Jump to content


Photo

FlightR and dateline


  • Please log in to reply
8 replies to this topic

#1 Hendrik

Hendrik
  • General Members
  • 14 posts
  • Auckland,
  • New Zealand

Posted 30 March 2016 - 03:33 PM

Hi there,
 
I am new to this forum. I am a phd student at Auckland University, New Zealand. One of my chapters is on wintering patterns of brown skua from the Chatham Islands using light-level geolocation. I was delighted to see Eldar's publications on FlightR early this year and started try and re-analyse my tracks using the FlightR package.
 
Unfortunately, I ran into a potential problem with some of the tracks.
My study site, the Chatham Islands, are as close as you can get to the international dateline (it runs right between the Chathams (Lon -176) and New Zealand (lon +170). I think I ran into a potential problem when a bird crosses the dateline: when setting up the spatial extent for the model, I would want to set the left boarder of the grid at a positive longitude in New Zealand (e.g., lon +170 ) . The right boarder would be a negative longitude (roughly –150, east of the Chathams). Unfortunately, when setting the grid in the spatial extent, R won’t let me do that. Has anyone encountered this problem before and do you know a way around it by any chance?
 
I was hoping that the problem can be fixed by changing the central meridian within the model.
 
Thank you very much for your support!
 
All the best,
 
Hendrik


#2 ebridge

ebridge
  • Society Members
  • 27 posts

Posted 31 March 2016 - 08:58 AM

Interesting problem. As a work around can you just shift all of your times forward 5 hours and calibration locations east 60 degrees? Just avoid the dateline altogether. Let me see if I can get Eldar to chime in.



#3 Eldar

Eldar
  • General Members
  • 47 posts
  • Netherlands,
  • Netherlands

Posted 04 April 2016 - 08:46 AM

  Hi Hendrik,

 Thanks for posting this issue.

 I have corrected make.grid function. Plotting is still a bit ugly but the rest should work fine.

Try

library(devtools)
install_github("eldarrak/FLightR@0.3.9") # note the version
library(FLightR)
A<-make.grid(170, -90, -170, 90, , c(-100, 100)) 

 Cheers,

   Eldar



#4 Hendrik

Hendrik
  • General Members
  • 14 posts
  • Auckland,
  • New Zealand

Posted 18 April 2016 - 12:44 AM

Hi Eli and Eldar,

 

thank you very much for your support with this, Eldar has solved the problem and its working perfectly fine now. Thank you both for your support!

 

All the best,

 

Hendrik



#5 yuhash

yuhash
  • General Members
  • 3 posts
  • Vancouver, BC,
  • Canada

Posted 02 June 2017 - 04:24 PM

Hello,

 

I am a GIS analyst with the Canadian Wildlife Service and am analyzing seabird GLS data (Ancient Murrelets) whose migratory routes cross the full E-W expanse of the Pacific Ocean (the first recorded for a seabird!).

 

I am currently re-analyzing the data in v0.4.7 and am wondering if the update provided in this post for FlightR v0.3.9 is included in this version. Is it possible to move the central meridian in the make.grid function such that the spatial reference isn't split within the Pacific Ocean?

 

Also, can anyone explain what the c(-100,100) parameter in the make.grid function does in the case described?

 

Thank you so much in advance!



#6 Eldar

Eldar
  • General Members
  • 47 posts
  • Netherlands,
  • Netherlands

Posted 12 June 2017 - 01:25 AM

   Hello yuhash,

 v0.4.7 has the same grid definition, so for try e.g. left=-90, right=90 and look at the plot that FLightR will produce, adjust boundaries if needed.

 In the additional parameters there are distance.from.land.allowed.to.use and distance.from.land.allowed.to.stay. 

 

distance.from.land.allowed.to.use sets limits for concurrence of you animal at twilight. Negative values are for the inland and positive for offshore. c(-100, 500) would mean that your animal cannot appear more than 100 km inland and more than 500 km offshore. distance.from.land.allowed.to.stay is the same but limits probability of being stationary (stay in the same place between two twilights). 

I would recommend running without limits first to check whether results are meaningful.

  Cheers,

  Eldar



#7 L Halpin

L Halpin
  • Society Members
  • 10 posts
  • Vancouver, Canada,
  • Canada

Posted 28 June 2017 - 10:02 PM

Hi Eldar,

 

I think Yuhash was referring to the c(-100, 100) in the work around for the date line crossing issue that you referred to earlier in this post:

 

A <-make.grid(170, -90, -170, 90, , c(-100, 100)) 

 

My understanding is that this part is not related to behavioural masks. Could you explain what the parameter “c(-100, 100)” does in order to assist with crossing the dateline issues?

 

Thanks!



#8 Eldar

Eldar
  • General Members
  • 47 posts
  • Netherlands,
  • Netherlands

Posted 29 June 2017 - 07:32 AM

 

 

My understanding is that this part is not related to behavioural masks. Could you explain what the parameter “c(-100, 100)” does in order to assist with crossing the dateline issues?

 

(-100, 100) is behavioural mask here and is not related to dateline at all.

make.grid functions has following parameters:

make.grid(left = -180,
          bottom = -90,
          right = 180,
          top = 90,
          distance.from.land.allowed.to.use = c(-Inf, Inf),
          distance.from.land.allowed.to.stay = c(-Inf, Inf),
          plot = TRUE,
          return.distances = FALSE, 
          probability.of.staying = 0.5)

So number 6 is distance.from.land.allowed.to.stay

 

    Hope this helps. Do not know why I specified it in the example above.



#9 yuhash

yuhash
  • General Members
  • 3 posts
  • Vancouver, BC,
  • Canada

Posted 29 June 2017 - 01:21 PM

Thank YOU both for the clarification! It's very much appreciated! :)






0 user(s) are reading this topic

0 members(s), 0 guests(s) and 0 anonymous member(s)