute for Ocean Exploration, Research and Technology. Detecting Oil/Gas Plumes. n.d. <http://cioert.org/flosee/detecting-oilgas-plumes/>. Scheuchl, Bernd , et al. "TerraSAR-X: Applications for Spaceborne High Resolution SAR Data." Anais XIV Simposio Brasileiro de Sensoriamento Remoto (2009): 7457-7464. <http://marte.dpi.inpe.br/col/dpi.inpe.br/sbsr@80/2008/11.18.05.02/doc/7457-7464.pdf>. The Ocean Portal Team. Smithsonian, National Museum of Natural History. 2010. April 2013. <http://ocean.si.edu/gulf-oil-spill>.