DGFI reference frame solution as contribution to ITRF2008

Transcription

1 COST Action: ES0701, Vienna, Austria, November 16-17, 2010 WG2: Velocity determination / reference frame realization DGFI reference frame solution as contribution to ITRF2008 D. Angermann, M. Seitz, H. Drewes Deutsches Geodätisches Forschungsinstitut, München

2 Contents 1. DGFI combination methodology for ITRF Results of DGFI solution 3. ITRF2008 accuracy evalution 4. Discussion regarding non-linear station motions

3 ITRF2008 input data sets Time series of station positions and Earth orientation parameters Techn. Service / TC Data Time span Constraints GPS IGS / NRCan Weekly sol Minimum VLBI IVS / IGG 24 h sess. NEQ None SLR ILRS / ASI Weekly sol Loose DORIS IDS / CLS - CNES-GSFC Weekly sol Minimum

4 ITRF2005 / ITRF2008 Kombinationsstrategie am DGFI Generation of constraint-free normal equations (NEQ) Epoch 1 VLBI NEQ SLR NEQ GPS NEQ DORIS NEQ Epoch 2 VLBI NEQ SLR NEQ GPS NEQ DORIS NEQ Epoch n VLBI NEQ SLR NEQ GPS NEQ DORIS NEQ Accumulation of time series Multi-year NEQ s VLBI NEQ SLR NEQ GPS NEQ DORIS GPS NEQ

5 ITRF2005 / ITRF2008 Kombinationsstrategie am DGFI Input: Datum-free normal equations (NEQ) Epoch 1 VLBI NEQ SLR NEQ GPS NEQ DORIS NEQ Epoch 2 VLBI NEQ SLR NEQ GPS NEQ DORIS NEQ Epoch n Multi-year NEQ s VLBI NEQ VLBI NEQ SLR NEQ GPS NEQ Accumulation of time series SLR NEQ GPS NEQ DORIS NEQ DORIS GPS NEQ The accumulation comprises Analysis of station position time series Identification of discontinuities Introduction of station velocities Combination of epoch normal equations

6 ITRF2005 / ITRF2008 Kombinationsstrategie am DGFI Input: Datum-free normal equations (NEQ) Epoch 1 VLBI NEQ SLR NEQ GPS NEQ DORIS NEQ Epoch 2 VLBI NEQ SLR NEQ GPS NEQ DORIS NEQ Epoch n Multi-year NEQ s VLBI NEQ VLBI NEQ SLR NEQ GPS NEQ Accumulation of time series SLR NEQ GPS NEQ DORIS NEQ DORIS GPS NEQ The accumulation comprises Analysis of station position time series Identification of discontinuities Introduction of station velocities Combination of epoch normal equations Inter-technique combination Station positions, velocities and Earth Orientation Parameters

7 ITRF2005 / ITRF2008 Kombinationsstrategie am DGFI Input: Datum-free normal equations (NEQ) Epoch 1 VLBI NEQ SLR NEQ GPS NEQ DORIS NEQ Epoch 2 VLBI NEQ SLR NEQ GPS NEQ DORIS NEQ Epoch n Multi-year NEQ s VLBI NEQ VLBI NEQ SLR NEQ GPS NEQ Accumulation of time series SLR NEQ GPS NEQ Inter-technique combination Station positions, velocities and Earth Orientation Parameters DORIS NEQ DORIS GPS NEQ The accumulation comprises Analysis of station position time series Identification of discontinuities Introduction of station velocities Combination of epoch normal equations The inter-technique comb. comprises Estimation of weighting factors Selection of local tie vectors Combination of techn.-specific NEQ Realization of the geodetic datum

8 Time series analysis and accumulation Time Series of GPS station YSSK, Russia (Sakhalin Seismic Belt) North East Height [mm] Hokkaido Earthquake , Magn. 8.3 Kuril Isl. Earthquake , Magn. 8.3

9 Time series analysis and accumulation Time series of GPS station HOFN, Iceland Antenna and Receiver Change Technique # Stations # Discontinuities % GPS SLR VLBI DORIS

10 Time series analysis and accumulation Time series of network scaling factor SLR VLBI

11 Inter-technique combination Distribution of co-location sites

12 Inter-technique combination Observation time spans for 4-techniques co-location sites Co-location sites VLBI SLR GPS DORIS Metsahovi, Finland Hartebeesthoek, S. Afr Goldstone, U.S.A Washington, U.S.A Monument Peak, U.S.A Santiago, Chile Tidbinbilla, Australia

13 32 (out of 41) VLBI - GPS co-location sites 3-D differences between local ties and space geodetic solutions-d Wettzell

14 29 (out of 37) SLR - GPS co-location sites 3-D differences between local ties and space geodetic solutions-d

15 40 (out of 42) GPS DORIS co-location sites 3-D differences between local ties and space geodetic solutions-d

16 9 (out of 18) VLBI - SLR co-location sites 3-D differences between local ties and space geodetic solutions-d

17 Observed discrepancies at co-location sites 3-D differences between ITRF2008 results and local tie vectors Co-locations < 5 mm 5-10 mm mm > 20 mm GPS VLBI (32) GPS SLR (29) GPS DORIS (40) VLBI SLR (9) Total (110)

18 Horizontal station velocities of DGFI solution

19 Vertical GPS station velocities

20 Vertical VLBI station velocities

21 Vertical SLR station velocities

22 Vertical DORIS station velocities

23 Comparison of vertical station velocities GPS DORIS VLBI SLR

24 Example: Time series of station positions SLR station Maidanak, Uzbekistan

25 Accuracy assessment of ITRF2008 Internal: Comparison of intra-technique solutions and combined DGFI solution R.M.S. Residuals DORIS GPS SLR VLBI Positions [mm] Velocities [mm/yr] External: Comparison of DGFI and IGN solutions R.M.S. Residuals DORIS GPS SLR VLBI Positions [mm] Velocities [mm/yr]

26 Comparison of DGFI solution with ITRF2008 Station coordinates residuals of all stations GPS VLBI SLR DORIS

27 Comparison of DGFI solution with ITRF2008 Helmert-transformation parameters, positions [mm] translation rotation scale

28 Comparison of DGFI solution with ITRF2008 Helmert-transformation parameters, velocities [mm/yr] translation rotation scale Long-term stability: GPS DORIS, VLBI, SLR 0.2 mm/yr mm/yr

29 Summary: ITRF2008 accuracy Comparison of DGFI solution with ITRF Transformation parameters: positions < 5 mm; velocities: < 1mm/yr - Station positions: VLBI 0.5 mm; GPS 1.5 mm; SLR 2 mm; DORIS: 3 mm - Station velocities: VLBI 0.1 mm/yr; GPS 0.2 mm/yr; SLR, DORIS: 0.8 mm/yr Limiting factors, challenges for further improvements: - Discontinuities: too many in particular in case of GPS - Distribution of high-quality co-location sites - Non-linear (e.g., seasonal) station motions are not considered

30 Non-linear station motions Major reasons are: Equipment changes => discontinuities Earthquakes, co- and post-seismic deformations => discontinuities, piece-wise linear functions Seasonal variations => not considered in ITRF2008

31 Seasonal signals in station positions Amplitudes of seasonal signals in GPS station heights

32 Effects of seasonal signals Example: GPS station Irkutsk, Siberia Δ velocities w.r.t. linear model Obs time span [yrs] Δ velocities [mm/yr] ± ± ± ± 0.5 If seasonal variations are neglected, there are... Errors in velocity estimations (e.g. Δt < 2.5 yrs) Systematic effects in the combined solution Biases in the alignment of epoch solutions and regional networks

33 Parameterization of seasonal signals Shape of seasonal signals can be approximated by sine/cosine annual and semi-annual functions Brasilia Ankara Estimation of annual signals in addition to velocities? Advantages: - Improved velocity estimation - Better alignment of epoch solutions Disadvantages / open questions: - More parameters (stability)? - Seasonal signal geophysically meaningful? - How to parameterize seasonal signals?

34 How to consider non-linear station motions? Different approaches shall be investigated: An extended parameterization of seasonal station variations e.g., by sine/cosine annual and semi-annual functions; Estimation of epoch reference frames to account for non-linear effects in station motions; Application of geophysical models and implementation of physical background models, e.g., for atmospheric and hydrological loading. External measurements, e.g., VLBI telescope deformation, GRACE-derived hydrology Estimation of geophysical model parameters within the TRF solution (e.g., Love numbers)