USE OF VERTICAL DIGITAL PHOTOGRAPHY AT THE BAYOU PEROT, LA SPILL FOR OIL MAPPING AND VOLUME ESTIMATION
| Autor: | Jon Sellars, Micheal L. Aslaksen, Jacqueline Michel, Charlie Henry, Chris Locke, Mark White |
|---|---|
| Rok vydání: | 2008 |
| Předmět: | |
| Zdroj: | International Oil Spill Conference Proceedings. 2008:127-130 |
| ISSN: | 2169-3358 2169-3366 |
| Popis: | The January 2007 release of over 8,000 barrels of a condensate crude oil from a damaged well in Bayou Perot, Louisiana resulted in intermittent oiling of remote mud flats and salt marshes over a 30 square mile area. NOAA'S National Geodetic Survey collected aerial vertical digital photography 17 days after the spill to assist in locating and quantifying areas of oiling. The effective pixel size was 34 centimeters, however, the data were processed to 40 centimeter resolution. Useful products were posted to the web within two days after acquisition. Standard supervised and unsupervised image processing techniques were used in conjunction with oblique photography and field knowledge to define the oiling signatures. Time constraints required that the classification be conducted on mosaiced, non-color balanced images (ideally each image would be classified independently to account for differences in illumination and/or processing). However, the strong visible signature of the oiled areas and ground-truth data from field surveys resulted in high confidence levels for several oil types which in turn were used to enhance the identification of the remaining classes. Five oil types were identified: Black (218,000 ft2), Red (81,000 ft2), Orange (154,000 ft2), Yellow (38,000 ft2), and Light Yellow (349,000 ft2) corresponding to the color and attributes of the oil. The total conservative estimate of oiled area was 840,000 ft2 or nearly 20 acres. Based on estimated thicknesses of the different oils, the total volume of oil present at the time of imagery acquisition was 3,330 barrels. This value was close to the actual amount of oil recovered over the time period between the date of imagery acquisition and the end of cleanup. |
| Databáze: | OpenAIRE |
| Externí odkaz: |
|