SCIENCE FOCUS: THROUGH A WATER COLUMN DARKLY, PART 3

What to Do Next

The problem of turbidity won't "settle" down anytime soon, and in the near future it may get even worse, as nutrient inputs and erosion from agriculture increase from Third World countries. In addition, global warming could even exacerbate the problem with increased rainfall, causing more floods, and sea level rise, causing increased coastal erosion.

One way to address the problem is with better sensors. SeaWiFS was primarily designed to accomplish the goal of observing the entire global ocean and determining the chlorophyll concentration everywhere to an accuracy of 35%. The SeaWiFS Project currently uses a global algorithm (the OC4 algorithm, described in More than Meets the Eye) that switches the bands used to calculate chlorophyll concentration as radiance changes, which adapts the algorithm to a range of conditions.

But this algorithm only goes so far. Optical oceanographers know that knowledge of regional water characteristics can be used to tune analytical algorithms to produce better results than a single algorithm used to analyse the entire world. So one way to address the problem of turbidity and Case 2 waters is to create a set of regional algorithms that overlapto cover problematic areas.

The next way to improve the analyses is to add more bands to the instrument. The optical properties of the various materials and conditions that create Case 2 water conditions interact to produce what the sensors observe. By adding more bands, algorithms employing these bands can "unravel" the tangled knot of optical characteristics and isolate the contribution of each. An example of how this might be done is the use of the 412 nm band found on SeaWiFS and MODIS. Gelbstoffe absorbs light strongly at this wavelength, much more than chlorophyll, so light absorption in this band can be used to calculate the concentration of Gelbstoffe.

Beyond the addition of bands, improving the optical capabilities of the instrument also aids the discrimination of optical properties in Case 2 waters. In overly simple terms, the "signal"due to chlorophyll in Case 2 waters is obscured by the optical "noise"of the other factors that have been described. By improving the optical performance of the instrument, the instrument will be more sensitive to the signal compared to the noise. The signal-to-noise ratio of the optical bands in SeaWiFS was more than 10 times better than its predecessor, the Coastal Zone Color Scanner, and MODIS is significantly improved over SeaWiFS. Furthermore, the width of the bands in MODIS is smaller than for SeaWiFS, which also helps to isolate the signal that is being measured.

The final factor that can be added is increased creativity by scientists studying the complex optical mixture found in Case 2 waters. Some researchers have employed iterative algorithms that test a variety of possible conditions until they converge on a single, hopefully accurate, answer. These algorithms take much more computational effort than the band ratio algorithms employed by the SeaWiFS Project, but they offer one of the only current ways to separate the contributions of various factors. Other groups have successfully improved atmospheric correction over turbid coastal waters by changing the assumptions that go into the global analytical algorithms. (See "Atmospheric Correction of SeaWiFS data for turbid waters" below.)

Looking to the future, the next major advance in Case 2 water analysis may be a hyperspectral ocean color sensor. A hyperspectral sensor, such as the Hyperion instrument on the Earth Observer-1 (EO-1) mission, collects data in many more bands than SeaWiFS or MODIS. Hyperion has 220 bands in the 0.4 to 2.5 µm wavelength range. To be effective, a hyperspectral ocean sensor would have to possess the same optical capabilities of SeaWiFS or MODIS -- and such an instrument has not been built yet.

ACKNOWLEDGMENT

Dr. Kevin Ruddick of the Ocean Colour Research Unit of the Management Unit of the North Sea Mathematical Models (MUMM) provided a review of this Science Focus! article.

Links

• The multicoloured North Sea
• Atmospheric correction of SeaWiFS data for turbid waters
• The Distribution and Dynamics of Suspended Particulate Matter in Belgian Coastal Waters Derived from AVHRR Imagery
• How to Monitor Coastal and Inland Water Quality
• Optical Relationships in Case II Waters and Their Application to Coastal Processes and Remote Sensing Algorithms
• Water Column Correction Techniques
• Classic CZCS Scenes, Chapter 10: River Plumes and Estuaries

<< Back

Return to Index