SCEC Award Number 18116 View PDF
Proposal Category Individual Proposal (Data Gathering and Products)
Proposal Title Development of merged GPS time series for the Community Geodetic Model
Investigator(s)
Name Organization
Thomas Herring Massachusetts Institute of Technology Michael Floyd Massachusetts Institute of Technology
Other Participants
SCEC Priorities 1a, 2a, 3e SCEC Groups CXM, Geodesy, SDOT
Report Due Date 03/15/2019 Date Report Submitted 03/19/2019
Project Abstract
 We have contributed to the development of the Community Geodetic Model (CGM) by designing and implementing software utilities to read, reformat and combine time series which are available from operation analysis centers. These analysis centers currently include GAGE; University of Nevada, Reno; USGS; JPL and SOPAC. We have led the efforts to implement two algorithms to fit time series to generate GNSS-derived velocities for the CGM. We have worked to develop a utility to use the latest release of Hector, which is now able to estimate such non-linear terms in the presence of temporally correlated noise. The utility consists of a shell script to read in a standard time series “.pos”-format which we have successfully petitioned the CGM to adopt, and to fit linear, seasonal and offset terms. The latter of these would be defined by the CGM discontinuities list, described above. We have generated initial combined time series analysis for the continuously operating GPS sites that will form the basis of and be distributed as part of GPS component of the SCEC Community Geodetic Model (CGM). We investigated techniques for the accurate and consistent recovery of secular velocities using a combination of time series from different analysis centers that make their products available for this purpose in the presence of time-dependent deformation, such as post-seismic decays. We have determined the weighting of individual contributions and transformed, though network rotation and translation, each of the input analyses into a common reference frame.
Intellectual Merit This work is a core piece of the production of the GPS component of the SCEC Community Geodetic Model. This product will eventually act as a resource for both the scientific and wider public community. Consensus for creation of time series and crustal velocities is required due to the current variety of projects that generate relevant products through a variety of approaches, which may result in subtle but significant differences in interpretation.
Broader Impacts Geodetic data products are now incorporated into probabilistic seismic hazard analyses (PSHAs), and the accuracy of these relies on an accurate, widespread and dense fault displacement and crustal velocity estimates. Beyond the scientific use of these PSHAs, they form a part of community education and outreach products, where they are visualized for public consumption.
Exemplary Figure Figure 2: Comparison of velocities estimated using three algorithms: tsfit (in default weighted least-squares mode), Hector and est_noise. These three algorithms generally agree well except, currently, at certain individual sites likely with short or noisy time series (random blue and white vectors in both figures around Los Angeles) and in areas affected by post-seismic motion (systematic white arrows between Salton Sea and US-Mexico border in right figure).