Erroneous overlap values?

[Storm Crews]

Hello all, hope you are well. I've been fitting joint AKDEs via the usual method:

AKDES <- akde(vultures_reproj[c(k,j)], FITS_object[c(k,j)], weights = TRUE, dt.plot = FALSE)

where vultures_reproj is a list of telemetry objects all on a shared assigned projection and FITS_object is a list of the individual ctmm models with identical index values to the vultures_reproj list.

This successfully fits the joint AKDEs on a shared grid in preparation for use of the overlap function. I'm fitting a large number of pairs and in some cases they seem to have no issue. However, one pair (or more) is definitely giving what I interpret to be impossible, with very high confidence interval values coming out of the overlap function (low: 0.89, est 0.95, high: 0.98) despite these two distributions possessing no physical overlap in space, separated by at least a few hundred kilometers completely. Is there any clear possible step at which I may have gone wrong here?

Thank you for any help or guidance you can offer! Can provide reproducible code privately if needed. 

Best,

Storm

[Storm Crews]

If helpful for additional context, please see here a plot of the AKDEs alongside their respective locations, as well as the CI output of the overlap function applied here.

Best,

Storm