As part of the Natural Heritage Program of the West Virginia Division of Natural Resources, I am interested in how we can reasonably determine from Motus detections we download via R/RStudio whether they are within geopolitical boundaries when a Motus station or antenna signal reach are near or crosses into other states? If the detection is made on an omni we can reasonably assume, but for some of our stations where signals can travel across state bounds from directional antennas, it’s not so clear, and so far all detections (at least from Hanging Rock (7878) as an example) seem to be from directional 5 and 9-element Yagis. I’ve gone through all the R Motus online exercises and don’t find the data intuitive to work with or to easily answer this question. Any guidance, R code, how to utilize signal strength, etc. would be appreciated! Determining this is not only important for the Natural Heritage Program but for the WV Bird Records Committee.
I would say that if the station is in the state, detections are in the state. It is likely a very rare detection where the tag is within 14km and not in the state. I cannot think of any combination of parameters to use that would not wrongfully exclude as many tags as are wrongfully included.
Also keep in mind that directionality of yagis is imprecise. If you look at the sensitivity pattern, tags may be detected on side lobes or even from behind.
I know I normally just lurk in this group, but I wanted to quickly chime in. I think caution does need to be used (for things like state records committees) depending on the location of the receiver and the animal being tracked. For migrating shorebirds, it was not uncommon to receive simultaneous or near-simultaneous detections on receivers with directional yagi antennas over 50 km away (and sometimes we did get these detections up to 150 km away on a cleaned/filtered Motus dataset). Shorebirds are high fliers during migration and receivers meant to track them are often in open coastal areas which increases detection range.
That being said, Mack, to my knowledge, there is currently not a straightforward/simple way to approximate the distance of the detection from a receiver.
I am unsure how state records committees take into account spatial error of other tracking technologies (e.g., Argos data). Maybe there is already a precedent (I would be interested to know that out of curiosity)
With the digital (i.e., CTT tags at 434 MHz) we regularly have instances where towers 80-120km or more apart from each other simultaneously detect the same transmission from the same tag. Therefore, the tags could be well beyond the border if the station is positioned near said border. I would recommend that detections from only a single directional antenna that points into an adjacent jurisdiction be flagged as “possible” for the state where the tower is located, but not a confirmed detection for the state.
This is clearly going to be a rare occurrence for species of concern, crossing near a boundary that matters, and this being the only record for that species. Most people are not going to dig into the antenna data to see which antenna is pointed out of the area. You could just ignore data from stations near boundaries, but that seems like the wrong kind of conservative.