Hydrogeological investigations at the Membach station, Belgium,
and application to correct long periodic gravity variations
M. Van Camp,1 M. Vanclooster,2 O. Crommen,3 T. Petermans,1 K. Verbeeck,1
B. Meurers,4 T. van Dam,5 and A. Dassargues3
Received 21 March 2006; revised 8 May 2006; accepted 6 July 2006; published 20 October 2006.
[1] A comprehensive hydrogeological investigation regarding the influence of
variations in local and regional water mass on superconducting gravity measurements is
presented for observations taken near the geodynamic station of Membach, Belgium.
Applying a regional water storage model, the gravity contribution due to the elastic
deformation of the Earth was derived. In addition, the Newtonian gravity effect induced by
the local water mass variations was calculated, using soil moisture observations taken at
the ground surface (about 48 m above the gravimeters). The computation of the
gravimetric effect is based on a digital elevation model with spatially discretized
rectangular prisms. The obtained results are compared with the observations of a
superconducting gravimeter (SG). We find that the seasonal variations can be reasonably
well predicted with the regional water storage model and the local Newtonian effects.
Shorter-period effects depend on the local changes in hydrology. This result shows the
sensitivity of SG observations to very local water storage changes.
Citation: Van Camp, M., M. Vanclooster, O. Crommen, T. Petermans, K. Verbeeck, B. Meurers, T. van Dam, and A. Dassargues
(2006), Hydrogeological investigations at the Membach station, Belgium, and application to correct long periodic gravity variations,
J. Geophys. Res., 111, B10403, doi:10.1029/2006JB004405.
1. Introduction
[2] Stable gravity observations over long periods can be a
useful tool for understanding geodynamic changes related to
Earth structure [Hinderer and Crossley, 2000], tectonics
[Francis et al., 2004], postglacial rebound [Williams et al.,
2001; Lambert et al., 2001] and natural as well as anthro-
pogenic subsidence [Zerbini et al., 2001, 2005]. However,
gravity is also affected by changes in environmental surface
mass. To obtain a reliable estimate of the geodynamic
signal, corrections for environmental mass changes are
routinely applied to the gravity time series. In the past,
the focus was on the effects of atmospheric mass move-
ments; for an excellent review of this problem please refer
to Spratt [1982], Merriam [1992], and Boy et al. [2002],
who applied routine corrections to continuous supercon-
ducting gravimeter time series.
[3] The influence of regional and local water storage
variations on gravity is more difficult to model. In contrast
to the atmosphere, where mass variations are mostly char-
acterized by long wavelengths (>500 km) [Merriam, 1992;
Neumeyer et al., 2004], changes from small changes in
water storage over very short wavelengths (<10 km) can
have a large and immediate impact on gravity observations.
Numerous investigations have been undertaken to address
the effects of local water storage variations on gravity
observations. Studies about the influences of local precip-
itation, soil moisture and groundwater are given by Ma¨kinen
and Tattari [1990], Crossley and Xu [1998], Bower and
Courtier [1998], Peter et al. [1995], Kroner [2001], Kroner
and Jahr [2006], Lambert and Beaumont [1977], Llubes et
al. [2004], Hasan et al. [2006], Imanishi et al. [2006], and
Abe et al. [2006]. The results indicate that the relationship
between local gravity variations and water storage depends
on very local geologic and hydrologic conditions, e.g., rock
porosity, vegetation, evaporation, and runoff rates.
[4] More recently, the relationship between regional longer-
period water storage signals, have also been investigated
over distances of several 100 km [van Dam et al., 2001;
Crossley et al., 2005]. While a large part of the observed
seasonal gravity change can be attributed to long-wave-
length changes in water storage, the variability in amplitude
and phase of the seasonal signal from superconducting
gravimeters in Europe, indicates the importance of a signif-
icant local component as well [Boy and Hinderer, 2006;
Neumeyer et al., 2006].
[5] The aim of this paper is to improve our understanding
of the dependence of gravity on extremely local changes in
soil moisture around the geodynamic station in Membach,
Belgium. Francis et al. [2004] investigated this issue when
JOURNAL OF GEOPHYSICAL RESEARCH, VOL. 111, B10403, doi:10.1029/2006JB004405, 2006
Click
Here
for
Full
Article
1Seismology, Royal Observatory of Belgium, Brussels, Belgium.
2Department of Environmental Sciences and Land Use Planning,
Universite´ catholique de Louvain, Louvain-la-Neuve, Belgium.
3Hydrogeology and Environmental Geology, University of Lie`ge,
Lie`ge, Belgium.
4Institute of Meteorology and Geophysics, University of Vienna,
Vienna, Austria.
5European Center for Geodynamics and Seismology and Natural
History Museum of Luxembourg, Walferdange, Grand Duchy of
Luxembourg.
Copyright 2006 by the American Geophysical Union.
0148-0227/06/2006JB004405$09.00
B10403 1 of 13

they compared seasonal and shorter-term gravity changes at
this location with predicted gravity changes from (1) mod-
eled water storage effects (1

1
spatial resolution and
monthly temporal resolution [see van Dam et al., 2001]),
(2) volume changes in two surface water reservoirs located
at 3 and 6 km, respectively, from the gravity site, and (3) local
rainfall. They found that the annual to monthly variations in
gravity tracked the reservoir changes well, but that gravity
changes with higher frequencies could only be explained by
local rainfall events (see Francis et al.’s Figure 4). With
respect to the water storage model, Francis et al. [2004]
focused on the annual signal and found that the amplitude of
the gravity changes predicted from the model was very
comparable to the observed ones in the SG gravity data, but
that study did not analyze the result further. On the other
hand the authors could not retrieve a reliable relationship
between the water level in the reservoirs and the observed
changes in gravity. They concluded that the installation of
monitoring probes above the gravimeter would help to
constrain a relationship between the local water saturation
in the ground and gravity change.
[6] We extend the work of Francis et al. [2004] by
including observations of soil moisture collected above
the superconducting gravimeter (SG). Using the inferred
water content and a digital elevation model (DEM) spatially
discretized into rectangular prisms, the Newtonian mass
effect on gravity is estimated. We find that with an im-
proved understanding of the hydrological effects in Mem-
bach, we are able to reduce significantly the short-period
scatter changes in gravity time series due to rainfall events,
as well as the seasonal variations. The removal of seasonal
and higher frequency effects from the gravity signal
improves our ability to monitor long-term gravity changes
induced by tectonics and postglacial rebound in the region.
2. Geophysical Station of Membach
[7] The Membach geophysical station is located in the
eastern part of Belgium. The station houses an accelerom-
eter, short-period and broadband seismometers and the
superconducting gravimeter, C021. The SG is installed at
the end of a 130 m long tunnel excavated in low-porosity
argillaceous sandstone. The gravity sensor is 48.5 m below
the surface, which is covered by primarily a deciduous
forest canopy. Continuous observations have been taken
since August 1995.
[8] The fundamental component of a SG consists of a
hollow superconducting sphere that levitates in a persistent
magnetic field [Goodkind, 1999]. An incremental change in
gravity induces a vertical displacement of the sphere. A
feedback voltage is induced to keep the sphere at a zero
position. This feedback voltage is proportional to the
gravity change. The SG provides relative gravity measure-
ments and the most common mode of operation is contin-
uously at a fixed location.
[9] Basic processing of the SG data includes editing and
correcting for steps, spikes and other instrumental distur-
bances (e.g., helium fills). The Earth tides, ocean loading
and polar motion effects are subsequently removed. The
Newtonian and loading atmospheric influences are cor-
rected by using a linear admittance factor of 3.3 mGal/
hPa (for details on the corrections, see Francis et al.
[2004]). The weak SG instrumental drift is controlled using
the absolute gravimeter (AG) FG5 202. Unlike the SG, the
AG is mobile and measures the gravity at the station many
times per year [Van Camp and Francis, 2006]. Being a
relative meter, the SG C021 is calibrated using the AG. The
AG is also used to remove the secular tectonic trend from
the data [Van Camp et al., 2005].
[10] Figure 1a shows the SG time series from Membach.
First, we observe a possible long-period (order of 20 years)
oscillation in the data, possibly due to hydrological effects.
Indeed, such a long-period signal can also be observed in
the predicted gravity effect calculated using water storage
estimate as well as in the reservoir capacities (Figures 1b
and 2a, see sections 3 and 4 for details). However, this is to
be confirmed when longer time series are available and
since we are primarily concerned with seasonal and shorter-
period responses to water storage, our results will not be
affected by removing this signal. Therefore we fit and
removed a sinusoid with a period of 20 years to the SG
data (red line in Figure 1a).
[11] Since June 2003 rainfall has been measured in situ
with a tipping bucket type rain gauge (Lufft 8353.01). The
tipping is counted and integrated in the data logging system
at a minute interval, the resolution is 0.1 mm and the
calibration is controlled at least yearly. Before 2003 rainfall
data were provided by daily measurements taken at the
Gileppe dam, 3 km away from the Membach station.
[12] In August 2004, four CS616 soil water content ref-
lectometers from Campbell ScientificTM [Campbell Scientific,
2002] were installed in the shallow partially saturated soil,
48 m above the station. These probes observe soil moisture
changes once an hour at 30, 45, 50 and 60 cm below soil
surface. The CS616 sensors are time domain soil reflectom-
eters (TDR) that measure the apparent dielectric constant of
soil. The sensor consists of two stainless steel rods connected
to a multivibrator. The rods act as a waveguide. An electro-
magnetic pulse travels the length of the probe rods and is
reflected from the probe ends traveling back to the probe
head. The reflection is detected and a new pulse is triggered.
The pulsing frequency in free air is about 70 MHz; however,
the velocity of the pulse is dependent on the dielectric
permittivity of the material surrounding the rods. The dielec-
tric number at a frequency of 1 MHz to 1 GHz is about 1 for
air, 3–4 for major soil minerals and 80 for water because the
polarization of water molecules takes time. Thus the dielec-
tric number of soil is highly dependent on the volumetric
water content (volume of water per volume of soil) in the
unsaturated zone, and consequently, the frequency. A data
logger measures the probe output period, which is empiri-
cally related to volumetric water content using a calibration
equation. The resolution of the probes is 0.1% volumetric
water content.
[13] Because of the temperature sensitivity of the probes,
negative temperature coefficient (NTC) thermistors are
installed at the same position of each probe and NTC data
are used to apply a temperature correction on measured soil
moisture data.
3. Regional Water Storage Effects on Gravity
[14] The residuals from the 20-year cycle are shown in
Figure 1b. We observe a very strong annual signal in the
B10403 VAN CAMP ET AL.: HYDROGEOLOGY AND GRAVITY IN MEMBACH
2 of 13
B10403