Nature Enhanced. To produce this image, Jonathan Chipman amplified visible blue, green and red wavelengths of reflected light to visually emphasize differences in reflectance properties of Green Bay and Lake Michigan. With this manipulation, high algae concentrations in Green Bay and Lake Winnebago produce deep yellow colors. In Lake Michigan, precipitated calcium carbonate produces swirling shades of blue. Land areas have been masked in black. Original image captured by Terra MODIS on Aug. 14, 2001.
UW SEA GRANT RESEARCH
Sampling Water from Space 8/13/2007
Scientist uses satellites, sunlight, and lots of math to measure water quality
By John Karl
The warm days of summer regularly trigger explosions of algae in Green Bay and other Great Lakes waters. The water turns green as pea soup, and the algae can be toxic to fish, pets, and people. Zillions of zebra mussels may be making things worse.
Using two satellites soaring 700 kilometers (435 miles) above Earth, remote sensing scientist Jonathan Chipman is seeking new perspectives on the unpleasant phenomenon. His work, funded by UW Sea Grant and other agencies, will soon help water quality managers and municipal, industrial, and agricultural officials better understand what causes excessive algal blooms and what can be done to control them.
Remote sensing cannot replace traditional, in-the-water measurements of water clarity or suspended solids, but it can tremendously multiply the value of those efforts, according to Chipman, a scientist at the Environmental Remote Sensing Center (ERSC) at the University of Wisconsin-Madison.
For example, logistical and economic considerations might limit sampling from a boat to a dozen locations on Green Bay every week. However, using those data to “ground truth” indirect measurements from satellites, remote sensing technology can make similar measurements from thousands of locations every day, covering the entirety of Green Bay, Lake Michigan or other large water bodies any day the skies are clear.
The technology essentially uses light to probe the water. The full spectrum of sunlight, or electromagnetic radiation, includes wavelengths both longer and shorter than visible light. When this radiation reaches the surface of Earth, the chemical composition of the surfaces it strikes — whether they’re lakes, forests, parking lots, or rooftops — determines how much of each wavelength is reflected back to space. Thus, anything in the water that absorbs some wavelengths of light and reflects others can be measured by analyzing the reflected light.
Limnologists refer to these light-absorbing and reflecting substances as “color-producing agents,” because they affect the apparent color of water bodies. In Green Bay, two of the most important color-producing agents are chlorophyll-related pigments, produced by algae, and particles of sediment that are suspended in the water. High concentrations of chlorophyll-a and total suspended solids (TSS) cause an increase in water turbidity and a decrease in water clarity (or Secchi depth).
The two satellites Chipman works with are part of NASA’s Earth Observing System. Named Aqua and Terra, they each orbit Earth 14 times every day, one passing over the upper Midwest in the morning; the other, in the afternoon. They carry sophisticated instruments that are like digital cameras sensitive to many wavelengths of light. The instruments divide the scene below them into millions of tiny regions, or pixels, and they record the radiation reflected from each pixel. The data for each pixel, tagged with its precise geographical location, is then transmitted to special receiving equipment on the roof of the UW-Madison Space Science and Engineering Center, home of ERSC.
That’s when Chipman’s challenges begin. He wants to analyze the energy reflected at the water’s surface, but dust, pollution, and water vapor in the atmosphere interfere with that radiation before it reaches the satellites’ sensors. So Chipman must, in a sense, mathematically decode the at-sensor radiation data to recover what that radiation looked like the moment it left the water’s surface.
Such mathematical filters have been developed by other scientists for other water bodies, but they do not work well for Green Bay. The chemistry of the bay’s water, and thus its optical properties, is substantially different from that of, say, the North Atlantic Ocean. It’s even quite different from the open waters of Lake Michigan.
“I’ve spent a lot of time producing algorithms that are reliable enough,” Chipman said.
A key part of the process involves using field measurements of water quality parameters — such as Secchi depths and chlorophyll concentrations — to “calibrate” a set of numerical models. These models take specific patterns of light measured by the satellites and correlate them with in situ measurements (made “in place,” from a boat or bouy) on the same date. To give a simplistic example, if a particular in situ sampling location (corresponding to a particular pixel in the image) shows a Secchi depth of 2.3 feet, then other pixels in the image with the same surface reflectance values must also have a Secchi depth of 2.3 feet.
Chipman describes this process as “training” the computer to interpret water quality parameters from the satellite imagery. Once the algorithm has been developed and tested, it can be used to estimate water quality at times and places where no field sampling occurred.
Several agencies and scientists provide the vital in situ data to Chipman, including the Green Bay Metropolitan Sewerage District and the Wisconsin Department of Natural Resources.
Chipman has so far produced a series of maps of Green Bay, from 2000 to 2006, that are color coded for Secchi depth, chlorophyll-a concentrations, suspended solids, and turbidity. He is working on ways to automate the production of such maps and to improve the accuracy of satellite estimates for times when no field data are available for comparison. His ultimate goal is to deliver near-real-time data and maps via the Web whenever the satellites can get a clear view of Green Bay or Lake Michigan.
From such data, maps of monthly and yearly averages can be produced, showing recurring trouble spots and, hopefully, improvements over time, according to Vicky Harris, UW Sea Grant water quality and habitat restoration specialist. Harris chairs an outreach committee charged with informing people about new daily limits being developed for the amount of phosphorous and suspended solids allowed to enter the Fox River in runoff from farm fields, urban storm water, and sewage and industrial wastewater discharges. The Fox is by far the largest source of such pollution to Green Bay, and that pollution is a main cause of the bay’s excessive algae blooms, according to Harris.
Maps of those blooms will demonstrate the problem much more effectively than mere numbers on a chart, Harris said.
“People will be able to see in a highly accessible way what phosphorous and other runoff pollution does to the bay and how extensive the problem is,” Harris said. “Hopefully, then they’ll be supportive of a solution.”
When it comes to water quality measurements, a satellite picture is indeed worth a thousand words – or numbers.