1. USM GPS RESEARCH ACTIVITIES Presented by: Technical support from: Funding provided by:

2. Sunil Bisnath, David Wells, Stephan Howden, and David Dodd Hydrographic Science Research Center, Department of Marine Science, The University of Southern Mississippi Hydrographic Science Research Center, Department of Marine Science, The University of Southern Mississippi USM GPS Workshop 2004 USM GPS Workshop 2004 16-18 March, Long Beach, Mississippi USM GPS RESEARCH ACTIVITIES

3. OVERVIEW OF GPS RESEARCH PROJECTS Design requirements for hydrographic GPS buoy Design requirements for hydrographic GPS buoy Evaluation of commercial Wide Area DGPS services Evaluation of commercial Wide Area DGPS services Creation of long baseline, carrier phase GPS database Creation of long baseline, carrier phase GPS database Improve GPS tropospheric modeling Improve GPS tropospheric modeling Common themes of research Common themes of research

4. DESIGN REQUIREMENTS FOR HYDROGRAPHIC GPS BUOY

5. NAVOCEANO BUOY EXPERIMENTS: PURPOSE To observe effectiveness of determining water level from GPS for tidal datum determination To observe effectiveness of determining water level from GPS for tidal datum determination U.S. National Ocean Service requires water level gauges to measure to precision of 10 cm (95%) U.S. National Ocean Service requires water level gauges to measure to precision of 10 cm (95%)

6. 2001 buoy deployment ref. stn. and 2002 buoy deployment 2002/2003 buoy deployment NAVOCEANO BUOY EXPERIMENTS: DEPLOYMENTS 5 km

7. UNFILTERED BUOY RTK GPS VERSUS TIDE GAUGE HEIGHT: APRIL 2003 ~15 km baseline WL float sol’n. 38.8 cm25.0 cm

8. FILTERED BUOY RTK GPS VERSUS TIDE GAUGE HEIGHT: APRIL 2003 ~15 km baseline 30 min. MA filter 9.5 cm 6.3 cm

9. BASELINE PROCESSING CONCLUSIONS TO-DATE Simple filtering greatly improves solutions Simple filtering greatly improves solutions PPK far superior than RTK for unfiltered data, but marginally better than RTK for filtered data PPK far superior than RTK for unfiltered data, but marginally better than RTK for filtered data Under certain conditions, commercial RTK GPS can on its own estimate water level at desired accuracy Under certain conditions, commercial RTK GPS can on its own estimate water level at desired accuracy DatasetBaseline 1  (cm) 95% (cm) 2002 ~ 50 m 2 3 2001 ~10 km 4 6 2003 ~15 km 610 Filtered, RTK solutions:

10. TRANSITION FROM EXPERIMENTAL TO OPERATIONAL BUOY Determine tidal datum:Capability shown by experiments. Determine tidal datum:Capability shown by experiments. Minimum sensor suite:Heave and tilt sensors recommended. Minimum sensor suite:Heave and tilt sensors recommended. Buoy types:Deployable from survey launch. Buoy types:Deployable from survey launch. Power management:Batteries rather than solar power. Power management:Batteries rather than solar power. Communication links:PPK or WADGPS rather than RTK. Communication links:PPK or WADGPS rather than RTK.

11. EVALUATION OF COMMERCIAL WIDE AREA DGPS SERVICES

12. WHY EVALUATE THESE SERVICES? High-precision, commercial, carrier phase-based Wide Area Differential GPS (WADGPS) services introduced High-precision, commercial, carrier phase-based Wide Area Differential GPS (WADGPS) services introduced Marine charting and navigation spatial resolution requirements becoming more stringent, particularly in shallow water Marine charting and navigation spatial resolution requirements becoming more stringent, particularly in shallow water Both manufacturers and users interested in independent, comprehensive evaluations Both manufacturers and users interested in independent, comprehensive evaluations

13. SERVICES EVALUATED C-Nav. C-Nav. –Developed by C&C Technologies, Inc. –State space approach Starfix-HP. Starfix-HP. –Developed by Fugro Chance Inc. –Multi-baseline approach SkyFix XP. SkyFix XP. –Developed by Thales GeoSolutions Group Ltd. –GeoSolutions since sold to Fugro –State space approach

14. STARFIX-HP STATIC TESTS: POSITION SOLUTION 25 JULY – 7 AUG. 2003 Max. = 341.2 Bias = 0.6 1  = 4.6 r.m.s. = 4.7 95% = 8.8 Max. = 81.5 Bias = 0.6 1  = 6.2 r.m.s. = 6.2 95% = 12.9 Max. = 334.4 Bias = 1.3 1  = 13.2 r.m.s. = 13.2 95% = 26.1 (All values in centimeters)

15. C-NAV STATIC INITIALIZATION TESTS: POSITION SOLUTION 19-21 AUG. 2003 convergence

16. SUMMARY OF RESULTS FOR ALL SERVICES Correction coverage area Global, less poles, or ~750 km from most major coasts Long-term static positioning precision and accuracy 5-10 cm hor. and 15-30 cm vert. (1  ) 15-20 cm hor. and 25-45 cm vert. (95%) Convergence period ~30 minutes Kinematic positioning precision and accuracy 5-10 cm hor. and 5-15 cm vert. (1  ) 5-25 cm hor. and 15-35 cm vert. (95%)

17. POTENTIAL USAGE FOR SERVICES Great potential: Great potential: –Greatly reduced baseline restrictions. –No base station infrastructure requirements. –Hydrographic surveys up to 1 st order, but not special order. Desirable improvements: Desirable improvements: –Shortened solution convergence period. –Precision. –Accuracy. –Quantified system integrity.

18. CREATION OF LONG BASELINE, CARRIER PHASE GPS DATABASE

19. COASTAL GPS DATABASE Why: Comprehensive database of diverse, high- quality coastal PPK data not available Why: Comprehensive database of diverse, high- quality coastal PPK data not available Ideal: Multiple baselines up to 200 km. GPS and weather data collected in: all weather conditions; various climates; all seasons Ideal: Multiple baselines up to 200 km. GPS and weather data collected in: all weather conditions; various climates; all seasons Use: Validate new tropospheric modeling approaches; aid in driving evolution of long baseline, marine differential models Use: Validate new tropospheric modeling approaches; aid in driving evolution of long baseline, marine differential models Access: Once assembled, available to GPS researchers Access: Once assembled, available to GPS researchers Size: Expected to eventually contain ~0.5 Tb of data Size: Expected to eventually contain ~0.5 Tb of data

20. FERRY DATA COLLECTION The Princess of Acadia HLFX FRED Bay of Fundy Bay of Fundy Long / short baseline pairs Long / short baseline pairs Spatial and temporal diversity Spatial and temporal diversity Temperate Temperate Collocation of GPS / met. sensors / NWP Collocation of GPS / met. sensors / NWP 1.5 – 200 km baselines 1.5 – 200 km baselines 12 month data collection 12 month data collection

21. BUOY DATA COLLECTION Northern Gulf of Mexico Northern Gulf of Mexico Long / short baseline pairs Long / short baseline pairs Single set of GPS geometries Single set of GPS geometries Sub-tropical Sub-tropical Collocation of GPS / met. sensors / NWP Collocation of GPS / met. sensors / NWP 10 – 100 km baselines 10 – 100 km baselines 12 month data collection + 12 month data collection + BUOY BASE

22. IMPROVE GPS TROPOSPHERIC MODELING

23. WHY CONCENTRATE ON THE TROPOSPHERE? Mis-modeling of differential troposphere seen as largest limiting error source in coastal PPK GPS Mis-modeling of differential troposphere seen as largest limiting error source in coastal PPK GPS Large spatial and temporal variability in differential delay, especially significant for difficult to model “wet” component Large spatial and temporal variability in differential delay, especially significant for difficult to model “wet” component  Improved means of modeling required  Improved means of modeling required  Methods for applying modeling improvements in processing required  Methods for applying modeling improvements in processing required

24. STRATEGY FOR IMPROVED TROPOSPHERIC MODELING 1.Review performance of existing GPS tropospheric delay models 2.Determine feasibility and potential of estimating residual tropospheric delay 3.Evaluate application of in situ meteorological sensor measurements 4.Analyze level of benefit of employing regional weather data 5.Customize subset of these strategies

25. IMPROVED PPK GPS ALGORITHMS To validate modeling improvements and to make them useful To validate modeling improvements and to make them useful Using commercial software: Using commercial software: –“DynaPos” from The XYZ’s of GPS, Inc. –“GrafNav” from Waypoint Consulting Inc. Enhancement of academic software: Enhancement of academic software: –“UNBRTK” from University of New Brunswick. –“USMOTF” under development at University of Southern Mississippi

26. EXAMPLE OF IMPROVEMENT: NOAA EXPERIMENTAL TROPOSPHERIC PRODUCT NOAA developing nationwide troposphere and ionosphere delay products NOAA developing nationwide troposphere and ionosphere delay products Tropospheric product based on available weather information and estimated tropospheric delay from GPS network Tropospheric product based on available weather information and estimated tropospheric delay from GPS network Input: user location and time Input: user location and time Output: wet and hydrostatic tropospheric delay Output: wet and hydrostatic tropospheric delay Southern Miss beginning an assessment of quality and utility of product Southern Miss beginning an assessment of quality and utility of product

27. NOAA zenith wet delay NOAA zenith wet delay 22 July 2003 22 July 2003 U.S. Northeast U.S. Northeast Zenith delay range: 10 to 30 cm Zenith delay range: 10 to 30 cm  Slant delay range: 10 to ~300 cm  Slant delay range: 10 to ~300 cm

30. COMMON THEMES OF RESEARCH

31. THEMES Require higher accuracy and integrity for GPS-based positioning in coastal areas – for individual projects and for integration of projects Require higher accuracy and integrity for GPS-based positioning in coastal areas – for individual projects and for integration of projects Tropospheric mis-modeling needs to be reduced to extend GPS baselines Tropospheric mis-modeling needs to be reduced to extend GPS baselines Need to define current capabilities Need to define current capabilities Need to improve performance Need to improve performance

32. DEFINITION OF COASTAL GPS ACCURACY Increasing baseline length Increasing lat., long., hgt. error

33. IMPROVEMENT OF COASTAL GPS ACCURACY Increasing baseline length Increasing lat., long., hgt. error