Observing Fennoscandian geoid change for GRACE validation

Transcription

1 Observing Fennoscandian geoid change for GRACE validation L. Timmen 1, O. Gitlein 1, J. Müller 1, H. Denker 1, J. Mäkinen 2, M. Bilker 2, H. Wilmes 3, R. Falk 3, A. Reinhold 3, W. Hoppe 3, B.R. Pettersen 4, O.C.D. Omang 4, J.G.G. Svendsen 4, O. Øvstedal 4, H.-G. Scherneck 5, B. Engen 6, A. Engfeldt 7, G. Strykowski 8, R. Forsberg 8 [1] Institut für Erdmessung (IfE), University of Hannover, Germany [2] Finnish Geodetic Institute (FGI), Finland [3] Bundesamt für Kartographie und Geodäsie (BKG), Germany [4] Department of Mathematical Sciences and Technology, Agricultural Univ. of Norway (NLH) [5] Onsala Space Observatory, Chalmers Univ. of Technology, Sweden [6] Norwegian Mapping Authority (Statens Kartverk) [7] National Landsurvey of Sweden (Lantmäteriet) [8] National Survey and Cadastre (KMS), Denmark Correspondence to: L. Timmen (timmen@ife.uni-hannover.de) Abstract Scandinavia is a key study region for the research of glacial isostasy, and, in addition, it offers a unique opportunity for validating and testing the results of GRACE. Over a period of five years, the expected life time of GRACE, a temporal geoid variation of 3.0 mm is expected in the centre of the Fennoscandian land uplift area, corresponding to a gravity change of about 100 nm/s². This is expected to be within the detection capabilities of GRACE. With terrestrial absolute gravimetry, the gravity change due to the land uplift can be observed with an accuracy of ±10 to 20 nm/s² for a 5-year period. Thus, the terrestrial in-situ observations (groundtruth) may be used to validate and test the GRACE results. Since 2003, absolute gravity measurements have been performed in Scandinavia at about 30 stations covering Norway, Sweden, Finland and Denmark. Four groups with FG5 absolute gravimeters (BKG, FGI, IfE, NLH) are engaged to survey the uplift network annually by a mutually controlled procedure. Nearly all absolute stations are co-located with permanent GPS stations. From the 2003 and 2004 comparisons between the instruments, an overall accuracy of ±30 nm/s 2 is indicated for a single absolute gravimeter and a single station determination. This is in full agreement with the project goal. 1 Determination of the Fennoscandian land uplift In Fennoscandia, the Earth s crust is rising continuously since the last glacial maximum due to the deloading of the ice. The region is dominated by the Precambrian basement rocks of the Baltic Shield and comprises mainland Norway, Sweden, Finland, the Kola Peninsula, and Russian Karelia. The maximum spatial extension is about 2000 km in northeast-southwest direction; see Fig. 1 for the approximate shape and location (after Ekman 1996). Geophysical approaches to study the postglacial rebound are associated with the evidence for past sea lev- 1

2 els, the knowledge or assumptions about the geometry of the ice sheets (thickness, position), and some Earth model parameters (lithosphere thickness, mantle viscosity). After Lambeck et al. (1998), the inverse solution for the sea level data includes both ice and Earth model parameters as unknowns. Despite the recent progress in understanding the underlying models, definite models for the isostatic rebound do not yet exist. Lateral rheological variations have to be taken into account to obtain a more realistic glacially induced uplift model (Kaufmann et al. 2000). Figure 1. Map of the postglacial uplift of Fennoscandia in accordance with Ekman (1996). To monitor and investigate the recent land uplift in Fennoscandia, various measurements were collected since 1892: mareograph records, geodetic levellings, and relative gravity measurements since With these observations, the capability of terrestrial point measurement techniques to determine the land uplift was proven along east-west profiles. According to Ekman (1996), these observations reveal a maximum orthometric height change of 1 cm/a in the northern part of the Bothnian Gulf (see Fig. 1) and show symmetry around the maximum, closely correlated to the former Late Pleistocene Fennoscandian Ice Sheet. The height change in the centre is associated with a maximum gravity change of 20 nm/s² per year. Based on these numbers, a geoid change of 0.6 mm/a has been derived for the central area (Ekman and Mäkinen 1996). An eustatic sea level rise of 1.0 mm/a has been deduced from the tide gauge observations (Nakiboglu and Lambeck 1991). Since 1993, permanent GPS stations were established in Scandinavia to implement a further geodetic method with several advantages compared to the classical techniques (permanent data acquisition, homogeneous point distribution, large extension of the measurement area, low cost, three-dimensional survey). In this respect, the project BIFROST (Baseline Inferences for Rebound Observations, Sea Level, and Tectonics) was based on the GPS technique and geophysical modelling, and has delivered a maximum height variation (with respect to an ellipsoid) of more than 11 mm/a (see Fig. 2, cf. Milne et al. 2001, Johansson et al. 2002, Scherneck et al. 2003). The location of the centre and the geometrical structure of the uplift process differ from the previous model, with no clear zero line and more regional structures being visible. 2

3 Figure 2. Measured rates at BIFROST permanent GPS stations after Scherneck et al. (2003). During the mission duration of GRACE (about five years), a temporal geoid change of 3.0 mm can be expected in the centre of the Fennoscandian land uplift area, corresponding to a gravity change of about 100 nm/s² ( 10 μgal). This is a clear secular gravity change of regional scale, and it might be a challenging task to detect this signal from satellite gravity data, in particular GRACE. After Wahr and Velicogna (2003), the land uplift causes a measurable signal in the GRACE observations. Early results from Tapley et al. (2004) confirm that this satellite mission is able to resolve geoid variations for a range of spatial scales down to 400 km for particular regions with large signals. They found that the error level in the 2003 solutions was in the order of 2 to 3 mm for spatial features of about 600 km. Considering the large extension of the land uplift area, Fennoscandia is a suitable application region for GRACE. Vice versa, the temporal gravity field change can be used for the validation of the GRACE results. Because the observation of the rebound signal is interfered with mass variations due to oceanographic, land hydrology and atmospheric processes, these effects have to be accounted for in the GRACE data analysis with appropriate mathematical approaches (e.g. Wiehl et al. 2005). Hence, the combination with other geological and geodetic measurements is inevitable. 2 Absolute gravimetry Besides the geometrical methods, terrestrial absolute gravimetry is a further geodetic technique to study the land uplift. In addition and complementary to the other geodetic measurements, absolute gravimetry have the following positive characteristics: accuracy ±20 to 30 nm/s² (10 nm/s² 1 μgal) for a single station determination, absolute monitoring of mass movements and vertical displacements (no problems with calibration and datum level), accuracy of absolute gravity net is independent of geographical extension, independent validation method for GPS, VLBI, SLR, and superconducting gravimetry, combined with geometrical methods, vertical surface deformations and subsurface mass movements can be separated. The benefit of absolute gravimetry has already been exploited in different research projects, covering areas of global or regional extensions, and monitoring variations caused by mass movements. The International Absolute Gravity Basestation Network (IAGBN) serves, among other purposes, to determine large scale tectonic plate movements (Boedecker and 3

4 Fritzer 1986). The recommendations of the Interunion Commission of the Lithosphere on Mean Sea Level and Tides propose the regular implementation of absolute gravity measurements at coastal points, 1 to 10 km away from tide gauges (Carter et al. 1989). The height differences between gravity points and tide gauges have to be controlled by levelling or GPS. In Great Britain, the main tide gauges are controlled by repeated absolute gravity determinations in combination with episodic or continuous GPS measurements (Williams et al. 2001). To determine crustal deformations by absolute gravimetry, secular gravity changes should be measured with a precision of about ±5 nm/s² per year. This can be achieved by annual measurements over five years. To exploit absolute gravimetry in combination with GPS for a pointwise validation of the GRACE results or to support the GRACE data evaluation by terrestrial gravimetry, the temporal variations of gravity disturbances (or gravity anomalies) are needed in accordance with the resolution of the GRACE data. Because of the longwavelength nature of the GRACE results, the terrestrial point results derived from absolute gravimetry have to be reduced for all local effects changing gravity with time. In this connection, a severe problem is the subsurface water mass movement (change of groundwater table, temporary water storage in clefts and crevasses). Such impacts are partly considered by the station selection. Moreover, by measuring every year the absolute gravity value at a station over a 5-year period, the impact of ground water variations is averaged out to a large extent within the computation of the linear gravity rate. In addition, observations of the ground water table in boreholes and in nearby wells during the absolute gravity surveys are taken and used to assess the disturbing impact. Furthermore, a second favourable averaging effect arises from deriving a spatial mean over a larger area with a few hundred of kilometres extension. For that reason, and to allow the determination of an asymmetric uplift model with possible regional structures, a rather large number of stations is occupied every year over the whole Scandinavian area. 3 The conversion of measured land uplift to geoid change Proceeding with an appropriate data set of terrestrial geodetic observations acquired over a period of at least five years, a model of temporal geoid changes in Fennoscandia can be derived, which is suitable for direct comparisons with the GRACE results. The calculation is based on a formula found by Hotine (1969), which is extended here for the conversion of temporal gravity and height changes to geoid changes: R 2γ N& = H ( ψ) g& + h& dσ, 4πγ σ r 1 1 (1) Ψ ( ) sin ln Ψ with H Ψ = sin N &, h&, g& are the temporal changes of geoid heights, ellipsoidal heights, and gravity; γ is the normal gravity; R is the mean Earth radius; r is the radius of computation point P, and Ψ is the spherical distance. This equation shows that the measured gravity changes from absolute gravimetry have to be combined with the ellipsoidal height changes from GPS to obtain geoid height changes. The accuracy of the observed gravity and height changes should be in accordance to each other (e.g., 20 nm/s 2 and 10 mm). 4

5 4 Project realisation In 2002, the Institut für Erdmessung (IfE) of the University of Hannover has received a new FG5 absolute gravity meter from Micro-g Solutions, Inc. (Erie, Colorado), which is a stateof-the-art instrument (Niebauer et al. 1995). This version replaced the older JILAG3 system at IfE, which was successfully employed in South America, Europe and China since In the meantime, a joint project for the survey of the land uplift in Fennoscandia was established. The Working Group for Geodynamics of the Nordic Geodetic Commission (NKG) serves as a platform to organise the project. Besides the IfE from Hannover, the following institutions are participating in the project: National Survey and Cadastre (KMS, Copenhagen/Denmark), Finnish Geodetic Institute (FGI, Masala/Finland), Bundesamt für Kartographie und Geodäsie (BKG, Frankfurt/Germany), Institute of Mapping Sciences, Agricultural University of Norway (NLH, Ås), Statens Kartverk (SK, Hønefoss/Norway), Onsala Space Observatory (Chalmers University of Technology, Onsala/Sweden), National Land Survey of Sweden (Landmäteriet, Gävle). The FGI procured a new FG5 at the beginning of 2003, and the NLH in spring The project realisation and strategy may be summarised as follows: absolute gravity determinations at 22 stations in 2003 with 3 gravimeters employed: BKG (), FGI (FG5-221), IfE (FG5-220), absolute gravity determinations at 24 stations in 2004 with 3 gravimeters employed: FGI (FG5-221), IfE (FG5-220), NLH (FG5-226), repetition measurements in the years 2005, 2006 and 2007, simultaneous GPS measurements (gravity stations are normally co-located with permanent GPS stations), geometrical connections between gravity stations and tide gauges (mainly by GPS), auxiliary levelling measurements to eccentres (control of local variations, ties between gravity and GPS stations), elaboration of reduction models (air mass movements, ocean tides, etc.), integration of already existing geodetic data sets (e.g., FG5 surveys by Wilmes et al. 2004), utilization of the project products (temporal changes of gravity disturbances, twodimensional model describing the geoid change) for GRACE validation. Figure 3. Integration of different geodetic techniques to survey the temporal gravity and geoid variations of the Fennoscandian land uplift area. 5

6 Figure 4. Observed absolute gravity stations in 2004 occupied by the absolute gravimeters FG5-220 (IfE), FG5-221 (FGI), FG5-226 (NLH). Figure 5. Station Skellefteå (Sweden) with an absolute gravity pier inside and permanent GPS antenna outside. Fig. 3 demonstrates the principles to combine different geodetic methods. Nearly all absolute gravity sites are co-located with permanent GPS stations, and also tide gauges are in the vicinity of coastal stations. The gravity points, GPS stations and tide gauges are connected locally, using terrestrial surveying techniques, like levelling, to monitor the local vertical variations. Fig. 4 shows the stations occupied in The employment of more than one absolute gravimeter allows simultaneous (parallel) observations in stations with two sites close to each other and control measurements on identical sites. This strategy increases the network reliability and accuracy because it helps to identify possible offsets of the instruments. To exclude uncertainties introduced by relative gravimetry (e.g. via the measured 6

7 vertical gravity gradient) into the absolute gravimetric results, the final absolute values are all related to a common height at m above the reference mark at floor level, cf. Timmen (2003). Figures 5 to 7 show typical conditions for the absolute gravimetric field work in Scandinavia. Figure 6. Determination of the local vertical gravity gradient with relative gravimeters to transfer the measured absolute gravity value to a common reference height. Figure 7. Absolute gravimeter FG5-220 of IfE installed at station Östersund. 5 Measurement accuracy of absolute gravimetry Table 1 shows the results of the comparisons of the IfE absolute gravimeter FG5-220 with the FG5-221 of FGI and the of BKG carried out during the absolute gravity campaigns in The comparisons were made by parallel registrations (Bad Homburg, Metsähovi, Vaasa) or double occupations (Onsala, Skellefteå, Trondheim, Trysil, Hønefoss). The decline of the observed g-values at the IfE reference station in Clausthal (Harz mountains) in 2003 can be explained by the decrease of the subsurface water content from spring to autumn 2003 due to a very dry season. The re-occupation of 4 sites by FG5-220 in August/September, observed already by in July, shows a deviating result for Hønefoss. During the 4-day occupation with the IfE instrument, all observations were disturbed by incoming microseismic noise, which was assumed to be of man-made origin. In March 2004 the IfE group (FG5-220) visited the new gravity laboratory of NLH in Ås to provide a reference for the new FG The result obtained with the new instrument one month later differs by -23 nm/s 2. A nearly simultaneous comparison of the FG5-220 to FG5-221 (FGI) in Vaasa in June 2004 yielded a difference of -14 nm/s 2. Comparisons at Metsähovi in May and in July 2004 agreed all within 20 nm/s 2 (online results). The overall discrepancy (RMS) of the comparisons of about ±20 to 30 nm/s 2 proves the high accuracy of the absolute gravimetric survey of Fennoscandia. 7

8 Table 1. Comparisons of the IfE absolute gravimeter FG5-220 with FG5-221 (FGI) and (BKG), and with Clausthal reference value (average in 2003). The difference is calulated as g(fg5-220) minus g(i). Date Station FG5-220 com- Diff. Remark Jan 03 Feb 03 Feb 03 May 03 Jun + Jul 03 Jun 03 Aug + Jul 03 Sep + Jul 03 Sep + Jul 03 (Sep + Jul 03 Aug 03 Aug 03 Aug 03 Aug 03 Oct 03 Nov 03 Nov 03 Clausthal Bad Homburg AA Bad Homburg BA Clausthal Onsala Clausthal Skellefteå Trondheim Trysil Hønefoss Metsähovi AB Metsähovi AC Vaasa AA Vaasa AB Clausthal Bad Homburg AA Bad Homburg AB pared with Mean 2003 Mean 2003 Mean 2003 FG5-221 FG5-221 FG5-221 FG5-221 Mean 2003 Mean RMS [nm/s 2 ] ) ±27 6 Summary and conclusions As a short status report, the ongoing project may be described as follows: The whole uplift network includes more than 30 absolute gravity stations located beside or in the vicinity of permanent GPS points. If possible, all stations will be occupied annually over a period of five years. In the first two years, absolute gravity determinations were performed at 22 stations in 2003 and 24 stations in 2004, employing 4 FG5 absolute gravimeters. Results derived from surveys with the IfE absolute gravimeter FG5-220 have been compared with results from the other participating instruments and with results from repetition measurements at the IfE reference station in Clausthal. The comparison leads to an rms discrepancy of approx. ±20 to 30 nm/s 2. The check of the instruments by parallel and reference measurements is essential. Especially for geodynamics projects more than one absolute gravimeter should be employed to increase the reliability of the results and to detect instrumental offsets. This procedure improves the absolute accuracy of the whole network. An overall accuracy of better than ±30 nm/s 2 is indicated for a single absolute gravimeter and a single station determination. This agrees well with the project aim to provide observed gravity changes in Fennoscandia with a sufficient accuracy to support the GRACE data evaluation. 8

9 Acknowledgements The research is supported by Deutsche Forschungsgemeinschaft (DFG, German Research Foundation), through the research grant Mu 1141/3-1. References Boedecker, G., Fritzer, Th.: International Absolute Gravity Basestation Network, Veröff. Bayer. Komm. für die Internat. Erdmessung der Bayer. Akad. d. Wissensch., Astron.- Geod. Arb. 47, München, Carter, W.E., Aubrey, D.G., Baker, T. et al.: Geodetic fixing of tide gauge bench marks, Woods Hole Oceanographic Institution Report WHOI-89-31/CRC-89-5, Woods Hole, Mass., U.S.A., Ekman, M.: A consistent map of the postglacial uplift of Fennoscandia, Terra Nova 8, , Ekman, M., Mäkinen, J.: Recent postglacial rebound, gravity change and mantle flow in Fennoscandia, Geophys. J. Int. 126, , Hotine, M.: Mathematical Geodesy, ESSA Monograph 2, U.S. Dept. of Commerce, Washington, D.C, Johansson, J.M., Davis, J.L., Scherneck, H.-G., Milne, G.A., Vermeer, M., Mitrovica, J.X., Bennett, R.A., Jonsson, B., Elgered, G., Elósegui, P., Koivula, H., Poutanen, M., Rönnäng, B.O., Shapiro, I.I.: Continuous GPS measurements of postglacial adjustment in Fennoscandia, 1. geodetic results. J. Geophys. Res. 107, B8, ETG 3, 1-27, Kaufmann, G., Wu, P. Li, G.: Glacial isostatic adjustment in Fennoscandia for a laterally heterogeneous Earth, Geophys. J. Int., 143, , Lambeck, K., Smither, C., Johnston, P.: Sea level-change, glacial rebound and mantle viscosity for northern Europe, Geophys. J. Int., 134, , Milne, G.A., Davis, J.L., Mitrovica, J.X., Scherneck, H.-G., Johannson, J.M., Vermeer, M., Koivuly, H.: Space-geodetic constraints on glacial isostatic adjustment in Fennoscandia, Science 291, , Nakiboglu, S.M., Lambeck, K.: Secular sea-level change, in: Sabadini, R. et al. (eds.): Glacial isostasy, sea level and mantle rheology, , Kluwer Acad. Publ., Niebauer T.M., Sasagava G.S., Faller J.E., Hilt R., Klopping F.: A new generation of absolute gravimeters, Metrologia 32, , Scherneck, H.-G., Johansson, J., Koivula, H., van Dam, T., and Davis, J.: Vertical crustal motion observed in the BIFROST project. J. of Geodyn. 35, , Tapley, B.D., Bettadpur, S., Ries, J.C., Thompson, P.F., Watkins, M.M.: GRACE measurements of mass variability in the Earth system, Science 305, , Timmen, L.: Precise definition of the effective measurement height of free-fall absolute gravimeters, Metrologia 40, 62-65, Wahr, J., Velicogna, I.: What might GRACE contribute to studies of post glacial rebound, Space Science Reviews 108, ,

10 Wiehl, M., Dietrich, R., Lehmann, A.: How Baltic sea water mass variations mask the postglacial rebound signal in CHAMP and GRACE gravity field solutions, in: Reigber, Ch., Lühr, H., Schwintzer, P., Wickert, J. (eds.): Earth observation with CHAMP, , Springer-Verlag Berlin Heidelberg, Williams, S.D.P., Baker. T.F., Jeffries, G.: Absolute gravity measurements at UK tide gauges. Geophys. Res. Letters 28, , Wilmes, H., Falk, R., Klopping, F., Roland, E., Lothhammer, A., Reinhold, A., Richter, B., Plag, H.-P., Mäkinen, J.: Long-term gravity variations in Scandinavia from repeated absolute gravity measurements in the period 1991 to 2003, proceedings (CD publ.) of the IAG symposium Gravity, Geoid and Space Missions, Porto,