Pros and Cons

  return to page 2

There is no question that satellites and computer models are opening the door to new ways of looking at ecological issues. While these new technologies and techniques offer a far greater range of information than has ever been possible before, there are still challenges to using satellite data effectively. For starters, remote sensing data need to be validated, or compared with other sources of information to see if there is agreement. Without proper validation, satellite data should not be taken at face value. Such ‘ground-truth’ information might come from ground-based sensors, higher-resolution remote-sensing sources, like aerial photographs, or researchers in the field.

“You still need to put effort into ground-truthing,” says Raxworthy. “You just can’t look at images from space and use computers to give you the answers. You still have to have a muddy-boots biologist trudging through the forests.”

  Photograph of Field Camp in Madagascar

Another consideration is the cost for imagery. As satellites and computers become more sophisticated, the costs are going down. Raxworthy recalls when he first started as a biodiversity researcher, the price of satellite imagery was beyond his meager budget. “In the mid-1980s, I would have given my eye-teeth to get satellite images of Madagascar,” he says. “Now I can download MODIS [NASA’s Moderate Resolution Imaging Spectroradiometer] images for free.”

In general, imagery from the newer and higher-spatial-resolution satellites is more expensive than lower-resolution imagery. Anyone can now buy a specialized image from the Landsat 7 satellite, currently run by the USGS, which offers 15-meter-resolution black-and-white images and 30-meter color images. These cost about 600 dollars per scene, down from the thousands of dollars for images from previous Landsat satellites. Handling satellite images requires special computer software and hardware tools, and while they are not negligible, these computing costs are declining.

Along with getting access to images and computers to process images, the next generation of researchers will need training to use these new tools. “It still takes a certain amount of knowledge to do this,” says Turner. Although many types of remote sensing data are in the research phase of development and currently are beyond the capabilities of most researchers, Turner says, “The article in TREE was a call to say, ‘Hey, think about this.’ This is especially true for the people who are just coming up in schools.”

The Future

In their overview article in TREE, Turner and colleagues point out that a perception problem exists within science communities that would benefit from these new tools. Though technology has advanced, contends Turner, many researchers cling to old standards, back in the days when satellites recorded data at spatial scales that were too coarse to be relevant to the needs of ecologists and evolutionary and conservation biologists. “Perception may be more the problem now than technology,” he says.

“It cuts both ways,” says Turner, describing how scientists in the field view remote-sensing technology. “On the one hand, people can be naïve and think it’s an eye in the sky, and we can zoom in and see all the elephants and count them.” On the other hand, some researchers do not recognize how realistic and applicable to their work these technologies might be. “There’s no question that improvements have happened in the last 15 years,” says Turner, but in trying to educate people about the new technologies, “we don’t want to oversell this thing.”


Although remote sensing data are exceptional for exploring isolated areas, the information must be verified. Chris Raxworthy’s team discovered several new chameleon species while validating computer models of species’ ranges that they developed with the aid of satellite data. (Photograph copyright Chris Raxworthy)

  Quickbird satellite image of elephant herd in Amboseli National Park, Kenya

Chameleon-hunter Raxworthy believes that in the future, as satellite data span greater area and longer time periods, coverage will become more diverse, and computer models will improve. He envisions a world where interactive programs will allow ecologists and biologists to choose any species they are working with, in any country, and simply input longitudes and latitudes into an interface. The computer would be able to spit out a model of distributions for that species, allowing researchers to ask ecological and evolutionary questions and get answers.

But today, the most pressing applications concern conservation issues. “Fifty years from now, the next generation of biologists will inherit a very different landscape than what we see right now. We’re the key generation, right now, to make the smartest decisions in terms of forest survival,” says Raxworthy.

“Time is our enemy, in this regard,” says Turner about biodiversity loss. “What is needed is more collaboration now among remote-sensing researchers and those working in biodiversity science and conservation. The tools are there. Let us hope the users will soon follow.”

  • References:

  • Turner, Woody; Spector, Sacha; Gardiner, Ned; Fladeland, Matthew; Sterling, Eleanor; and Steninger, Marc, June 2003: Remote sensing for biodiversity science and conservation, TRENDS in Ecology and Evolution, 18 (6).
  • Conference on Biological Fingerprinting: Using Remote Sensing for Improved Modeling and Monitoring of Biodiversity, December 3-5, 2001, People Center of the American Museum of Natural History, New York, NY.
  • Raxworthy, Christopher J.; Martinez-Meyer, Enrique; Horning, Ned; Nussbaum, Ronald A.; Schneider, Gregory E.; Ortega-Huerta, Miguel A; and Peterson, A. Townsend, 2003, Predicting distributions of known and unknown reptile species in Madagascar, Nature, 426, 837–841.

Satellite data is usually unable to detect individual animals, but some extremely large individuals are visible from space. Conservationists in Amboseli National Park, Kenya, use data from the Quickbird satellite to track the movement of elephant herds, which appear in the left side of this image as bright ovals edged with dark shadows. (Images courtesy Amboseli Elephant Research Project, based on data copyright DigitalGlobe.)