Introduction

On April 20, 2010, an oil drilling platform slightly larger than an American football field called Deepwater Horizon exploded roughly 61 kilometers (41 miles) off the Louisiana coast, resulting in human casualties and the dispersion of over 3.19 million barrels of crude oil into the Gulf of America. To this date, it is considered one the world’s largest oil spills with disastrous effects that could last decades. However, the environmental tragedy spawned new innovations and research in the use of Advanced Spaceborne Thermal Emission and Reflection Radiometer (ASTER) data to study oil spill thickness.

Science Objectives

During the Deepwater Horizon tragedy, federal agencies dispatched over 300 near-real-time remote sensing derived data to emergency personnel for assessing, tasking, and planning response. One of the first images to capture the spatial extent of the oil spill was collected by Terra Moderate Resolution Imaging Spectroradiometer (MODIS) at 1630 CDT on April 22, 2010. Other optical sensors involved in emergency response included ASTER, which is also aboard the Terra platform. Optical sensors such as Terra MODIS and Terra ASTER provide images with multiple spatial coverage over an area during an event. Terra and Aqua MODIS capture the entire Earth every one to two days and have a swath width that ranges from 250 meters (m) to 5600 m. Due to its repeat coverage and large footprint, MODIS can provide a spatial synoptic overview of an oil spill’s extent to responders in the field, while Terra ASTER, a higher resolution (15m) sensor, can be tasked to capture the disaster region by qualifying emergency personnel and can be available for download within 48 hours of acquisition.

Major Findings

A study by Garcia-Pineda et al. (2020) relied on Terra ASTER expedited reconstructed unprocessed data (AST_L1AE) to validate classification of oil thickness to support observations of oil emulsions, or droplets of oil in water, by two satellite synthetic aperture radars (SAR) around MC20, a research drilling platform not too far from the Deepwater Horizon explosion. The study’s aim is to deliver classification of oil thickness to provide a quick turnaround of near-real-time data to emergency responding vessels. The classification of oil thickness is challenging since crude oil undergoes a physical process in saltwater that transforms its structure; the difficulty is compounded by sun glint as well as surface wind conditions. Therefore, the oil thickness is heterogeneous at the pixel level, with each pixel containing different thicknesses. 

The authors used a simple machine learning algorithm called maximum likelihood classification (MLC), a classification technique that is very popular in the remote sensing community and especially in the land cover community. However, the application of MLC to classify oil thickness with Terra ASTER data is a relatively new approach. It requires robust training samples to present a probability measure of each pixel belonging to a particular class. By applying MLC, the authors were able to extract two classes of thickness. The classification of oil spill thickness is important because there is “actionable” and “non-actionable” oil. Actionable oil is thick, which could be cleaned up effectively, while non-actionable oil is more difficult to manage and contain. This timely information is important to the field responders in order to take appropriate actions.

The Deepwater Horizon was, and still is, a tragedy with ecological impacts that will take decades to recover from. However, due to this catastrophic event, new and innovative approaches to observe and monitor oil spills are being developed. More than twenty years ago, the design concept of Terra ASTER was innovative because of its high-resolution multispectral capabilities along with two telescopes that provide nadir and backward viewing angles. Today, with advances in remote sensing coupled with scientific computing, satellite sensors like Terra ASTER can be used to help in cases of disaster response and mitigation of both natural and man-made disasters.

References

Brekke, C., and Solberg, A.H.S., 2005, Oil spill detection by satellite remote sensing: Remote Sensing of Environment, v. 95, no. 1, p. 1–13. doi:10.1016/j.rse.2004.11.015.

Garcia-Pineda, O., Staples, G., Jones, C.E., Hu, C., Holt, B., Kourafalou, V., Graettinger, G., DiPinto, L., Ramirez, E., Streett, D., Cho, J., Swayze, G.A., Sun, S., Garcia, D., and Haces-Garcia, F., 2020, Classification of oil spill by thicknesses using multiple remote sensors: Remote Sensing of Environment, v. 236, art. no. 111421.  doi:10.1016/j.rse.2019.111421.

McNutt, M.K., Camilli, R., Crone, T.J., Guthrie, G.D., Hsieh, P.A., Ryerson, T.B., Savas, O., and Shaffer, F., 2012, Review of flow rate estimates of the Deepwater Horizon oil spill: Proceedings of the National Academy of Sciences of the United States of America, v. 109, no. 50, p. 20260–20267. doi:10.1073/pnas.1112139108.

Streett, D.D., 2013, NOAA'S satellite monitoring of marine oil, in Liu, Y., Macfadyen, A., Ji, Z.-G., and Weisberg, R.H., eds., Monitoring and modeling the Deepwater Horizon oil spill—A record-breaking enterprise: Washington, D.C., American Geophysical Union, p. 9–18. doi:10.1002/2013EO200006.

Image References

Granule IDs

• AST_L1A_00305012010164340_20200706101315_28968
• AST_L1A_00305012010164331_20200706101315_28970

Data Citation

NASA/METI/AIST/Japan Spacesystems, and U.S./Japan ASTER Science Team (2001). ASTER Level 1B Data Set Registered Radiance at the Sensor [Data set]. NASA EOSDIS Land Processes DAAC. Accessed 2020-06-30 from doi:10.5067/ASTER/AST_L1B.003.

Author's note, February 20, 2025: Place names in this article were updated to reflect Executive Order No. 14172 and Secretary Order 3423.