Seals were captured and tagged in April or May of 2007 and 2008 following the methods of Jeffries et al. (1993) at three sites: Padilla Bay, Bird/Belle Rocks, and Protection Island. In 2009, seals were captured on Protection Island. In 2007-2008, animals were tagged with time-depth recorders (TDR; Wildlife Computers, model Mk-9 or Mk-10F) and satellite tags. The TDR tags were placed on the dorsal midline of the animals and the satellite tags were placed on top the head. In 2009, animals were tagged with a combined satellite-linked TDR and Fastloc GPS receiver (Wildlife Computers, model Mk10AF). These instruments were epoxied to the animals on the dorsal midline so that satellite tags would be exposed to the air when the back of the seal reached the surface.
Position transmissions were received via the ARGOS satellite network. Tags transmitted locations daily. TDR sensors were programmed to sample every 10 seconds. TDR tags were equipped with a VHF transmitter to allow for the device to be recovered when it was shed during the animal's annual molt.
The Argos system provides 2 position estimates (lat/lon and lat2/lon2). Argos usually picks the correct lat/lon pair (of the two it generates), but occasionally it does not. When working with these data, one of the first steps is to check the lat/lon pairs to see if swapping out the lat/lon pair for the lat2/lon2 pair improves the data and is more biologically reasonable.
The Argos positioning system uses the following system for classifying location quality. These codes are used in the loc_q_flag column. Standard locations are those with > 4 uplinks from the tag; auxiliary locations are those with 4 or less uplinks from the tag.
loc_q_flag codes (according to Ward et al.):
Standard locations:
3 = 68th percentile predicted accuracy < 150 m
2 = 68th percentile predicted accuracy 150 - 350 m
1 = 68th percentile predicted accuracy < 1,000 m
Auxiliary locations:
0 = 4 uplinks, with > 1,000 m predicted accuracy
A = 3 uplinks, with no predicted accuracy
B = 2 uplinks, with no predicted accuracy