- Commonly used in home range analysis
- Sum of bivariate normal curves centered on each point
The kernel estimator is commonly used in home range analysis, typically with radiotelemetry data. Itís conceptually simple: you drop a bivariate normal curve on each animal location, and sum across these curves, as suggested by the dotted yellow line in the figure at the right. You could do this in your kitchen by laying a map of animal locations on the table and pouring a quarter cup of sugar on each dot. The resulting three-dimensional surface, shown here for Blandingís turtle radiolocations, represents the probability of an individual being found at any point.