July 20 July 20 July 19 July 17 July 16 July 16 July 16 July 15 July 13 July 12 July 12 July 11 July 10 Click here for more mission logs Meet the Team Mission Home Page

Paulo Maurin
University of Hawai'i

Dena Deck
Bellflower Unified School District, California

In consultation with Joe Chojnacki
NOAA Coral Reef Conservation Program

When soldiers from Napoleon’s army found the Rosetta Stone, it was a breakthrough discovery. Carved in ancient Egypt, it contained pieces of a message in known languages and also a language that had been dead for centuries. Without any link to other known languages, historians had been unable to decipher this language until the stone was found, which provided the necessary clues to translate it. Modern day ocean mappers are looking for their own Rosetta Stone that will allow them to link backscatter data to other ecological information.

A map integrating backscatter map with bathymetry, showing the seafloor in rich detail. (Photo: NOAA Fisheries Coral Reef Ecosystem Division)

Our ship, the NOAA ship Hi`ialakai, has a set of three sonars that, when used in conjunction, can provide accurate data about the seafloor. When emitted by a sonar, a “ping” comes back bringing two pieces of information with it: travel time and strength. The two-way travel time (the time it took from emission, bouncing off the seafloor and return back to the ship), coupled with the measured velocity of sound in the specific water location where the ship is traveling in, gives mappers a bathymetric view of the seafloor, revealing the depth of each of its points. (See “Painting the Seafloor” article.)

A backscatter map, indicating substrate characteristics. Dark areas represent a harder seafloor, while lighter areas are indicative of a soft, sandy bottom. (Photo: NOAA Fisheries Coral Reef Ecosystem Division)

A second piece of data obtained from each ping is the strength of the signal. When sound hits a surface, above water or below, some of it is absorbed and the rest bounces back in what we experience as an echo. The strength of this echo depends on the hardness of the material that the sound is bouncing from. This is a very convenient fact of nature that is used when mapping to compliment the bathymetric map that provides the depth. The acoustic hardness of a substrate, or ocean bottom, affects the strength of the ping coming back to the sonar. In a real sense, the loudness of the echo changes if it is bouncing off sand or rock. Sand, being soft and full of small holes in between grains, will absorb quite a bit of sound. A more solid surface like a rock will provide a bigger echo for each ping that hits it.

Backscatter showing on top, each of the blue dots shown across represent points were image data was taken to ground-truth the backscatter map. Red dots have the image showing below it, and the topography profile at the bottom. Note how images show first a mix of hard and soft substrate, a sandy bottom, and a hard substrate, each shown on backscatter as gray, light, and dark areas, respectively. (Photo: NOAA Fisheries Coral Reef Ecosystem Division)

This strength of the signal coming back is called “backscatter” and provides mappers with a second view of the seafloor. While bathymetry is a measure of the depth, backscatter gives us a clue about the nature of the seafloor being mapped. Since coral reefs, with their calcium carbonate, provide a much harder surface than a sandy sea bottom, the two will appear differently in the backscatter map. Values of intensity range from low intensity, showing up as white and representing soft, sandy bottom, to high intensity, represented as dark areas for harder substrate in the backscatter gray scale map.

When the backscatter map shows up binary data – white and black – it is easy to infer on the type of substrate being mapped. The challenge is presented with all of the gray areas in the map. Does light gray represent coarse sand? Is dark gray indicative of sand over rocks, or thousands of coral polyps? Or maybe just rock covered by sand? Every shade of gray has a value that can indicate a type of substrate.

A diver armed with a camera is towed from a boat, obtaining many pictures that will be used to groundtruth mapping data. (Photo: NOAA Fisheries Coral Reef Ecosystem Division)

Backscatter alone cannot give you these answers. With so many variables present in the mapping process, data needs to go through a “ground-truthing” process, or compared to visual observations of the sites. To do this, researchers collect video, photographs and perform actual dive observations of many of the sites that are mapped. These video and images need to be analyzed by a person. It’s a tedious process that cannot be automated - it requires having a person able to classify types of substrate from watching hour after hour of video data or many photographs. And all of these data needs to be “geo-rectified,” or coupled with GIS information to know exactly where each video segment and photograph was taken. Sometimes the payoff for “groundtruthing” backscatter is unexpected: wrecks or rich coral beds can be discovered.

We do not have yet a backscatter “signature” for each type of substrate, or sea bottom, yet. This would be the Rosetta Stone of mapping, a development which will allow mappers to correctly identify some of the ecological characteristics of each area mapped. For instance, mappers are working towards refining their backscatter analysis to allow them to tell apart live coral from bleached ones.

The NOAA Coral Reef Conservation Program has built a pilot data set from the French Frigate Shoals, consisting of large amounts of video footage, observations, and other data. They are in the process of compiling all of this information with their backscatter maps they have for the area, and study how they relate, trying to find meaning to each gray area in these maps.

When mapping, additional and unexpected discoveries can take place. Sometimes what we think of as featureless terrains are revealed to have rich topographies. In 2004, an ocean area off the island of Oahu in Hawai`i, thought to be featureless and plain, was discovered to have sand dunes and ridges, providing important habitat to the marine fauna. Interpretation of backscatter data has improved in quality over the years, and when combined with videos and photographs, remote characterization of sea floor habitats becomes possible.