GGP Agreements and Standards
GGP Home Agreements Purpose / Goals Mailing Address / Email List Links
Newsletters Publications / Online Stations Glossary Maps

GGP Agreements and Standards

Data Involved in GGP

GGP is a project to improve the quality and quantity of vertical gravity signals at the Earth's surface that are recorded by existing and future SG installations. Through the GGP, SG installations are encouraged to record the gravity signal to minimum specified standard and to make their data available on a mutually agreed upon format to other GGP Groups and to the scientific community in general.

The recommendations that follow are the result of intensive, and sometimes lively, discussions on the various possibilities for recording gravity signals. The input to these decisions was representative of those persons present at the GGP meetings.

The most important signal to be recorded is gravity so this gets the most attention. For most GGP purposes the atmospheric correction to gravity is required and so the local measurement of pressure is also necessary. At most sites the effect of groundwater, often associated with rainfall, is an important contributor to gravity variations over periods of days to years. SG groups are therefore urged to record rainfall and groundwater level at their sites as auxiliary data. Finally each station is asked to supply a log file of important events that might affect the data for each month.

Following large earthquakes, GGP groups are asked to send earthquake data for a period of 1 week following the origin time of the earthquake. This data is intended to be used for free oscillation studies of the Earth.

Most SG groups monitor many other variables at their stations. These are either

  • variables reflecting local environmental conditions (e.g. humidity and temperature) or
  • those reflecting the status of the instrument (e.g. helium level and tilt feedback signals).

At the present time, SG stations are not required to send this data to ICET.

Data Recording Parameters

This section describes the manner in which the basic signals are recorded.


  1. For two purposes:

    (a) seismological studies of the Earth's normal modes, and

    (b) to permit data repair before decimation,

    the gravity output should be recorded at high rate, defined as 1s-10s sampling with 1s or 2s being the most common

  2. The precision of a single measurement is limited by the DVM conversion and should be the equivalent of 0.1 ngal. If the full bandwidth of the signal is to be recorded (for example through the tide output) , this requires a DVM with 7.5 digits
  3. The timing accuracy for each gravity sample should be 10 ms.

    Note that a semi-diurnal tidal wave (M2 of amplitude say 100 mgal) has a maximum slew (derivative) of 0.0145 mgal / s. At a sensitivity (gravity error) of 0.1 ngal, this slew translates into a timing (epoch) error of 6ms. In practice a 10 ms error is therefore not unreasonably stringent.
  4. This 1s or 2s (or 10s) data is regarded as GGP Raw Data.
  5. The individual SG groups can decide (a) from which GEP outlet (TIDE, MODE, 1s) the gravity signal was to be sampled, and (b) the exact characteristics of the gravity anti-aliasing filter prior to sampling.

GWR has designed a new gravity tide electronic board with two options for the gravity signal to be sampled at either 1s or 2s. These new boards define the standard for sampling high precision gravity (Newsletter #5)


The question of 1s versus 10s sampling rate has been actively discussed by a number of individuals, including Mansinha, Crossley, Richter, Zürn and Hinderer. The consensus is that a 10s sampling rate is sufficient for all geodetic purposes, and most normal mode studies, provided that a suitable anti-aliasing filter is used (see above) that does not cut off frequencies of seismological interest. With sampling at 1s or 2s, it is easier to provide a flat amplitude response for the anti-aliasing filter at 'short' normal mode periods of 10s of seconds.

There is no a priori reason for every GGP group to record the gravity at the same high rate or through the same filter, provided the resulting 10s data (earthquake data for seismic studies) and 1m data (GGP Data for all other studies) are produced to a consistent standard (i.e. signal to noise ratio, frequency band).


  1. The atmospheric pressure should be recorded at a minimum sampling interval of 10s at an accuracy equivalent to better than 0.1 mbar.
  2. Ideally the pressure should be recorded at the same high rate as the gravity to permit easy correction of the high-rate gravity data for atmospheric loading.
  3. The characteristics of the pressure anti-aliasing filter are to be determined by the individual SG group.

Data Exchange Format

  1. Bearing in mind the need to balance the large amount of raw data collected by the GGP versus the need to allow flexibility in future processing, 1 min samples of gravity and local atmospheric pressure are the standard GGP data for exchange.
  2. Separate files for each calendar month should be prepared as follows :
    • FILE 1 - date, time, gravity, pressure (each minute)
    • FILE 2 - date, time, water level, rainfall, tilts, temperature, humidity (each minute, or as available)
    • FILE 3 - date, time, log file information (giving all the known disturbances, He refills, gaps etc., as necessary)
  3. All times are to be quoted as UTC (including leap seconds). All variables are to be quoted in the units in which the measurements are made, with the appropriate calibration factors given.
  4. Usually the signals are recorded in volts. For other purposes, the units to be adopted are mgal (gravity), mbar (pressure), m (rainfall, water level), °C (temperature) and SI for all other variables.
  5. The three data files for each month, as defined above (FILES 1-3), constitute the definition of GGP Data.
  6. Considering the widespread use and availability of the ETERNA program for tidal analysis (Wenzel, 1994b), the ETERNA, or more precisely the PRETERNA, data format (Wenzel, 1994a) is to be adopted for data exchange as outlined by Wenzel (1994c). In this format, for example the gravity and pressure data (FILE 1) would appear as -
    Header information
    (station parameters, comments and parameters of the data acquisition,
    scale factors, etc.)
    date      time    gravity (volts) air pressure (volts)
    yyyymmdd  hhmmss  ggg.gggggg      ppp.pppppp

    As indicted by Wenzel (1994c) ``Step corrections (to be added to all subsequent data of the channel) may be input via a code 77777777 in columns 1..8 and a data gap may be input via a code 66666666 in columns 1..8. The end of the data series is marked by a 99999999 in columns 1..8''.

  7. GGP groups should be prepared to make available the GGP Raw Data to other GGP participants, when (see below), and if so requested.

Data Filenames

GGP filenames are standard DOS-compliant for portability; they are constructed month-by-month from the following elements:

[SC] [YY] [MM] [RC] . [EXT]

each [ ] on the left is two characters: [SC] is the Station Code, [YY] is the year (despite the Y2K problem!), [MM] is the month and [RC] is the Repair Code (see below). The extension [EXT] on the right is 2 or 3 characters.

Station Year Month Repair Code   File Extension Content
CA 89 12 00 . GGP gravity / pressure
CA 89 12 00 . AUX auxiliary data
CA 89 12 00 . LOG log file
CA 89 12 00 . ZIP compressed version of all files


Repair Code

The data repair code depends on the data treatment and decimation (Newsletter #5)

(a) No repair

Code significance
00 raw data, decimated but untreated

(b) Repair done on raw data, before decimation to 1 min

Code significance
01 gaps and disturbances filled with synthetic signal
02 as 01 + offsets adjusted

(c) Repair done on data after decimation to 1 min

Code significance
11 gaps and disturbances filled with synthetic signal
12 as 11 + offsets adjusted

(c) One hour data, decimated from 1 min

Code significance
h1 data processed by ICET
h2 data processed by user

Earthquake Filenames

We suggest the following: [SC][YYMMDD].[EXT]

where, in addition to the abbreviations given above, DD is the day of the month on which the earthquake focal time occurred (UT). The extension [EXT] is as follows:

S1   gravity and pressure together, 1 sec sampling
S2   gravity and pressure together, 2 sec sampling
G1   gravity alone, 1 sec sampling
P1   pressure alone, 1 sec sampling
G2   gravity alone, 2 sec sampling
P2   pressure alone, 2 sec sampling

e.g. ST980325.S2 for an earthquake beginning on 25 March 1998, in which the gravity and pressure are together, 2 sec sampling.

Data File Format and Headers

Gravity and Pressure

The complete header for the gravity and pressure file should be as follows:

Line Text (a20) Parameter 1 Parameter 2
Line 1: Filename [name of file] (a20)  
Line 2: Station [name of station] (a20)  
Line 3: Instrument [name of instrument] (a20)  
Line 4: Phase Lag (deg/cpd) [phase lag] (f10.4) [error] (f10.4)
Line 5: N. Latitude (deg) [latitude] (f10.4) [error] (f10.4)
Line 6: E. Longitude (deg) [longitude] (f10.4) [error] (f10.4)
Line 7: Height (m) [height] (f10.2) [error] (f10.4)
Line 8: Gravity Cal (mgal/V) [g calibration] (f10.4) [error] (f10.4)
Line 9: Pressure Cal (mbar/V) [p calibration] (f10.4) [error] (f10.4)
Line10: Author [email address of author] (a40)  
Lines 11 other information    
C**** end of header    

As an example of the above (Newsletter #6a):

Filename                  ST970910.GGP
Station             Strasbourg, France
Instrument                    GWR C026
Phase Lag (deg/cpd)    0.1500   0.0100             nominal
N Latitude (deg)      48.6220   0.0010             estimated
E Longitude (deg)      7.6840   0.0010             estimated
Height (m)            180.0000  1.0000             estimated
Gravity Cal (mgal/v) -792.0000  1.0000             measured
Pressure Cal(mbar/v)  200.0000  1.0000             nominal
Author              (
yyyymmdd hhmmss   gravity(V) pressure(V)
19970901 000000  0.075913  0.420192

Auxiliary Data

For the auxiliary data, it is not necessary to repeat the complete header used for the gravity/pressure file:

Filename                  ST970910.AUX
Station             Strasbourg, France
Instrument                    GWR C026
Calibration (m/v)      1.0200   0.0100              estimated
Author              (
yyyymmdd hhmmss   water level(V)
19970901 000000  5.170252

Log File

The log file header and format can be quite simple, but should follow the style of the other files:

Filename                  ST970910.LOG
Station             Strasbourg, France
Instrument                    GWR C026
Author              (
yyyymmdd hhmmss   comment
19990705 050330  15 microgal offset of unknown origin
19990712 220610  power loss due to lightening strike

Earthquake Data

Earthquake data if possible should follow the format of the 1 minute GGP Data, with each value attached to a time stamp.

Filename    : CA930712.S2
Station     : Cantley, Canada                
Instrument  : TT70 GWR 012                   
N Latitude  : 45.5850                        
E Longitude : 284.1929                       
Height m    : 269                            
Gravity Cal (ugal/v) -792.0000  1.0000             measured
Pressure Cal(mbar/v)  200.0000  1.0000             nominal
Gravity     : -63.94    microgal / V         
Author      : Merriam (
yyyymmdd hhmmss   gravity(V) pressure(V)
19930712 000000  -0.5061047  0.420192

Data Exchange and Availability

(1) GGP groups are free to make individual any arrangements with other scientists regarding the sharing of their data.

(2) GGP groups may choose, or may be required by their funding charter, to provide all data free to the general public. Some of the GGP data sent to ICET will therefore become immediately available to everyone.

The following agreements exist within GGP to provide a limited time for the individual SG groups to process and analyze their data, following which in two step the data is made available to everyone. Notice these are maximum delays only, many groups are making their data available before the deadlines.

  1. Each GGP group will record and process data for each month at 1 minute sampling.
  2. Within 1 calendar year, these data files, plus any auxiliary data and log files, must be sent to ICET in the recommended format. Thus data for July 1997 must be sent to ICET by 1 August, 1998; data for August 1997 must be sent by 1 September 1998, etc.
  3. Between the first and second years after collection of the monthly data, all GGP groups who have contributed data will have access to the GGP data from ICET.
  4. After a further year, ICET will put these 1 minute data files on Open Report, available to all scientists and the general public.
  5. After each year of collection, ICET will prepare a CD-ROM containing all the GGP data for the previous year.
  6. The Syowa station in the Antarctic, due to the extreme effort needed to collect the data and maintain the instrument, will have 1 additional year in which to send data to ICET. Thus data for July 1997 from Syowa must be sent to ICET by 1 August 1999.
  7. Following large earthquakes, GGP groups should prepare earthquake data files from their GGP Raw Data and send these files to ICET.
  8. For special requests, GGP groups should be prepared to send their GGP Raw Data to requesting scientists.

Authorship of Papers using GGP Data

Details of the arrangements for the authorship of papers involving GGP data have yet to be finalized. In the interim, scientists are expected to follow normal protocols and to acknowledge GGP as the source of their data in any publications using that data.