Satellite Remote Sensing on the High Seas
Utilized by oceanographic researchers for decades, satellite data is now also used in an increasing number of commercial applications, ranging from locating favorable fishing areas to tracking dangerous ice, currents and sewage effluent.
By Jan Svejkovsky

Field monitoring of ocean environmental conditions is a very difficult and expensive enterprise. Because the oceans undergo constant change from currents and biological processes, even multiple boats and buoys often cannot provide enough synoptic measurements to resolve conditions existing in a given area. Satellite remote sensing offers a highly cost effective alternative for monitoring ocean processes on different time and space scales. Utilized by oceanographic researchers for decades, satellite data is now also used in an increasing number of commercial applications, ranging from locating favorable fishing areas to tracking dangerous ice, currents and sewage effluent. However, the marine environment imposes several unique logistical problems which must be addressed by all successful marine remote sensing services.
      Long before the first satellite was ever launched into space, ingenious tuna fishermen were using remote sensing in search of their elusive prey. "Tuna towers" - baskets erected precariously high above the ship's decks were manned by fishermen who spent long hours scanning the vast ocean surface for signs of fish. Although a school of tuna could sometimes be seen directly, often the fish were located indirectly by sighting areas of turbulence caused by their swimming just under the surface, weed-lines and floating logs which tend to attract fish and wave patterns indicative of current edges. Efforts to expand the search perimeter led to using hot air balloons, helicopters and, by the early 80s, satellites. The utilization of satellite images by ocean-going fishing fleets was probably the first wide-scale commercial application of marine satellite remote sensing.
      Satellite oceanography provides an effective means to measure numerous environmental variables, the most common ones being ocean surface temperature, water color, turbidity, currents and waves. Applying satellite ocean sensing as a routine, operational tool, however, carries several difficulties which are unique to the ocean realm. The first of these is linked to the ocean's dynamic nature. Currents, some flowing at 9 miles per hour or more, move vast quantities of water, causing constant changes in temperature and turbidity patterns. Plankton blooms and similar biological events can cause increases in primary productivity from less than 1 to 40 mg/m3 within 48 hours over hundreds of square miles. Effective environmental monitoring of such conditions therefore requires a very high resampling (i.e. satellite revisit) frequency. Quite literally, yesterday's imagery quickly becomes worthless history for many applications. The second problem is the inherent difficulty of delivering the time-sensitive satellite information to the user in the field. This complication may be minimal if the end-user is a coastal water district office or oil drilling platform. It becomes significantly more important if the user is a fishing boat 3000 miles at sea. Even with the most sophisticated option - an INMARSAT satellite telephone - transferring a 1MB image from a land-based analysis facility to an ocean-going vessel can be a logistically challenging and costly task. Finally, a recurring problem with ocean remote sensing is that good quality satellite data is often simply hard to obtain for a particular region of interest. Many parts of the world's oceans are out of range of established ground stations. Alternately, equipping a vessel with its own data downlink requires sophisticated stabilized antennas and expert data processing staff for reception of all but the very low resolution weather satellite image formats. Then there is the problem of clouds and other atmospheric contamination. While one cloudfree image every week or two may be perfectly acceptable for many terrestrial applications, such data frequency is only marginally useful over the ever-changing ocean.
      Despite these difficulties, a handful of specialized companies have begun to successfully exploit market opportunities in marine remote sensing. One of these, California-based Ocean Imaging, has offered commercial remote sensing services to world fishing fleets since the mid-eighties and has since expanded its monitoring services to include offshore oil drilling, coastal engineering and waste disposal applications. Their fish-finding tricks are no longer reserved for large commercial vessels - even the occasional "weekend warrior" can now use satellites to help find that once-in-a-lifetime trophy.
      As is the case with the tuna tower approach, satellite data is not used to spot fish directly. Instead, satellite imaging is employed to identify areas with a particular combination of environmental parameters known to be preferred by a certain fish species. For example, albacore tuna tend to favor regions with clear, blue water in the 59 to 65¡F range. Specifically, they tend to aggregate and migrate along the edges of such waters, making the boundaries between clear and murky, more plankton-rich areas especially attractive. However, the complexity of relationships between physical and biological parameters precludes such fish-finding analysis to become a simple unsupervised classification exercise. At Ocean Imaging, knowledge gained through years of fisheries research as well as close interactions with the various fishing fleets are used to better define spots with the best fishing potential. Since many environmental variables affecting fish distributions cannot be discerned with available satellites, oceanographic expertise must often be relied upon to secondarily deduce non-measurable quantities from the satellite data.
      The need for high frequency image updates limits the types of satellite systems useful for ocean applications. The most commonly utilized sensor is the Advanced Very High Resolution Radiometer (AVHRR) carried aboard the NOAA polar orbiter series. This sensor provides 1 km spatial resolution mutichannel thermal and single channel visible imagery. With two NOAA satellites usually functioning simultaneously, four image updates per day are possible, even more at higher latitude locations. The geostationary GOES satellites also provide useful thermal information at low latitudes and the soon to be launched SeaWiFS ocean color sensor will provide high sensitivity information on chlorophyll and suspended sediment with a daily repeat coverage.
      For most commercial applications, the company provides its customers with a fully processed, calibrated product in the form of a digital image or a graphic chart. The company has developed its own end-user software for use with the digital analyses. The SeaView software was especially designed for marine applications, incorporating such features as GPS interface input to display the ship's track and position directly over the received satellite imagery. Simplicity and ease of use was also an important part of the software design, since vessel crews are often not highly computer skilled. The satellite image analyses are updated twice daily via cellular or satellite phone link.
      Although the color satellite image products provide the most detail and accuracy, Ocean Imaging also generates specialized graphic charts to get information to specific users. The black and white charts show locations of temperature, color and current boundaries as lines of various thicknesses along with representative temperature values and fishing spot recommendations. The charts do not preserve the full detail of the original digital image, but are often easier to disseminate to users. One very popular service automatically provides sportfishermen twice-weekly satellite fishing charts through their home or office fax machine. Charts also allow the company's analysts to better combine cloudfree information from several satellite orbits and to interpolate the available data based on known oceanographic principles, thus generating an artificially cloudfree product. This is especially useful for providing services over notoriously cloudy ocean regions. From December through March, Ocean Imaging transmits by radio wefax satellite-derived temperature and current charts to American and New Zealand fleets targeting albacore tuna in the central South Pacific. With much of the fishing grounds located within the permanently cloudy subtropical convergence zone, data from every cloud free opening is utilized for the three weekly updates. "Sometimes we literally squeeze blood out of a stone," says Mark Hess, Ocean Imaging's chief technical analyst.
      Applied in the field, the satellite ocean analyses are very effective. " Back in the early 1980s, before Ocean Imaging came to us, we just had to drive," says John Gibbs, a San Diego based swordfish and tuna fisherman. "We'd team up with other boats, roam around and share information. Now we can head directly to the most promising spots or send different boats to several areas to check them out." The decreased search time and increased fishing efficiency result in savings of thousands of gallons of fuel as well as increased catches. Some large Mexican tuna boats, which spend a month or more at sea per trip, have actually begun to take on less fuel prior to each voyage knowing that these analyses will significantly reduce their operating expenses.
      As useful as satellite oceanography is for finding fish in the ocean, the same technology is also being employed to preserve those fishing resources for the future. Environmental variations, both short and long term, are often reflected in satellite sensed patterns of ocean temperature and primary productivity. Analysis of trends in a satellite image series can be used to identify habitat changes that may affect the migration, size or recruitment of a particular fish stock. Fisheries management decisions can then be made to protect it from possible overfishing. Under this premise, Ocean Imaging has recently begun work on a grant from the fishermen's own American Fishermen's Research Foundation to better understand environmental effects on migration trends of the North Pacific albacore.
      While fisheries is one obvious industry to which satellite remote sensing offers significant benefits, the same environmental information finds use in other applications. Oil companies involved in offshore drilling exploration need information on the location and strength of currents and eddies which can potentially damage equipment and hinder field work. Currents are also of obvious importance to sanitation districts which discharge effluent into the sea near heavily populated areas. Ocean currents can be effectively monitored from AVHRR and ocean color data by measuring the displacement of small thermal or turbidity patterns between successively acquired images. Through a development grant from NASA's EOCAP program, Ocean Imaging is presently working to increase the type of water quality parameters which can be obtained from existing and future satellites and to apply the developed techniques commercially to new markets. One such application is the daily generation of satellite-derived water clarity analyses for use by sport divers to locate dive sites with the best existing underwater visibility. Another involves the use of satellite sensing by coastal sanitation districts to monitor environmental conditions around their outfalls more efficiently. NASA's assistance through the EOCAP program is especially useful for allowing small companies design and develop specific remote sensing products for the numerous emerging marine market sectors.
      Satellite ocean sensing stands to especially benefit from the upcoming new sensors and satellite communication technology. Global wireless communication networks will considerably simplify the delivery of analyses and image products to vessels at sea. New high-resolution and hyperspectral sensors with more frequent revisit times will increase the utility of remote sensing for small-scale coastal monitoring applications. When used in conjunction with the AVHRR, Synthetic Aperture Radar (SAR) imagery (which is not hindered by cloud cover) is already showing good promise for monitoring ocean currents under all-weather conditions. In capabilites as well as applications, marine remote sensing will remain a growing field.

About the Author:
Dr. Jan Svejkovsky is the president and founder of Ocean Imaging. He may be reached at 619-792-8529.

Back