Skip Navigation | ANU Home | Search ANU
The Australian National University
Research School of Earth Sciences
Earth Physics

Atmospheric Pressure Loading: Application in the GAMIT software

From version stdrel.10.2 onwards it is possible to apply atmospheric pressure loading at the observation level within the GAMIT software. This is done in two steps

    Application of non-tidal deformation

    The non-tidal deformation is computed and applied in the CM reference frame (centre of mass of the solid Earth plus fluid loads). Since the GPS satellites are sensitive to geocentre motion (the translations between the centre of mass of the solid Earth, CE, and CM), the CM frame is the most appropriate reference frame. All reductions of GPS observations are actually performed in the CM reference frame.

    Binary files are available from 1990 to 2005 (up to September only) in which the deformation in NEU is provided every 6 hours on a 2.5 x 2.5 degree grid. The values are interpolated (bi-linearly) to the actual coordinates of each station. A new program, GRDTAB, performs the interpolation and generates a file, atmlod.DDD, that is subsequently read by MODEL. The grid files are generated by Tonie van Dam (ECGS).

    Application of tidal deformation

    The S1 (diurnal) and S2 (semi-diurnal) atmospheric "tides" also be applied. The amplitude and phase are extracted in program GRDTAB for each site from a (binary) global grid file for the north, east and up components. These values are appended to the u-file and are subsequently applied in MODEL.

Comments on ATM Loading

Tregoning and van Dam (2005) investigated the application of tidal and non-tidal atmospheric pressure loading (ATML) at the observation level or as a daily-averaged correction. They showed conclusively that applying ATML at the observation level (ie the way that we have set it up within GAMIT) produces significantly better reduction in height WRMS than applying daily-averaged corrections. Therefore, we advocate that users activate this new feature within GAMIT.

Tregoning and van Dam (2005) also showed that it is not yet possible to produce a truly "non-tidal" ATML model by removing the diurnal and semi-diurnal atmospheric "tides" using the approach of Ponte and Ray (2002). In fact, applying the tidal models of Ponte and Ray (2002) and the associated "non-tidal" ATML yields a worse solution than just applying the ATML convolved from the raw NCEP pressure data - data that contains a partial (but not complete) sampling of the S1 and S2 tides.

The global grids of ATML provided with stdrel.10.21 are therefore the "partial-tide" versions of ATML. That is, the raw NCEP pressure data have been convolved without any attempt to remove the partial tides. The convolved pressure data of the tidal models of Ponte and Ray (2002) are also provided. Users can choose to apply just the tidal deformation if they wish but - naturally - should not apply this as well as the "partial-tide" versions of the ATML because the tidal component would be applied twice.

Figure 1. Reductions in WRMS of station heights. a) Non-tidal at the observation level relative to daily-averaged non-tidal. b) raw NCEP ATML (containing stronger partial tides) at the observation level relative to tidal and non-tidal component (both at the observation level). A positive value means that the first-named solution has a smaller height WRMS than the latter solution (ie a positive reduction). From Tregoning and van Dam (2005).

Figure 1a shows that station heights have smaller WRMS values when ATML is applied at the observation level. This provides the justification for making the effort to apply it! Figure 1b shows that solutions using just the raw NCEP pressure data (ie make no attempt to remove the partial tides) yields station heights with smaller WRMS values than from the approach of applying separately the tidal and non-tidal components. While the latter is intuitively the better approach, at present the limitations of the tidal models seem to make this worse than one would expect.

All of the above is explained in Tregoning and van Dam (2005). Please contact Paul Tregoning if you have any specific questions.

Papers relating to ATML coded within GAMIT

Ponte, RM and R. Ray, Atmospheric pressure corrections in geodesy and oceanography: A strategy for handling air tides Geophys. Res. Lett., 29, doi:10.1029/GL016340, 2002.


Tregoning, P. and T. van Dam Atmospheric pressure loading corrections applied to GPS data at the observation level Geophys. Res. Lett., 32, L22310, doi:10.1029/2005GL024104, 2005.
 pdf  

Tregoning, P. and T. van Dam.
The effects of atmospheric pressure loading and 7-parameter transformations on estimates of geocenter motion and station heights from space-geodetic observations. J. Geophys. Res., 110, B03408, dio:10.1029/2004JB003334, 2005.
 pdf  
 

(top of page)

How to activate ATML within GAMIT

The non-tidal ATML is activated with a single line in sestbl. :

Apply Atm loading = Y ; Y/N for atmospheric loading; need atmdisp.YYYY grid file

The tidal component of the modelling is turned on by activating the 6th bit of the bit-mapping for the tides applied :

Tides applied = 47 ; Binary coded: 1 earth 2 freq-dep 4 pole 8 ocean 16 pole tide 32 atmosphere (not yet coded) default = 15

    Additional Files Required

    You will need the non-tidal loading grid files as well as the grid file for the tidal loading. The (binary!) files available at present are listed below. Note: each atmdisp file is ~180Mb.

    For the moment, the files are accessible through the GAMIT/GLOBK distribution web page:

    http://www-gpsg.mit.edu/~simon/gtgk/gtgk_download_form.html

    (select ATM loading grid files as the "desired product for download")

    The files available are:


      atmdisp_cm.YYYY : computed in the CM frame for years 1990-2005
      atmdisp_ce.YYYY ; computed in the CE frame
      atmdisp_cf.YYYY : computed in the CF frame
      atmasc_c?.YYYY.Z : compressed ascii versions of the above files. These files are about 50 Mb each.

    Each of the atmasc_c?.YYYY files needs to be uncompressed then expanded into a binary atmdisp file by running the program ATMCOMP. There is a loss of accuracy of < 1 mm in the atm loading computations in this process. If this is an acceptable level, downloading the compressed ascii versions will be around 3 times faster than the full binary versions of the files.

    ATM tides

    Below is a link to the atm tidal pressure data of Ponte and Ray (2002) convolved into site deformation. As shown by Tregoning and van Dam (2005), the application of the model improves the height WRMS for most sites near the equator but tends to add noise at other sites and actually degrades the results. Users can make their own decision as to whether to use this model or not.

    atl.grid   (this is only provided in the CM frame)

    In addition, you will need updated files in gg/com that are provided in stdrel.10.2 and subsequent versions (currently stdrel.10.21)

    Other Recent Changes

    The following recent modifications can be invoked in GAMIT processing:

      Lunar Eclipse Modelling:
      Lunar eclipses = Y ; model lunar eclipses on the satellite SRP parameters

      IERS2003 solid Earth tide model:
      Etide model = IERS03 ; IERS92/IERS03 (default IERS92)

      Ambiguity resolution by autcln:
      Choice of Observable = LC_AUTCLN ; L1_SINGLE/L1&L2/L1_ONLY/L2_ONLY/LC_ONLY/LC_AUTCLN

      NOTE: you will also need to add a line " lc_autcln" to your autcln.cmd file

      SV Antenna Model = ELEV ; turns on sv antenna phase centre patterns

      NOTE: you MUST have the igs_05.atx file (or IGS_test05.atx) for this to work

    Is it worth the effort?

    Below are time series of height estimates at Davis, Antarctica for 2004. The four plots are (in order from the top):

      1. No atmospheric pressure loading corrections
      2. Daily-averaged correction applied (CF frame) after transformation to ITRF2000
      3. Daily-averaged correction applied (CM frame) before transformation to ITRF2000
      4. Atmospheric pressure loading corrections applied at the observation level.






    Bonne courage et bon bricolage ..... :-)

     

    Publications Relating to Atmospheric Pressure Loading

    1. The effects of atmospheric pressure loading and 7-parameter transformations on estimates of geocenter motion and station heights from space-geodetic observations.
    Tregoning, P. and T. van Dam.
    J. Geophys. Res., 110, B03408, dio:10.1029/2004JB003334, 2005.
     pdf  
     

    2. Atmospheric pressure loading corrections applied to GPS data at the observation level
    Tregoning, P. and T. van Dam
    Geophys. Res. Lett., 32, L22310, doi:10.1029/2005GL024104, 2005.
     pdf  

    Other publications in 2005

    1. The impact of tropospheric mapping functions based on numerical weather models on the determination of geodetic parameters,
    Boehm, J., P.J. Mendes Cerveira, H. Schuh, P. Tregoning
    IAG Proceedings, Cairns, Aug. 2005, submitted Sep 2005
     pdf  

    2. Deformation in the Jura Mountains (France): First results from GPS measurements,
    Walpersdorf, A., S. Baize, E. Calais, P. Tregoning and J.-M. Nocquet,
    Earth Planet. Sci. Lett. , submitted April 2005 .

    3. Effect of post-seismic deformation on Earth Orientation Parameter estimates from VLBI observations: a case study at Gilcreek, Alaska.,
    O. Titov and P. Tregoning,
    J. Geodesy, doi:10.1007/s00190-005-0459-9, 2005.
     pdf  

    4. Tectonic interpretation of aftershock relocations in eastern Papua New Guinea using teleseismic data and the Arrival Pattern method.
    Tregoning, P., M. Sambridge, H. McQueen, S. Toulmin and T. Nicholson.
    Geophys. J. Int.,160(3), 1103-1111, 2005.
     pdf  

    A full list of publications from geodetic and tectonic studies can be found here


    Return to Top of Page

    Return to RSES