Joint analysis of historic leveling data and permanent GPS observations for subsidence determination over the Netherlands R. Grebenitcharsky, R. Hanssen and P. Ditmar [R.Grebenitcharsky@tudelft.nl,R.F.Hanssen@tudelft.nl, P.Ditmar@tudelft.nl], DESCRIPTION: A joint analysis of historic leveling data together with permanent GPS vertical observations is conducted. Conclusions are drawn about the absolute and relative vertical movements over the Netherlands. This analysis is used to formulate a joint model for subsidence determination using leveling, GPS permanent observations and InSAR. PERMANENT GPS DATA DATA: Time series of ellipsoidal heights for 13 permanent GPS stations in the Netherlands and neighbor countries are used. We use time series from global solutions for stations KOSG, WSRT and BRUS and time series for all local baselines w.r.t. KOSG from weekly EUREF solutions. EFFECT OF PERIODICS ON VERTICAL RATE RESULTS: For local time series a wavelet decomposition up to 12 level is applied. It allows to investigate and remove the effect of periodics less than 39 years and to estimate the remaining trend. The final vertical rate estimations for the EUREF stations is the sum of the global vertical rate for KOSG and the relative rates w.r.t. KOSG. COMPARISON BETWEEN RATES GPS AND LEVELING UNITS:[CM/CENTURY] * Background counters by TNO – National Geological Survey Agreement between the rates due to periodics larger than 39 years and trend and the rates determined by leveling (TNO – National Geological Survey) is observed. HISTORICAL LEVELING DATA DATA: Four leveling campaigns between first order bench marks all over the Netherlands are used (1930s, 1950s, 1970s, 1990s). RELATIVE VERTICAL VELOCITIES (from west to east) RESULTS: The relative vertical velocities between bench marks are determined. Areas with significant relative rates (pink boxes) are shown. One relatively stable area (blue box) in the south-east and a large area in the south-west with significant relative vertical velocities exist. RELATIVE VERTICAL VELOCITIES AND ACTIVE QUATERNARY FAULTS ( * after Worum et al., 2005, Quat. Sc. Rev. # 24) In the stable area significant relative velocities can be associated with existing faults A NEW STEPWISE LEAST-SQUARES ADJUSTMENT APPROACH Absolute velocity estimations from relative velocities from leveling & InSAR and relative & absolute velocities from GPS allows a better representation of vertical rates. Vertical changes of WSRT w.r.t. KOSG RAW DATA (green), PERIODIC COMPONENTS (blue), UNITS [MM] approximation level vertical rate -0.07mm/year 1th level vertical rate -0.00mm/year 2th level vertical rate 0.00mm/year 3th level vertical rate -0.01mm/year 4th level vertical rate 0.01mm/year 5th level vertical rate 0.09mm/year 6th level vertical rate -0.16mm/year 7th level vertical rate -0.45mm/year 8th level vertical rate -0.11mm/year 9th level vertical rate -0.53mm/year 10th level vertical rate 0.03mm/year 11th level vertical rate -0.07mm/year 12th level vertical rate -0.06mm/year 1997 1998 1999 2000 2001 2002 2003 2004 2005 APPROXIMATED TREND (red) : -0.19mm/year TREND FROM RAW DATA (black) : -1.31mm/year mm/y + Relative velocities from leveling plus their variance- covariance estimations Relative& absolute velocities from GPS Relative vertical velocities from InSAR Surface displacements GPS Leveling Difference KOSG -0.60 -0.90 -0.30 WSRT -2.50 -2.00 +0.50 BORK -12.30 -5.10 +7.20 TERS -8.90 -10.50 -1.60 DELF -9.20 -7.20 2.00 EIJS +3.30 +7.00 3.70