Hendrik Posted November 10, 2018 Share Posted November 10, 2018 Hi Eldar, I am currently helping a friend to analyse data from mottled petrels breeding on Codfish Island, South of New Zealand. She used migrate technology tags, the same model (C-250) as I used on skua. Her light data looks very clean and she has a few days of 'rooftop' callibration from the study site. When running the model, it always comes out with a track that makes sense given what is known from previous tracking of the species. However, there are no confidence areas displayed on the lat lon graph (looking at the data, Min.lat, Max.lat and Medianlat/Lon values are all the same for each location). Switching "known last" to False makes them appear in the very end of the track, but they are still missing for the rest. We have tried different individuals, but the problem persists. I also tried using a calibration from my skua data (Chatham Islands), but the same problem occurs. Have you had a similar problem before, and do you think it could be due to issues with the callibration? I would be very thankful for any suggestions. Thanks a lot in advance! All the best, Hendrik UPDATE: I have tried doing the calibration from onbird data as estimated by the 'plot_slopes_by_location' function. This now results in confidence intervals on the tracks throughout. Could it be that the onbird data is simply better suited in this case? Many thanks in advance! Hendrik Link to comment Share on other sites More sharing options...
Eldar Posted November 15, 2018 Share Posted November 15, 2018 Hi Hendrik, Yes, we had this issue already with some other roof top calibrations. I must admit that on bird works better. The calibration consists of the two main parameters - slope (mean+- sd) of measured irradiance vs expected irradiance. Mean should be the same, but the sd differs between the roof top and the on bird. Hope this helps, Eldar Link to comment Share on other sites More sharing options...
Hendrik Posted November 25, 2018 Author Share Posted November 25, 2018 Hi Eldar, thanks a lot for your comments. I will keep the on bird calibration in mind for future analysis! It improved things a lot with this data set! Again, many thanks and have a nice Sunday! Hendrik Link to comment Share on other sites More sharing options...
Recommended Posts
Archived
This topic is now archived and is closed to further replies.