SNARF Working Group - Report of the First SNARF Workshop

Plan for producing and testing SNARF

 

 The SNARF data product is intended to address a range of solid earth deformation goals, including optimization for regional deformation studies, in addition to providing a direct correction to NAREF, ITRF 2000 and other commonly used reference frames and plate models. SNARF will include two sets of stations well-known position and velocities on the North America plate or, for reasons of geometric strength, stations such as those on the Pacific plate.  From this set of primary stations, a subset of "zero order stations" will be selected.  These will be stations believed to best characterize rigid block North America plate motion.  The reference frame will be determined by minimizing the velocities of the zero order stations.  As development and testing evolves, other parameters, such as the amplitude and phase of annual and semi-annual corrections for individual stations within the reference frame, may be added.

 

One of the central goals of SNARF is its ease of use by a cross section of scientists of differing expertise; thus its design and implementation should support this goal.  The primary SNARF data product will consist of a SINEX file and published rotation parameters between SNARF and other common reference frames; the position and velocity solution will also be provided in stacov format.  A flat ASCII file will include documentation such as criteria for inclusion of each station and other needed descriptive material. Instructions for its implementation will be provided.  As additional parameters are provided (such as phase and amplitude solutions for annual and semi-annual signals; optimization of vertical deformation reference), tools for generating daily SINEX and stacov files that incorporate these in predictions will also be developed and distributed with the data product.

 

Thus, the plan for producing and testing a SNARF baseline or beta version (SNARF 0) is to develop a set of stations of well-known position and velocity, and to test sets of a priori (e.g., geologic) and/or statistical criteria for selection of the subset of zero order stations.  Some zero order stations may even be selected from the interior of the Pacific plate, where a robust and standard independent plate motion correction can be applied to the GPS velocity determination.  The set of all primary stations shall initially be drawn from existing high quality time series of the IGS, CORS, and existing regional arrays.  Ultimately, SNARF will include all PBO backbone stations that meet data continuity and quality criteria.

 

Testing of the application of SNARF will be undertaken by participating data analysis labs at the regional arrays in order to assess its robustness, ease of implementation and for optimization for crustal deformation (including GIA) studies.  Strategies for testing will focus on stability of velocities and improvement of uncertainties relative to the defined North America datum.  It will also be tested for stability or improvement to short baseline determinations using criteria such as strain, stability or improvement on vertical deformation rates, and optimization for regional stabilization.

 

 

Random thoughts and other notes:

Robustness of velocity field relative to "rigid plate".

Baseline analysis relative to daily stabilization,

Rigid plate - how meaningful does it need to be?

      Improvement to precision?

      Trade offs on strain?  (vertical vs. horizontal)

      Actual plate motion vs. mathematically proxied blocks

      Precision on vertical - never mind reference frame for vertical (e.g. absolute velocities)

 

 

Testing of SNARF baseline version 0.

 

Testing:

Stabilize to a SNARF:  velocity model.

 

Products:

Set of time stamped positions and velocities.  Sinex format.

Standard in sinex format. Procedure (or software) for generating .stacov files also included.

Published position and annual velocities.

Daily frame files.

 Does frame include time variant (at least periodic) deformation positions?

 

 

 

Goals of testing:

Robustness of velocity field relative to "rigid plate".

Baseline analysis relative to daily stabilization,

Rigid plate - how meaningful does it need to be?

      Improvement to precision?

      Trade offs on strain?  (vertical vs. horizontal)

      Actual plate motion vs. mathematically proxied blocks

      Precision on vertical - never mind reference frame for vertical (e.g. absolute velocities)

 

 

 

Daily SINEX fiducial free solutions

 

 

 

Developing and testing incremental versions of SNARF.  Initial baseline version to be produced by Tom Herring.  Testing by existing regional data analysis labs.  SINEX file, flat acii file with one line per site, criteria for inclusion.  Well-defined connection to ITRF.  Pacific plate sites.

 

Zero order sites:  stable north America - stations for which the velocity is minimized

Primary sites:  PBO backbone and selected existing (but moving sites) characterized by linear velocities,

Secondary sites:  Other stations for which it makes sense to publish velocities and positions.  [NAREF, regional]

 

Transformation parameters between SNARF v. x and ITRF 2000.

 

Continual improvement, version control.  Protocols for adding and removing sites.

 

Building in GIA components:  feedback to modeling.

 

Baseline version should be SINEX and a standard tool for translating available in stacov

A list of instructions for implementation

 

 

Testing:

 

Stability of solution

Effect on time series

      Improvements of to statistics (like RMS) at sites not included in Improvements to structure/resolution of time series

 

Ease of use

 

--

M. Meghan Miller

Interim Dean

College of the Sciences

Central Washington University

Ellensburg, WA    98926

 

509/963-1866

509/963-1977 (fax)