GAMIT版本更新至10.6

这次更新有较大的变动，也将更适应目前的数据处理需要。以下为更新说明：

=--------------------------------

Release Notes for GAMIT/GLOBK 10.6 


Last updated 12 June 2015
    
----------------------------------------------------------------------


Release 10.6 consolidates incremental updates made since release 
10.50 of July 2013. As noted in the release announcement, this 
release has extensive structural and format changes to accommodate 
GNSS SVs other than GPS.  Of the files in the usual processing flow,
four have changed formats:  svnav.dat, dcb.dat, g-files, and c-files.
The new version will read old-format g-files and dcb.dat files, but
NOT svnav.dat or c-files. We will continue to update old-format 
svnav.dat and dcb.dat files and put them into incremental updates
for another 6 to 12 months with extent 'gps' for user who do not wish 
to move to 10.6 immediately.  There was a flaw in our old-format 
creation of dcb.dat from the AIUB P1C1.DCB files in the time-tagging
and in not correctly accounting for different SVs with the same
PRN numbers during the month.  This error will degrade wide-lane
ambiguity resolution in some cases.   Finally, with the new code,
you'll find differences in the print files for arc and model, but 
unless you read these with your own scripts, these changes will be 
seamless.   
                                       
Whenever you process data beyond the date of the software release, you
MUST update supporting files.  For GAMIT:  svnav.dat and antmod.dat for 
changes in PN assignments; dcb.dat for P1/C1/P2 code biases that affect 
wide-lane ambiguity resolution; station.info for antenna changes at
reference sites; and ut1/pole files for current EOP information. Of these
files, only the EOP files are automatically updated by sh_gamit.  For
GLOBK, you should keep current itrf08_comb.apr and itrf08_comb.eq to
account for improved coordinates and antenna changes and earthquakes at 
reference sites.  Both the GAMIT and GLOBK files are always current in 
incremental_updates/tables and included in any incremental_update tar-file.
                                   
All GAMIT-GLOBK scripts that produce plots (e.g. time series and velocity 
maps) use Generic Mapping Tools (GMT) to do so. The release of GMT 5 in 
November 2013 has necessitated major alterations to the syntax of GMT 
commands in many scripts. The updated scripts, which use GMT 5-compatible 
syntax, are contained in com/ and will be the only ones maintained and 
developed beyond this release of GAMIT-GLOBK. We therefore recommend that 
all users now download and switch to using GMT 5 because the updated scripts 
are NOT backwards-compatible. The old scripts, which may be used with GMT 4 
or prior, are now contained in the com_preGMT5/ directory. Users must 
prepend this directory to their path in order to continue to use GMT 4, e.g. 
$PATH is 
...:$HOME/gg/com_preGMT5:$HOME/gg/com:$HOME/gg/gamit/bin:$HOME/gg/kf/bin:...


For plotting time series, we are transitioning from using VAL and mb 
files produced by ensum and multibase and used by sh_plotcrd or sh_baseline
to PBO-format pos-files produced by tssum and used by sh_plot_pos.  tsview
can still read both mb-files and pos-files.    sh_glred will accept either
'E' (for ensum/mb) or 'T' (for tssum/pos).  


sh_gamit now writes the sky plots and phase vs elevation angle plots
into a /figs directory instead /gifs, and the files are no png rather
than gif. 


glist can now take multiple apr and eq files and accept use_site commands, 
so that it can fully check a globk solution.  


To be consistent with IGS standards, we have replaced the Bar-Sever model 
Kouba model for satellite yaw, removing the need for the ascii y-file 
generated by ARC.  




Other changes and comments:


Modified memory management in globk to handle >20,000 parameters.  See
globk.hlp and glorg.hlp for other features added in recent versions.


Kalman filter added to tsfit (see tsfit.hlp) 


Changed the conversion of sp3 files to g- or t-files to use the IERS2010
dirunal and semi-diurnal EOP so as to be consistent with the rotations
in MODEL (differences of 2 mm in orbit position).


Introduced and made default IGRF12 for the magnetic field used to compute
2nd order ionospheric corrections. Thanks to Seiichi Shimada of NIED for
this change.
   
Changes to sh_plot_pos: Switched position of north and east component plots 
and fixed long-standing bug with not plotting discontinuities when specifying 
rename- and/or eq-files and, implicitly or explicitly, using "NONE" as tsfit 
command file.  Added additional width to time axis in figures to prevent
first and/or last time series points plotting beneath y-axes, and added -w
option to control fraction of time series length to add (default 0.1 = 10% on
both ends).


Changed sh_update_eop to use wget or curl rather than ftp or ncftp to 
download EOP files from USNO to avoid frequent ftp hang-ups.  The new
version supports extent 'usno' (Bulletin A finals.data from 1992, updated 
weekly) and 'usnd' (Bulleton A finals.daily, last 3 months only, updated 
daily). We no longer distribute Bulletin B since the Bulletin A value are
as accurate and updated more frequently.


Replaced hyphen ("-") with underscore ("_") for output org-file name
produced by sh_glred when run in combination mode (using -ncomb) to
avoid clash when being read by another GLOBK program, e.g. ensum,
bcsum, etc. Output combined org-file name now of the form
globk_[expt]_[YYDoY]_[YYDoY].org rather than globk_[expt]_[YYDoY]-[DoY].org.


Correct the format for reading ATMOSMAP values from a VMF1 list file.
 
Fix bug in gamit/lib/read_antex.f that caused PCO and PCV values to
be zero if the frequency codes are written 'G 1' and 'G 2' instead
of 'G01' and 'G02". This affects the following antennas using NOAA
calibrations: TPSPG_A1_6+GP, SOK_GSR2700ISX, THA800961+REC, THA800961+RTK,
THA800961RECUHF, THA800961RTKUHF, NOV702_3.00, JNSCR_C146-22-1, TRM_R6,
TRM_R8, JAV_TRIUMPH-1. 
  
Altered sh_makeeqdef event naming logic so that all earthquakes found
in catalogs are numbered sequentially and are only given two-character
IDs (e.g. see below entry) after they are selected as potentially 
influencing a site in the . 
    
Modfified glinit so that when the make_svs command is used, the satellite
antenna offsets from the latest run (run time value) are adopted as the 
apriori antenna offsets.  Needed when merging new processing (IGS08) with early
MIT global files (e.g. in IGS05). 


Add general relativity corrections to the orbital integration, changing 
scale and clock slightly but little affect on positions.


Fixed bugs in the mean pole used in the ocean pole tide correction and
in equate decoding in glorg.  Updated model output in the SINEX files.  


Fix long-standing bug causing loss of a 32nd satellite. 
  
Fixed bug in kf/glinit/read_eq_file.f that could result in the end time 
of rename being incorrectly set for open ended renames. Fixed bug in 
kf/utils/stinf_to_rename.f that could result in misssed renames after 
the -post date. 
   
Increase max_sites in /kf/utils/merge_apr.f to accommodate the current 
itrf08_comb.apr.


Added "-H" option to "find" commands in install_software to follow
symbolic links. 
 
Increase the PID in the c-file name to 6 digits for large machines.


New script sh_pbor_vel to create PBO-format velocity files from GLOBK 
output.             


Fix bug in sh_org2vel that causes some sites to be missed.
 
Replace itrf00.apr with itrf08_comb.apr as default for sh_setup. 


Added new scripts sh_cats and sh_hector in com/ to analyse noise charac-
teristics of time series (using pos-files).  
                                   
Updated glorg to allow multiple site selections in equate/unequate/force 
commands. Fixed typo in .pos file headers. 


Changed sh_makeeqdef "-pos " option to "-f " to
avoid confusion with PBO-format pos-files; changed event ID sequencing
scheme from purely numeric to purely two-character (01...99, then A1...A9,
B1...B9 etc. until Z1...Z9, then 1A...1Z, 2A...2Z etc. until 9A...9Z);
added option to use a translation table (e.g. see ~/gg/tables/eq.transtbl)
to edit automatic ID to user-friendly alternative. All automatic IDs now
contain at least one numeric character, while manual IDs are recommended
to be two alphabetic characters, as always.  Updated USGS earthquake search 
URL and made minor changes to event (re) naming logic in sh_makeeqdef. 


Updated coefficents for the solid-Earth pole tide to be consistent
with IERS 2010 standards.


Add libration terms to UT1.  
 
Updated tsfit, tscon and tsjumps to allows ~ (home directory) in file
names.  


Add a script sh_glist2cmd to invoke existing program glist2cmd, which 
creates from a glist a use_site list, an input file for unify_apr, and 
(new) an input site list for sh_makeqdef. 


Add the EPN cumulative solution to the ITRF2008 apr and eq files. Fixed bug 
causing bogus entries for 73 sites when EPN or NGS steps are inconsistent 
with those from ITRF2008 and IGS08. See tables/renames.bad for possible 
errors between early 2014 and June. 


Fix problem with poorly determined average of the east component of the
atmospheric load for a site close to the pole. 


sh_org2vel: Updated longitude/latitude format to five decimal places, in
keeping with org-file. 
     
ftp_info: Modified for correct igni address for IONEX files; add
PANGA, EPNCB, and TrigNet sites. 


Added additional width to time axis in figures made by sh_plot_pos to prevent
first and/or last time series points plotting beneath y-axes, and added -w
option to control fraction of time series length to add (default 0.1 = 10% on
both ends).  Corrected expected residuals file name to plot after running tsfit. 
  
svnav.dat: Correct start date for PN06/SV49---important bug affecting
data for epochs 2013 12 19 thru 2014 4 3.  Fix typo in start date for 
PRN30/SVN49 introduced in 121220 update.  
  
Fix bug in rotation derivatives in GLOBK.  


Fix format error in svpos.


Move rotation and translation estimates to be next to PMU estimates
in GLOBK print file.


Fix gamit/utils/merge_rinex.f to handle more than 10 observables. 


Updated wild card use in site names in mar_neu command (needed for
change in sh_gen_stats script). 


Edited name of temporary file in install_software to avoid accidentally
overwriting any libraries/Makefile.config.new when performing (and
succeeding) with automatic search for X11LIBPATH and X11INCPATH. 


Fixed bug in sh_get_orbits that caused IGS orbits from the CDDIS or
SOPAC archive to be operational (igsf) rather than reprocessed (igs1)
prior to GPS Week 1000 (1999/3/7).  


Updated tscon uto handle new Measures XYZ format.  


Fix problem with computing GRACE seconds.
    
Update gamit/utils/autecl.f to allow deleting post-eclipes data for
Block II satellites.  


Fix end-of-year problem with dcb.dat. Thanks to S. Shimada of NIED,
 
Fixed bug in MODEL, introduced in December 2012, which caused GPT/GPT2 values 
to be used for mapping functions and dry zenith delays even if VMF1 was
specified.  


Fix version number in sh_gamit.  King 130727.


Fix typos is globk_comb.cmd and glorg_long.cmd templates in gg/tables.
 
Fix bugs in fitting GAMIT's model to a broadcast ephemeris (sh_bcfit).
  
New option -subdir for sh_get_rinex to speed up uncompression when there
are many RINEX files.  


Update kf/Khandler/eq_name_chng.f so that _XPS/_XCL sites can not be 
renamed by later renames.  RESET needed to rename _X?? sites if
saved to binary hfile. 
 
Updated grdtab for modification to CoM correction to S1/S2 tidal loading 
and correct detection of dates out of time range of map.grid files.  


Added lunisolar tables for 2015; Updated leap.sec for June 2015 leap second.  
 
Allow reading of a time-of-year-dependent grid for atmospheric tidal loading
(not yet recommended by the IGS).


Add glinit/solution_inf.f,  glist/glist.f, and htoglb/cresnx_file.f
to /kf to fix problem with null string. 


Add the Earth-radiation, antenna-radiation, and 2nd-order ionospheric
model nmmes to the c-file and h-file.  


Various features and updates made to the suite of globk and tsxxx programs.  
See help files for new features.  tscomp added to suite.  .


Increase the dimensions for IONEX files. 
 
Increased dimensions in kf/utils/netsel.h   Herring 140718
  
Enhancements to cplotx including adding color. 


Sped up search for earthquake/rename displacements, and corrected output for 
field overflow in org-file and multiple/duplicate renames in sh_exglk. 
      
sh_upd_stnfo: corrected bugs in kf/htoglb/mstinf.f in using IGS SINEX and 
GIPSY sta_svec files.  Add a '-nogaps' option to close gaps in erroneous 
entries in IGS log files.  


sh_glred: Fix remaining bugs in allowing use of PBO-style h-files (.glb extent).


Fixed long-standing typo in format statement in gamit/makexp/makexp.f
producing a compile error with gfortran 4.6ff (but not earlier compilers).  


TRACK: Fix minor problem with noise_tol.  Fix bug in wide-lane-reference 
combinations (read_all_rinex.f 130730).


New Receivers: TIASAHI PENG3100-R2, JAVAD TRIUMPH_1, JAVAD TRIUMPH_1M,
JAVAD TRIUMPH2,MAG111406, NOV OEM4-G2, ITT 3750300, CHX90D-OPUS, 
ASHTECH PM-500. 


New Antennas: TIAPENG3100R2, STHCR3-G3 , JAVTRIUMPH_1M, JAVTRIUMPH_1MR, 
JAVTRIUMPH_LSA, Antcom ACC123CGN_XN, TRM57971.00 SCIT, LEIAX1202A, 
LEIAX1202A, TPSCR.G3 SCIT, MAG111406, ASH701945A NONE, TRM55970.00, TRMR6,
JAV_TRIUMPH-1R, CHX90D-OPUS, JAVTRIANT_A, JAVTRIANT_B,ASH700829.3 NONE,
JAVTRIANT_A, JAVTRIANT_B.




** 10.50 vs 10.41
  
We have provided and will keep up-to-date, ITRF2008 coordinates
sites in prior releases of the ITRF as well as new sites that have 
been and will be added for IGS processing.  The primary file, in the 
ITRF2008 NNR frame is itrf08_comb.apr; and there are secondary files, 
itrf08_comb_[plate].apr, created by rotating this file to the frames of 
the major plates using the Euler poles estimated by  Altamimi et al. 
[J. Geophys. Res., 117, B07402, doi:10.1029/2011JB008930,2012].  (Note 
that the plate-referenced files that we created 2012-01-06 and put into 
incremental_updates/tables were based on preliminary vectors that changed 
in the final publication, so users who downloaded these should replace 
them by the current files.) These files should always be used in conjunc-
tion with the rename file itrf08_comb.eq to assure that the 8-character 
site name assigned to your data matches the coordinates for the appropri-
ate period.  See 'apr_and_eq_files.txt' in the download documentation 
directory for a complete explanation of how the combined files were 
constructed. 
                     
There are improved templates in ~/gg/tables for GLOBK runs: 
globk_comb.cmd, glorg_comb.cmd, globk_long.cmd, glorg_long.cmd.  
These reflect current best practice and take advantage of the 'option' 
feature of the globk, glorg, and sh_glred command lines so that the 
same command file can be used for both repeatabilities and velocity 
solutions without editing the file itself.  The standard southern 
California example has been reworked to use these file and should 
now be easier to execute and to use as a template for your own 
processing.  Also, there is a new detailed recipe for processing
and analyzing survey-mode (sGPS) measurements, 'sGPS_recipe.txt',
to be found in the documents directory of the download site and
also on the web as part of the presentations from the recent
workshop held at UNAVCO (Boulder13).  


The GPT/GMF models for zenith hydrostatic delay and mapping functions
have been replaced by the new GPT2 model (Lager et al., Geophys. Res.
Lett., ) which is based on a more refined numerical weather model, 
includes semi-annual as well as annual variations, has better spatial 
resolution, and accounts better for height differences through a seasonally 
dependent lapse rate.  In the new model both the ZHD and mapping funtions
are derived from a table, gpt.dat, automatically linked into the experiment
and day directories from ~/gg/tables.  If you specify GPT or GMT in
your sestbl. and have gpt.dat in ~/gg/tables, you will automatically
get GPT2 (the old GPT is invoked only if gpt.dat is missing).


GLOBK now has the ability to remove non-tidal atmospheric loading
(ATML) based on the average of the day given in the h-file.  Since
GLOBK can not yet read an grid file to apply the ATML corrections
directly, if you want to investigate their effects, the best strategy
is to apply the ATML in GAMIT and then optionally retain or
remove the corrections using the new command -appload_mod in the
GLOBK commaand file.  The preferred ATML files for GAMIT are those
that have been filtered to remove sub-daily terms (atmfilt_cm.YYYY).
In principle, the sub-daily thermal tides should be applied in GAMIT,
but none of the current models, including a time-dependent recently
developed by Ray and Ponte, has been shown to reproduce accurately
the tidal loading, so the present recommendation of the IGS for
the next reprocessing is to omit these terms.


Liz Petrie of the University of Newcastle has coded a model in ARC for 
the accelerations on the GPS satellites due to antenna radiation and 
Earth radiation.  These models are invoked by setting 'Earth radiation 
model = NCELE1', now default in the sestbl. and 'fixdrv'.  As part of 
this modification, we have reworked many of the subroutines in ARC to 
be more modular, added (backward-compatible) models to the ARC batch file, 
and coded sh_get_orbits and sh_sp3fit to read the orbital models from 
from a sestbl.  This latter change assures that the orbital models used 
for creating g- and t-files from sp3 files will be consistent with the 
models used for the data processing.  For sh_gamit processsing, the 
sestbl. in ../tables is automatically linked into the /igs directory. 
If the sestbl is absent, hard-wired models will be used.                    
       
Three model conventions have been updated to be consistent with those
used by the IGS: mean pole and short-period EOP are changed from IERS 
1996 to IERS 2010, and the gravity field from EGM96 to EGM08.  These 
changes are important for EOP estimation but have neglible effect on 
positioning.


A new script sh_plot_pos will read PBO-style pos files rather than _mb
files for plotting time series.  GLOBK now generates the 'PBO' lines
in the print files automatically (no need to use PBOP in the print options).
We have retained sh_plotcrd and the use of _mb files for this release,
but they will eventually go away.  The Python script ts_plot.py, introduced
in Release 10.40 to read pos files will also be abandoned because it required 
too many non-standard Python libraries (see below).
                                 
The new script sh_plot_elmean, together with additions to the autcln summary
files allow plotting of mean residuals over several days as a function of 
elevation angle or satellite nadir angle, providing more diagnostics for bad 
antenna models or tracking environments.  These can be used instead of or
in conjuction with the individual-day sky plots and phase vs elevation plots
produced by sh_oneway from the autcln DPH files when '-pres ELEV' specified
with sh_gamit.