Refinements to the National Spatial Reference System

1. Refinements to the National Spatial Reference System A New Nationwide Adjustment of GNSS Monumented Control National Adjustment of 2011 and New NGS Datasheet Format Great Lakes RegionHeight Modernization Consortium Meeting October 14, 2011 ● Columbus, OH Michael L. Dennis, RLS, PE Geodesistmichael.dennis@noaa.gov

2. A (very) brief history of NAD 83 • Original realization completed in 1986 • Consisted (almost) entirely of classical (optical) observations • “High Precision Geodetic Network” (HPGN) and “High Accuracy Reference Network” (HARN) realizations • Most done in 1990s, essentially state-by-state • Based on GNSS but classical stations included in adjustments • National Re-Adjustment of 2007 • NAD 83(CORS96) and (NSRS2007) • Simultaneous nationwide adjustment (GNSS only) • New realization: NAD 83(2011) epoch 2010.00

3. Introducing…NAD 83(2011) epoch 2010.00 • Multi-Year CORS Solution (MYCS) • Reprocessed all CORS GPS data Jan 1994-Apr 2011 • 2264 CORS & global stations • NAD 83 computed by transformation from IGS08 • National Adjustment of 2011 (NA2011) • New adjustment of GNSS passive control • GNSS vectors tied (and constrained) to CORS NAD 83(2011) epoch 2010.00 • Approximately 80,000 stations and400,000 GNSS vectors • Realization SAME for CORSand passive marks • This is NOT a new datum! (still NAD 83)

4. Why a new NAD 83 realization? • Previous NAD 83(CORS96) needed many improvements • NSRS improvements achieved with the MYCS include: • Consistent coordinates and velocities from combined solution • Aligned with most recent realization of global frame (IGS08) • IGS08 epoch 2005.0 (previous aligned at epoch 1997.0) • NAD 83 epoch 2010.0 (previous epochs of 2002.0 and 2003.0) • Major processing algorithm, modeling, metadata improvements • Absolute phase center antenna calibrations • Highly accurate and consistent CORS coordinates and velocities determined using Best Available Methods • Needed because CORS network is foundation of NSRS

5. Why a new national adjustment? • Optimally align passive control with new CORS • Over 700 projects submitted since 2007 project • Also observations for Hawaii, other Pacific islands • Get network and local accuracies on all stations • Including future submitted projects • More consistent results in tectonically active areas • More current data, better tectonic modeling • Better computations and analysis techniques • E.g., improved outlier detection, weighted constraints • Incorporation of lessons learned from previous national adjustment

6. Approach • Used a Helmert blocking strategy • Approximately 80,000 points and 400,000 GNSS vectors • Outlier detection (for rejecting vectors) • Initially using 5 cm threshold • Method for vector rejection • Rejection by downweighting vs. removal • Other challenges: • Transformations and tectonic tribulations • CORS complications • Constraint conundrums

7. Stations included in NA2011 79,106 stations 78,186 passive 920 CORS

8. Transformation & Tectonic Tribulations • NAD 83(2011) transformation tools? • NAD 83(2011)  (NSRS2007/CORS96)  (HARN) • Under study by NGS • When is North America not North America? • Not all stations on the North American tectonic plate • Pacific plate: Hawaii and American Samoa • Mariana plate: Guam and CNMI • Caribbean plate: Referenced to North American plate • Can relate frames (and epoch dates) with HTDP

9. Tectonic Motions From Sella et al., 2002

10. What’s in a name? That which we call a datumBy any other name would smell as sweet… • NAD 83(2011) epoch 2010.00 • “2011” is datum tag  year adjustment complete • “2010.00” is “epoch date” (January 1, 2010) • Date associated with coordinates of control station • Frame fixed to North American tectonic plate • Includes California, Alaska, Puerto Rico, and US Virgin Islands • NAD 83(PA11) epoch 2010.00 • Frame fixed to Pacific tectonic plate (Hawaii and American Samoa) • NAD 83(MA11) epoch 2010.00 • Frame fixed to Mariana tectonic plate (Guam and CNMI)

11. CORS Complications • All MYCS coordinates are to the ARP • But not all positioning referenced to ARP • Some to L1 Antenna Phase Center…depends on model used • Some to “monuments” (e.g., eccentric points, GRPs) • Two different “types” of MYCS coordinates • Computed. Determined from simultaneous solution • Requires at least 2.5 years of data (otherwise position modeled) • Include formal accuracy estimates (standard errors, σ) • Modeled. Determined using HTDP velocities • No formal accuracy estimates • A few CORS not included (36 former COOP CORS) • No data available for use in MYCS

12. Constraint Conundrums • Only MYCS CORS constrained • “Rigid” constraints (σ = 0.01 mm) • Apply to all 3 components of all constrained points • This method used for NSRS2007 • Variable weighted constraints • Use actual MYCS error estimates • σ range 1 – 140 mm; mean 11 mm; median 8 mm • Only available for computed MYCS coordinates • Modeled MYCS errors to be determined some other way • Initial estimate of σ = 50 mm used; likely will refine

13. Preliminary Results • NA2011 free (unconstrained) adjustment • 4150 projects • 79,080 stations (912 are CORS) • 371,147 vectors used • Standard error of unit weight = 1.25 • NSRS2007 free (unconstrained) adjustment • 3411 projects  22% more in NA2011 • 67,693 stations (688 are CORS)  17% more in NA2011 • 283,691 vectors used  31% more in NA2011 • Standard error of unit weight = 1.28

14. Preliminary Results • NA2011 constrained adjustments • Rigid constraints (σ = 0.01 mm) • 912 CORS (ARPs, L1, mons)  Std error unit weight = ~1.49 • 640 CORS (ARPs only)  Std error unit weight = ~1.38 • Variable weighted constraints • 845 CORS (ARPs, L1, mons)  Std error unit weight = 1.26 • 639 computed ARPs + 81 non-ARPs (σ = 1-140, avg 11, median 8 mm) • 125 modeled ARPs (σ = 50 mm assigned) • 67 CORS non-ARPs not fixed because associated ARP fixed • NSRS2007 constrained adjustment • Standard error of unit weight = 1.38 • 673 CORS rigid 3D (σ = 0.01 mm), 5 rigid 2D (σh = 100 mm) • 7 CORS loose 3D (σ = 100 mm), 3 unconstrained

15. How fixed should fixed control be? • Question: Use rigid or variable weighted constraints? • For variable, use MYCS σ values directly, or scale them? • Concern: Actual MYCS σ values may be too “loose” as constraints • Rigid vs. weighted constraints: How much do they move? • Max ~7 cm horiz and ~5 cm vert; mean ~0 cm (±1 cm) • 185 have shift > 3 cm (half of these in CA) • Expect overall coordinate change about same as MYCS • Horizontal: Mean ~2 cm (±8 cm), median ~0 cm • Vertical: Mean ~ -1 cm (±2 cm), median ~ -1 cm • This is for change in realization and reference epoch • NAD 83(CORS96) epoch 2002.00  NAD 83(2011) epoch 2010.00

17. 3772 stations 12,039 enabled vectors 961 rejected vectors

18. 3772 stations 12,039 enabled vectors 961 rejected vectors

19. Related Tasks & Deliverables • GIS tools for display and analysis of results • Create point, vector, and vector residual features • Error ellipses, spatial analysis, provide in common formats • Release in “Readjustment Distribution Format” (RDF) • aka “Bluebook” format (b-files) • New version of NGS program “ADJUST” • Network and local accuracies • Improved output • Graphical User Interface!

20. When will it all be done? • National Adjustment of 2011 • Goal: Complete by end of 2011, publish early 2012 • Deadline for submitted projects was Aug 31, 2011 • Final data added to network this week • Multi-Year CORS Solution • Officially released coordinates September 2011 • OPUS (Online Positioning User Service) • Provides solutions referenced both to previous (CORS96) and new (MYCS) coordinates • Dual solutions will be available until NA2011 complete • New geoid model…

21. Near-future plans & possibilities • New hybrid geoid model (GEOID12) • Use NAD 83(2011) epoch 2010.00 ellipsoid heights on NAVD 88 benchmarks • Might also use OPUS-Database results on NAVD 88 BMs • Plan release same time as NA2011 • New NGS Datasheet format, also released with NA2011 • May perform national vertical adjustment • Constrain vertically to NAVD 88 benchmarks • Perform as simultaneous nationwide adjustment • GNSS-derived orthometric heights • NAD 83(2011) ellipsoid heights with GEOID12 • NOT a readjustment of NAVD 88 leveling

22. More information… geodesy.noaa.gov

24. Announcing…A New NGS Datasheet Format • Update to new Datasheet version (8.00) • Changed location, length, and text for many fields • Added new fields, deleted fields, augmented existing fields • Implemented by end of calendar year 2011 • Will add announcement and prototype to NGS web site soon • Summary of content changes • Added country (e.g., USA) where control station located • Hyperlinked vertical datum designation to datum web page • Ortho height epoch date, if applicable (e.g., subsidence areas) • Note for geoid model used on Ht Mod stations if not current geoid • Network and (median) local accuracies • Horizontal and ellipsoid height accuracy at 95% confidence (per FGDC) • Includes link to detailed accuracy info, list of all local accuracies • Superseded Ht Mod ortho heights indicate geoid model used

25. Announcing…A New NGS Datasheet Format

26. Announcing…A New NGS Datasheet Format

27. Announcing…A New NGS Datasheet Format

28. Announcing…A New NGS Datasheet Format

29. Announcing…A New NGS Datasheet Format

31. Questions? National Adjustment of 2011 New NGS Datasheet Format