The next step is to transfer quasi-observation data files with various formats into the QOCA data files with the uniform QOCA qob format. The software to perform such a job is mk_qob. To run the mk_qob, you need to prepare a driver file. The driver file of many QOCA software packages has similar style. The first column is used for the comment symbol. If the first column is filled by any non-blank character, this line is considered as the comment line. The colon ":" is used as the separator. The left side of the colon is the command. The right side of the colon is the argument. Please look at the template of the mk_qob driver file by clicking the above "Preview" button with the "driver file" option.
When you have the driver file ready, the remaining procedure is very easy. The runstring of mk_qob is:
mk_qob [driver file]
If you simply type "mk_qob" without the driver file name, the program will output the above runstring on the screen to remind you the required arguments.
Arguments of the driver file:
in_list file: input file which gives a list of names of the input and output qob data files. Note the input and output files are not necessarily in the same directory of the driver file. You can specify the absolute or relative path of the i/o files (Please see the template). The first line is the total number of the input files. Starting with the second line, the first field is the input file name, the second field is the output file name. The fields are separated by a space.
From version 1.24, mk_qob accepts compressed file format for GIPSY stacov file and GAMIT h-file. If in the mk_qob.list file, user gives:
2 Hfile1.Z Hfile1.qob Hfile2 Hfile2.qobthe mk_qob automatically recognizes that the Hfile1.Z is the compressed file and uncompress the file first.
rename file: This is an optional argument. Please see the template of the rename file. The first line is the number of sites to be renamed. After the first line, the first field is the original site name, the second field is the new site name. The following three fields are the off-set for x, y, z meters. If you comment out the option of the rename file or keep the argument as blank, the software will use the default value (no rename site).
unused site list (drop_site) file: This is an optional argument. The default is to use all sites. The first line of the file is the number of dropped sites. Starting with the second line, each line contains one station name.
deconstraint file: This is an optional argument. The default is no deconstraint. This file will provide information about your desired level of the loose constraints. If the original quasi-observation file is tightly constrained, we want to loosen the constraint so that the transfered quasi-observation data become loosely constrained. In this case, we need to activate this option. Since this option requires some advanced knowledge, we will introduce the details in the advanced class.
swap_byte: This is an optional argument. The default is "no". When the input file is a binary format file (GLOBK glf and BERNESE neq file), the byte order is platform-dependent. In general, the byte order in Linux and PC machine is in reverse order of that in SUN and HP machine. When the binary file is generated in Linux machine and we want to transfer it to QOCA format file in SUN or HP machine (or vise versa), we should activate the swap_bype option. Currently, only glf format swap_byte function is available. We will add swap_byte function to neq file in the future.
in_opt: xyz or uvw or xyzuvw. The xyz denotes to transfer the coordinate quasi-observation only. The uvw implements a transfer of the velocity quasi-observation only. The xyzuvw transfers both coordinate and velocity quasi-observations. The default is xyz.
scaling factor: The default is no scaling (1.0). We do not activate this option in this released mk_qob version, because we can rescale the qob covariance matrix easily in the QOCA main program.
output qob type shift (it_shift): This is an optional argument. Currently it can be used only for SINEX file. SINEX file usually contains site positions and velocities. If we want to output baseline vector qob file insead of site position qob file, we can activate this option. In this case, we put word "baseline" like the following:
data type shift (it_shift): baseline
The default is not shift the data type.
in_style: This is a mandatory argument. You must specify the format of your input data file. The accepted formats are: gipsy, glf, sinex, qoca, fonda, gamith, bernese, gsi_bern. The gipsy stands for GIPSY stacov file; glf means GLOBK (and GAMIT) output binary h-file, sinex is the SINEX format file, qoca is the QOCA qob file, fonda is the FONDA output data file, gamith is the GAMIT output ascii h-file, bernese is the BERNESE software output neq file (either binary or ascii format), and gsi_bern is the BERNESE software output neq file but with the special site name convention of GSI. Currently, the fonda option is not included in this mk_qob release code. Please look at the templates to get an impression of the various formats (glf and neq format is a binary file, so that it can not be displayed. The size of gamith format file is too large, we do not display it either).
infmt: The default is to use the embeded formats. In most cases, we should use the default format. Only in some special cases, will we need to specify the new format explicitly.Current mk_qob does not accept mixed format. That means if your specified in_style is gipsy, all your input data files in the in_list file should be in gipsy stacov format. If you have some data files with other formats, you should put these data files in another in_list file and create another mk_qob driver file to run the mk_qob separately.
In the output qob file, the first line is the time tag of the quasi-observations. Current QOCA accepts the following time tags:
ymdh: year, month, day, hour, minute, second.
year: year in decimal format. Since the year is not a uniform concept (some year has 365 days, some year has 366 days), old QOCA versions used an artificial 'year' with uniform 365.25 days. The time tag was counted by the days from 1990.0. This time tag has uniform 'year', but is not consistent with convential time tag from other programs. From version 1.28, the time tag is changed to year+(day of the year)/(year length). Hear the year length = 365 days for regular year and 366 days for leap year. QOCA does not distinquish if the ob files are from early mk_qob or later mk_qob. It is recommended to re-run mk_qob for these early qob files to get consistent time tag.
juld: Julian day.
The second line of the qob file contains 2 entries (some version has three entries): station number and pmu parameter number (third entry is satellite orbit parameter number).
The next lines are station names (one station one line) and pmu parameters (one parameter one line) (if the orbit parameter number is not zero, there will be orbit parameter number, one parameter one line).
After these lines will be one line, which contains observation type, station number, pmu number (some version has orbit number). Here the observation type = 33 represents station coordinates in Cartetian frame (xyz), type = 35 represents station coordinates and velocities in Cartetian frame (xyz and vx, vy, vz).
The rest lines are quasi-observations and their covariance matrix.