GOTM simulations can be forced by observational data, that is actually measured or artificially computed data. The options are described in the associated namelist file, obs.nml, as well as in section 6, Working with observed data in GOTM, of the documentation [GOTM scientific documentation, http://www.gotm.net/index.php?go=documentation].

For most observational forcings, three _methods are implemented:
0: zero
1: constant, analytical or tidal
2: from file

This HOWTO describes the format of the input files which need to be provided if _method = 2. Besides the data itself, the ASCII files have to contain the time stamps and, for profile data, the associated depth information.

The time format is yyyy-mm-dd hh:dd:mm, i.e. with date and clock time separated by a space and the three date components as well as the three clock time components separated by single characters. The time stamps do not need to be equidistant, i.e. the time step, dt, of the observational data is allowed to vary.

Files for non-profile data (0d) comprise a table of two or more space- or tab-separated columns. The first column contains the time stamps and the following columns contain the data.
Example:
1999-07-05 01:38:24       1.6411
1999-07-05 01:42:44       1.7143
1999-07-05 01:47:03       1.7847
<snip>
[GOTM devel test case liverpool_bay/zeta.dat]

Files for profile data (1d) comprise time-based data sets, each of which consists of a single line, containing the time stamp and depth metadata, followed by a table of two- or more space- or tab-separated columns. The depth metadata consists of the number of entries in the vertical and an integer representing the direction of the entries, i.e. 1 for upward (from bottom to surface) and any other integer, commonly 2, for downward entries. It is allowed to vary from data set to data set, i.e. over time. The first colum of the table contains the depth (negative, zero at top) and the following columns contain the data. The depth does not need to be equidistant.
Example:
1976/04/06 00:00:00  50 1
   -143.55    35.104939
   -140.65    35.104012
   -137.75    35.103081
<snip 45 lines>
     -4.35    35.060390
     -1.45    35.059460
1976/04/07 00:00:00  50 1
   -143.55    35.104939
<snip>
[GOTM devel test case flex/sprof.dat]

The following list summarises the type of data (0d: non-profile, 1d: profile), the number of data columns (exclusive of time for 0d or depth for 1d) and the respective variable(s) for each input file:

s_prof_file
	1d; 1; salinity
t_prof_file
	1d; 1; temperature
ext_press_file
	0d; 3; ???
	data interpreted as ext_press_mode:
		0: surface elevation gradient
		1: current meter measurement at a given height
		2: vertically averaged current speed
int_press_file
	1d; 4; dsdx, dsdy, dtdx, dtdy (horizontal salinity and temperature gradients)
extinct_file
	0d; 3; ??? (extinction coefficient?)
	note: only if _method = 0
w_adv_file
	0d; 2; w_height, w_adv (depth, vertical advection velocity)
zeta_file
	0d; 1; sea surface elevation
wave_file
	0d; ?; ??? (wind waves' significant wave height, mean zero-crossing period, mean direction?)
vel_prof_file
	1d; 2; u, v (horizontal velocities)
e_prof_file
	1d; 1; dissipation rate
o2_prof_file
	1d; 1; oxygen
	data in o2_units:
		0: saturation (%)
		1: mmol/m3
		2: ml/m3
bio_prof_file
	1d; 4; ???? (biogeochemical profile, NPZD?)


version 2011-06-17 by Elisabeth Schulz (IOW)