-- For my original question and pictures of elephants wearing GPS equipped anklets, go here. --
I'm experimenting with the use of GPS anklets to track the movement of zoo elephants. This has proved endlessly complex, but I'm focused on one particular question now: is ArcGIS basemap imagery orthorectified?
My elephants wear GPS anklets that collect data points every 5 seconds. My goal is to cleanup this raw data as best as I can, without going overboard. Some of the steps I've taken are applying a variety of filters (NSAT>=4, HDOP<2, min and max movement thresholds based on elephant speeds) and keeping track of the times when the elephants are inside structures so that I can remove those points from the dataset. When I map the remaining points, some are still falling outside of the exhibit boundaries.
Originally, I had planned to map the data on the basemap, draw a polygon on top of the elephant exhibit, and use the clip function to remove points falling outside the exhibit. I've heard that I'm operating on the flawed assumption that the basemap provides an accurate view of the exhibit, when in fact, various distortions (camera tilt, local topography) may be skewing the imagery. I've also read that ArcGIS buys it's imagery from various sources, all of which probably do some orthorectification, but that there is no way to know the accuracy of the map at any particular place. On the other hand, my research may be operating on such a small scale that these distortions, if they exist, are unlikely to cause any problems.
Any solution would need to be easily implemented and widely applicable, as this data is being collected at zoos across North America.