New Madrid and surroundings - GPS Results (updated July 2010)
Position Time Series and Residual Velocities

A USGS/NEHRP-funded project
Collaborative Research, P.I.s E. Calais (Purdue University)
and C. DeMets, University of Wisconsin, Madison


Click on a GPS site location or name in the map or table below to see its position time series. Each dot represents a 24-hour position estimate with its associated formal error. A (slightly outdated) summary of the processing procedure can be found here. Gaps in the time series result mostly from data loss because of equipment or communication breakdowns. Statistics listed with time series include white noise only. Time series have been detrended.


The GPS sites shown on this map are operated by NOAA/FSL (BLMM, CNWM, OKOM), CERI (CJTR, CVMS, HCES, MACC, MAIR, MCTY, PIGT, PTGV, RLAP, STLE), and CORS (MEM2). Seismicity from CERI New Madrid Earthquake catalog.


The table below gives position time series in ITRF2008, the detrended time series shows positions after removal of a constant velocity derived from a linear fit to the raw position time series.

ARBT raw detrended N E U
ARHR raw detrended N E U
ARLR raw detrended N E U
ARPG raw detrended N E U
ARPG raw detrended N E U
ARMH raw detrended N E U
CJTR raw detrended N E U
CNWM raw detrended N E U
CVMS raw detrended N E U
HCES raw detrended N E U
LCHS raw detrended N E U
MACC raw detrended N E U
MAIR raw detrended N E U
MCTY raw detrended N E U
MEM2 raw detrended N E U
NMKM raw detrended N E U
NWCC raw detrended N E U
OKOM raw detrended N E U
PIGT raw detrended N E U
PTGV raw detrended N E U
RLAP raw detrended N E U
STLE raw detrended N E U


Comparison between successive horizontal velocity estimates for the New Madrid area (velocities with respect to stable North America, reference frame implementation procedure as in Calais et al. (2005)). Velocities with respect to stable North America keep decreasing (data scatter decreases) as data time series become longer. The two sites with the worse quality position time series (RLAP and PIGT, large number of outliers, large amount of non-white noise, large seasonal signals) have the largest residuals.


Average residual velocity at the New Madrid CGPS sites (with respect to a rigid plate model) and associated standard deviation (RMS scatter of the residual velocities) as a function of number of years processed. Residual velocities in the NMSZ decrease as a function of converging toward zero, with smaller and smaller uncertainties, but still including zero.