Assessing the Accuracy of Long-Term Subsidence Derived from Borehole Extensometer Data Using GPS Observations: Case Study in Houston, Texas

Abstract

This study investigated long-term land subsidence derived from borehole extensometer and Global Positioning System (GPS) observations from Addicks, a suburb located in West Houston, Texas. The Addicks borehole extensometer was built in 1974 by the USGS. Two GPS stations were installed at this site in 1993 and 1996, respectively. One GPS antenna (ADKS) is mounted on the inner pipe of the extensometer borehole, which is firmly anchored 549 m below the land surface. Another GPS antenna (PA05) is mounted on a permanent antenna pole anchored 6 m below the land surface. The horizontal distance between these two GPS antennas is 50 m. Continuous GPS data collected from the deeply anchored GPS antenna indicate that the bottom of the borehole has been stable over the past 20 years (1993–2012). Hence the compaction derived from the borehole extensometer data represents the total subsidence at this site, which should be approximately equivalent to the vertical displacement (subsidence) recorded by the shallowly anchored GPS antenna (PA05). In this study, the vertical displacement time series recorded by the shallowly anchored GPS antenna are regarded as the true values to assess the accuracy of the subsidence time series recorded by the extensometer. The authors’ analysis indicates that the difference between the cumulative subsidence measured by the extensometer and the GPS antenna (PA05) is less than 5 mm for a 17-year span (1996–2012). The RMS of the residuals of the monthly subsidence time series derived from the extensometer and GPS antenna (PA05) data over 13 years (2000–2012) is 2 mm.