* README for the GAMIT/GLOBK standard example

*   Last updated by rwk 180604 09:40 UTC

This example serves the dual purpose of testing your installation

and guiding an new user through the basic steps to get time 

series and velocities from GAMIT/GLOBK processing.  It replaces

the Southern California example from earlier releases in order 

to add GNSS other than GPS.  If you are an experienced user,

you may want to simply execute the batch file 'runtest' to 

complete all steps without intervention in about 90 minutes.

In fact, running just one day or one GNSS for 2018 will likely

confirm a correct installation.  If you are a novice, however,

we strongly recommend that you execute each of the commands 

manually and peruse the results until you understand what that 

step accomplished.

The example is set up to 1) download RINEX 3 files from CDDIS for

10 continous stations in Europe and environs for 3 days in 2016, 

2017, and 2018 (note that if long names were used for the RINEX 3

files, they will be renamed to the older short-names as part of

the download); 2) conduct phase processing (sh_gamit) for GPS-only 

or GPS, Glonass, Beidou, and Galileo; 3) compute and plot daily

repeatabilities (sh_glred) for each year and combine the daily

h-files (.glx) into a single H-file for each year (.GLX); and

4) compute repeatabilities and velocities for the three years 

together (globk/glorg).  The only input file provided in the 

example is sites.defaults, which is always specific to the network

you are processing.  All of the others are copied or linked from

gg/tables and will work for most networks. 

The structure established by the example has three GAMIT processing 

directories, named (arbitrarily) by year (/2014, /2016, /2018), 

each of which has below it a /rinex, /tables, and /gsoln directory 

specific to that year.  At the top level there is a processing (/vsoln)

directory and a /tables directory for the multi-year GLOBK solution.

The steps described assume that you have downloaded and linked the

ocean tidal loading grid (e.g. otl_FES2004.grid, linked to otl.grid

in gg/tables) and have internet access while processing; if you do 

not have these, see Notes 3 and 4 below.  In directory /check_files 

are saved copies of the q-files and .org files for each day, and 

the .org files and postscript files for the multiyear repeatablities 

and velocities. 

Before you start, make sure that you have constructed the paths and

aliases described in the installation README.  The example may be run 

from any directory on your system, preferably the place you intend to 

process your own data, not under gg.  

STEP 1:  Run GAMIT for the three days from 2018.

In the example/2018 directory type

    sh_setup -yr 2018  

The 2018/tables directory will now contain links to most standard 

files in gg/tables and copies of these files for process.defaults, 

sestbl., station.info (complete MIT version), and autcln.cmd.  The 

sites.defaults file was already in the directory and therefore not 

overwritten by sh_setup. 

Examine sites.defaults to note that it has  been set up to ftp from 

a remote archive (CDDIS by default) RINEX files for 10 IGS stations 

and 'xstnfo' is set to avoid any automatic update of station.info 

during processsing. Note also that the sittbl. copied from gg/tables 

is set up to impose moderate constraints on IGS core stations to support 

ambiguity resolution in GAMIT; the four included in the example are 

more than sufficient.  

Edit process.defaults to change the 'mailto' to your own email address 

to receive the sh_gamit summary file.  (If left null, it reverts to

'whoami', so this change may not be needed.)  Note that 'aprf', used

to initialize the lfile. for GAMIT points to itrf14_comb.apr, which

has been copied by sh_setup to the /2018/tables directory. 

If you have not downloaded or copied from an earlier distribution the 

(large) grid file for ocean tidal loading and do not wish to use it, 

change 'Tides applied' in the sestb. from '31' to '23'.

Construct a small, experiment-specific station.info file by using the 

following  procedure in the 2018/tables directory:

    sh_upd_stnfo -l sd 

will create 'station.info.new', using from the MIT station.info only 

the sites listed in sites.defaults. (This step will take a while since

the MIT global station.info file is so long.)  After checking, rename it 

to 'station.info' (overwriting the no-longer-useful MIT station.info).

(In your own processing, if you have stations that are not in the MIT

station.info file, you can add their entries from the RINEX headers

by runing in /tables 'sh_upd_stnfo -files ../rinex/*.18o'. (See Section

2.4 of the Intro manual for what to do if the RINEX header entries are 

non-standard.)                                     

GAMIT currently supports processing of each GNSS separately, with the day 

directories for each created with a single-letter added to designate the 

system, e.g. 095G, 095R, 095E, 095C.  If you wish to test full GNSS 

capability, type at the /2018 level

 sh_gamit -expt eura -gnss G -s 2018 095 097 -pres ELEV -orbit igsf -copt x k p -dopts c ao  >&! sh_gamit_2018G.log 

 sh_gamit -expt eura -gnss R -s 2018 095 097 -jclock sp3 -pres ELEV -orbit comf -copt x k p -dopts c ao  >&! sh_gamit_2018R.log 

 sh_gamit -expt eura -gnss E -s 2018 095 097 -pres ELEV -orbit comf -copt x k p -dopts c ao  >&! sh_gamit_2018E.log 

 sh_gamit -expt eura -gnss C -s 2018 095 097 -pres ELEV -orbit comf -copt x k p -dopts c ao  >&! sh_gamit_2018C.log 

(With bash use '> sh_gamit.log &' for the redirect.)

For GPS-only, omit the -gnss so that the day directory names will be the 

day-of-year only.

A summary file should be emailed to you as each day completes execution.  

Check these files for number of stations (6), Postfit RMS (3-15 mm, none 0), 

postfit nrms (~0.2), ambiguity resolution  and coordinate adjustments (< 30 cm).

For this large network, the 2018 ambiguity resolution is 80-90% for GPS and 

Galileo, 40-50% for Beidou, but near zero for Glonass due to poor orbits),

a condition also reflected in the sky plots in the /figs directory.  Allowing

the Glonass orbits to adjust (RELAX mode in the sestbl.) will improve the

fit signficantly.  To view the result of the fit of GAMIT's orbit model to 

the input orbit, see the rms files in /igs. To view the complete editing statistics 

see autcln.prefit.sum and autcln.post.sum in the day directories; for the 

least-squares adjustment, see qeuraa.[ddd].

Optionally, remove the x, k, and p files from the day directories to save space:  

 sh_cleanup -d 2018 095 096 097 -dopts x k p  

Notes: 

1. For large or complex data sets, the utility sh_get_times can be helpful in 

determining the days and session spans to be processed.

2. In creating station.info for your own experiments, it is important to check 

it after updating from the RINEX headers unless you are sure these headers are 

correct. In processing, the station.info entries always override whatever is 

in the RINEX header or the x-files.  An alternative way of creating entries 

for survey-mode sites is to use interactive program 'make_stnfo', then use

sh_up_stnfo to merge this file with the one created from the MIT station.info 

file for continuous sites.  The survey-mode file will have a shorter form of 

the station.info format, but this will be converted when it is merged with 

the continuous file, which should be listed as the reference (-ref) in the 

call to sh_upd_stnfo. 

3. The example is set up to use ocean tidal loading ('Use otl.grid = Y' in the 

sestbl.), which requires you to have previously downloaded into gg/tables an 

OTL file from the anonymous ftp directory on everest.mit.edu and to have linked 

this file to otl.grid also within gg/tables.  The IERS/IGS standard model is 

otl_FES2004.grid (730 Mb).  You may, however, substitute the smaller (45 Mb) 

otl_CSR4.grid, or you may turn off ocean tidal loading by setting 

'Tides applied = 23' and 'Use otl.grid = N' in the sestbl.  (Note that the 

links to the other grid and list files (met.grid, met.list, map.grid, etc.) 

can remain empty for running the example and for most processing.

4. If you want or need to run the example without having internet access while 

running, you can pre-load the RINEX, navigation, and orbit files into the

 /rinex, /brdc, and /igs directories, respectively.             

STEP 2: Run GLOBK to get repeatabilities and a combined H-file for the span  

At /2018:

  sh_glred -cmd

to get the globk.cmd and glorg.cmd files copied from gg/tables to 2018/gsoln,

then 

  sh_glred -s 2018 095 2018 097 -expt eura -gnss G R E C -opt R H G T >&! sh_glred.log

The script as commanded will translate the GAMIT ascii h-files in each day 

directory to GLOBK binary h-files (.glx) and put them into the /glbf directory

(H option); create a gdl file for each day listing the h-file, run GLOBK for 

each day (G option) using globk.cmd and glorg.cmd; and generate time series plots

(T option using pos files, program tssum and sh_plot_pos).  

For this test the globk/glred and glorg command files created in /gsoln 

by 'sh_glred -cmd' will work without editing; however, in general there are 

changes you may need to make, for example, to define the reference frame with 

translation-only rather than tranlsation and rotation, and to change the list 

of reference (stabilization) sites.  

The daily plots combining the GNSS measurements for each day, may be found in 

gsoln/plots_2018_095-2018_097.  You can view the *.org files to see what globk

did, paying attention to the stabilization iteration (what sites were removed),

the chi2 increments in the 'EXPERIMENT LIST', and the statitics of the 

stabilization ('POS STATISTICS'). In your own runs processing many days, it

is convenient to get a quick summary by grep'ing on *.org with 'USED' (from

the chi2 list) and 'POS STAT'.

Next combine the days to get a single H-file for the 3-day span to use for the 

multi-year repeatabilities (time series) and velocities. 

 sh_glred -s 2018 095 2018 097 -expt eura -gnss G R E C -ncomb 3 -globk_cmd_prefix COMB -opt G >&! sh_glred_comb.log

STEP 3: Repeat Steps 1 and 2 for 2016 and 2014:

  At /2016: sh_setup -yr 2016 

  At /2016/tables, edit process.defaults for mailto 

  At /2016/tables:  sh_upd_stnfo -l sd  ; mv station.info.new station.info

  At /2016:

   sh_gamit -expt eura -gnss G -s 2016 101 102 -pres ELEV -orbit igsf -copt x k p -dopts c ao  >&! sh_gamit_2016G.log 

   sh_gamit -expt eura -gnss R -s 2016 101 102 -jclock sp3 -pres ELEV -orbit comf -copt x k p -dopts c ao  >&! sh_gamit_2016R.log 

   sh_gamit -expt eura -gnss E -s 2016 101 102 -pres ELEV -orbit comf -copt x k p -dopts c ao  >&! sh_gamit_2016E.log 

   sh_gamit -expt eura -gnss C -s 2016 101 102 -pres ELEV -orbit comf -copt x k p -dopts c ao  >&! sh_gamit_2016C.log 

   sh_glred -cmd 

   sh_glred -s 2016 101 2016 102 -expt eura -opt H G T >&! sh_glred_2016.log 

   sh_glred -s 2016 101 2016 102 -expt eura -gnss G R E C -ncomb 2 -globk_cmd_prefix COMB -opt G >&! sh_glred_comb.log

  At /2014: sh_setup -yr 2014

  At /2014/tables: edit process.defaults for mailto 

  At /2014/tables:  sh_upd_stnfo -l sd  ; mv station.info.new station.info 

  At /2014

   sh_gamit -expt eura -gnss G -s 2014 113 114 -pres ELEV -orbit igsf -copt x k p -dopts c ao  >&! sh_gamit_2014G.log 

   sh_gamit -expt eura -gnss R -s 2014 113 114 -jclock sp3 -pres ELEV -orbit comf -copt x k p -dopts c ao  >&! sh_gamit_2014R.log 

   sh_glred -cmd 

   sh_glred -s 2014 113 2014 114 -expt scal -opt H G T >&! sh_glred_2014.log

   sh_glred -s 2014 113 2014 114 -expt eura -gnss G R -ncomb 2 -globk_cmd_prefix COMB -opt G >&! sh_glred_comb.log

We've specified only 2 days in 2016 and 2014 to save run time, and we've included only GPS 

and Glonass for 2014 since there were too few Beidou and Galileo satellites available at 

that time to obtain an accurate solution. 

STEP 4: Run GLOBK to get 3-epoch (4-yr) repeatabilities and velocities

The key user-specific controls for this step (also incorporated in the sh_glred 

runs within each year) are the list of sites to be used in defining the reference 

frame, and the a priori coordinates for these sites,  Here we use the same sites 

and coordinate file (itrf14_comb.apr) as in the single-year solutions, but this 

may not always be the case.  For the multi-year repeatabilities, we will use

the combined H-files (.GLX) created in the year /gsoln directories. Working

in /vsoln, thet a list of these for globk:

 ls ../????/gsoln/H*GLX >! eura.gdl

For large or complex data sets, it's helpful at this point to run glist which will 

check for blunders and give you a list of all the sites used and their spans.  Program 

glist2cmd can then be helpful in establishing a use_site list.  It may also be desirable 

to aggregate the data within each year, both to save time in the multi-year combination 

and to provide more representative long-term statistics.  The procedure for doing this 

is given in the file sGPS_recipe.txt in the documentation directory in the ftp area or 

web site.  

 \rm globk_replong.org globk_replong.log 

 glred 6 globk_replong.prt  globk_replong.log eura.gdl  globk.cmd   >&! glred.out  

Unlike in sh_glred, the rm command may be needed here since glred will 

concatenate the new log and org files with any previous files of the 

same name.  The globk_replong.prt will not be created since we have set 

'prt_opt NOPR' in globk.cmd. 

Create the pos files for plotting:

tssum . mit.final.itrf14 -R globk_replong.org 

Create the multi-year plots:

sh_plot_pos -f *.pos -r -t NONE -u -t1 2014-001 -t2 2018-180 

Get velocities from the 4-year span:

 \rm globk_vel.org globk_vel.log 

 globk 6 globk_vel.prt  globk_vel.log eura.gdl  globk.cmd VEL >&! globk.out  

where the 'VEL' token at the end tells globk to uncomment the lines in 

globk_long.cmd that begin with 'VEL'.

Check the globk_vel.org file under EXPERIMENT LIST for the chi2 increments 

(< 1.0) in stacking the 3 H-files and the stabilization statistics in the

'VEL STATISTICS' and 'POS STATISTICS' lines.  For this network and the 

ITRF 2014 apr file, the 9 stabilization stations fit the reference frame

in velocity at the level of ~1 mm/yr horizontal and 5 mm/yr vertical with

uncertainties about half of the rms.   

Plot the velocities using

 sh_plotvel -ps euratest -f globk_vel.org -R-60/50/-5/55 -factor 0.5 -arrow_value 10  -page L                                                         

(Type 'sh_plotvel' without arguments to see the full range of options for 

producing more attractive maps.)

------------

To start over from scratch, run sh_clean_test, once for each year and 

once for the velocity solution. (Type the name of the script without 

arguments to see the documentation.)