South Carolina Virtual Reference Station Network Centimeter Positional

South Carolina Virtual Reference Station Network - Centimeter Positional Accuracy in Real Time for South Carolina DMT ‘ 07 May 22, 2007 Lewis Lapine South Carolina Geodetic Survey

Technology

Miniaturization of Components

The Global Positioning System Three Segments 2. ) Space Segment 3. ) User Segment 1. ) Control Segment Master Control Station Colorado Springs Control Segment Monitor Stations Diego Garcia Ascension Is. Kwajalein Hawaii

Space Segment n n n 29 Satellites 6 orbital planes inclined 55 degrees to equator 4 or 5 SVs in each plane 20, 200 Km high orbits (12, 550 Miles) Orbit twice daily (11 hr 58 min)

Control Segment Schriever AFB

Typical Future Geologist At Work User Segment

Calculate your position One measurement narrows down our position to the surface of a sphere 4 unknowns : Latitude Longitude Height Time Need 4 equations 12, 000 miles We are somewhere on the surface of this sphere.

Calculate your position, cont’d Second measurement narrows it down to intersection of two spheres Intersection of two spheres is a circle

Calculate your position, cont’d Third measurement narrows to just two points Intersection of three spheres is only two points In practice 3 measurements are enough to determine a position. We can usually discard one point because it will be a ridiculous answer, either out in space or moving at high speed.

Calculate your position, cont’d Fourth measurement will decide between the two points Fourth measurement will only go through one of the two points The fourth measurement allows us to solve for the receiver clock bias.

Integer Ambiguity Resolution Yields Survey Quality Baselines Dl = First Partial Wavelength N = Integer Ambiguity Dl N D Distance D = Nl + Dl

Initialization From The Phase Measurements Dl 12 9 3 Continuous Carrier Phase Measurement 6

Good Satellite Geometry PDOP

Multipath

GPS Signals n Variable signal paths and piercing points

GPS Post Processed or RTK Uses Baselines From A Known Position Determined 33 23’ 30. 195065” N 079 o 02’ 33. 948394” W -12. 445 m o Known 33 23’ 28. 607434” N 079 o 02’ 41. 161474” W -12. 637 m o

Real Time Kinematic - RTK

Ambiguity Resolution for RTK l

Limitations of Classical RTK Survey n Limited range from single reference station Potential gross error in establishing reference station No integrity monitoring Dependency on single reference station Productivity loss Security Communications n Power supply n n n

RTK PPM Error vs. Baseline Length

What is a CORS?

Reference Stations…defined n n Located at a precisely known position Records GPS data for later use • Post Processing n Generates GPS corrections for immediate use • Real - time, broadcast or dial - in • Results degrade with distance from Reference Station • Useable range can be from 10 km to 500 km (RTK vs DGPS)

Components of a CORS site Monumentation Hardware Photo Courtesy UNAVCO Communications Software

Progress?

Virtual Reference Stations Infrastructure - The future for your industry

Component Monitoring: Server n Server: Central Data Management

RTKNet VRS - Data Flow Reference Station data streams back to server through LAN or Internet

RTKNet VRS - Data Flow Roving receiver sends an NMEA string back to server using cellular modem - VRS position is established VRS NMEA - GGA

RTKNet VRS - Data Flow Server uses VRS position to create corrected observables and broadcasts to rover M R+ VRS CM T R C r o

RTKNet VRS - Data Flow Rover surveys as in normal RTK survey but getting data as if coming from the VRS

VRS - How does it work? n Network Processes • Integrity monitoring • Geometric correction • Correction for tropospheric errors • Correction for ionospheric errors • Ambiguity resolution • Consistency check • Delivery of VRS corrections for each mobile‘s location

Network Multiple Base Station Solution Let B 1, B 2 B 3 be base stations, R be the rover B 1 B 2 R B 3 The indexes r = 1, 2, 3 are for bases, r=0 stands for the rover Four RTK engines run simultaneously: 1. (1, 2) 2. (2, 3) 3. (3, 1) 4. (1, 0)+(2. 0)+(3, 0)

Probability of Certainty Field Test Results Time (secs)

Modeling gradients of the ionosphere Changes of satellite geometry lead to changes in mapping function Changes of piercing coordinates and dynamical changes in ionosphere lead to time dependence of expansion parameters Gradients

Modeling gradients of the ionosphere Changes of satellite geometry lead to changes in mapping function Changes of piercing coordinates and dynamical changes in ionosphere lead to time dependence of expansion parameters Gradients

Kalman filter – part 1: observation equation Observation, measurement State vector measurements from N stations N ambiguities N stations N multipaths N stations 3 parameters 1 ionosphere Design matrix mapping relative distance to function reference piercepoint

Components of the state vector multipath (centimeters) ambiguity (meters) 5 1. 0 4 6. 0 5. 8 2. 5 2. 0 3 0. 5 2 5. 6 1. 5 1 5. 4 1. 0 0 0. 0 -1 -2 -0. 5 -3 5. 2 5. 0 4. 8 0. 5 -4 -1. 0 6000 15: 00 12: 30 differences between ambiguities are constant ionosphere parameters 8000 10000 0 Japan, day 1 -0. 5 Japan, day 2 (shifted by 4 minutes) 12000 14000 15: 00 1600012: 30 correlated uncorrelated days between stations 15: 00 activity of ionosphere about to decline Use double differences together with code measurement and tropo model to resolve the ambiguity (FAMCAR)

Advantages of VRS n Eliminates dependency on single reference station • Graceful degradation should a R. S. fail n n Uses established communications Establishes a single coordinate system

The South Carolina Virtual Reference Station Network

SC Prototype VRS Network 10 m k 0 ~ ~120 km

Test Results from the SC Original VRS Prototype Network 0. 150 ft = ~5 cm

Scatter Gram of Differences With Logical Rejection of 5 Points 2 – Five Minute Sessions RMSE(2 -D) = 0. 020 m @95% 0. 0200 0. 05 ft = 1. 5 cm 0. 0150 0. 0100 0. 0050 -0. 020 -0. 015 -0. 010 0. 0000 -0. 005 0. 000 Series 1 0. 005 -0. 0050 -0. 0100 -0. 0150 -0. 0200 Axis Units in Meters 0. 010 0. 015 0. 020 0. 025

Computation of Horizontal Positional Accuracy RMSE (2 -D) @95% n Local Network Accuracy (NGS) = 0. 020 m (equivalent relative accuracy = 1: 400, 000) n Centering Error = 0. 003 m n Contract Specification for VRS = 0. 012 m n Combined RMSE = (0. 0202 + 0. 0032 + 0. 0122)1/2 = 0. 024 m

Scatter Gram of All Differences 2 – One Minute Sessions RMSE(2 -D) = 0. 024 m @95% 0. 025 0. 020 0. 015 0. 010 0. 005 -0. 030 -0. 020 -0. 010 0. 000 0. 010 -0. 005 -0. 010 -0. 015 -0. 020 -0. 025 Axis Units in Meters 0. 020 0. 030 Series 1 0. 040

USERS RS Data Quality Control RS Data Da t af or Us DDa atat afof or r UUs seer srs er BA D s ! Quality Control

SC VRS Network Construction

Questions?