@prefix ns0: . @prefix skos: . ns0:altLabel [ ] ; skos:broader ; skos:prefLabel "GUSREX"@en ; skos:inScheme ; skos:definition """Gulf Stream Recirculation Experiment (GUSREX) and Line Experiment SOFAR Float Data 1980-1982 by M.A.Kennelly & T.K. McKee This also pertains to Gulf Stream (GS), Long Range (LR) and Ring (RI) Experiments DATA PROCESSING Processing at the University of Rhode Island Initial data processing and float tracking were done at URI under the supervision of H. T. Rossby. This consisted of subdividing the ALS records into individual float files. The signals for an individual float were extracted from the ALS data, and the floats were tracked in two steps. First, the three ALSs which best surrounded a float geographically were used to track it hyperbolically, giving a series of positions and the time that the float emitted the tracking signal. These signaling times were used to obtain an initial offset and drift rate for the float's clock. Then, the best pair of receivers were used to track the float with circular (range-range) navigation using the clock corrections and speed of sound corrections. The tracking procedure during GUSREX was identical to that used for previous SOFAR float experiments and described by Spain, O'Gara, and Rossby (1980). The three daily times of arrival (TOA) of each SOFAR float's acoustic signal were low-pass filtered and smoothed using a least squares polynomial fit averaged over 11 observations (3 2/3 days) which corresponds to a half power of 1.5 days. Float positions were obtained from the smoothed TOAs using range- range tracking. The final float trajectories were then smoothed with a filter similar to that used on the TOA's. A "master position" tape containing indi- vidual float trajectories smoothed to 3 positions per day was prepared and sent to Woods Hole Oceanographic Institution (WHOI) and this was used for our sub- sequent processing and data reduction. For each position, a smoothed velocity and 48 hour average of temperature and pressure are included. Temperature and pressure, telemetered from the float on alternate days (Spain et al., 1980) appear with the positions for those days. These data have been presented in two informal reports prepared by the URI SOFAR Float Group (Line Experiment FLoats 1980-1981, Line Experiment FLoats 1980-1982, Preliminary Data Report). Processing at Woods Hole Oceanographic Institution In order to use the float data to make reasonable calculations of eddy kinetic energy, variances, & vertical velocities, the data were careful- ly reviewed for correctness. Tests on various methods of smoothing, subsamp- ling, and filtering the data were performed to determine how the data set would be processed. Analysis of the results indicated that the best basic data set would include one position for each day, since the smoothing already done on the tracks had eliminated frequencies higher than a day (inertial and tidal). The velocity components would be derived from the daily fixes using a cubic spline curve fit. This method produced velocities that are different from the URI velocities which are smoothed independently of the positions. The 48 hour average temperature and pressure, transmitted by the float on alternate days were selected rather than the smoothed temperature and pressure provided with the URI format. Basic processing was accomplished using a series of routines written by Roger Goldsmith and Terry McKee, known as the FLOATER programs. The initial program, REFORM, reads and reformats the data into VAX ASCII files, one for each float, in FLOATER format. In this form, the data is easy to access, man- ipulate, edit and back up. The raw files were then passed through a sub-sampler, SSFDIF, that subsampled the first of the original 3 daily positions and recal- culated the east and north components of velocity based on a 24 hour interval. First differences between consecutive temperature and pressure values were calculated. Listings of these derived values were produced along with prelim- inary trajectory and time series plots. These were reviewed to identify and eliminate erroneous data. Unreasonably high speeds were used to identify bad positions. Radical changes in temperature that were not accompanied by a similar change in pressure (or vice versa) usually indicated a bad value. Temperature and pressure values that drifted outside the range of the sensors were also defined as bad. These points were removed by an editing program, FLEDIT, and flagged as gaps in the data. Gaps of less than ten days duration in position, temperature and pressure were then linearly interpolated. Daily values of temperature and pressure were interpolated from the bi-daily values recorded. Files with gaps of greater than ten days in position information were broken into sub files (GU109, GU154). These series were plotted and reviewed once again for incongruities. Program FFSPLINE then fit a cubic spline to the float positions and recomputed the east and north components of velocity based on the bias and slope of consecutive positions."""@en ; a skos:Concept . skos:prefLabel "G - I"@en ; a skos:Concept ; skos:narrower .