Counting Animals from Space: Chapter Two Transitions from Captivity to Wild Places
Scott
Bergen, Wildlife Conservation Society, sbergen@wcs.org
(Presenting)
Eric
Sanderson, Wildlife Conservation Society, esanderson@wcs.org
With the advent of commercially available high spatial resolution space based imagery, there have been several successes using this remotely sensed data for identifying and enumerating wildlife, Pacific walrus (Burn and Weber 2002), Orca and humpback whales (Alibeah pers comm), and African elephants (African Elephant Trust 2002). Other researchers have not had success with emperor penguins (Sanchez and Kooyman 2004). Our recent study has quantified the type of fauna able to be detected and identified within a range of habitat types (Bergen and Sanderson, in review).
Verifying mobile wildlife in wild condition for validating the use of high spatial resolution satellite imagery poses unique logistical challenges, since it is necessary to explicitly locate individual animals in space and time. Using data collected at the National Elk refuge in Jaskson, WY, we show how ground census procedures document the location of over 5,000 elk and bison occurring on the refuge on February 12th, 2006. These ground census procedures use digital photography that can cross-reference digital ground photography with individual elk in the satellite imagery. This has allowed us to examine different ground factors and there influence on reflectance spectra.
Animal’s mobility can be an asset in identifying individual animals. In a study from Ruaha National Park, Tanzania, we show how change detection techniques can be used to identify animals. This method distinguishes individual animals from objects having similar size and reflectance spectra characteristics. These change detection algorithms performed between aerial digital photography and high spatial resolution satellite imagery have distinguished individual animals from cryptic background ground environments.