Release Notes for GAMIT/GLOBK 10.5 Last updated by rwk 12 August 2014 ---------------------------------------------------------------------- Release 10.41 consolidates incremental updates made since release 10.40 of November 2010 and adds several new features not included in the incremental updates. 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 The GPT/GMF models for zenith hydrostatic delay and mapping functions have been replaced by the new GPT2 model (Lager et al., Geophys.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.grid, automatically linked into the experiment and day directories from ~/gg/tables. If you specify GPT or GMT in your sestbl. and have gpt.grid in ~/gg/tables, you will automatically get GPT2 (the old GPT is invoked only if gpt.grid 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 = NCEL1', 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. Keeping an up-to-date svnav.dat and antmod.dat file is <> to correct processing. Failure to do so will often impose the wrong radiation- pressure model, SV antenna offset, and yaw model on the satellite. We will notify users any time these files are changed, but regular checking of the incremental_updates/README file is advised. Changes since 10.40 are: - Launch of SV 35 (Blk IIA) as PRN 1 (date: 110606) - Reassignment of PRN 1 to newly launched SV63 (Blk IIF) (110718) - Reassignment of PRN 30 from SV 31 to SV 35 (both Block IIA (110816) - Update the maximum yaw rates for Block II and Block IIA satellites using the weighted average values determined by JPL 1996-2008 (110824) - Reassignment of PRN 24 from SV 24 (Blk IIA) to SV 49 (Blk IIR-M) - Reassignment of PRN 24 from SV 49 (Blk IIR-M) to SV 32 (120315) - Reassignment of PRN 24 from SV 32 (Blk IIR) to SV 37 (Blk IIA) (120426) - Reassignment of PRN 24 from SV 37 (Blk IIA) to SV 49 (Blk IIR-M) (120806) - Launch of SV 65 (Blk IIF), replacing SV 49 (Blk IIR-M) as PRN 24 (121008) - Reassignment of PRN 27 from SV 27 (Blk IIA) to SV 49 (Blk IIR-M) (121018) - Reassignment of PRN 30 from SV 35 (Blk IIA) to SV 49 (Blk IIR-M) (130510) - Reassignment of PRN 27 from SV 49 (Blk IIR-M) to SV 66 (Blk IIF) (130516) Other changes and commments: ** 10.41 vs 10.40 New receivers: TRIMBLE NETR3, R7 GNSS, and SPS852 ; LEICA GG02PLUS, GS08PLUS, GS09, GS12, GR25, and GS25; NOV OEM6; Ashtech PF500 and PF800;, NAVCOM SF-3040 SF-3050; Topcon TPSGR5; JAVAD TR_VS; SEPT POLARX4, POLARX4TR, and POLARXS; SOK GSR2700 RSX; SOUTH S82T GNSS and S82V GNSS; STONEX S9II GNSS; IFEN NX_NTR_303_D, NX_NTR_500_D, NX_NTR_600_D, SX_NSR_RT_200, SX_NSR_RT_300, SX_NSR_RT_400, SX_NSR_RT_401, SX_NSR_RT_402, SX_NSR_RT_700, and SX_NSR_RT_800. RNG FASA+, ITT 3750300, GEOMAX Zenith10, GEOMAX Zenith20 receivrs New antennas: TRM_R6, TRM_R8. TRM55970,TRM59900.00, TRM59900.00, and TRMAG25; NOV533 (compact), NOV702_2.02 and NOV702_3.00; ASH701023.A and ASH70123.L1; LEIGS08, LEIGS12, LEIGG02PLUS, and LEIGS08PLUS;, NavCom NAV_ANT3001BR (AT1675-20W); JAV_TRIUMPH-1, JAVTRIUMPH_VS, JNSCR_C146-22-1 and four JPSREGANT; TPSHIPER_II, TPSHIPER_II+10, TPS NET-G3A, and TPSGR5; SEPCHOKE_MC; Navxperience NAX3G+C; Stonex STHS82_7224V3.0 and STXS9SA7224V3.0; Altus APSAPS-3; Rusnavgeoset RNG80971.00; ITT3750323; GeoMax GMXZENITH10 and GMXZENITH20. Correct the values in hi.dat used to convert slant height to direct height for the TPSHIPER_PLUS antenna. Fix typos in hi.dat entries for SEN14R, SENS49R, SENS96, SENS96, SEN96R, TPS721, TRSXDS, and TRSLDR antennas; add dimensions for Leica LEIAR25 antennas. Updated Makefile.config.new (to be renamed to Makefile.config) for new Linux releases. Creating an earthquake file: The script sh_makeeqdef has been updated to use hypercenter locations from the EHB bulletin (via ISC) up to the end of 2008, and from the standard ISC bulleint thereafter. Update sh_get_hfiles for changes in the SOPAC archive ftp. Fix bug in reading more than 36 SVs from a RINEX file. Fix the p-file and h-file printout for SP EOP. Update sh_get_orbits to use repro1 orbits when down loading igsf from CDDIS. These orbits are available week 730-1459. Updated grdtab to remove S1/S2 tides from non-filtered grids. Change mapping of ietide for h-files: mean-pole-tide removed is now bit 4 (8), rather than bit 5 (16). Other minor changes to 'solve'; see sversn.f. convertc: Updates to check reasonableness, allow more file types, fix bug in reading vel file created by velrot with longer lines, and reverse the order from lat/lon to lon/lat for GEO or GEP files. sh_get_nav, sh_get_orbits: Add wget option. sh_PBS_gamit: changed "mailto" variable to "mail_to because" of some conflict/problem returning blank argument to -M option in qsub command, which stalls qsub. sh_eq: Removed from /com since made obsolete by eq option of sh_exglk. sh_metutil/metutil: Increase dimensions for z-file entries to allow 10-second sampling over 24 hours. sh_gamit: Add '-alpha off' and '+antalias' to ImageMagic convert to avoid transparency and pixelation in converting sky plot images from Postscript to gif; change default aprf to itrf08.apr. Updated hfupd.f and hfupd_ut.f to find renamed sites in station.info for updating. Updated template sestbl. to fix typo for 'Mag field'. Updated read_pmu_inp.f to account for GPS-UTC time difference. Fixed format "errors" in tsfit.f and tssum.f. Fix bug in computng average hydrological load for h-file. Added displace folder to updates. Allow by-passing uncompress when downloading RINEX files: /com/sh_get_rinex. Fixed bug in grdtab reading OSO list files with slightly different formats; correct comments for order of tokens on the grdtab command line. Copy from gamit/grdtab: grdtab, get_otl_list. Added "GCR APTOL" warning to email summary from sh_gamit. sh_get_orbits/sh_sp3fit/ngstot: GLONASS SVS can be can be present (but are ignored); add ESA finals (ESAF) as an orbit option; fix bug with SV id is blank rather than 'G'. sh_glred: add options to select the command file and to input a prefix string for the globk OPTION command; add ability to search for binary files with extent 'glb') (PBO, e.g.); fix problem with 'word too long'. Changes to better handle CYGWIN installations. Copy com/unimake and libraries/Makefile.config.new (and rename to Makefile.config after checking for compatibility with local settings). Updated sh_link_rinex and sh_PBS_gamit to use -expt option for links and copies. Updated sh_get_rinex to allow >1000 site downloads. Error with Word too long in original version. Update globk/autcln to implement new features (see help files) and updated pole tide definitions. Makefile.generic in gen_util updated. New script to compare a list of RINEX files with one or more station.info files to check for station.info.completeness, /com/sh_stinfo_list. Changed the USNO ftp command in gg/tables/ftp_info to use passive transfer; necessary since 1 February 2012. Updated kf/ctogobs/proc_phsfin.f to output mean residuals by elevation in 1-deg bins. ELMEAN lines. Bins with no data list 99.9 values. Updated frame_to_fra.f with help option and plate to use the feature. Allow automatic uncompressing of bzip2 files. Copy from com: sh_crx2rnx, sh_get_rinex, sh_gamit. Fix bug in getting the average atmospheric loading value into the h-file. Update 'profile' utility to handle changed vel-file format. Corrected "find" commands in install_software with trailing slashes to follow linked directories successfully when verifying or searching for libX11 and Xlib.h. Allow reading of ANTEX files with 0.5-degree grid spacing. Allow processing from c-files to check model changes. Read met values from a RINEX met file using free-format (work-around for improperly written RINEX v2 files). gamit/model/read_metrnx. Added NONE command file name feature to tsfit.f (in help but not implemented) . Change to model/read_metrnx.f to allow reading of mis-formatted RINEX met files from Matera. TRACK: Added new relative humidity option to GPT; fixed bug in velocity apriori coordinates; updated help file; new script sh_track_to_kml; account for model differences between L1 and L2; fix bug in reading the last epoch of an IONEX file; trap better bad clocks and data; add support for C2 range data. trackRT: TRACK updates plus fixed timing problem with trackRT and BNC2.5; updated handle BNC 2.6 feed format. Update DLAPACK routines to current versions to avoid a bug effective with gfortran 4.6 when run with optimization (now default unless -mno-align-double is sset). Fix bug in kf/svpos/svdiff.f when more than 12-channels of data. sh_upd_stnfo/mtinf: Fix bug with in and out station.info files with rx files; fix -rxo option; increase dimensions to work with current SOPAC station.info files. sh_setup: Change to copy in whatever is linked was station.info in ~/gg/tables, now nominally station.info.mit. Rename station.info.sopac.newfmt to station.info.sopac. Increase maximum channels for Sokia receivers from 14 to 24. Updated gg/tables/ftp_info for new server and directory structure at TU Wien for VMF1 file. Fixed conflict with useof :A option in kf/glist/glist.f. Updated RMS/NUM format in autcln for more than 9999 obs per satellite. Use 'tr' instead of potentially unavailable 'dos2unix' to remove extraneous carriage returns in RINEX files. Fix bug in sh_gamit for case of pre-existing SP3 files. Fix bug in gamit/lib/read_antex.f that causes a bogus SV antenna offset for PN30/SV35 (first time a PRN has been reused for the same SV Block). Update tables/otlcmc.dat for additional terms in the NAO_99b model. Fix problem with sh_glred when running a single day. sh_exglk: Cleaned up extraction of apriori coordinate lines. sh_get_orbits: Changed default eop_ser from bull_b to usno. sh_makeeqdef: New script to create globk earthquake file from http://neic.usgs.gov/ catalog. The command 'reset' in the eq_file will reset all of the input site extents to _GPS. Added a white-noise process to the mar_scale command so that some process noise can be applied between to h-files at the same time, useful if they were generated with differeent scales, e.g., ITRF2005 for IGS repro1 and ITRF2008 for data processed using itrf08_comb.apr. Allow 'getrel' to work with newer GLOBK velocity formats. Updated kf/utils/stinf_to_rename.f to handle current scheme of using only a 4-character ID for the initial site name. Fix serious bug in ARC affecting the SV yaw during eclipse. Fix bug in failing to use C2 when both C2 and P2 indicated with the RINEX header but P2 zero for SVs that have C2. Copy gamit/lib/rrinex.f. Kogan/King 110420 Fix bug in failure to apply center-of-mass correction (CMC) for ocean tide models when translating sp3 files to t-files; fix bug in writing MIT sp3 files. Add 'status' message for CMC corrections in both ngstot and ttongs. Copy from gamit/orbits: oversn.f, ngstot.f, rsp3hd.f, ttongs.f, wsp3hd.f, and sdtrit.f. Kogan/King 110414 Fix bogus header in ANTEX file. King 110322 Allow use of IERS2010 model for short-period EOP (Ray model still default). Copy cversn.f, pmut1_oceans.f, and Makefile.generic from /libraries/comlib; lversn.f and srotat.f from gamit/lib; fversn.f and mdmake.f from gamit/fixdrv. Herring/King 110223 Fix htoglb bug in making ephemeris file (occurs only with Intel compiler). Herring 110202 Change default setting in sh_plotvel to be compatible with GMT4. Ferhat/King 110119 Fix problem with setting day-loop span in sh_gamit with some operating systems. Shimada/King 110117 Copy into /tables itrf08.apr, IGS08_101117.apr (current IGS) (from IGS08.ssc), and IGS08_101117.eq (GLOBK rename, from soln_IGS08.snx) files. The eq fiile has names matching those in both itrf08.apr and IGS08_101117.apr; note, however, that the IGS files may have discontinuities later than those accounted for in itrf08.apr, and that the IGS08 apr file will not include coordinates for early segments used in ITRF2008. King 110106 Allow sh_find_data to work over a year boundary. Copy from /com. Shimada/McClusky/King 110106 Add the 2011 lunisolar tables to incremental_updates/tables/. (luntab and soltab were in the 10.4 distribution, but nutabl. was missing). King 110104 Remove bogus test on t-file year > 2010. Copy gamit/fixdrv/armake.f. King 110101/110103 Allow more up to 20 observables in RINEX files. Updated glist to output of sites by time all on one (long) line. Updated treatment of coordinates in renames globk (fixed problem with XCL renames with earthquakes that have been previously applied). Khandler/eq_name_chng.f Correct a bug in sh_gamit when linking IONEX files with multiple experiments. Add slant heights for the TRM39105.00 (Zephyr w/o ground plane) to hi.dat. Correct the echo of the atmospheric tidal loading model in the p-file. ** 10.4 vs 10.35 Introduction of a real-time kinematic module, trackRT. See help/trackRT.hlp for a complete description including instructions for installation. Changes to the code and svnav.dat to use an appropriate radiation-pressure model and antenna offset for Block IIF satellites (only PN 25 thus far). If you need to combine new processing with MIT or SOPAC global h-files that used the old models set ' apr_svs prn_12 100 100 100 10 10 10 0R ' to unlink the radiation pressure models. Both MIT and SOPAC will rerun these days to remove the problem. Two new programs, tscon and tsfit, are available for prototyping GLOBK solutions. They allow you to convert PBO, REASON, and other time series to GLOBK formats and to generate in an iterative but rapid fashion appropriate models for discontinuities and earthquakes in continuous time series prior to making long globk runs. See Chapter 4 of the GLOBK Reference Manual and the help files for tscon and tsfit. There is a new example, better designed for instruction on setting up processing with new data and incorporating a multi-year solution with GLOBK. The c-file and h-file formats have changed. C-files from prior releases cannot be used, but the code remains compatible with older GAMIT (ascii) and GLOBK (binary) h-files. The impetus for this change was to better document models used in the processing, particularly for receiver and satellite antenna models. Higher-order ionospheric (HOI) corrections (2nd and 3rd order, but not bending) are available in GAMIT, as described by Petrie et al. [J. Geophys. Res., 115, B03417, doi:10.1029/2009jb006677, 2010]. Using these terms requires downloading daily IONEX files containing total electron content (TEC) at 2-hr intervals from CODE into the [expt]/ionex directory and linking these into the day directory using '-ion' in the sh_gamit command file. To apply the effects in processing, set 'Ion model = GMAP' and 'Mag field = ITRF11' in the sestbl. See the new Chapter 8 of the GAMIT Reference Manual for more details. Thanks to Liz Petrie of Newcastle U (UK) for most of the work on this, and to Andrzej Araszkiewicz the Military U of Technology (Poland) for important contributions. The procedure for creating and keeping an updated station.info file has been better documented and enhanced by changes to sh_upd_stnfo. See the description on p. 27 of the Introduction to GAMIT/GLOBK manual, or the README of the new example. GLOBK will now accept multiple eq_files, allowing you to configure separate files for earthquakes, breaks, and editing (_XPS, XCL). If an eq_file entry uses the 'break' command rather than the 'rename' command, globk will assign the extent. A new RENM print option causes the full rename list to be output. glist now also allows for multiple eq_files, separated by colons. We are now providing in the GRIDS directory on everest.mit.edu atmospheric loading files that have been filtered to remove the diurnal and semi-diurnal thermal tides (S1, S2) using the procedure described by Tregoning and Watson (J. Geophys. Res., 14, B09403, doi:10.1029/2009JB006344, 2009). These can be used in conjunction with atl.grid (also on everest) comprising the IERS standard S1,S2 tides (Ponte and Ray, Geophys. Res. Lett.,29, 2153, doi:10.1029/ 2002GL016340, 2002) by renaming the filtered files (atmfilt_cm.YYYY) to the standard name for unfiltered files (atmdisp_cm.YYYY) in gg/tables, and setting 'Use atl.grid = Y' and 'Tides applied = 63' in the sestbl. Based on feedback from users, we hope to eventually discontinue distribution of the unfiltered grids. Note that earlier versions of the filtered files, from either ANU or MIT, were written incorrectly; further, there was a bug in the 'model' routine that read the sub-daily ATL from the u-file, causing the north and up components to be reversed. Any use of ATL should use the new grids and Release 10.4. There is a new plotting script, ts_plot.py, that reads PBO-style time- series files (.pos), created by tscon or tssum, rather than the mb_-files created by ensum. We plan to transition to the .pos format eventually since it is more complete and flexible, but this will happen slowly and only when we're sure that users can easily install all of the libraries needed to run ts_plot.py: Python V2.7 ( http://www.python.org/download/ ); Numpy ( http://new.scipy.org/download.html ), optionally scipy; matplotlib ( http://sourceforge.net/projects/matplotlib/files/matplotlib/matplotlib-1.0/ ) and egenix-mx-base-3.1.3 ( http://www.egenix.com/products/python/mxBase/ ). Changes to tsview to enhance performance when viewing and editing large numbers of time series; recording of the reasons for edits when creating entries for eq_files. There are updates to autcln and solve to better handle single-frequency data. The default in autcln now requires the user to include 'L1only' in the autcln.cmd file to process single-frequency data. Pseudorange resolution of wide-lane ambiguities (LC_AUTCLN) is now recommended for networks larger than 50 km and data from January 1994 (except with codeless receivers), the date from which empirically determined differential code biases (DCB) for each SV have been determined (dcb.dat) by AIUB. A new algorithm, implemented for Release 10.34, is more robust than the earlier one and gives some weight to the ionsophere-constrained phase WL. The defaults controls should work well for most networks with recent data, but you may want to test increasing the 'ion constraint' value (from the default .01 to .02) under maximum ionspheric conditions (2000-2002), or decreasing the value (to .005) for old data under minimum ionospheric conditions (1994-1998). There are now IGS RINEX files with as many as 11 obervables, requiring an update of the code (rrxhed and rrinex in gamit/lib) and remaking the executables with MAXOBT=11 in gamit/includes/makex.h. IGS ANTEX files begining with igs05_1499.atx have more comment lines than we originally coded (50). Set MAXTXT=1000 in gamit/includes/dimpar.h. and remake gamit. Prior to an October 2008 incremental update to 10.34, failure to make this change will result in a cryptic failure in suggesting a misnamed receiver or antenna. Release 10.4 will recognize ANTEX Version 1.4. Fix a bug in time argument for GPT; affects the annual signal in heights at the level of 0.2-1.0 mm. Thanks to Francois Fund (ESGT) for catching this and Paul Tregoning (ANU) for testing it. Fix bugs and deficiencies in sh_upd_stnfo/mstinf to obtain a better match of entries with RINEX files and to avoid multiple entries continuous stations when not needed. Add the ability to get a station list from sites.defaults and/or the /rinex directory, and the ability to use wildcards to get multiple IGS log files. Fix bug in distinguishing between Receiver Type' and 'Receiver SN', and 'Antenna Type' and 'Antenna SN' in reading station.info files (thanks to Olivier Charade (INSU/CNRS), allow more complete updating of station.info entries mid-session. Fix two bugs in fixdrv, both introduced in 10.35: 1) full-solution observable erroneously set when quick_pre observable = LC (allowed for backward compatibility) rather than LC_ONLY (preferred); 2) check on too-tight a priori coordinate constraints failed in sh_sigelv. Thanks to Liz Petrie (Newcastle) for catching both of these. Fix bug in sh_get_nav when the day number has leading zeroes, needed for CDDIS downloads with tscsh (one line of code missed in fixes applied to most scripts for 10.35 release). Thanks to S Shimada of NIED for the catch and Mark Murray for additional cleanup. Fix bug in kf/glfor/get_max_glbd.f that resulted in wrong dimension for the partial derivative array when earthquakes logs and/or apr_rot commands used. Updates to 'track' to account for millisecond range jumps. Fix bug in creating station.info entries with sh_upd_stnfo when the compiler option for floating point is not real*8. Thanks to Mark Murray (New Mexico Tech) for catching this. Fix bug in utility org2stab to get a stablization list from a velocity solution. Fix bug in sorting median process-noise values in sh_gen_stats. Correct a bug in sh_update_eop after detecting an invalid EOP series. Fix failure to link binary for utility ne2azel Add gradients to the output file of sh_metutil; trap an empty z-file from having no good observations at a site. Copy gamit/utils/metutil. Thanks to Mark Murray (New Mexico Tech). Modify sh_get_met to work with either Z or gz files, and for multiple days; update the met directories for CDDIS. Thanks to Kalev Rannat (Tallin U). Add ~60 receivers and ~10 antennas to rcvant.dat. Update to current IGS ANTEX file (igs05_1597.atx) and create igs05_1597_plus.atx by adding PCV models from NGS for 8 antennas not in the IGS file. Correct misspellings or official changes for 6 receivers or antennas. Changes to unimake and Makefile.config to facilitate installation under CYGWIN. See the notes 'CYGWIN' under updates/documentation. Thanks to Semi Ergintav (TUBITAK) and George Vergos (AUTH) for notes and testing. Add to sh_gamit the ability to invoke sh_metutil to get ZWD and PW (met-file) from an input z-file and o-file, and to archive, compress, or delete z- and met-files at the end of the run. Add script sh_met2pbo to concatenate multiple met-files. Fix header entries and format in writing SP3-C files. Update Hatanaka RINEX compression routines to version 4.0.4, fixing function name conflict with getline in gcc 4.4 and cleaning handling of line-feed at end of file. Add output lines in Euler-pole utility conveuler. Modify format in utility confpol to convert N,E sigmas and correlations to confidence ellipses. New utility check_siteid to compare 4-character IDs from glist and several external formats to detect conflicts. Fix bug in reading number of data types in simulation model. Fix compiler flags for Intel 9.0. Fix minor compile errors ** 10.35 vs 10.34 sh_gamit (and sh_get_orbits) allow the use of MIT orbits from global reprocessing 2000-present, typically a factor of two better than the operational IGS Final orbits prior to 2004 and without the scale discontinuity at Week 1400 (November 2006); see http://kg4-dmz.gfz-potsdam.de/igsacc/repro/igsacc_final.html. sh_glred (and sh_get_hfiles) allow the use of combined binary h-files from MIT or SOPAC reprocessing. The options for SOPAC sub-network ascii h-files now include 'igsall' which will fetch all of the global sub-nets available on the day. (Note that it is not legitimate to use only a subset of the igs? h-files.) See the revised Chapter 4 of 'Introduction to GAMIT/GLOBK' and the presentation 'Reference_Frames.ppt' from the March09 Advanced Workshops on the GAMIT/GLOBK documentation web page for a discussion of strategies for establishing the reference frame for time series and velocity solutions. Allow for 10 or more observables in a RINEX file. New versions of tsch now enforce a standard (apparently always present) that numbers with leading zeroes are treated as octal. This causes problems with GAMIT day number lower than 100 in a dozen shell scripts. Updates to 'track' to better handle short broadacast empheris and to resolve some issues with independent bias flags. Using broadcast ephemeris in processing: Update sh_bcfit to check for validity of dates and to generate a session.info file if it does not exist. New sh_gamit option to block updating of experiment l-file (useful to avoid clashes when running in a cluster environment). New utility (org2stab) for an generating unweight list for hfupd and a stab_site list for glorg from a globk/glorg velocity solution. New utility (grw) for generating quickly sig_neu downweight commands while viewing a time series. Change directory structure for SOPAC h-files and OLG RINEX files, and add BKG met files to tables/ftp_info. Changes to sh_gamit make the downloading and use of orbit files and IONEX files more robust. CVIEW: Fix bug in reading in x-files; fix bug in screen display with OSX and some other laptops. Fix bug in sh_gamit when using IGS ultra-rapid orbits. Fix bug iin identifying l-file coordinates to update when the site is fast-moving (e.g. on ice). Fix bug in sh_org2vel when plate name > 7 characters. Fix bug in sh_oneway for GMT formating Fix bug in creating x-files with Topcon firmware > 3.0. Fix bug in assignment of number of zenith or gradient parameters from an input interval (one too few in recent releases). Thanks to Ian Thomas of Newcastle for the catch. Fix bug in sh_sp3fit, which incorrectly used IGS92 rather than EGM96 gravity field for the final t-file. New 'fixdrv' option: Allow skipping of the loose solution ('skip loose = y') to avoid poorly conditioned matrix with very short baselines. Allow of VMF1 ZHD with GMF; works now if UFL first 'met source' option and either map.grid or map.list is present. Fix bug in reading ANTEX files with more than 50 comments. The command-line coordinate conversion program (a replacement for 'tform' in most cases) has been updated to output longitudes as either -180 to +180 or 0 to 360 and to fix a bug in outputting spherical l-files. Add to sh_gamit summary the names of sites removed from processing by the minxf control in processing.defaults. Cleanup of 'solve': changes in the information written to the q-file by, remove some screen print; initialize 'avgload' variable; remove obsolete single-difference code. Add site code to error messages in reading ANTEX files Support 'gz' with Hatanaka-compressed RINEX files. tables: Updates to svnav.dat and antmod.dat for SV49/PN01; it is essential to make this update for data acquired after the launch (March 2009). Add luni-solar tables for 2009. New receivers: NOV4G2 (NOV OEM4-G2), LC123G (LEICA GX1230GG), LC902G (LEICA GMX902GG), LC1220 (LEICA GX1220), SEPX2E (SEPT POLARX2E), SEPX2C (SEPT POLARX2C), SEPAX1 (SEPT ASTERX1) SEPAX2 (SEPT ASTERX2), ALTPS3 (ALTUS APS-3), LEICA GRX1200+, LEICA GRX1200+GNSS, LEICA GX1210, LEICA GX1210+, LEICA GX1220+, LEICA GX1230+, LEICA GX1220+GNSS, LEICA GX1230+GNSS, NOVV3R (NOV OEMV3-RT2) New Antennas: TPSG3A (TPSG3_A1),TR5980 (TRM59800.00), SOKGSR (SOK_GSR2700ISX), AER167 (AERAT1675_29, NOV702 (NOV702), LEICA AX1201, LEIAX1203+GNSS, LEIATX1230+GNSS, SEP220 (SEP_POLANT+), SEP29S (EP_POLANT+_GG), TRMZGS (TRM57971.00). Correct typos in TOP72110 and SOK_RADIAN_IS entries. ** 10.34 vs 10.33 sh_gamit (and sh_get_orbits) allow the use of IGS ultra-rapid orbits that include predictions, facilitating near-realtime generating of coordinates or atmospheric parameters. Fix a serious bug in 'model' whereby the pole tide, if requested, was returned as zero or undefined. The effect is up to 2 mm in horizontal and 5 mm in heights for a global solution, less than 1 mm horizontal for interstation differences up several thousand km. The problem is avoided if you omit the pole tide in GAMIT and set 'app_ptid all' in GLOBK. (You cannot simply set the control in GLOBK because GLOBK will not apply the tide if the h-file indicates that it has been applied in GAMIT.) tables: Fix error in svnav.dat in setting start dates for PN29/SV57 and PN07/SV48, both erroneously set for 2007 10 20 instead of their actual launch dates. There is no problem with the PN29 error since the previous SV having this PN was dead by 2007 10 20 (day 293), but for PN07, using the old files will result in errors for the period between day 293 and the demise of the old PN07 on day 345. Thanks to Sigrun Hreinsdottir of U. Arizona for catching this. Add percentage of ambiguities resolved to sh_gamit summary file New receivers: SOK GSR2600, JPS EGGDT, SEPT POLARX2E, TPSHIPER_LITE TPS-HIPER-GGD (same as TPS HIPER-GD). New antennas: Micropulse WAAS 2225NW, TRM41249USCG; fix name of TPSCR.G3 to conform with IGS standard. Correct SLAGP to SLHGP for ASH700228A,B,C in hi.dat. Handle correctly site renames--either explicit for instrument changes or implicit for earthquakes---within a session ('model'): use only data from the longer of the two spans. There is a new coordinate-conversion utility 'convertc' which will take as input a GLOBK apr-file (Cartesian) or a GLOBK velocity file or glist file (geodetic), and output an apr-file, velocity file, or KML file (for Google Earth). of the functions formerly performed by tform; Track: Fixes for milli-second clock resets that are not correctly accounted for in the RINEX files. Small bug fixes and some format changes. Increased size to allow 12-hrs of 2-Hz data. Plate-referenced apr file for ITRF 2005: The files distributed with 10.33 were in error, having used the TIRF 2000 poles to rotate itrf05.apr to the major plates rather than the new poles from Altamimi et al. [2007]. The new files are correct, and also have site extents (_1Ps, 2PS, etc) corresponding to the GLOBK convention used in itrf05.apr and corresponding to the names used in itrf05.eq_rename. Corrected versions of itrf05_aust.apr, itrf05_pcfc.apr, itrf05_soam.apr, itrf_noam,apr, itrf05_eura.apr, and itrf_anta.apr. Updated algoritm for clock estimation in 'autcln'. Fix bug in setting DCBs for non-Trimble receivers when AS off. Occured only prior to DCB availability, so no effect except a rare * printed in x-file. Wild cards now allowed for all GLOBK commands. Fix bug in making sky plots with sh_gamit. New feature in sh_plotcrd to select stations by region. Fix bug in locating RINEX and h-files on the local file system. Increase the dimensions of MAXCHN to account for receivers with more than 20 channels. Change in Apr., Unc., and Int. formats of globk/glorg print files to allow velocities of -999.99 m/yr (motion on ice). Fix a problem reading an integer elevation cutoff from the sestbl (should be used only with LC_HELP). Thanks to Rossana deMarco of CNR-IRPI. ZHD in 'model': There is a bug that causes a fatal when RNX UFL is the met heirarchy and the RINEX values are bogus. The bug is complicated by the presence of a redundant and unused file in the distribution. To fix the problem, copy /gamit/model/get_met_source.f and Makefile.generic, remove from your /gamit/model directory the file get_ion_source.f, Makefile, and model_lib.a; then regenerate the Makefile (unimake -m model from the /gamit level) and remake model. Thanks to Liz Petrie at U Newcastle for catching this. 'yawtab': Fix bug in call to oversn.f, effect is only to miss version numbers. Mac OSX Intel: Fix inadvertent carriage return in /libraries/Makefile.config. ** 10.33 vs 10.32 Update svnav.dat and antmod.dat (igs05_1461.atx) for SV55/PN15 (25 Oct 2007) and SV57/PN29 (21 Dec 2007). Lunisolar tables for 2008; IAU 1980 nutations replaced by IAU 2000 nutations. H-files created with the old model will be automatically rotated to the new model. sestbl: Change 'Antenna model' to AZEL and 'SV antenna model' to ELEV to use absolute PCVs; sittbl: Loosen suggested constraints to relect preference for ambiguity resolution. Fix bug for OTL interpolation for sites between 0 longitude and the first grid point (0.25 deg for FES2004, 0.5 deg for other grids). 1 Fix bug in reading and using ATML or VMF1 values on the last day of the year. New receivers: LEICA GRX1200GGPRO, TPS NETG3, TPS MAXOR, TPS EUROCARD, NOV OEMV3 New antennas: LEIAT504GG, TRM4600, TRM41249USCG, NOV702GG, NOV702GL, NOV702 rev.4 Replace 'compress' but 'gzip' in scripts. Velocity histograms (sh_velhist): Add capability to set a minimum as well as maximum sigma and a geographicbox. Enhancements to facility cluster computer: In /com, sh_PBS_gamit, sh_gamit, sh_sp3fit, sh_get_orbits. Add worst, best, median in calculating random walk for'real sigma' algorithm in sh_gen_stats, sh_enstat: guess_rcvant.dat: Added missing carats (^) to entries for the TOPCON/TPS receivers and antennas. TRACK: Fixed 10-hz epoch problem and reading gamit zenith delay estimates. sh_sp3fit: Fix bug reading sp3c files (old code wrong but works ok with g77 but not gfortran). Thanks to Demitry Balandin (Inst Geophysics Urals, RAS) for finding this. Fix bug with merging station.info files (sh_upd_stnfo, mstinf2). New RINEX compresssion routines provided by Yuki Hatanaka (GSI-Tsukuba) to handle RINEX Version 3.0. Fix header comments for sh_metutil (correct units for temperature and ZHD). Using Az-El PCV corrections: There was an error in interpolating the ANTEX tables when both elevation- and azimuth-dependent corrections are used. On rare occasions, depending on the details of the model and distribution of observations, the bug could generate up to 10 mm error in heights and zenith delays, though usually the effect is negligible. Thanks to Wang Min of CEA for the catch. Using GPT for a priori pressure and temperature. There was a bug in the time argument used to read the GPT model; the change in values, however, is very small and has a negigible effect on estimates of heights or zenith delays. Copy gamit/model/atmdel.f. Thanks to Peter Smythe of ANU for the catch. Fixed problem with converting mutiple SINEX files with htoglb. sh_rx2apr (svpos); Fix problem using RINEX Version 1 (missing 'G' for SV name). Change to autcln to improve postfit editing for old data. Change the default gravity field for 'arc' to EGM96' (fixdrv). Add gravity and nutation models to the h-file. Fix bug in 'xtorx' with new-format station.info. ** 10.32 vs 10.31 Installation: 1) For the Intel compiler the link for libraries/comlib/execute_intel.f should point to execute_hp.f, not exectute_g77.f; 2) for OSX 8.8, Makefile.config updated to read 'OS_ID Darwin 5200 8800'; 3) for gfortran, restore a missing link for csubs.f in gamit/clean. Ambiguity resolution with LC_AUTCLN: changes in autcln will improve resolution of WLs signficantly, both before and after the introduction of DCBs (April 2001). Changes in solve will reduce or eliminate cases of failed inversions due to problems removing dependent biases. Ambiguity resolution with LC_HELP: WL resolution in the presence of high ionosphere is improved by rescaling the WL bias sigmas by the rms deviations for each baseline rather than by the overall nrms of the ionosphere-constrained solution. Narrow-lane ambiguity resolution : With both LC_AUTCLN and LC_HELP, deviation-scaling of the NL sigmas is also implemented. The effect on resolution is not large, but the statistical summaries now printed in the q-file provide useful diagnostics for tuning ambuiguity resolution. (Changes also to sh_gamit to pick up the new print for the mailto summary file. VMF1: Changes to allow multiple formats of VMF1 station-list files. autcln.cmd: The template gg/tables had an extraneous blank before 'POST lc_autcln' that should be removed to use LC_AUTCLN in GAMIT. METUTIL: Fix two bugs using RINEX met files: 1) add correctly an extra value at the end to allow proper interpolation; 2) change temperature units to K. Copy gamit/utils/metutil.f. Ashtech choke ring antennas: Bug in setting the aliases for these antennas for purposes of calculating direct height from slant height, results in a direct height error of about 1 mm for an antenna mounted 1 m above the mark. New receviers: Added Leica GMX902 and Trimble R7 and R8. Creating new-style station.info: Add self-documentation for program gamit/hi/conv_stnfo) that converts an old-style to new-style station.info file. sh_plotvel: Better diagnostics. RINEX compression: Change sh_get_rinex to use sh_crx2rnx instead of sh_uncompress, and sh_uncomopress to use gunzip instead of uncompress. GLORG: Output option (MIDP) which references solutions to the midpoint epoch of the combined data; should not be used when large stochastic position estimates are present. ** 10.31 vs 10.30 hi.dat: The ASHDMG antennas was misnamed ASHGD3; fixed errors in slant range dimensions for the ASHDMG, ROGSNR, and ROGRMA antennas. None of these errors will affect runs with DHARP as the height measurement. Atmospheric delay models: Fixed bug in GPT; made reading of a RINEX met file or a u-file ZHD (VMF1) more robust. Met utilities: Fixed problem with RINEX met files and improved interpolation for sh_met_util. gcc/g77 3.4.x bug: Added a temporary patch to a lapack routine causing 'solve' to stall in an infinite loop when compiled with level 3 optimization with gcc/g77 3.4.x and possibly 2.95. kf/track : Fixed problem reading RX files with internal comments with more than 12 lines rd_rinex_obs.f. Deleting eclipse data: Has not worked properly for some time, but now fixed in gamit/utils/autecl.f. Problems compiling and using ocean loading with the HP f90 compiler.. Download the entire incrementals to get Makefile changes, system-dependent routines, and several GAMIT routines with uninitialized variables. Initialization problem for 'idatum' in fixdrv. Sky plots not created by sh_gamit because sh_make_sky_gifs looks for 'autcln postfit' in the sestbl, but this command is no longer necessary; fixed sh_make_sky_gifs. Fix intel compiler flags for 'solve'. Thanks to S. Shimada (NIED/Tsukuba) for catching this. SV-dependent data reweighting did not work with (default) 'AUTCLN Postfit = R' option. Copy gamit/fixdrv/bmake.f. Thanks to Wang Min (CEA/Beijing) for catching this. Add traps for # SVs on g-file or nav-file > maxsat. Copy gamit/fixdrv/satnmr.f and gamit/makexp/checke.f hi.dat: Correct entries for SLBCR for ROGSNR and ROGRMA. Thanks to Sigrun Hreinsdottir of U. Arizona for catching this. OS-dependent routine for HP: The HP requires a special version of libraries/comlib/systime.f so so we have re-introduced OS-dependent versions; also we need an HP version of kf/gen_util/execute.f. Copy cversn.f, Makefile.generic, and the systime_HHH.f version you need from /libraries/comlib; and Makefile.generic and the execute_HHH.f version you need from kf/gen_util. MAKEX problem opening and closing hi.dat. Copy makex.f and closem.f from gamit/makex. svnav.dat: Update for PN12/SV58. Copy svnav.dat or svnav.dat.igs, depending on whether you want to maintain backward-compatibility or IGS-consistency for SV antenna offsets when using a relative PCVs (with absolute PCVs, these values are overridden by antmod.dat). Note that SOPAC used an old svnav.dat for processing of PRN 12 between 2006 12 15 and 2007 01 13, so if you are combining SOPAC H-files with your own processing, you must either exclude PRN 12 from your processing or allow the radiation-pressure parameters to adjust independently in the SOPAC and your own h-files. This is easily accomplish by 'apr_svs prn_12 100 100 100 10 10 10 0R'. As described in the 10.2 Release Notes, a similar situation pertains to PN22/SV22 as a result our error in listing its block number. Its radiation pressure parameters should be decoupled from all SOPAC or MIT H-files created between 1993 and 2003. GLORG: Fixed I*8 declaration for memory allocation on 64-bit machines when plate estimate used. HTOGLB: Sinex routines changed to read and deconstrain center of mass and satellite anttena offset estimates. Added codes for IGS05_1404 antmod file to SMod_to_full.f. sh_rx2apr: svpos/svdiff now work properly with the C2 observable. Apriori coordinates: GAMIT/model now has the ability to use the full 8-character station name when reading a GLOBK-style L-file together with a GLOBK eq_file. The link scripts will now attempt to link into the day directory, via the [expt]/tables directory a file named 'eq_rename', which you may link (in [expt]/tables or gg/tables to a particular rename file. In the gg/tables directory distributed with 10.31, theeq_rename file is linked to a file derived from the steps documented for IGS stations by IGS Analysis Center Coordinator Remi Ferland (NRCan) and distributed as IGS-Discontinuities.snx. This file is incomplete even for IGS stations but may serve as a starting point for GAMIT users. Trapping bad a priori coordinates: If you overconstrain a station's coordinates in the GAMIT run, you will get an elevated nrms and degraded ambiguity resolution. There is now a test for this in 'solve', based on an adjustment more than twice the apiori constraint. The script sh_sigelv in the batch file now checks for the requisite lines in the q-file, and if a station has a large adjustment in the first ('p') solution, its constraints are relaxed for the final ('a') solution. N-file reweighting: Because the N-file is now always written (for trapping bad a priori coordinates), the sestbl command for invoking phase-error reweighting is changed from 'Use N-file = Y' to 'AUTCLN reweight = Y'. However, the old command will still work in 10.31. ** 10.30 vs 10.21 Speed of computation: New inversion routines have been added to SOLVE, improving the speed by about 30%. Note that gfortran will gain another 10% compared to g77, and the Intel compiler about 25%. sh_gamit.log: All of the GAMIT batch file status, warning, and fatal messages are now written into the log file, and more traps have been added to stop the run once a fatal is encountered. This should make it easier to find the source of a failure. Fix logic to 'cfmrg' and 'solve' to allow more than 30 SVs. This change is necessary if you use the full complement of SVs currently distributed with IGS orbits. Fix sh_gamit problem resulting in deleted rinex files when making Hatanaka compressed RINEX at the end of the run. Copy /com/sh_gamit. Add aliases for JNS versions of Topcon 'zero-centered' antennas. Copy gamit/lib/ant_alias.f. AUTCLN: Fixed problem with postfit editing bias flags, added additional calls to trim one way to remove bad data and fixed P1/P2 residual output, and changed default # of channels from 14 to 15 to accommadate newer receivers (also fixable by setting 'max_chan 15' in autcln.cmd.) MAKEXP: Fixed problem when a station.info entry begins later than the session start and on a separate day. Avoid problem when station.info has no entry for data in a RINEX file linked into the day directory but not used in the session: rstnfo now prints only a warning if called by makexp; if there is a true omission, makex will stop with a fatal. SOLVE: Fixed bug in printing observable/ambiguity type in q-file, Fix bugs when estimating a single zenith delay. Fix bug (introduced with incrementals of 060227) in assigning gradient apr values; thanks to Sunil Bisnath of CfA and Mathilde Vergnolle of U. Luxembourg. Fix a bug in reading the batch file and writing the q-file header when there are more than 50 sites; thanks to Kirill Palamartchouk of TU Crete. GLORG: Added eq_dist option for estimates of logarithmic parameters; fix some bugs in earthquake reporting. Plotting time series: The new print option PBOP in glorg will generate output lines in the format adopted by the U.S. Plate Boundary Observatory (PBO); plots are generated by new program tssum. GLBTOSNX: Improved output of renamed sites. Allow true C2 observables by not changing 'C2' to 'P2' unless the data are earlier than 2004 (needed for Trimble serial P-code data). sh_merge_rinex: Now allows input of the sampling intervel. Thanks to Seiichi Shimada of NIED for the fix. sh_make_rinex: The script to run to run TEQC on raw data files to create RINEX had a bug that prevented mixing of Trimble data files with those from any other receiver type; now fixed. SP3-C files: Fix option (-u) in sh_sp3fit for excluding SVs, thanks to Wang Min of CEA Beijing. Fix problems when flags are needed, thanks to Mark Murray of New Mexico Tech. Fix bug in naming g-file with 2- or 3-day fits. PCV models: Updated versions of the IGS01 relative and IGS05 absolute PCV models in gg/tables: igs_01_antex11.pcv and igs_05.atx. The igs_05.atx file has a model for the new PRN 17 (SVN 53) as well as revisions of additions for about 50 ground antennas. TRACK: Updated to handle sp3c format and new IIRM satellites; small changes to better handle bad pseudorange data. Fix non-working -u option (exclude SVs with unknown accuracy code) in sh_sp3fit. Thanks to Wang Min of CEA for the catch. McClusky 060306 Missing data from Leica GRX1200: Fix omission of SVs in upper channels when an incorrect firmware version is used in station.info; thanks to Wang Min of CEA Beijing. sh_baseline: Allow turning off of error bars and setting the tick and labels of the y-axis. sh_upd_stnfo: Fixed bug in updating station.info from a RINEX file when more than 12 SVs listed in an epoch line; thanks to A. Zubovitch of RAS Moscow. Bogus antenna heights: Fix bug that set HI=0. when using hi.dat with a missing height code. unify_apr: Better handling of sites when they appear multiple times in pos_mode and vel_mode. merge_apr: New program to merge two a priori files with no duplicate entries. Fix non-POSIX standard calls to 'head' and 'tail' command in > 100 scripts (old form does not work with Fedora Core 5). Use gzip/gunzip instead of compress/uncompress in locating and creating RINEX files. (Still more scripts to change for this.) Copy from /com: sh_find_data and sh_make_rinex.