Using Remote Sensing for Mapping and Counting Animals

Ad:

Recent events, such as concern over diminishing polar bear populations, have brought to the public’s attention the use of remote sensing data in observation of wildlife, in particular counting the number of animals in remote locations. While satellite data have been used, now, with low costs for various unmanned aerial vehicles (UAV) and their easy use, a growing number of studies have focused on using UAVs to detect and count animals.

Counting Animals with UAVs Can Be More Accurate Than Ground Surveys

Thermal and infrared imagery, particularly from UAVs, has helped improve counting by relating animals’ heat and visibility signatures to given reflectance. Ground observations and GPS data have been used in conjunction to help validate observations and geolocate detected animals such as white tailed deer in one study. Supervised classification can then be applied to create signatures that are then applicable for wider areas where less control might be available (i.e., lack of ground observations).[1]

In one study, it was shown that UAVs can be about 43-96% more accurate than ground or human-made observations in counting wildlife. In fact, UAVs are also better at detecting fake animals by focusing on signatures animals give. In this case, the study looked at animals that often crowd closely together, specifically seabird colonies. The results do suggest that for fine-scaling monitoring, UAVs are likely to be more powerful than either human observation methods or even satellite-based remote sensing.[2] In another study on marine seals, it was found that UAVs, using thermal imagery and auto-classification based on detected signatures, led to about 95-98% correspondence to human counts. In other words, the accuracy was similar. Given the low cost of maintaining and flying UAVs, they were a far more economical instrument to use.[3]

Aerial vantage of a replica seabird colony compared with the ground counter's viewpoint. Source: Hodgson et al., 2018.

Aerial vantage of a replica seabird colony compared with the ground counter’s viewpoint. Source: Hodgson et al., 2018.

Despite the fine-scale benefits of UAV counting, there are problems. The main critique of detecting and counting animals has been that studies have mostly focused on relatively small areas, such as small parks or enclosures, whereas there is a greater need to count over very wide areas. In fact, this is one reason why it has been difficult to determine how well some endangered or threatened species are doing, such as polar bears. There is a greater need to expand studies, enable long-term monitoring, and better develop automated or semi-automated methods and classifications so signatures could be counted accurately. In effect, this is still an area that needs further research and work.[4]

Using Satellite Data to Map Species

One area where satellite-based data have demonstrated utility, perhaps ironically, is in smaller species, particularly counting phytoplankton growing in lakes and water systems. Using the Medium Resolution Imaging Spectrometer (MERIS) platform, which absorbs phycocyanin at around 620 nm, the system can detect sensitivity and robustness of phytoplankton growth. This makes satellites such as the Envisat, an older but still utilized system, useful for large area surveys. In effect, satellite-based data have so far been better at conducting such large area survey relative to UAVs, as most UAVs operated by biologists likely have limited range and flying time.[5]

While there is still a lot of potential growing room for remote sensing to improve in counting animals, there are limitations that could prevent the field from growing rapidly or improving to the point where it can fully replace human observation and counting. Primarily, the main limitation is the ability to fly UAVs continuously and for long distances. While satellite ofter this possibility, UAVs available to most scientists are not capable. With costs going down for more sophisticated UAVs that can automatically fly over wide areas, then we are likely to see improvements in remote sensing techniques for detecting and counting animals.

References

[1]    For more on the technique for counting white tailed deer, see:  Chrétien, L.-P., Théau, J., & Ménard, P. (2016). Visible and thermal infrared remote sensing for the detection of white-tailed deer using an unmanned aerial system: Detection of White-tailed Deer Using an UAS. Wildlife Society Bulletin, 40(1), 181–191. https://doi.org/10.1002/wsb.629.

[2]    For more on the accuracy of UAVs, or remote piloted aircrafts (RPAs), in counting, see:  Hodgson, J. C., Mott, R., Baylis, S. M., Pham, T. T., Wotherspoon, S., Kilpatrick, A. D., et al.  (2018). Drones count wildlife more accurately and precisely than humans. Methods in Ecology and Evolution, 9(5), 1160–1167. https://doi.org/10.1111/2041-210X.12974.

[3]    For more on the seal counts and accuracy, see:  Seymour, A. C., Dale, J., Hammill, M., Halpin, P. N., & Johnston, D. W. (2017). Automated detection and enumeration of marine wildlife using unmanned aircraft systems (UAS) and thermal imagery. Scientific Reports, 7, 45127. https://doi.org/10.1038/srep45127.

[4]    For more on the benefits and limitations of some recent methods and applications of using remote sensing to detect and count animals, see:  Hollings, T., Burgman, M., van Andel, M., Gilbert, M., Robinson, T., & Robinson, A. (2018). How do you find the green sheep? A critical review of the use of remotely sensed imagery to detect and count animals. Methods in Ecology and Evolution, 9(4), 881–892. https://doi.org/10.1111/2041-210X.12973.

[5]    For more on the phytoplankton counting study, see:  Lunetta, R. S., Schaeffer, B. A., Stumpf, R. P., Keith, D., Jacobs, S. A., & Murphy, M. S. (2015). Evaluation of cyanobacteria cell count detection derived from MERIS imagery across the eastern USA. Remote Sensing of Environment, 157, 24–34. https://doi.org/10.1016/j.rse.2014.06.008.

See Also


Advertising



Like this article and want more?

Enter your email to receive the weekly GIS Lounge newsletter:

Advertising