107 The following parameters allow various adjustments of the data in terms of refer-ence frames, grid patterns, and temporal averaging.
reference: ADCP pro les are stored in CODAS as pro les of velocity relative to the transducer. They are converted into absolute velocities as needed when plotting and analyzing the data. (Note that only ONE way of referencing is possible). The options as they appear in the control le are shown in Figure 40.
reference: { choose only one of the following: }
< reference_bins <5> to <20> | none |
final_ship_ref | nav_position | nav_velocity | bottom_track ]
Figure 40: Options for the reference: parameter.
Several reference methods are available:
1. reference_bins: Each pro le is referenced to the mean velocity over the speci ed bin range. A bin range over which good data are available is normally selected, such as bins 5 - 20 (for 8 m depth bins representing the layer from roughly 50 to 170 m.
2. final_ship_ref: The ship speed calculated from combined ADCP and navigation data. We consider this our best estimate of the ship speed.
3. nav_position: In this and in the following option, we use navigation data that were recorded together with the raw ping data in the pingdata le (variables loran.lat, loran.lon, loran.vel and loran.dir in the naviga-tion structure). For nav_position, the reference velocity is calculated by dierencing the position data. Note this option is rarely ever used and only makes sense if navigation positions were recorded at the end of the ensembles.
4. nav_velocity: Uses loran.vel and loran.dir for referencing. It is rarely used.
5. bottom_track: Uses the recorded bottom track data.
rotate: Allows for rotation of the velocity components. There are two possi-bilities: (1) the data may be rotated by a speci ed angle (the sign convention is such that the the coordinate system is rotated counterclockwise by the speci ed
108
amount) (2) the data are rotated into ship's coordinates, with the zonal (x) direction becoming starboard, the meridional (y) direction becoming forward.
For example, this would facilitate analysis of the Gulf Stream o Hatteras if the ship is steaming with the current and the y-axis points in the forward direction.
The navigation output (the.nav le) is not aected by the rotate parameter.
ctd: dbname: In this option, the default ADCP bin depths calculated from the default sound speed (1470 m/s) will be recalculated based on the in situ sound speed pro le. The sound speed is calculated from the speci ed CTD database the CTD pro le on or after (in time) to the current ADCP pro le is selected. dbname: refers to the name of the CODAS database of CTD data.
A separate manual for loading CTD data into a CODAS database is available upon request from the authors.
A CTD database must also be speci ed if any vertical regridding based on density intervals is selected (see below).
regrid: Refers to regridding the ADCP pro le in the vertical. This option contains many suboptions as show in Figure 41. Choose one option from each suboption delimited by greater-than and less-than signs (<..>).
regrid:
< integrate | interpolate | centered_average | average | per_grid_interval >
< depth | svan | pot_svan | sigma_theta | sigma_t >
{ one of the last 4 only if the ctd: option has been selected }
< grid number= <50> origin= <16> increment= <8> |
grid_list number= < 5> boundaries: <20> <40> <60> <80> <100> > ]
Figure 41: Options within the regrid: parameter.
1.
statistical method for regridding in the vertical
< integrate |interpolate |centered_average | average |per_grid_interval >
Several of the regridding schemes are based on vertical integration, such as
integrate, average, centered_average, and per_grid_interval. For the last three options, the integral is divided by a suitable length. For av-eraging, this length is always the depth interval over which the average was taken. For per_grid_interval, the length is the grid variable interval. If the variable is depth, then this is the same as the average. If the variable is other than depth (i.e. one of the density options) per_grid_interval
109 divides the depth integral by the grid variable interval so the result is the transport per unit width with respect to the grid variable.
average, per_grid_interval, integrate, and integrate_top all start by integration from one grid point to the next. For integrate_top, the rst grid point is set to 0m, so it extrapolates to the surface. averagethen divides this depth integral by the depth interval over which the average was taken.
centered_average and interpolate put the grid point in the mid-dle of the increment while the rest average over the increment.
centered_average is the average value in a region surrounding the grid point. interpolateis simple linear interpolation from the original grid to the user-speci ed grid.
Examples ofregrid:options can be seen as used withcontour(Appendix A.2) and with the vector (Appendix A.5) output options.
2.
vertical-variable
< depth |svan |pot_svan |sigma_theta | sigma_t >
The depthis the bin depth array data. One of the last four options can only be chosen if the ctd: option was applied.
3.
vertical regridding scheme
< grid number= <50> origin= <16> increment= <8> |
grid_list number= < 5> boundaries: <20> <40> <60> <80> <100> >
You specify the new vertical grid either by the three valuesgrid number,
origin, andincrement, or by enumerating theboundariesand specifying how many as grid_list number.