August 31, 2011
Nearly 3 million lobster traps off the coast of Maine bring in 78 million pounds of lobster annually, adding $228 million to the state’s economy and making the industry’s sustainability a hot topic. To track the number of traps that are in use at any given time, the fishing industry marks the traps with buoys. However, counting thousands of buoys in a constantly moving ocean can be challenging, to say the least. With the traditional method of counting the buoys from a boat by hand or with GPS equipment, buoys are often double counted or missed entirely. Another method is to take photographs with a 35-mm camera mounted in a single-engine airplane, but this method is also unreliable.
Certain that there had to be a better way, the Maine Department of Marine Resources (MeDMR) contracted with KAPPA Mapping Inc., based in Bangor, Maine, in 2009 to evaluate the use of photogrammetry. The project area was in Tenants Harbor, a small lobstering community located along Maine’s rocky coast between Portland and Rockland. The objective: To establish baseline data in a specific 3.5 square nautical mile study area using a time-series analysis of the data. “We wanted to match buoy locations with catch rates to determine how trap numbers affect the lobster population,” says Carl Wilson, MeDMR Marine Scientist.
KAPPA’s work included developing flight plans, georeferencing imagery, and capturing data using softcopy photogrammetric workstations. But generating an accurate buoy count would require more than basic photogrammetry--the team would need to apply some creative techniques to handle the numerous complexities of the project.
One of the team’s first challenges was to determine the photo scale. Lobster buoys are approximately 1.5 feet in length with a tail on the end, so the team needed to acquire imagery that would allow them to see, with confidence, the lobster buoy in the water.
Team members determined that a photo scale 1"=660' to 1"=800' would allow them to cover the ocean area and also capture land features. The land features would allow them to measure pass points using photo-identifiable features, and would also allow them to scale control from existing USGS quadrangle sheets. However, after further study, the team was concerned that they would not be able confidently identify lobster buoys at those altitudes. They revised their flight plans to fly at a lower photo scale of 1"=600', which has a flying height of 3,600 feet above the water surface.
Team members also determined that scanning the film at 14 microns would enable them to have a ground sample distance of 0.33 foot, which would allow them to discern features greater than 0.5 foot. They knew that stereoviewing would provide the best opportunity to extract data from the imagery because stereo detail is much clearer than the detail of individual photos.
KAPPA contracted with James W. Sewall Co., headquartered in Old Town, Maine, for the imagery acquisition, which was to be flown under specific conditions including low tide, calm seas, early morning, and overcast skies (to minimize or eliminate sun glare on the water). The preferred schedule for this time series analysis was one week between flights. The team developed a flight plan and then waited for the dates to arrive.
In the summer of 2009, 10 overlapping photographs were captured during calm seas, low tide, and overcast skies on July 10 (pilot), August 6, August 13, August 20, and August 26. All but one dataset provided usable imagery; in the August 26 dataset, a heavy reflection from the sun combined with choppy water prevented many of the buoys from being seen.
Working closely with MeDMR’s Wilson, the mappers developed a hierarchy of confidence levels that were applied to all datasets: Buoys were labeled certain, probable or possible. Where feasible, the secondary floatation (toggle) was also captured and defined. Buoy points and attributes were then translated and delivered in an Esri ArcView personal geodatabase.
Although the use of photogrammetry to locate features is not a new technique, KAPPA’s application of the technology allowed the team to ensure that what was being seen in stereo was a buoy. For positioning, the firm used the navigational-grade GPS coordinates that are printed on the photograph during the instance of exposure. Usually this information is trimmed off the contact print or scan, but the team specifically requested that it be kept on for this project.
To complete the aero-triangulation process, the team matched up buoys in adjacent photos and used the navigational-grade GPS photo positions to aerotriangulate the imagery. Next, they leveled the stereomodels on the buoys and then located them. Geo-positioning of the data eliminated double counting and ensured that buoys were not missed during the visual quality control process.
The project yielded surprising results. MeDMR estimated that about 3,000 buoys would be located on a given day, but KAPPA located approximately 9,000 buoys for each dataset. On Aug. 6, 2009, more than 10,450 buoys were counted in the 3.5 nautical square mile area. “This is the first time that we have been able to fully visualize the Maine lobster fishery as it happens on the water,” says Wilson.
With an accurate buoy count, scientists can begin to study whether an equal number of lobsters can be caught with fewer traps to yield efficiencies in fishing operations and more sustainable fisheries. “We have always had individual data, but this experiment marks the first time we can map the entire activity of the fishery,” Wilson says. “The utility of aerial mapping of lobster gear has a direct impact on lobster science and management discussions.”
Exploring Emerging TechnologiesThe success of the Tenants Harbor buoy counting project enabled MeDMR to pursue funding for a much larger project that began in the fall of 2010. The project is being managed by KAPPA Mapping with aerial data acquisition provided by Keystone Aerial Surveys, headquartered in Philadelphia. A similar methodology is being used to locate buoys along the entire length of the Maine coast. The purpose of locating these buoys is to identify vertical lines in the water (the rope or chain that connects the buoy to the lobster trap), which can impede the movement of whales in deeper ocean waters and present shipping channel conflicts.
This new project is allowing KAPPA Mapping to explore the use of advanced technologies such as the Vexcel UltraCAM camera, a large-format digital camera that collects true color, color infrared, and black and white images at the same time. The camera is integrated with airborne GPS and inertial measurement unit (IMU) equipment that provides highly accurate geo-positions for each image. The images can be captured at 16-bit, which provides increased bit-depth for more visual data. Challenges have included a difficult location (the project site is outside the fishing exemption line, which means it is more than 3 miles out to sea), fickle fall weather and rough seas (bouncing buoys). Despite these hurdles, the firm has been able to collect nearly 40,000 buoys to date.