THE NODC STANDARD SUBSET OF SHIPBOARD ADCP DATA The NODC provides shipboard ADCP data in two forms: 1) a high-resolution set in CODAS binary format consisting of currents and ancillary parameters at the sampling interval with which the data were recorded and processed and 2) a standard subset of absolute currents at hourly and 10 m intervals which is available as ASCII or NetCDF. If absolute currents are not available due to lack of navigation, then relative currents are provided. If NODC receives data sets that are not of high resolution, then the standard subset may be of a slightly different timing sequence than hourly and the depth spacing may be other than 10 m. For instance, if a data set is received as gridded averages in space, then a nominal time will be assigned. Similarly, if a data set is received with 8 m depth spacing, then this interval is used in the subset. The standard subset includes the hourly means and standard deviations of ship velocity components and the transducer temperature, when available. A description of the methodology and format of the standard subset is given below. METHODOLOGY The standard subset will consist of absolute currents if the ship navigation was satisfactory. Otherwise the currents will be given relative to the mean over a reference layer, usually 50 to 150 m. In the vertical, each velocity profile is linearly interpolated to a regular grid at integral multiples of 10 m. The shallowest level is the first multiple of 10 m that is greater than or equal to the depth of the first good bin of the recorded data, typically a function of the instrument configuration and the transducer depth. The deepest level is that at which at least 50% of the ensembles over the course of the cruise have a Percent Good greater than 30%. For each depth bin, the Percent Good is the percentage of individual pings in an ensemble with valid data. As noted above, for data sets received at NODC in a reduced form (ie., not high-resolution), the original depth interval is maintained. In time, averages are calculated for one-hour intervals centered on the hour. The time column in the ASCII subset gives the mean of the times of the ensembles that fall into the interval. Hence the mean may be slightly off the integral hour, even when ensembles are evenly available over the entire hour. The mean time may be even less centered in the case where data are only partially available over the interval because instrument recording was turned off (e.g., at port calls). In the case where data are totally unavailable over the entire interval so no mean time can be calculated, a "placeholder" entry is made and assigned a time value exactly at the middle of the interval, i.e., on the hour, followed by bad value flags for the longitude, latitude and velocity fields. It is left to the user to manipulate the time column if a truly even time grid is desired. At each level, a bad value flag is given if less than 50% of the ensembles in the hour have valid data. Bad value flags (99999) for velocity components also occur at depth layers where less than 50% of the ensembles in the interval contribute to the average because of low percent good and other editing criteria, or in the case of absolute currents, occur over the entire depth range because ship velocity was not available as a reference. The transducer temperature is provided as a mean and standard deviation over the hour. The ship velocity is provided as means and standard deviations of the components over the hour and is derived from the final absolute ship velocities of each ensemble as stored in the CODAS high-density database. The ship velocity information allows one to see at a glance if the ship was on-station, underway, or both over the hour. FORMAT DESCRIPTION ====> HEADER RECORD sac_id=xxxxx yr_base=yyyy start_lev=ddm num_lev=ddd type < > denotes optional e.g. sac_id=00001 yr_base=1993 start_lev= 20m num_lev= 45 absolute depth_int= 8m where sac_id is a unique identifier for the cruise (NODC bookkeeping) yr_base is the year inwhich the cruise begins start_lev is starting level (meters) num_lev is the number of levels type means either "absolute" or "relative" currents depth_int is the depth interval or spacing between levels and is left blank if equal to 10 m (default) If a data set is received at NODC in a reduced form (ie., not high- resolution), then a comment will be appended to the header to define the timing scheme and the depth spacing if it is other than hourly or 10 m. A complete documentation file which describes the cruise, instrumentation, processing, and other pertinent comments for the given cruise will be available separately. ====> DATA RECORD The format is designed for ease of application; therefore, the date-time and positions are provided in decimal form as a single field. Each data record will be fixed length although the length may vary from cruise to cruise due to differences in the number of depth bins. The order of each record is as follows: With a template such as: ddd.fffff dddd.dddd ddd.dddd tt.t tt.tt suu.u su.uu svv.v sv.vv uuuuu vvvvv .. e.g. 350.00002 157.9365 6.9120 28.9 0.01 -4.6 0.11 -3.4 0.09 419 177 .. field template comment ddd.fffff: decimal day; e.g. 0.50000 is 1200 UTC on January 1 of the year given by yr_base; always present dddd.dddd ddd.dddd: longitude and latitude decimal degrees; lon positive east, lat positive north; missing flag= 1E38; e.g. -145.1667 is 145 deg 10' W tt.t tt.tt transducer temperature mean and standard deviation (deg C); missing = 1E38 suu.u su.uu svv.v sv.vv absolute ship velocity components means and standard deviations in m/sec; missing = 1E38 uuuuu vvvvv current components in mm/sec; bad or missing = 99999; e.g. 121 is 12.1 cm/s