Postseismic deformation of the Andaman Islands following the26 December, 2004 Great Sumatra–Andaman earthquakeJ. Paul,1A. R. Lowry,2R. Bilham,3S. Sen,4and R. Smalley Jr.1Received 20 June 2007; revised 20 August 2007; accepted 5 September 2007; published 13 October 2007.[1] Two years after the Great Sumatra-Andamanearthquake the 3.1 m WSW coseismic displacement atPort Blair, Andaman Islands, had increased by 32 cm.Postseismic uplift initially exceeded 1 cm per week anddecreased to <1 mm/week. By 2007 points near Port Blairhad risen more than 20 cm, a 24% reversal of coseismicsubsidence. Uplift at eight GPS sites suggests a gradualeastward shift of the coseismic neutral axis separatingsubsidence from uplift. Simulations of the GPS postseismicdisplacements as viscoelastic relaxation of coseismic stresschange and as slip on the plate interface indicate that slipdown-dip of the seismic rupture dominates near-fielddeformation during the first two years. Postseismic slipbeneath the Andaman Islands released moment equivalentto a magnitude Mw 7.5 earthquake, and the distributionsuggests deep slip in the stable frictional regime acceleratedto catch up to the coseismic rupture.Citation: Paul, J., A. R.Lowry, R. Bilham, S. Sen, and R. Smalley Jr. (2007), Postseismicdeformation of the Andaman Islands following the 26 December,2004 Great Sumatra – Andaman earthquake, Geophys. Res. Lett.,34, L19309, doi:10.1029/2007GL031024.1. Introduction[2] The December 26th 2004, Mw=9.3Sumatra-Andaman earthquake ruptured 1600 km of subduction thrustat the east boundary of the Indian plate. In addition to severalmeters of southwestward coseismic displacement (Figure 1,inset), Global Positioning System (GPS) instruments mea-sured 60 cm of coseismic uplift at Diglipur in the NorthernAndaman Islands, 84 cm subsidence at Port Blair, 7 cmsubsidence at Havelock island, and 34 cm uplift at Hut Bay[Freymueller et al., 2007]. Remote sensing and visualinspection of shoreline changes provide evidence for bothuplift and subsidence, ranging from +40 to 70 cm, whichMeltzner et al. [2006] characterize as separated by a hingeline. Others have used the term pivot line. Because thecoseismic surface flexure is not truly a planar tilt, in thisarticle we refer to this line of zero uplift as a neutral axis.Postseismic processes discussed here have translated thisline eastward (Figure 1).[3] We collected GPS measurements of postseismic de-formation at eleven sites in the Andaman Islands (Figures 1and 2). We initiated continuo us GPS recording at Port Blair(CARI) three weeks after the earthquake, Havelock (HAV2)in January 2007, and on Little Andaman (HUTB) inDecember 2006. Other sites were occupied for periods ofseveral days at intervals of several months. The data affordsufficient spatial and temporal sampling to characterizenear-field postseismic deformation.[4] Three deformation processes are likely candidates forpostseismic response to coseismic stress change: (1) poroe-lastic relaxation moderated by flow of interstitial fluids[Peltzer et al., 1998]; (2) viscoelastic rel axation of themantle [Rundle, 1978]; and (3) aseismic slip in velocity-strengthening frictional conditions [Tse and Rice, 1986].Surface deformation caused by these processes can looksimilar, such that the effects of one process may bemismodeled using the physics of another [Thatcher andRundle, 1979; Fialko , 2004]. Laboratory deformationexperiments leave little doubt that all three processescontribute to transient deformation after earthquakes, butthe relative contribution of each should depend on temporaland spatial scales of measurement. Where data are ade-quately sampled in both time and space, all three processesmay be required to fit the observations [Freed et al., 2006].However examples of great earthquake postseismic defor-mation sampled densely in both time and space are few.Consequently questions remain as to the roles of these threeprocesses in the earthquake cycle, and even whether theyplay a similar role on all fault zones or in subsequent eventson the same fault zone.[5] In this article we combine campaign and continuousGPS data to examine end-member models of postseismicdeformation. The first two years of near-field Andamandeformation predominantly reflects postseismic slip down-dip of the coseismic rupture, consistent with simulations offrictional slip dynamics.2. Data and Analysis[6] GPS measurements were initiated at six sites (CARI,WNDR, HAVE, RUTL, RMNG and MHRT) within two tosix weeks after the earthquake. We collected three to fiveepochs of data at seven sites during the first two years(Figure 2). All of the campaign GPS monuments consist ofsteel pins drilled into rock. At continuous sites, monumentsconsist of 4 m concrete pillars that rise to 2 m above ground.At 2 m depth, the pillars are anchored to a 30 cm thick, 4 4 m reinforced concrete foundation. Data analysis at theUniversity of Memphis used G AMIT/GLOBK. Twelveregional IGS sites were used to determine the ITRF2000reference frame.[7] In the first two years beginning 20 days after themainshock, station CARI moved 7.5 cm south, 31 cm westGEOPHYSICAL RESEARCH LETTERS, VOL. 34, L19309, doi:10.1029/2007GL031024, 2007ClickHereforFullArticle1Centre for Earthquake Research and Information, Memphis, Tennessee,USA.2Department of Geology, Utah State University, Logan, Utah, USA.3Department of Geological Sciences, University of Colorado, Boulder,Colorado, USA.4Society for Andaman and Nicobar Ecology, Middle Point, Port Blair,India.Copyright 2007 by the American Geophysical Union.0094-8276/07/2007GL031024$05.00L19309 1of6and rose 23 cm (Figure 2). All of the campaign sitesexhibit uplift and SW to WSW motion, but magnitudes varyby up to 20% and azimuths differ by as much as 36°.Motion at Mount Hariot (MHRT) and Rangachang (JBNK)is very similar to that at nearby CARI. Wandoor (WNDR)and Rutland (RUTL) have smaller, more westerly displace-ments than sites further east. HAVE and continuous siteHAV2 both exhibit a large southward component of motion.Ramnag ar (RMNG) moved the least among t he well-sampled sites.[8] Prior to the earthquake, four epochs of campaign GPSmeasurement were collected at a site near CARI in 1996 –1999 (CARO, subsequently destroyed). Our reanalysis ofCARO data indicates 9.3 ± 1.8 mm/yr right-lateral obliqueconvergence with the Indian plate [Paul et al., 2001]. Howthis relates t o India–Burma relative plate motion dependson unknown coupling of the two, and