Johannes Boehm, 15 May 2010
---------------------------
last update: 19 Aug 2016 (GRAD added)
contact: johannes.boehm@tuwien.ac.at
This readme file contains information about troposphere delay parameters which
are provided with the project >>GGOS Atmosphere<< at the Institute of Geodesy
and Geophysics at the Vienna University of Technology.
At the moment, three different types of parameters/models are provided:
1. Vienna Mapping Functions 1 (VMF1)
2. Empirical Models (like GMF or GPT)
3. Other Parameters (z200, Tmean, LHG, GRAD)
>> 1. Vienna Mapping Functions 1 VMF1
In the last years, troposphere mapping functions have been developed which are
based on data from numerical weather models, e.g. from the European Centre for
Medium-Range Weather Forecasts (ECMWF) [http://www.ecmwf.int]. The Vienna
Mapping Functions 1 (VMF1) are relying on empirical equations for the "b" and
"c" coefficients of the continued fraction form, whereas the "a" coefficients
are determined from rigorously raytraced mapping functions at 3 degrees
elevation. The VMF1 are provided on a global grid (2.5 x 2.0 degrees) as well
as at selected sites with a resolution of 0.25 degrees for recent data.
Recommendation: Use the gridded VMF1 data for GNSS or DORIS analysis.
Important: Please mind that the gridded hydrostatic "a" coefficients are valid
for zero height, whereas the site dependent values of "a" are valid for the
actual station height. Thus, for VMF1 given on the grid you need either the
subroutine vmf1_ht.f [http://ggosatm.hg.tuwien.ac.at/DELAY/SOURCE/vmf1_ht.f] or
vmf1_grid.m [http://ggosatm.hg.tuwien.ac.at/DELAY/SOURCE/vmf1_grid.m], whereas
for the site dependent VMF1 you need vmf1.f
[http://ggosatm.hg.tuwien.ac.at/DELAY/SOURCE/vmf1.f].
Reference for VMF1:
Boehm, J., B. Werl, and H. Schuh (2006), Troposphere mapping functions for GPS
and very long baseline interferometry from European Centre for Medium-Range
Weather Forecasts operational analysis data, J. Geophys. Res., 111, B02406,
doi:10.1029/2005JB003629. Download from AGU:
[http://www.agu.org/journals/jb/jb0602/2005JB003629/2005JB003629.pdf].
Warning: The coefficients to determine 'c' in Table 1 of the JGR paper are
wrong. Please use the correct Table 1 as provided in the corrected version of
the paper: [http://ggosatm.hg.tuwien.ac.at/DELAY/DOC/boehm_etal_3rd.pdf].
The source code on the webpage has always been correct.
>> 1a. Gridded VMF1
Gridded VMF1 coefficients are available in two different formats. In both
formats, the coefficients "a" are given on a global grid with 2.0 degrees
sampling from north to south and 2.5 degrees sampling from west to east, and
for each parameter there are four files per day, i.e. at 0, 6, 12, and 18 UT,
and they are stored in yearly directories. In addition to the "a" coefficients,
the hydrostatic and wet zenith delays are provided on the grid, too. Their values
are in m. These zenith delays correspond to the ellipsoidal heights given in the
file orography_ell [http://ggosatm.hg.tuwien.ac.at/DELAY/GRID/orography_ell].
The routine vmf1_grid.m [http://ggosatm.hg.tuwien.ac.at/DELAY/SOURCE/vmf1_grid.m]
automatically reads the gridded VMF1 files, time-interpolates from the NWM epochs
to the specified mjd, transforms the "a" coefficients as well as the zenith delays
to the desired height, bilinearly interpolates from the grid points to the desired
location and outputs the respective mapping factors and zenith delays.
The routine vmf1_ht.m [http://ggosatm.hg.tuwien.ac.at/DELAY/SOURCE/vmf1_ht.m], on
the other hand, outputs only mapping functions. It is to be applied just like
vmf1.m, with the suffix "_ht" meaning the transformation of the hydrostatic "a"
coefficients from zero to h_ell.
The VMF1 from analysis data of the ECMWF are made available with a
latency of less than 34 hours. You can find an example for the calculation of
the VMF1 at [http://ggosatm.hg.tuwien.ac.at/DELAY/DOC/example.pdf].
Warning: Up to 30 June 2009, the grid point: lat = 30 deg, lon = 87.5 deg
(Himalaya) is wrong. Replace it by a neighbouring grid point.
VMF1 grid files are available in a 'standard format' at
[http://ggosatm.hg.tuwien.ac.at/DELAY/GRID/STD]. Files with the hydrostatic "a"
coefficients are starting with the characters "ah" and those containing the wet
coefficients with "aw". The next five digits show the year (two digits) and the
day of the year (three digits). The file extension shows whether the file is
for 0, 6, 12, or 18 UT. (Example: "aw05193.h018" contains the wet coefficients
on July 12, 2005, at 18 UT.) For past years, the compressed tarfiles are to be
used. The first line in each file is a header showing the values in degrees for
north, south, west, east, spacing north-south, spacing west-est. The rest of
the file contains the parameters in latitude bands going from north to south,
and from west to east within each band. Both the hydrostatic and wet "a"
coefficients have to be divided by 1.d8 .In addition to the "a" coefficients,
the hydrostatic and wet zenith delays are provided on the grid, too. Their
filenames are starting with "zh" and "zw", respectively.
VMF1 grid files are also available in new row-based format, e.g. as used by the
Bernese software package: [http://ggosatm.hg.tuwien.ac.at/DELAY/GRID/VMFG].
>> 1b. Forecast Gridded VMF1 (VMF1-FC Grid)
Grids with the VMF1 (and hydrostatic zenith delays) are also determined from
forecast data of the ECMWF, but in this case a password is needed. Please ask
Johannes Boehm for username and password if you want to use the data for
scientific (non-commercial) applications. (Please provide a short description.)
Missing values are not filled at a later time. The VMF1 from forecst data of
the ECMWF are made available at about 8 UT for the following day (0, 6, 12, 18
UT).
Forecast VMF1 grid files are available in the standard format at
[http://ggosatm.hg.tuwien.ac.at/DELAY/GRID/STD_FC] and in the new row-based
format at [http://ggosatm.hg.tuwien.ac.at/DELAY/GRID/VMFG_FC].
A closer description is provided in the reference:
J. Boehm, J. Kouba, and H. Schuh (2008), Forecast Vienna Mapping Functions 1
for real-time analysis of space geodetic observations, Journal of Geodesy,
doi:10.1007/s00190-008-0216-y,
to be downloaded at [http://dx.doi.org/10.1007/s00190-008-0216-y].
>> 1c. VMF1 Site
The coefficients of VMF1 are provided for IGS (GPS), IVS (VLBI), and IDS (DORIS)
stations, and the station lists were initially taken from Goddards's atmosphere
pressure loading service [http://gemini.gsfc.nasa.gov/aplo/]. The ellipsoidal
coordinates (name, latitude, longitude, height, domes number) for the current
station lists can be found in the first column of the table below. The VMF1 from
analysis data of the ECMWF are made available with a latency of less than 34
hours.
The list of ellipsoidal station coordinates can be found at
VLBI: [http://ggosatm.hg.tuwien.ac.at/DELAY/SITE/vlbi.ell ]
GPS: [http://ggosatm.hg.tuwien.ac.at/DELAY/SITE/gnss.ell ]
DORIS: [http://ggosatm.hg.tuwien.ac.at/DELAY/SITE/doris.ell]
and the data at
VLBI: [http://ggosatm.hg.tuwien.ac.at/DELAY/SITE/VLBI ] (since 1979)
GPS: [http://ggosatm.hg.tuwien.ac.at/DELAY/SITE/GPS ] (since 2005)
DORIS: [http://ggosatm.hg.tuwien.ac.at/DELAY/SITE/DORIS] (since 2005)
Recommendation: Use the gridded VMF1 data for GNSS or DORIS analysis.
Each line of the time series consists of the following information:
Station name, modified Julian date, hydrostatic coefficient "a", wet
coefficient "a", hydrostatic zenith delay in meter, wet zenith delay in meter,
mean temperature in Kelvin (to convert the wet zenith delay into precipitable
water), pressure at the site in hPa, temperature at the site in degree Celsius,
water vapour pressure at the site in hPa, and the approximate orthometric
height. The hydrostatic and wet coefficients "a" are then input parameters to
the Fortran subroutine vmf1.f at
[http://ggosatm.hg.tuwien.ac.at/DELAY/SOURCE/vmf1.f], which then yields the
hydrostatic and wet mapping functions. The Matlab equivalent vmf1.m is provided
at [http://ggosatm.hg.tuwien.ac.at/DELAY/SOURCE/vmf1.m].
Warning: Up to 1 Dec 2014, the VLBI station ARECIBO had the wrong longitude.
It had 66 Eastern longitude instead of 66 Western longitude.
>> 1d. VMF1-FC Site
A password is needed for any parameters determined from forecast data of the
ECMWF. Please ask Johannes Boehm for username and password if you want to use
the data for scientific (non-commercial) applications. (Please provide a short
description.) The coefficients of VMF1-FC (from forecast data) are provided for
IVS (VLBI) stations. Missing values are not filled at a later time. The VMF1
from forecst data of the ECMWF are made available at about 8 UT for the
following day (0, 6, 12, 18 UT).The data can be found at
[http://ggosatm.hg.tuwien.ac.at/DELAY/SITE/VLBI_FC].
>> 1e. Do you want to determine your own VMF1?
If you run your own ray-tracing programs on profiles from numerical weather
models, you can determine the VMF1 coefficients with the Fortran subroutine
mkvmf1.f at [http://ggosatm.hg.tuwien.ac.at/DELAY/SOURCE/mkvmf1.f]. The
resulting coefficients have to be applied with the subroutine vmf1_ht.f
[http://ggosatm.hg.tuwien.ac.at/DELAY/SOURCE/vmf1_ht.f].
>> 2. Empirical Models
>> 2a. Global Mapping Functions GMF
There is a 'backup' mapping function for VMF1 called Global Mapping Function
GMF. Similar to NMF, the input parameters are day of year, latitude, longitude
and height. The Fortran code is available as gmf.f at
[http://ggosatm.hg.tuwien.ac.at/DELAY/SOURCE/gmf.f]. Thanks to the Bernese
group, also their Fortran implementations of GMF are available:
[http://ggosatm.hg.tuwien.ac.at/DELAY/SOURCE/gmf.f_hu]: gmf.f_hu and
[http://ggosatm.hg.tuwien.ac.at/DELAY/SOURCE/gmf_deriv.f_hu]: gmf_deriv.f_hu
(with derivatives). Also available is the Matlab version gmf_f_hu.m at
[http://ggosatm.hg.tuwien.ac.at/DELAY/SOURCE/gmf_f_hu.m].
Reference:
J. Boehm, A. Niell, P. Tregoning, and H. Schuh (2006), Global Mapping Function
(GMF): A new empirical mapping function based on numerical weather model data,
Geophysical Research Letters, Vol. 33, L07304,doi:10.1029/2005GL025546.
Download from AGU at
[http://www.agu.org/journals/gl/gl0607/2005GL025546/2005GL025546.pdf].
>> 2b. Global Pressure and Temperature GPT
An empirical model to determine pressure and temperature from the station
coordinates and the day of the year has been developed. The Fortran and Matlab
source codes are available at
[http://ggosatm.hg.tuwien.ac.at/DELAY/SOURCE/gpt.f]: gpt.f
[http://ggosatm.hg.tuwien.ac.at/DELAY/SOURCE/gpt.m]: gpt.m
[http://ggosatm.hg.tuwien.ac.at/DELAY/SOURCE/gpt.f_hu]: gpt.f_hu (Bernese)"
This version of GPT is based on the vertical extrapolation of the pressure as
introduced by Berg (1948).
Reference:
J. Boehm, R. Heinkelmann, and H. Schuh (2007), Short Note: A global model of
pressure and temperature for geodetic applications, Journal of Geodesy ,
doi:10.1007/s00190-007-0135-3. Download at
[http://dx.doi.org/10.1007/s00190-007-0135-3].
>> 3. Other Parameters
>> 3a. Height of the 200 hPa pressure level
The height of the 200 hPa pressure level which is the input paramter for the
hydrostatic IMF is provided on a grid and and can be found at
[http://ggosatm.hg.tuwien.ac.at/DELAY/ETC/Z200].
Please mind that starting with doy 184 in 2006 the format of the header line
has changed. (The first line in each file is a header showing the values in
degrees for north (south), south (north), west, east, spacing north-south,
spacing west-est. In both cases the gridded file is starting in the north.)
>> 3b. Mean temperature Tmean
The mean temperature of the atmosphere (in Kelvin) which can be used to convert
the wet zenith delays into precipitable water is provided on a grid and and can
be found at [http://ggosatm.hg.tuwien.ac.at/DELAY/ETC/TMEAN].
These values correspond to the ellipsoidal heights given in the file
orography_ell [http://ggosatm.hg.tuwien.ac.at/DELAY/GRID/orography_ell].
Please mind that starting with doy 184 in 2006 the format of the header line has
changed. (The first line in each file is a header showing the values in degrees
for north (south), south (north), west, east, spacing north-south, spacing
west-east. In both cases the gridded file is starting in the north.)
>> 3c. Linear horizontal gradients LHG (Böhm and Schuh, 2007)
Hydrostatic and wet linear horizontal gradients can be found at
[http://ggosatm.hg.tuwien.ac.at/DELAY/ETC/LHG].
These horizontal gradients are provided for the same stations where VMF1 Site
is available. The time series are starting on Jan 1, 2006 and the columns are:
station name, mod. Julian date, hydrostatic north, hydrostatic east, wet north
and wet east gradient in mm. We have found reduction factors that need to be
applied for the hydrostatic gradients of 0.53 and for the wet gradients of
0.71. However, we are looking forward to your experience with these gradients
and reduction factors. You can find more details in the
Reference:
J. Boehm, and H. Schuh (2007), Troposphere gradients from the ECMWF in VLBI
analysis, Journal of Geodesy, doi:10.1007/s00190-007-0144-2.
Download at: [http://dx.doi.org/10.1007/s00190-007-0144-2].
>> 3d. Horizontal Gradients GRAD (Landskron et al., 2015)
New hydrostatic and wet horizontal gradients can be found at
[http://ggosatm.hg.tuwien.ac.at/DELAY/ETC/GRAD].
These horizontal gradients are provided for all active VLBI stations between
1999 and 2014. They have the same format as the LHG (Böhm and Schuh, 2007)
and thus can be applied equally, without the reduction factors though.
Considered are all stations and epochs 5 NWM epochs before and after a real
measurement, according to the IVS. The files are available yearly.
You can find more details in the
Reference:
D. Landskron, A. Hofmeister, J. Böhm (2015), Refined Tropospheric Delay Models
for CONT11. International Association of Geodesy Symposia, DOI:10.1007/1345_2015_56
Download at: [http://dx.doi.org/DOI:10.1007/1345_2015_56] .
>> Registration and Acknowledgements
Registration: There is no password with the VMF1 time series determined from
analysis data of the ECMWF. However, a password is needed for any parameters
determined from forecast data of the ECMWF. Please ask Johannes Boehm for
username and password if you want to use the data for scientific
(non-commercial) applications. (Please provide a short description.)
Acknowledgements: We would like to thank the Austrian Zentralanstalt fuer
Meteorologie und Geodynamik ZAMG for providing us access to the ECMWF data.