SNARF Working Group - Report of the First SNARF Workshop

SNARF: Theory and Practice, and Implications
Thomas Herring
Department of Earth Atmospheric and Planetary Sciences, MIT
tahmit.edu     http://www-gpsg.mit.edu/~tah

Extended Abstract

The development of a reference system and reference frame for North America needs to consider the following issues (see Blewitt paper for distinction between system and frame). Ideally, the reference system for North America would be a dynamically consistent system in which for example the equations of motions of satellite could be integrated after the system is rotated into an inertial system. Such a system would imply that the relationship between this system and the system in which the gravity field coefficients are given is known. The realization of such a system would probably require globally distributed sites to ensure that the origin of the frame is consistent with first degree harmonic terms in the gravity field. Another approach to defining a North American reference system is a kinematic one in which the system is defined to have zero velocity over the portion of North America that is thought to represent the stable plate. In the realization of this system other non-zero velocity sites could be included if the motions of these sites are known (or reliably estimated) relative to the stable plate. Ideally a SNARF reference frame would satisfy both types of systems.

An initial version of SNARF can be defined by estimating the positions and velocities of a group of sites on the North American plates and then realizing the frame using those sites that do not appear to move relative to each other. The internal consistency of such a frame is between 0.5 and 0.7 mm/yr, horizontally, when data between 1996 and 2004 are used to determine the velocities. The most difficult problem to solve is the day-to-day realization of the SNARF frame. Analysis of the time series of some of the sites in the SNARF definition shows that there are systematic deviations from linear motions. These deviations can be partly explained by atmospheric pressure and water loading deformations of the Earth. The atmospheric pressure loading can be determined with high temporal frequency (4 cycles per day) but water-loading models are only available for monthly averages. Almost certainly there could be large deviations with sub-monthly periods. These non-secular variations limit the accuracy with which daily frame realizations are possible. Models for the non-secular deviations can be developed and caution then is needed in carefully defining the meaning of station positions. The other complication in daily realizations is data quality at the sites used to realize the frame. Instrument failures and local site effects such as snow on radomes and antennas will affect the frame realization if the sites are retained in the analysis. These classes of problems can be minimized if a large number of stations are included in the reference frame. In this case, stations that are deemed problematic on a specific day or over an interval of time can be removed from the frame realization, without adverse effects on the realization. Ambiguity resolution can also strengthen daily frame realizations but ambiguity resolution approaches often rely on high-quality pseudo range data being available which is not always the case.

     Comments or questions about this page? Send e-mail to Will Prescott prescottunavco.org.

Last modified: 2019-12-24  02:12:54  America/Denver