GPS data analysis and results from the Geodesy Advancing Geosciences and EarthScope (GAGE) project

Monday, 15 December 2014
Thomas Herring1, Michael Floyd2, Robert W King2, Timothy Ian Melbourne3, Walter Michael Szeliga4, Mark H Murray5, David A Phillips6, Christine Maria Puskas7, Frances M Boler7, Charles M Meertens8 and Glen S Mattioli8, (1)Massachusetts Institute of Technology, Cambridge, MA, United States, (2)MIT, Cambridge, MA, United States, (3)Central Washington Univ, Ellensburg, WA, United States, (4)Central Washington University, Ellensburg, WA, United States, (5)New Mexico Institute of Mining and Technology, Socorro, NM, United States, (6)UNAVCO, Boulder, CO, United States, (7)UNAVCO, Inc. Boulder, Boulder, CO, United States, (8)UNAVCO, Inc., Boulder, CO, United States
We report on the methods and results from the GPS data analysis part of the NSF Geodesy Advancing Geosciences and EarthScope (GAGE) project operated by UNAVCO. Current analyses include GPS data from the 1100 Plate Boundary Observatory (PBO) sites and approximately 700 other sites used to densify the network in some locations and to provide spatial extent on the North America and Pacific plates. Analyses from 1996 to present collectively include approximately 2050 unique sites over the 18-year period. The results from these analyses are updated daily for rapid solutions (~1 day latency) and weekly for initial analyses using IGS final orbits (2-3 week latency). Analyses are also run with 12- and 26-week latencies to add sites that were not available (either telemetry failures or manual download sites) during the initial analyses. Raw data are processed using two different GPS analysis programs: GAMIT at New Mexico Tech and GIPSY at Central Washington University. Combined results are then produced with GLOBK at MIT. All results are available through the UNAVCO web site in the form of time series and velocity fields in the NAM08 (Altamimi et al.’s, 2012, ITRF2008 North America Euler pole) and IGS08 frames. Daily SINEX files are provided in fiducial free and NAM08 frames. Event files are generated within a few days of earthquakes in the GAGE analysis region that generate co-seismic displacements greater than 1 mm. The median weighted root-mean-square (WRMS) scatters of combined position time series are less than 1 mm in north and east (NE) and 4 mm for vertical (U) over monthly durations. For all results processed thus far (~18 years of data for the longest running sites), WRMS scatters of the position residuals about linear trends, with offsets for earthquakes and antenna changes removed, are ~1.5 mm NE and 4.5 mm U. The top 10% of sites have short period scatters (month duration) of 0.5 mm NE and 1.9 mm U, while the long-term scatters increase to 0.8 mm NE and 3.3 mm U. The noisiest sites are generally in volcanic areas and/or affected by snow and ice on antennas. We will present results from GAGE analyses in terms of secular rates across the Pacific/North America plate boundary and non-secular signals arising from earthquakes (co- and post-seismic deformation) and other natural and human-induced processes.