Geodetic measurement of tectonic deformation in the southern Alps and Provence, France, 1947–1994

https://doi.org/10.1016/S0012-821X(98)00065-XGet rights and content

Abstract

Active deformation at the boundary between the Eurasia and Africa plates varies in style. The belt between the Alpine mountain range and the Mediterranean Sea, for example, differs markedly in its western and eastern parts. In the western part, around southeast France, the mountains are higher, but the seismicity lower, than in the eastern part, around northern Italy and Greece. Yet the inter-plate convergence rate of 6 mm/yr varies by less than 15% between these two areas. To better understand the behaviour of this complex plate boundary, we use geodesy to map the spatial distribution of the deformation. In this paper, we focus on southeast France, a tectonic crossroads between three different domains (Alps, Ligurian Sea, and Massif Central) which exhibits a moderate level of seismicity. Here, the geodetic measurements imply low rates of horizontal deformation. By combining historical triangulation measurements mostly from 1947 to 1983 with Global Positioning System (GPS) surveys in 1993 and 1994, we estimate the rate of angular shear in triangular subnetworks covering the study area. The estimated strain rates in thirteen of nineteen triangles are smaller than their (1 standard deviation) uncertainties of about 0.1 microradian/yr. This value bounds the rate of deformation for a 100-km wide zone in Provence, between Marseilles to the south and the Ventoux massif to the north. The geodetic estimates place an upper bound of 1 to 2 mm/yr on the slip rates of two seismically active structures, the Durance fault and the Nı̂mes fault, assuming a fault zone ∼20 km wide in each case. We also find strain rates as high as 0.20±0.07 microradian/yr in three subnetworks near the epicentre of the magnitude 5.3 Haute-Ubaye earthquake in 1959, in a region which includes the higher summits. This may be interpreted either as pure shear with compression oriented NE–SW across this region or right-lateral simple shear along NNW–SSE-trending faults. Given that this earthquake is the only one of its magnitude recorded in the study area during the time interval spanned by the geodetic measurements, we infer that most of the geodetically observed deformation occurs aseismically. On the whole, the geodetic results suggest that the rate of north–south shortening across the 100-km wide study area is not more than 1 or 2 mm/yr, in agreement with kinematic models based on other types of geophysical data. Since this deformation represents only a small part of the convergence rate of 6 mm/yr predicted by the NUVEL-1 model for the entire boundary between the African and Eurasian plates, the remaining deformation must be accommodated elsewhere. This study illustrates how a careful analysis of historical geodetic data can measure the rate of tectonic deformation, even where it is low, and thus meaningfully bound geophysical models.

Introduction

The theory of global plate tectonics makes two assumptions which are only approximately correct. The first is that the lithosphere behaves as a set of rigid caps (plates) rotating on the surface of a sphere. The second is that their rate of relative rotation is constant on time scales of the order of a million years. Both these approximations break down in plate boundary regions, where episodic earthquakes on faults produce deformation which is neither rigid nor constant in rate. In these regions, the complexity of the faulting, topography, and seismicity suggests that a better mechanical understanding requires observations at length scales shorter than the characteristic size of a continent and at time scales shorter than the recurrence interval for large earthquakes. This poses the problem of how the crustal deformation is distributed in time and space. Precise geodetic measurements can help improve our understanding of the shortcomings of these approximations mapping the rate of deformation on the continents.

In approaching this problem, we note a serious trade-off between spatial and temporal resolution. We choose to favour the former at the expense of the latter in this paper. Increasing spatial resolution requires a geodetic network with closely spaced stations. To save time and expense, we choose to remeasure an existing triangulation network by GPS rather than install a new set of points and wait for them to move. These two choices lead to a map of deformation at the spatial scale of several tens of kilometres and a temporal scale of four decades. Our approach contrasts markedly with a network of continuously recording GPS receivers, which samples the deformation field, much more coarsely in space (several hundred kilometres between stations) but much more frequently (position estimates once per day). As a consequence, however, our approach inherits two (minor) drawbacks from the triangulation surveys. The existing networks are regional, not global, because triangulation requires visibility along the line of sight between stations, thereby degrading precision over long distances. Furthermore, triangulation measurements are not as precise as GPS, primarily because of optical refraction in the lowermost atmosphere. The major advantage, of course, is that our approach can estimate the rate of deformation without waiting any longer. In this paper, we use the fine-scale approach, favouring spatial resolution, to focus on a small, but active, part of a complex plate boundary.

In the western Mediterranean, Africa and Eurasia are colliding with a convergence rate of 6.2±0.5 mm/yr in the direction N17°W ± 9°, assuming two rigid plates and constant velocity over the last three million years 1, 2. A relative lack of seismicity, combined with space geodetic measurements, suggest that northern Europe is currently deforming quite slowly 3, 4. The central and eastern parts of the Mediterranean region, however, exhibit higher rates of deformation, in the presence of sustained seismicity, particularly in Greece 5, 6, 7, Turkey [8], and Italy [9]. The western part of the Mediterranean coast of Europe, on the other hand, appears to be less active. Here, the most pertinent geodetic observation involves the Satellite Laser Ranging station at Grasse (near Nice) in southeast France (Fig. 1Fig. 2). Its velocity is 3±2 mm/yr of southward displacement with respect to northern Europe [4] (Fig. 3). In the northern Alps, geodetic data have been interpreted in terms of some 4 mm/yr of horizontal shortening in the Jura range 10, 11 and 3–5 mm/yr of shortening between the Belledonne range and the subalpine chains [12]. Such displacement rates, when distributed over small geodetic networks, imply strain rates in excess of 2×10−7/yr. These rates are startling because they are as high as those observed around the active trace of the San Andreas fault system in California [13].

In this paper, we map crustal deformation in southeast France during this century. The study area lies at the crossroads between three different tectonic domains: (1) the high, active Alpine mountain range; (2) the relatively undeformable Massif Central; and (3) the relatively malleable offshore oceanic basin of the Ligurian Sea (Fig. 1). In the study area, the level of seismic activity is moderate (Fig. 2). The largest known and most recent earthquake is the Lambesc event of 1909 with intensity IX (on the MSK scale [14]), attributed to the Trévaresse thrust fault 15, 16. Several large (intensity VIII–IX) earthquakes also occurred near Vésubie, north of Nice, in 1564, 1618, and 1644 [16]. The largest earthquakes recorded instrumentally are the 1959 magnitude 5.3 Haute-Ubaye event and the 1963 magnitude 5.6 event in the Ligurian Sea 16, 17. The geodetic network spans the Provence domain (Fig. 3), which is characterized by strike-slip faults striking NE–SW (Durance, Nı̂mes, and Cévennes) and thrusts trending E–W (Ventoux, Lure, Lubéron, Trévaresse, and Costes). In the northeastern part of the study area, the Embrunais–Ubaye region includes folds and faults striking NNW–SSE (Durance–Serenne–Roburent). The eastern edge of the study area includes the outer folded parts of the Digne and Castellane nappes. Field observations and palaeostress determinations from fault slip data analysis suggest that the tectonic regime in the study area is characterized by north–south compression 18, 19, 20. In this paper, we examine the historical geodetic data set, the longest available instrumental geophysical time series, to measure the rate of tectonic deformation.

The objectives of this study are four in number. First, to validate a geodetic technique for measuring low rates of crustal deformation from a heterogeneous triangulation survey from the late 1940s. Second, to estimate the fraction of the inter-plate convergence accommodated in our study area. Third, to estimate the amount of deformation accommodated in the absence of earthquakes. Fourth, to test the kinematic consequences on short time scales of geophysical models intended to explain long-term tectonic deformation in the area.

Section snippets

Data selection

We analyze two types of geodetic data at 17 geodetic sites measured at least twice by triangulation or GPS (Table 1). The early surveys are first- and second-order triangulation observations performed by the French national survey agency, Institut Géographique National (IGN), primarily during campaigns in 1947–1952 and 1981–1983 with a few measurements between 1887 and 1931. The early first-order campaigns before 1947 used azimuth circles and reduced the eccentric observations to the primary

Estimation procedure

To estimate the strain rates, we use the forward modelling network deformation analysis software package to adjust the network 28, 29. For each station, we can estimate as many as four parameters: the two horizontal components of both the position and velocity vectors in a so-called `free network' solution. The velocity does not vary with time.

We pay particular attention to the occupation history of each benchmark. For example, we do not estimate a velocity for a station with only a single

Geodetic estimates of strain rate

To evaluate the robustness of our solutions to the estimation strategy, we have performed a suite of sensitivity tests on the following issues: (1) a priori values of the vertical component of station position; (2) correcting for the deflection of the vertical where possible; and (3) choice of fixed stations or a `free network' adjustment. None of these perturbations alters the estimated values of γ̇1 and γ̇2 by more than about a third of their stated uncertainty [22].

We have rejected one

Interpretation

The maximum shear strain rate γ̇ is significantly different from zero (with 68% confidence) in only six of nineteen subnetworks (Table 2). The uncertainty is typically about 0.1 μrad/yr, which corresponds to 1 mm/yr of shearing motion across a zone 10 km wide.

To visualize the shear strain rates, we plot them on a map in Fig. 3. We group the subnetworks into three categories according to the values of the maximum shear strain rate γ̇ and the azimuth θ of the compressive principal axis of the

Conclusions

We have shown that a rigorous analysis of historical data in a tightly spaced geodetic network can provide useful geophysical information, even in areas of relatively low rates of deformation. Although most of the geodetic triangles yield strain rate estimates which are not significantly different from zero, they tend to support the kinematic model proposed previously on the basis of other geological and geophysical observations 18, 19. Eight of nineteen geodetic subnetworks indicate

Acknowledgements

We wish to thank Danan Dong for generously sharing and supporting his Forward Modelling Network Deformation Analysis software. Georges Ferhat and Vincent Deschaux tirelessly carried batteries to mountain tops. In addition, we thank all the members of the `GPS-Alpes' working group [23] for collecting the 1993 GPS data under difficult conditions as well as the surveyors of the Institut Géographique National (IGN). We appreciate the conscientious efforts of IGN in recovering the historical data.

References (45)

  • R. Noomen et al.

    Crustal deformations in the Mediterranean area computed from SLR and GPS observations

    J. Geodyn.

    (1996)
  • C. DeMets et al.

    Current plate motions

    Geophys. J. Int.

    (1990)
  • C. DeMets et al.

    Effect of the recent revisions to the geomagnetic reversal time scale on estimates of current plate motions

    Geophys. Res. Lett.

    (1994)
  • D.E. Smith et al.

    Horizontal crustal motion in the central and eastern Mediterranean inferred from satellite laser ranging measurements

    Geophys. Res. Lett.

    (1994)
  • H. Billiris et al.

    Geodetic determination of tectonic deformation in central Greece from 1900 to 1988

    Nature

    (1991)
  • A. Rigo et al.

    Active normal faults in the Gulf of Corinth (Greece): microseismic and geodetic observations

    Ann. Geophys.

    (1994)
  • A. Rigo, Étude sismotectonique et géodésique du Golfe de Corinthe (Grèce), France, Thèse de Doctorat de 3ème cycle,...
  • M.B. Oral et al.

    Global Positioning System offers evidence of plate motions in Eastern Mediterranean

    Eos

    (1995)
  • E. Mantovani et al.

    Tectonic interpretation of large scale geodetic measurements (VLBI, SLR) in the Central Mediterranean region: constraints and uncertainties

    Ann. Geofis.

    (1995)
  • F. Jouanne et al.

    Present-day deformation of the French northwestern Alps/southern Jura mountains: comparison between historical triangulations

    Geophys. J. Int.

    (1994)
  • F. Jouanne et al.

    Present-day vertical displacements in the north-western Alps and southern Jura mountains: Data from leveling comparisons

    Tectonics

    (1995)
  • J. Martinod et al.

    Present-day deformation of the Dauphiné Alpine and Subalpine massifs (SE France)

    Geophys. J. Int.

    (1996)
  • K.L. Feigl et al.

    Space geodetic measurement of crustal deformation in central and southern California, 1984–1992

    J. Geophys. Res.

    (1993)
  • S. Medvedev, W. Sponheuer, W. Karnik, Seismic intensity scale version 1964, Iena, 1964, 48...
  • J. Vogt, B. Cadiot, J. Delaunay, G. Faury, J. Goguel, B. Massinon, D. Mayer-Rosa, C. Weber, Les Tremblements de Terre...
  • J. Lambert, Q. Levret-Albaret, Mille ans de séismes en France: catalogue d'épicentres, paramètres et références,...
  • G. Ménard, Structure et cinématique d'une chaı̂ne de collision — les Alpes occidentales et centrales, Thèse d'Etat,...
  • P. Combes, La tectonique récente de la Provence occidentale: microtectonique, caractéristiques dynamiques et...
  • J.F. Ritz, Evolution du champ de contrainte dans les Alpes du Sud depuis la fin de l'Oligocène, implications...
  • S. Rebaı̈ et al.

    Modern tectonic stress field in the Mediterranean region: evidence for variation in stress directions at different scales

    Geophys. J. Int.

    (1992)
  • M. Lejeune, Mémorial de la Nouvelle Triangulation Française, Institut Géographique National, Paris,...
  • G. Ferhat, Etude des déformations tectoniques actuelles des Alpes du Sud et la Provence à partir de la comparaison de...
  • Cited by (42)

    • Risk faults in stable crust of the eastern Bohemian Massif identified by integrating GNSS, levelling, geological, geomorphological and geophysical data

      2020, Tectonophysics
      Citation Excerpt :

      An implementation of Global Navigation Satellite System (GNSS) technology into geodesy during the last few decades has allowed researchers to detect changes in the land surface precisely (Bock and Melgar, 2016; Segall and Davis, 1997). The GNSS data used in geodynamic studies improved the knowledge of the crust (Ferhat et al., 1998; Grenerczy, 2002; Grenerczy et al., 2000; Hefty, 2007; Hefty et al., 2010; Kapłon et al., 2014; Kierulf, 2017; Mantovani et al., 1995; Pesci et al., 2010; Ziegler and Dèzes, 2007; Zubovich et al., 2010). A number of epoch or permanent GNSS networks were founded in tectonically active areas characterised by recent earthquakes.

    • Was the Trévaresse thrust the source of the 1909 Lambesc (Provence, France) earthquake? Historical and geomorphic evidence

      2001, Comptes Rendus de l'Academie de Sciences - Serie IIa: Sciences de la Terre et des Planetes
    View all citing articles on Scopus
    View full text