Neural network analysis determination of oil slick distribution and thickness from satellite Synthetic Aperture Radar, April 24 - August 3, 2010

One hundred and sixty six Synthetic Aperture Radar (SAR) images collected from Radarsat-1, Radarsat-2, TerraSAR-X, CosmoSKY-MED 1-2-3-4,ENVISAT, ALSO01 and ERS02 satellites from April 23 to August 2 2010 were analyzed to quantify the distribution of floating oil discharged from Deepwater Horizon (DWH). Raw satellite data were rendered into 8 and 16 -bit geotiff formats when required. The Texture Classifier Neural Network Algorithm (TCNNA) and Oil Emulsion Detection Algorithm (OEDA) were applied to identify the extent of floating oil sheen (~ 1 um) and the areas with thicker patches of emulsion (~ 70 um), respectively. The DWH data are gridded (5x5 km cells) and include the percentage of water covered by thick and thin oil, and total percentage of water covered from April 24 to August 3 2010. Aerial dispersant (liters) and burning areas for cells is included. A subset of 60 SAR images were analyzed using the OEDA. The TCNNA was separately used to analyze 176 SAR images collected over the Gulf of Mexico prior to 2010. Data from prior to 2010 is presented as Google Earth map layers (KMZ) showing TCNNA outlines of oil slicks, oil slick origins, predicted seep zones and average area of floating oil (10x10 km grid). This data is published in: MacDonald et al. 2015. Natural and Unnatural Oil Slicks in the Gulf of Mexico, Journal of Geophysical Research - Oceans.