Q&A: How Ursa Space uses SAR to detect oil spills

AUTHOR

The oil spill off the coast of Mauritius in August demonstrated how satellite-based, synthetic aperture radar (SAR) can be used to assist clean-up efforts.

Ursa Space’s team of SAR experts began analyzing imagery shortly after the bulk carrier MV Wakashio struck a coral reef on July 25.

For the next few weeks, we provided regular updates on the size and location of the oil spill, as the slick spread up and down the pristine shores of a rich marine ecosystem.

The results of these efforts were discussed in previous blogs, but one angle was left out.

How exactly did we use SAR to detect an oil spill?

To answer that question, let’s turn to Brian Barris, Senior Imaging Specialist at Ursa Space, and Corey Miller, the company’s Head of Research and Development.

Blog 0925 Q&A.png(If you need a quick SAR 101 lesson, Here’s an infographic).

Geoffrey Craig: What exactly are you looking for in a SAR image to detect an oil spill?

Corey Miller: Oil changes the surface properties of the water, typically resulting in a calmer surface than an area without oil on the surface.

When a flat surface is very smooth, it acts almost like a mirror to SAR energy, causing the energy to bounce away from the satellite sensor. The energy returned, or backscatter as it’s known, is of a lower value.

The lower value means that in a SAR image, an oil spill will often appear as a dark spot relative to the area around it.

GC: Let’s take a look at the SAR image below, taken August 11. It captures the vessel, the coastline and the oil spill, which is colorized red for easier visualization.

Ursa-ICEYE-Wakashio-Oil-Spill-081120-GIF-1.gif

GC: The oil spill appears as relatively dark, like you said, but there are also dark patches not colored red, which means they are oil-free. Can you explain that?

CM: You have situations where the water surface is very still, even if you don’t have oil in it. There might be no wind, so the water is very calm, like in a lake, reservoir or protected bay.

Because the radar signature depends on the water surface properties, in the real world, you get look-alikes. Even though there is a dark blob in the SAR image, it doesn’t necessarily indicate the oil slick. You have to account for this.

Brian Barris: That’s something we ran into with the Mauritius oil spill. The fact that there was enclosed water added complications. The water between the reef and land can be calm. There were days when that was more or less an issue.

CM: Because the wreck happened on a reef, the oil spill was in a protected bay, which meant you’re likely going to have oil look-alikes. One of the challenges was figuring out whether the dark areas in an image were likely oil or calm water.

We got lucky the first few days because there were favorable wind conditions, making it more obvious whether there was oil in the water or not. Later, it looked like the whole bay was covered in oil, and the next day, none of it was covered. This was an artifact of changing environmental conditions.

GC: What do you mean by “favorable wind conditions”?

CM: There’s a sweet spot where the wind isn’t blowing too much or too little. That way, even in bays, you’ll get some surface roughness from the wind. In those types of weather conditions, you’re more likely to have confidence that dark spots are, in fact, oil, rather than just calm water.

You can also follow the path where the oil has traveled, and gain some confidence in the ability to distinguish between oil and oil-free areas.

GC: Given these challenges, is SAR a better choice than optical imagery to detect an oil spill?

BB: Overall, oil spill detection is still a good demonstration of SAR capabilities and our analytical capabilities.

Oil changes the water surface structure and that’s what SAR is good at measuring. You’re directly measuring the phenomenon you’re trying to detect.

In an optical image, you aren’t going to necessarily see a black layer floating on the water surface. The colors you see depend on the angle of the sun and the sensor, as well as the sun and cloud converge, which will determine how much glean you have.

With optical, unlike SAR, what you see isn’t nearly as much a direct cause of the oil spill.

GC: Thanks for your time.

For more information, read our story map detailing Ursa Space’s oil spill detection capabilities using SAR.

Check back here for more updates on how satellite imagery data can reveal insights about the physical world.

You can also sign up for a free evaluation of our data.

#TAGGED