Satellite Imaging of Oil Slicks – A Primer

We get a lot of questions from folks interested in our work using satellite images to detect and monitor oil spills around the world. The Montara spill off Australia last year, and the ongoing BP / Deepwater Horizon spill in the Gulf of Mexico, are striking examples of how this technology can help us investigate and illustrate what’s happening far out to sea and in remote locations.

Like all data sources, satellite imagery has its strengths but also some important limitations. Few imaging satellites (the ones taking pictures of Planet Earth) are “turned on” all the time, so images are not necessarily available. Usually somebody has to contact the satellite operators – some operators are government agencies, some are commercial for-profit businesses – and request that images be collected over an area of interest. Often, you’ve got to pay to have this done. NASA makes images from their taxpayer-supported systems, including MODIS, available for free, but satellite images from private vendors can cost thousands of dollars each.

Imaging systems that operate at visible to infrared wavelengths of light, like the MODIS system we’ve used so often, can’t see through clouds, smoke, dust or haze. And oil slick imaging is sometimes dependent on the sunglint pattern, which varies considerably from one image to the next, and is also affected by wind and wave conditions on the water. Radar imagery gets around some of these problems, but NASA doesn’t operate any radar satellites so the cost can be prohibitive.

For all of these reasons, we haven’t been able to produce good images of the BP oil slick every day (NASA just published an excellent illustrated article on this topic). But at SkyTruth we have acquired good images often enough to illustrate the enormity of the spill and inadequacy of our initial spill response efforts; provide the first estimate of the spill size and rate that made any sense; to identify oil making landfall along the Alabama coast before it was being acknowledged by officials; to show clear entrainment of the spill in the Loop Current while officials were actively denying it; and to detect small but chronic leaks from other damaged wells, raising the related issue of inadequate plugging and abandonment.

This spill has also provided a unique opportunity to collect imagery from multiple different remote-sensing systems, both satellite and airborne, working at visible to infrared to microwave wavelengths, over a long period of time under a wide range of weather and illumination conditions. A systematic analysis of this dataset will yield a much better understanding of how imagery can be used to accurately measure and monitor oil pollution events in the future. We’re looking for funding opportunities to conduct such an analysis.

Because as long as we continue to produce and transport oil offshore, there will be a next time.

Hopefully not too soon.

BP / Gulf Oil Spill – Radar and MODIS, July 19, 2010

The cap on BP’s infamous Macondo oil well in the Gulf of Mexico is still shut, but small leaks have reportedly appeared on the seafloor around the well site. This is troubling because it suggests that the well casing is damaged and leaking somewhere below the seafloor. According to Coast Guard Admiral Thad Allen, one of those leaks is actually coming from an older abandoned well nearby – he goes so far as to say:

“it’s not unusual to have seepage around the old wells”

The AP recently did a story (featuring SkyTruth, among others) on the fact that there are 27,000 abandoned wells in the Gulf, and that abandoned wells on land leak so frequently that there is an ongoing need to re-plug them. I guess Admiral Allen confirms what the AP suspected – that this is also a problem with offshore wells. Who knew? Now we all do.

BTW, we wish folks would only use the term “seepage” when talking about natural oil and gas seeps on the seafloor, not human-caused leaks.

MODIS / Aqua and CSK radar satellite images taken on July 19 show oil slicks and sheen spanning about 7,868 square miles. This is almost twice as large as the area of slicks observed on satellite imagery from July 14, but still a lot smaller than it’s been on previous imagery.

MODIS / Aqua satellite image from July 19, 2010

Oil slicks and sheen appear through a complicated assortment of clouds and haze on the MODIS image, taken at about 2pm local time on July 19, 2010. An area of anomalous ocean color (dashed line marks its eastern edge) appears to mirror the eastern edge of the area covered by surface oil slicks. This may be an indication of changed water chemistry in the area affected by the spill – possibly due to oxygen depletion as a result of the elevated levels of methane (natural gas) dissolved in the water.

The edge of the ocean-color anomaly seen on the MODIS / Aqua image is shown for reference on the CSK radar images taken a few hours later that same day:

COSMO-SkyMed radar satellite images (black-and-white) taken July 19, 2010, superimposed on MODIS image from the same day (color). CSK images courtesy CSTARS.

BP / Gulf Oil Spill – Stopped (For Good?)

Screen capture from Skandi ROV2 live spill cam, 12:42am EDT, July 16, 2010

Finally, after 87 days, the leak from BP’s Macondo well in the Gulf of Mexico is fully stopped. All the valves on the new sealing cap have been closed and an “integrity test” is being conducted on the well. If the pressure steadily increases in the well, that’s a good thing: it would mean there are no leaks in the wellpipe and casing below the seafloor. In that case BP will keep the valves closed, effectively shutting off this catastrophic spill. If the pressure doesn’t build in the well it could mean there are leaks below the seafloor, and BP will re-open some of the valves and the spill will resume. In any event, the only permanent solution is a successful relief well that fills the damaged well with cement.

MODIS / Aqua satellite image, July 14, 2010

This MODIS / Aqua satellite image, taken on July 14, shows that the area of oil slicks and sheen appears greatly reduced: slicks cover approximately 3,786 square miles (9,805 km2) on this image. Radar satellite images taken on July 11 and July 12 confirm this smaller slick area.

Persistent, moderately strong winds over the past few days (ranging from 7-20 miles per hour) may have dispersed thinner portions of the slick over much of the region.

BP / Gulf Oil Spill – Radar Comes Through Again

This satellite radar image taken by Envisat’s ASAR sensor at 10:44pm local time on July 7 shows a large patch of oil extending north from the site of the leaking Macondo well, and an area of small slicks along the Mississippi shoreline:

Envisat ASAR radar image (black-and-white) taken July 7, 2010. Image courtesy of CSTARS.

A large dark area extending from west of Mobile Bay to beyond Panama City may include patchy oil slicks and sheen, as seen in this area on previous days. But it is also an area of calm winds; the surface wind speed was measured at Buoy 42012 at 1 meter per second, gusting to 2 m/s, at the time this image was acquired. That’s on the low-end threshold for oil slick detection with radar imagery.

This infrared GOES weather satellite image taken within minutes of the radar image shows the skies are mostly clear in the area, with no sign of rainfall:

BP / Gulf Oil Spill – July 4th Weekend

With oil continuing to billow into the Gulf of Mexico from BP’s failed Macondo well, the holiday weekend brought little cause for celebration, and no break for the folks working hard to clean up the oil coming ashore and to plug the leaking well. Oil was reported for the first time in Lake Pontchartrain, and the discovery of tar balls on some Texas beaches means this spill is now directly impacting all of the Gulf states.

A RADARSAT-2 satellite image taken July 2 shows oil slick and sheen still spread across a large area of the Gulf in the wake of Hurricane Alex, which brought large waves and strong gusty winds to the area last week:

RADARSAT-2 satellite image taken July 2, 2010. Image courtesy of CSTARS.

MODIS satellite images acquired on July 3 and July 4, while impaired once again by clouds, showed portions of the oil slick in a few areas. The July 3 image shows patches of slick along the Louisiana coast reaching west beyond Vermilion Bay:

MODIS/Aqua satellite image acquired on July 3

It also shows a neat little circular pattern formed by a cluster of natural oil and gas seeps; the small slicks that form at the ocean surface above these deepwater seeps appear to be caught up in a clockwise gyre (a rotating surface current).See all the images in our Deepwater Horizon Oil Spill gallery.