[IGSMAIL-3341]: GPS Time Series Data Set

[Michael B Heflin]

******************************************************************************
IGS Electronic Mail      21 May 10:02:33 PDT 2001      Message Number 3341
******************************************************************************

Author:  Michael B. Heflin

An updated GPS time series data set is now available at URL:

http://sideshow.jpl.nasa.gov/mbh/series.html

Recent improvements include:
-------------------------------------------------------------------------------
1.  Time series for 616 GPS sites:  IGS, SCIGN, CORS, and NBAR networks
2.  Ten year time span:  91JAN22 through 01APR28
3.  Consistent reference frame for all results:  ITRF2000
4.  Includes 939 reprocessed historical days
5.  Updated position and velocity estimates
6.  Geocenter and scale time series
7.  Polar motion and length of day time series
8.  Updated maps
9.  New .x-files for GIPSY users

What scientific questions can this data set address?
-------------------------------------------------------------------------------
Global plate motion
Post-glacial rebound
Interseismic deformation in Southern California
Earthquake displacements
Seasonal loading variations
Angular momentum exchange between the solid Earth, oceans, and atmosphere

How many people does it take to generate this data set?
-------------------------------------------------------------------------------
One person to retrieve the data
One person to compute precise GPS orbits and clocks based on 42 sites per day
One person to post-process >500 sites per day

Why is the ITRF2000 reference frame important?
-------------------------------------------------------------------------------
Use of a single frame ensures continuous time series without artifacts
Reduced translational offset and drift relative to Earth's center of mass
Reduced offset and drift of scale relative to other geodetic techniques
Reduced rotational offset with respect to historical polar motion series
Reduced rotational drift with respect to no net rotation of Earth's crust

Why was historical data reprocessed?
-------------------------------------------------------------------------------
Updated analysis strategies improve the quality of historical results

How good are the GPS velocity estimates?
-------------------------------------------------------------------------------
The formal errors provided are optimistic.  Comparisons with VLBI, SLR, and
ITRF2000 indicate WRMS agreement of:

             N     E     V
SLR        1.6   1.7   3.2  mm/yr
VLBI       0.8   0.8   2.0  mm/yr
ITRF2000   0.8   0.7   1.5  mm/yr

Precision was estimated by dividing the data into segments and computing the
scatter of velocity estimates.  Improvement is observed to be inversely
proportional to the time span of data T in years according to:

N 3 mm/yr / T
E 4 mm/yr / T
V 9 mm/yr / T

Daily precision for other parameters:
Polar Motion       0.1 mas
Length of Day      0.02 msec/day
Geocenter          1 cm X, 1 cm Y, 2 cm Z
Scale              0.4 ppb