7 Feb 1999

Data Title: Fitted Ice Position, Velocity files from Ocean Hut GPS
PI McPhee
Version 1, dated 7 Feb 1999
Date available via anonymous FTP 7 FEB 99
Files: 12 ascii files, one for each month: named tpv_oct.txt to tpv_sep.txt
Contents:
Each file contains the following columns
  1. time in format yyyymmddhhmm (UT)
  2. time in decimal days of 1997 (day 1.5 = noon UT 1 Jan 97)
  3. latitude degrees
  4. longitude degrees, west negative
  5. eastward velocity, cm/s
  6. northward velocity, cm/s
  7. interpolation indicator (1 if enough gps fits in the 12-hr fit
     window to perform the complex demodulation fit, 0 if lack of data
     requires linear interpolation of adjacent complex demodulation
     coefficients).

Method: 1 min average GPS positions were recorded continuously at the
  Ocean Hut, some distance from the ship. Gaps were filled and/or
  supplemented by 5 min average positions from the Met Hut, supplied by
  Ola Perrson. Data were quality checked, then fitted via a complex
  demodulation scheme (described by McPhee, M.G. 1988. Analysis and
  prediction of short term ice drift, Transactions of the ASME, J. of
  Offshore Mech. and Arctic Engineer., 110, 94-100), using a moving
  12-hr window to calculate fit coefficients every 3 hr. Time series
  are evaluated at half hour intervals.

Errors: A rigorous error analysis has not been done. Caveat emptor. The
  complete file of gps fixes used in the fitting is available by
  special arrangement: contact miles@wolfenet.com.
  There are significant data gaps from the breakup periods in February
   and March-- use column 7 to locate them.
  Interpolation in gaps is applied to complex demodulation phasors,
   not velocity, hence the gaps may exhibit possibly spurious inertial
   oscillations (see, e.g., the record from day 82-85 of 1998)