Application Areas and GPS Results

Principles of the Global Positioning System OVERVIEW • Run through HW 03. • Final lecture: Application area and GPS results – Analysis of contemporary motions in the Tien Shan region of central Asia. – Example of evolution of a network – Issues faced when generating multi-year (decade) velocity model. 1 Homework 03 Results from analyses: Phase -2225432.8807 -4676994.9505 3711598.8548 Rinex file -2225431.6719 -4676995.2141 3711599.9580 L1 only -2225401.05 -4677074.26 3711596.32 LC data -2225395.55 -4677062.77 3711585.48 Diff BKAP -2225432.72 -4676994.79 3711599.45 Erwan -2225468.944 -4676973.873 3711593.862 Lili -2225402.870 -4677022.626 3711614.283 Central Asia Analysis • Network – Development of network – Transition from “pins” with tripods to IVTRAN fixed-heights monuments – Continuous stations • Analysis methods – Pre-continuous station analysis – Merger with global GPS analyses • Motion relative to Eurasia – Methods of used to realize a Eurasia fixed frame • Vertical motions – Vertical motions as a function of existing topographic heights 2 Network • Development of Network – First measurements made in July 1992; just after start of International GPS Service (IGS). Since then measurements each year and often multiple campaigns per year. – The current network has over 1000 sites, many of them being recovery marks for the main stations (typically 3 per location) – Network extends from Kyrgyzstan to the Kazakh Platform – In 1995, the first continuously operating stations installed. Currently, 9 continuous sites with 2 IGS sites (POL2, SELE) Regional view of network Zoom Eurasia Fixed Frame realized using ITRF2000 50% confidence ellipses shown 3 Monument evolution • Original monuments were mainly steel pins in bedrock which required tripod setups • Starting in 1995: IVTRAN designed fixed-heights were installed. Simplified setup. Three recovery marks installed for each original monument. • After 1995, measurements made on the recovery marks rather than original marks. • With the original marks and fixed height marks, for many stations two independent velocity estimates (results shown later) Analysis methods • Prior to 1995; selected IGS from Europe, Asia and Australia included in the analysis of regional data • After 1995 when POL2 became an IGS site and was routinely included in the IGS analysis; only local sites included in the MIT analysis • Using GAMIT/GLOBK; regional analyses are included in the Scripps Institution of Oceanography IGS (SOPAC) analysis including orbit improvement (more important in early data). ftp://garner.ucsd.edu/pub/hfiles • Campaigns are combined with SOPAC h-files for velocity field and repeatability analyses. 4 Analysis Method • Types of analyses performed with combined files: – Velocity field analysis. Multiple types • Global: Global selection of sites along with regional sites. Eurasian frame then defined using the ITRF2000 Eurasian rotation pole. • Eurasian: Only sites from Eurasia included. Eurasian frame defined to minimize velocities of stable Eurasian sites • In both analyses: Option to force velocities at nearby sites (<0.5 km separation) to be the same (some exceptions) – Repeatability analysis: • Each campaign is rotated/translated to best align as a function of time with positions and velocities from one of the velocity analyses. IVTRAN Monuments Rod length 355 mm Monument components Installed monument (SST antennas used in Central Asia 5 Statistical analysis • Since the network contains both continuous sites and campaign sites, we include statistical process noise in the velocity analyses to account for temporal correlations in the time series. • Each site has random walk process noise (2 mm)2/yr • Without process noise, velocities of continuous sites would have sigmas of < 0.1 mm/yr; with process noise sigmas are between 0.6 and 1.0 mm/yr e.g. Site Random Walk (mm/yr) N E U White noise (mm/yr) E N U POL2 0.5±0.6 3.0±0.6 -0.5±1.1 0.2±0.1 3.3±0.1 -1.1±0.2 SELE 0.9±0.8 3.7±0.8 -2.5±1.2 1.7±0.1 3.8±.0.1 -2.1±0.2 Frame Realization • Realization of Eurasian frame; Two methods used – ITRF2000: Used global distribution of sites(52) on many plates to rotate/translate frame onto ITRF2000. ITRF2000 Eurasia pole used to rotate to Eurasia fixed frame • Fit to ITRF2000 (52-sites): • Horizontal RMS 0.8 mm/yr; √c2/f 1.2 • Vertical RMS 1.8 mm/yr; √c2/f 2.4 – Eurasia only: Used 14 sites on stable Eurasia to define frame • Fit to Eurasian sites only(14-sites) • Horizontal RMS 0.5 mm/yr; √c2/f 0.8 • Vertical RMS 1.8 mm/yr; √c2/f 2.3 • Notice c2 is <1 for Eurasia but >1 for global; sigmas depend on size of region considered. 6 ITRF2000 Residuals Horizontal RMS 0.8 mm/yr 52-sites Eurasia only residuals Horizontal RMS 0.5 mm/yr 14 sites Postglacial rebound model at NYAL, ONSA 7 Difference between Realizations Zoom RMS Difference 1.3 mm/yr Set by systematic -1.3 mm/yr E -0.7 mm/yr N If removed 0.1 mm/yr RMS Local Frame realization • Another method for frame realization is the use of local stations (1200x1800 km region) in forming velocity solution. • The frame is then found by rotation/translation to best match the Eurasia only realization • In this method, effects of global scale systematic errors should be reduced. • RMS difference of velocities is 0.4 mm/yr (√c2/f 0.2) and only slightly changed by estimating rotation/translation between fields. 8 Local Frame Realization Differences Zoom Differences between co-located sites • Of the approximately 400 sites with highquality velocity estimates, 104 sites are collated between old and new monument styles. • Comparison: Analysis Horizontal RMS 3D-RMS Global 1.9 mm/yr (√c2/f 1.0) 2.5 mm/yr (√c2/f 1.1) Regional 1.9 mm/yr (√c2/f 0.9) 2.4 mm/yr (√c2/f1.1) 9 Collocated velocity estimates Zoom 50% confidence intervals Differences are consistent with sigmas Motion in Eurasia Frame Motion from difference analyses Best Estimates: Site East mm/yr POL2 0.5±0.6 SELE 0.9±0.8 North mm/yr 3.0±0.6 3.7±0.8 AZOK 1.2±0.9 1.1±0.9 CHUM -0.3±1.0 0.4±1.0 10 Complete Velocity Field Zoom Field dominated by NorthSouth compression Also EastWest extension to East 20 o 0 o 0 l 76 -77 l 75 -76 l 74o-750 North Velocity (mm/yr) 15 Profile of North Velocities 10 5 0 40.0 40.5 41.0 41.5 42.0 42.5 Latitude (deg) 43.0 43.5 44.0 11 10 o f 42-43 East Velocities as function of Longitude East Velocity (mm/yr) f43-440 5 0 -5 68 70 72 74 76 Longitude (deg) 78 80 82 8 Height Rates versus Height 6 Vertical Velocity (mm/yr) 4 Regression of height rate versus height yields values between 1.4-1.9 (mm/yr)/km 2 0 -2 -4 l76o-77o o o o o l75 -76 -6 -8 l74 -75 0.5 1.0 1.5 2.0 2.5 Station Height (km) 3.0 3.5 4.0 12 Conclusions • North of the Tienshan the motion of sites relative to Eurasian appears to be less than 1 mm/yr • Within the Kyrgz Tienshan ~13 mm/yr North converge • East west extension across the range front of ~3 mm/yr • Height rate range of ~4 mm/yr • Median RMS scatter of position estimates 1.4-1.8 mm horizontal and 4.6 mm height. • Consistent statistics are obtained with random walk process noise of (2 mm)2/yr. Tools • Most modern GPS analyses now contain hundreds of GPS sites • Current programs are velview and tsview. 13