look here for gdh2000.html

look here for gdp2001.html

Sten's programs and data sets are described here:

file:///home/sten/Bi4asm/algorithm2.txt
 

How to ...
   create/append gd (geodetic datum) and derive rates in a regional frame:

The basic workhorses are ~/gps/mstcs and ~/gps/moregd
 

Both Procedures need environment parameters. The latest environment is
usually saved in last.env . You should consider to keep saving environments
if you made important changes / finally found one that works for you:
   dumpenv
Procedure gpsenv will filter out those parameters that are relevant for the
processing. If you update mstcs (and other ~/gps scripts) you will have
to update gpsenv so that the book-keeping also comprises the new parameters.

Re-install this environment. You can hilite and drop it on the shell prompter
using the mouse. But you might prefer to do it the command way:

so you can have this condensed environment on file for editing.
You want to see guidelines for "HOW TO devise the environment" ?
You'll find such a section below.

IT IS IMPORTANT THAT YOU REALLY GO TO THAT DIRECTORY !

Comment Since we've got the rights to write on $SD, we can do
uncompressdir there. In case there are compressed stacov's
that operation will warrant against space expense.
We use a script ~/gps/SD to give us the actual $SD
Thus:

cd ~/gps
emacs many-moregd (to set actual dates)

cd $SDIR ; <-- Really, go there !!!
;
; many-moregd -do [-a] `the-sites-of itrf_95dec01` # ??? -do ???
many-moregd `the-sites-of itrf_95dec01`
many-moregd `the-sites-in $DDIR .gd.ep`
many-moregd ...

     If the year and time span is set inside many-moregd using the environment ( dts=$GD_YEAR )
     then the following tcsh source saves you a lot of time:
source ~/gps/all_gd.env

   eventually misses can be added by e.g.
setenv GD_PERIOD "96 oct"
many-moregd.sh -do [-a] `the-sites-of itrf_95dec01`

cd ~/gps
many-gd2ren [-2f] `the-sites-of ...`

Two frames (compensate orbits between itrf93 and itrf94):
use option -2f

( Two frames and 6 parameters (i.e. with translation) will
probably be no good since horizontal rotational motion
component is difficult to discriminate from geocentric
rotation. It only works for the vertical part since
this motion is never reproducible by geocentric rotation. )

Three frames, esp. Heflin frame as the primary reference, ITRF frames
used only to compensate "kinks" in orbit evolution due to ITRF-transitions.
Two variants exist, gd2ren-3f and -03f.
In the latter, all ITRF stations are used to determine the rot/trans
parameters. The first parameter of Namelist &param n_rot_trans
specifies whether translation is switched (6) on or not (3)
of the basic reference frame (the Heflin frame)

There is a general version using a newer main program, where the number of frames can be extended.
gd2ren-0mf  ("multi-frame").  Use option -0mf  in  many-gd2ren

(a) time series and lin. regression
     edit the file (Upto_95/)offsets.dat

many-gd2ren -2f `the-sites-of itrf_94may10`               \|/
cd ~/tap; many-urtap-ss `the-sites-of itrf_94may10`     -->X<--
cd ~/gps                                                  /|\

           variants:
many-urtap-ss `cat ~/gps/the_clim_sites`
setenv TAPVARIANT .NoClim
many-urtap-ss `cat ~/gps/the_noclim_sites`
 

(aa) (optionally:) sliding straight line sections
 many-slopes -s `the-sites-of ...`

(a) cont'd
nice many-gdplot-ts [`the-sites-of ...`]
nice many-gdplot-ts .NoClim `the-sites-of FINNNET`

(b) GPS rates versus tide gauges and levelling.

For collecting rates information from the *.prl files, use procedure ~/gps/bin/fira -s
It replaces ss_rates_data and it has better formatting. -s option is for short format.
Jim Davis needs a strict format with extensive print, then not with the -s option.

    You need tide gauge results etc. The tide gauge file name is specified
    by means of
setenv TGGRATES filename

    The default /home/hgs/gps/ekman_rates_up is set in script ra4ra
    For GPS data, take
jmj_rates.dat
    (see below how to prepare -> "Instructions to process jmj rates table")
    or
ss_rates_data -1 .ra > ss_rates_2f.dat
    or
ss_rates_data -1 .ra | fgrep -v .NoClim > ss_rates_2f.dat
    and ...
setenv RATESTBL ss_rates_vs_rates.dat
rates_vs_rates_data -sf ss_rates_2f.dat
emacs $RATESTBL
    there'll be bad values + bad label locations
rates_vs_rates_plot -s $RATESTBL
 

(b') HOW.TO collect rates from a number of sources:

Note that the open output pipe allows collecting rate results from a number of
sources. If we for instance have a second solution for a subset of stations we can
collect a temp selection and then choose the latest addition for each site:

(b'') HOW.TO request geoid-augmented tide-gauge rates:
cd ~/Maxwell
tangvecs prof/ '>GEOID>'
    will create ~/Maxwell/prof/Bifrost-geoid.xyz thru file unit 33.
setenv GEOIDRATES ~/Maxwell/prof/Bifrost-geoid.xyz
    The default is empty, i.e. no geoid rates will be applied. Define a new
    file name to be sure:
setenv RATESTBL ss_rates_vs_drtgg.dat
cd ~/gps
rates_vs_rates_data -sf ss_rates_03f.dat
rates_vs_rates_plot -s $RATESTBL
 

(b''') HOW.TO narrow the search to TAPVARIANTs:

rates_vs_rates_data will yield results for a wide range of sites.
You may wish to devise a narrow set of sites and TAPVARIANTs to
include in the plot. To faciliate reprocessing you could prepare a file
"the_sites_4_rates" containing

(c) HOW.TO plot vectors on map :
ss_rates_data -2 .ra > ss_rates_2f.hrv
ss_rates_data -2 .ea >> ss_rates_2f.hrv
ss_rates_data -2 .no >> ss_rates_2f.hrv
emacs ss_rates_2f.hrv # because there'll be bad values
sshrrv.out
    or
sshrrv -f ss_rates_2f.hrv stacov=\'$SDIR/comb/\*.stacov\'
cd maps
swemap -s euro -o hrv2f.ps -e igs swe -T taphor tapver
    To get it in Mercator:
setenv REGION -12/35/40/72r
setenv PRJCT m1:35000000
swemap -o hrv2f.ps -e igs swe -T taphor tapver

      HOW TO  scale GPS rate errors after the fact

If the data comes from Jan and his Igor,
a table with rates may look like the following (jmj-rates.gd):

gd2ren can process this data after some additions:

The instructions in gd2ren.ins for processing jmj-rates-new.gd look like
those below, cf. > Instructions to process jmj rates table

*** If you don't see any arrow, then you probably have input columns to mean
*** long lat ... and forgot q_swap_lola=.true. (The standard is a confusing
*** lat long height due to Jan or Jim or whoever wrote the Igor script.)

How to create Rates table x y z:
awk '{printf "%.4s %s\n", $1,$2}' ss_rates_2f.dat |\
xargs -l1 lola-of-site > vertical_rates.xyz
    or
awk '{printf "%.4s %s\n", $2,$3}' ss_rates_vs_rates.dat |\
xargs -l1 lola-of-site > vertical_rates.xyz
    ... and plot color+isolines
cd maps
clr_vrates ~/gps/vertical_rates.grd
    The grd file does not exist; it will be prepared by clr_vrates.

How to create Haze grid:
/home/hgs/gps/mindist.out -X0,40,1 -Y50,75,1 \
-F/home/hgs/gps/vertical_rates.xyz |\
xyz2grd -R4/36/52/72 -I1/1 -G/home/hgs/gps/distance.grd

     observe that -X -Y do loop increments (1 1 in example) must fit the
     -I increments in xyz2grd;
     and that the -X -Y ranges in mindist must be equal to or larger than
     the -R range in xyz2grd to avoid storage of NaN's.

     ... and plot color+isolines+haze
cd maps
clr_vrates ~/gps/vertical_rates.grd ~/gps/distance.grd

    The files *_rates.dat should look like this

UMEA.ra: 10.45 0.7
----------------------
SITE.co: rate sigma

    You might need awk to create it:
awk '{if (NR>1)printf "%s%s%6.2f %6.2f\n",$1,".ra:",$12,$13}' \
jmj-rates-970828.gd >! jmj_rates.dat

    If you need a new label file (to indicate the locations that went into
    to color contour plot), follow this example:
cd ~/gps
cat jmj_rates.dat | xargs -l1 -i label4site {} 12 0.0 4 1 \
> maps/jmj-labels.dat
    or
awk '{print $2,$3," 12 0.0 4 1 ",$1}' jmj-rates-970828.gd \
>! maps/jmj-labels.dat

cd maps
clr_vrates -o jmj_cvr.ps \
~/gps/jmj_vertical_rates.grd ~/gps/jmj_distance.grd jmj

We have an experimental version for perspective mesh-plot
clrlf_vrates ~/gps/vertical_rates.grd ~/gps/distance.grd
This one needs a lower grid resolution 50/50, so the intermediate
grd files of this job and clr_vrates are incompatible.

HOW TO devise the environment.

    The stacov files you want to process are typically named
/gps2/SWEPOS_RESULTS/ITRF_HEFLIN_93/93aug23_itrf_vf.stacov(.Z)
    Then:
setenv SD /gps2/SWEPOS_RESULTS/ITRF_HEFLIN_yy
    File names contain "itrf", therefore
setenv ITRF yes
    The "_itrf" comes late in the file name, therefore
setenv SWAP_ITRF yes
    "_itrf" is suffixed by "_vf", therefore
setenv VF _vf
    ( If ITRF = no the file name would be itrf_vf_93aug23.stacov,
VF = vf_ (!))
    Instead of setting ${ITRF} and ${SWAP_ITRF} you can directly control
    the strings that prepend (${I1}) and append (${I2}${VF}) the date
    part of the file name.
    The temporary directory holding the uncompressed stacov files (or
    the symbolic links) is called ${SDIR}/temp, therefore
setenv SDIR_T temp

    ( ${SDIR_T} is used by ~/gps/moregd (~/gps/moreltu);
    you can collect
moregd data by cycling through stages, e.g.
setenv SDIR_T temp_93
moregd SITE 93 aug sep oct nov dec
setenv SDIR_T temp_94
moregd SITE 94
    etc.)

The reference file is called ${SDIR}/${REFF}.stacov
You will have to copy or link that file.
The data directories are ${TAPDIR} for ~/tap/urtap results
and ${DDIR} for other results. You will need a directory
$DDIR/PS for the postscript plot files.

${STATISTICS_TYPE} is an option passed to gipsy statistics.)

You might need site aliases. If you want to combine e.g. WTZR and
WETB in one file called WTZR*,

 ___________________
|                   |
| badies            |
|___________________|

> all_*.stacov and all_*.gd
comb/all_plus_KUUS.[stacov|gd]:
In the current bad situation the stacov file is a mock-up to include
a couple of stations for which no STA information concurrent with the
majority of sites (i.e. with the sites in all_93_94_95_itrf93.stacov)
is available. As a violation of gipsy rules, no covariance parameters
appear. New sites in comb/all_plus_KUUS.[stacov|gd] are
KUUS POTS BRUS KEVO KIVE ROMU
 

 ___________________
|                   |
| goodies           |
|___________________|

> the-sites-of ...

This procedure usually looks into a particular stacov file to find out
the actual sites. There are a few short-cuts (SWEPOS, IGS, XIGS, FINNNET).
Many times there is a demand for more elegance and flexibility. Consider
to use procedure the-sites-of-files. Here's an example:

ss_rates_data -2 .ra -v \'RAAH WTZR \' \
.NoClim \'BORA BRUS POTS RIGA `the-sites-of FINNNET` \' >! ss_rates_03f.hrv
 

> make a site coordinate subset file from a stacov file
    (coordinates only, useful for olfg site files etc)
echo " nn PARAMETERS that date" > subset.sta
the-sites-of -l FINNNET |\
xargs -i fgrep {} itrf_97may25.stacov >> subset.sta
 

> days in the solutions:
grep -c '199[3-6].' $DDIR/*.gd.ep

echo " " >! epstats.lst
foreach file $DDIR/*.gd.ep
epstats $file >> epstats.lst
end

> Interpolate GPS-determined rates at PSMSL stations ?
    Use the grd file behind the famous hgs color plot and
    code a temp script called proc:
tail -1 | awk '{print $7,$1,$2}' |\
  xargs -n3 gcdist -i vertical_rates_i.xyz -C | sort -nr | tail -1

foreach site ( `awk '{printf "%s " ,$7}' maps/psmsl_a-labels.dat` )
foreach? fgrep $site maps/psmsl_a-labels.dat | proc >> bifrost_rates_at_psmsl_i.dat
foreach? end

   If alias proc is preferred, the awk script must be placed in a file.

> Nearest GPS-determined rates at PSMSL stations ?
    change the proc script: file name is now vertical_rates.xyz
    i.e. the original Bifrost result.
    Collect in another file: proc >> bifrost_rates_at_psmsl.dat

> Making a new version of Heflin-style ...
     sta_pos file (like ~/IGS/heflin.ssc)
     to stay consistent with Jan's analysis:

cd ~/IGS
sta_pos2ssc > like-jmj.ssc

    (won't have sigmas, however. Until it has, the ssc file can't be used !)
    sta_pos files: look into
       /gps1/gipsy_r2/gipsy-oasis/sta_info
       /gps4/NETWORK/gipsy/gipsy_r5/gipsy-oasis/sta_info

Instructions to process jmj rates table

(a) for color isolies
awk '{if (NR>1)
print $2,$3,$12}' jmj-rates-970828.gd |\
grep -v '^ ' >! jmj_vertical_rates.xyz

(b) gd2ren for arrows
JPL>
21 O jmj-rates-?.gd input
* changed heflin to jpl in the following 5 lines
31 B maps/jplhor-vecs.dat
32 B maps/jplhor-sigs.dat
33 B maps/jplver-vecs.dat
34 B maps/jplver-sigs.dat
35 B maps/jplhor-labels.dat
36 B maps/jplver-labels.dat
41 B jpl-rates.dat output
Q
 &PARAM
 q_rate_results=.true., q_swap_lola=.true.
 q_print=.false.
 q_plate_motion=.true., def_plate='EURA', q_itrf=.false.
 pm_comment=' '
 vel_itrf='jpl',plate='EURA'
 q_all_plates=.false., select_plates='EURA '
 q_all_sites=.false.
 q_all_techniques=.false., select_techniques='GPS'
 iun_plates=70, fn_plates='/home/hgs/IGS/gps_domes_plates.lst '
 iun_itrf=71, fn_itrf='/home/hgs/IGS/heflin.ssc '
 n_select_sites=9, n_rot_trans=3
 select_sites(1)='10402', select_sites(2)='14201'
 select_sites(3)='10302', select_sites(10)='13212'
 select_sites(5)='10403', select_sites(6)='13504'
 select_sites(7)='10317', select_sites(8)='10503'
 select_sites(9)='12734', select_sites(4)='13407'
 xlatl=46.5, xlonl=30., dlatl=.5, dlonl=+.3
 xlatv=48.0, xlonv=30., dlatv=0., dlonv=+.5, verdir=90.
 size=200.,sample=1.d-3
 sizv=75.,samplv=5.d-3
 &END

|----------------------------------|
| gd2ren-03f generated .ins file |
|----------------------------------|
980825>
22 * o /geo/hgs/ts_itrf_heflin/WETB.gd
22 o /geo/hgs/ts_itrf_heflin/WTZR.gd
23 o /geo/hgs/ts_itrf_heflin/comb/all_plus_KUUS_itrf93.gd
24 b /home/hgs/gps/Upto_97/vfh_94_93/WTZR.gd.ep
25 b /home/hgs/gps/Upto_97/vfh_94_93/WTZR.ss.info
26 b /home/hgs/gps/Upto_97/vfh_94_93/../offsets.dat
71 O ~/IGS/itrf94.a.ssc
72 O ~/IGS/itrf94.b.ssc
73 O ~/IGS/itrf94.c.ssc
q
&param
q_time_series=.true., try=-1.d0
site='WTZR', alias='WETB'
scale=1000.
q_print=.false.
q_plate_motion=.true., def_plate='EURA', plate='EURA'
q_itrf=.true.,.true.,true.
pm_comment='Reduced motion = NUVEL1-a'
itrf_comment='Motion w.r.t. JPL-ITRF94-93N/IGS(EURA) '
q_three_frames=.true., q_zero_frame=.true.
frame_switch_date=1996,9,1
frame3_switch_date=1995,6,1
vel_itrf(1)='JPL'
vel_itrf(2)='File : ITRF94.SSC'
vel_itrf(3)='ITRF93N VELOCITY FIELD '
q_all_plates=.false.,.true, select_plates='EURA ',' '
q_all_sites= .false.,.true.
q_all_techniques=.false.,.false.,.false.,.true.
select_techniques='GPS', 'GPS', 'GPS', ' '
iun_plates=60,70,-1
fn_plates(1)='/home/hgs/IGS/gps_domes_plates.lst'
fn_plates(2)='/home/hgs/IGS/iers.site '
fn_plates(3)=' '
iun_itrf=61,71,81
fn_itrf(1)='/home/hgs/IGS/heflin.ssc'
fn_itrf(2)=' '
fn_itrf(3)='/home/hgs/IGS/itrf93.ssc '
n_select_sites=7,0 n_rot_trans=3,6,6,6, q_rot_only=.false.
select_sites(1,0)='10402', select_sites(12,0)='14201'
select_sites(2,0)='10302', select_sites(3,0)='10503'
select_sites(4,0)='13504', select_sites(5,0)='13407'
select_sites(6,0)='10317'
select_sites(7,0)='14001'
select_sites(8,0)='12734'
select_sites(9,0)='10403'
select_sites(10,0)='13101', select_sites(11,0)='13212'
n_delete_sites=1,0
delete_sites(1,0)='10503S999'
&end

Comments:
try=-1. is a remnent of the trial-and-error testing of
the relative motion reduction.
The primary reference frame is heflin.ssc
The secondary frames are itrf94 and itrf93
We use all sites to derive global rotation and translation parameters
(q_all_sites(1)=.true.)
The primary frame is constructed by seven stations on the EURA plate
(q_all_sites(0)=.false.)
There is no translation component in this frame (n_rot_trans(0)=3).

• ~/Maxwell/all-reb-bl_s18.dat - model
• $TAPDIR/bl_rates - observed
• $DDIR/bl.info - sites, lengths, orientation

 

 
 
 
 
 
 
 
 
 
 
 
 
 

We usually run the superscript 'many-bll' or the blanalysis script.
'blanalysis.out blanalysis.ins' is used for ... (?) ...
but primarily for the '>BI>' (generate Baseline Information) part.

Follows a working example
setenv DDIR ~/gps/Upto_97/Bl
setenv TAPDIR ~/tap/Upto_97/Bl
setenv REFF /home/hgs/gps/comb/all_plus_KUUS_itrf93
       or
source latest-bl.env

      It might be necessary to link stacov files:
cd $SDIR
cd temp..
mstcs -i 93 sep oct nov dec
 

cd ~/gps
blanalysis.out      program to compute daily baselines from stacov
                        Time-series are output to the path given in namelist variable
                    TSPATH [$DDIR] in binary mc-form.
blanalysis          shell script
blanalysis.ins
nice many-bll       superscript

/home/hgs/gps/p/ms/blanalysis.f   is a replacement for  /goa/bin/statistics -ltu  computing baseline length.

mblanalysis.out     program Maxwell model
mblanalysis.ins     to prepare ~/Maxwell/all-reb-bl_s18.dat from ~/Maxwell/all-rebound_s18.dat
                    If you need a new set of sites fitting your file catalog, consider

                        ls Upto_97/Bl3/WETB* |\
                           sed 's|.*/....\(....\).*|\1|' |\
                            awk -v q="'" 'BEGIN{printf " sites="};
                                          {printf "%s%s%s,",q,$1,q};
                                          END{printf "\n"}'

blanalysis.out '>BI>'   to prepare bl.info
cp bl.info $DDIR

blanalysis -c 1997,01,01 - temp97 190

to continue/update from a certain date.
If we must redo the total, rm $DDIR/*.mc first.
If new sites are to be added, the only option at present is to process
everything from the beginning again.

!! there is a problem with the current data set (1997/spring 1998):
!! The epoch and t0 are set on the midnight AFTER the first day.
!! Should be epoch = 1993 09 01 and t0 = 12.d0

cd $TAPDIR
fgrep 'Linear 1/[year]' *.prl |\
awk '{print $1,$6,$7,$8}' |\
sed 's/\.le\.prl://' > bl_rates.dat
= ~/tap/collect-bl_rates >! $TAPDIR/bl_rates.dat
 

* Make the big table
cd ~/gps
cut -c 1-8 $DDIR/bl.info |\
fgrep -v '___' | xargs -i ~/gps/collect-bl {} > blanalysis.rsl
    This is compised in the single command
collect-blanalysis >! blanalysis.rsl

* Make summary plot
cd ~/gps
blplot-rsl [#M]
    where #M is a number to allow time-series to have up to #M missing samples
    (default=500)

* Make a plot
cd ~/tap
setenv BLOUTS .false.
urtap-bl SIT1 SIT2
cd ~/gps

* plot time series
blplot-ts SIT1 SIT2

* Make slopes
cd ~/gps
slopes -b mc SIT1 SIT2
blplot-ts SIT1 SIT2 pn

* Plot a lot:
   Example: all baselines to Hassleholm
cd ~/tap
all-urtap-bl +O -s HASS `the-sites-of SWEPOS` SAAR TROM WETB
cd ~/gps
for-all-baselines -c blplot-ts -b -s HASS \
`the-sites-of SWEPOS` SAAR TROM WETB > ! temp.proc
source temp.proc