The Panagopoula Landslide
Research Studies by S. El Bedoui, Geoazur, UNS, Nice, France.

Wednesday 2 December 2009 by Thomas Lebourg

traduction [English ] [français ]

The monitoring system is based on the positioning of a targets network installed in the area as shown in Figure 1. The purpose is to acquire the position variations of the reflectors starting from the same reference stations (S1, S2), considered as fixed during the survey. We also use two referenced targets to improve the angular measurement (C1 and S3). The maximum distance measured by topographic survey is arising between 235 m and 880 m from S1.

1. Instrumentation: Strategy and Protocol

Equipment necessary for monitoring displacements is constituted by:
- a TDA 5005 (Leica) electronic distance meter, of very high accuracy (infra-red distance measurement),
- targets (Leica GMP104) materializing the surveyed points.

The measurement stations are located at the eastern side of the active zone. They consist of a fixed concrete pillar of 0.5 m of diameter and 1.50 m high. These stations allowed the repetition of measurements starting from the same positions throughout the survey. Targets are installed on both sides of the upper fault. The S3 target is located to the south of the faults mapped. Then, other targets are located on the northern part which is affected by recent landslides.

The aim of this positioning strategy is the measurement of a hypothetical tectonic component and the gravitational one. All targets are sealed on the ground, except for C1 and C7, directly fixed on metal structures where stability was preliminarily checked. The positioning measurement of each target is based on: (i) two measures of distance (horizontal and along the slope) between the electronic distance meter and the target, and (ii) horizontal and vertical angles. The measurement of the horizontal angle requires the installation of an angular reference to direct the circle for each measurement session. One target is located in a stable zone (in the southern foot of the slope) and relatively far away from the measuring site (respectively 800 and 900 m) allowing high accuracy of angular measurement.




Figure 1: Topometric Survey Network

During the measurement operations, each target is measured 20 times with a double reversal movement in order to correct the errors related to horizontal and vertical collimations due to the instrument. The instrumental accuracy, noted Δ, (starting from the manufacturer data) is calculated using the formula of Barrel and Sear (1939). Table 1 presents the theoretical accuracy calculated on the positioning. We then calculated the "real" accuracy (noted Δc, Table 2) by using the Barrel and Sear’s formula. These are based solely on the measures carried out on the field. It relates to the dispersion of the results and either to the data manufacturers: the measured precision is given by the standard deviation associated with the repetition measurements on the same point. Table 2 synthesizes the results obtained for each target, based on the series of measurements carried out in April (optimal conditions of measurements). The accuracy is certainly underestimated, in particular because the atmospheric conditions vary during the time period necessary for the measurement sequence.

Table 1: Theoretical Accuracy for each Target

 Table 2: Measured Accuracy for each Target (based on April 2006 session)

2. Results over a one year Survey

Figure 2 presents the results obtained on a one year chronicle (5 series of measurements). For each target, we represented the variations of position according to axes X (East) and Y (North). Taking into account the fact that the C1 target is the angular reference, it is not represented in the figure. Therefore its position is considered as constant. The weak variations of its position were corrected and assigned for all targets. Displacements according to the X axis (E-W direction) are negligible for all the targets (except C4 and S3), because the variations are included in the positioning accuracy according to this component.

With regard to target 4, considerable variations are highlighted. This target is placed on a topographic complexity, a secondary peak directed N040, which could explain the displacement variations. For the other targets, an annual trend was calculated by linear regression, showing a general displacement towards the west (included in the accuracy interval) corresponding to the slope axis (N050).

The S3 target shows more complex behaviour. Over the one year survey, displacement is clearly N200 oriented. S3 being located on a slope crest which is N090 oriented, topographic measurements seem to clearly indicate an anti-slope movement of about 13 mm y-1. However, the accuracy gap does not allow a clear identification of actual anti-slope movement or a virtual movement related to an instrumental error. Since this trend affects only S3 and does not appear for the other targets, it is strongly probable that this movement exists effectively.

 Figure 2: Results for East and North Components over a one year Survey

3. Discussion on Gravitational and Tectonic Components

3.1. Gravitational Movements

Two types of displacements are characterized:
- a) N040/050 trend affecting all targets ("long term"), with displacements about 10 mm y-1,
- b) seasonal effect (February and June) on targets 4, 6 and 8 ("short term").

The most important movements are recorded on targets 3, 6 and 7 (approximately 15 mm y-1). They correspond to sectors presenting structural complexities (tectonic and stratigraphic). The strong residual movements are recorded on the targets located in the purple limestone unit (C4 and C6), where the morphological indices were recognized as intense and recent. In all the cases, the periods when the residual displacements were recorded correspond to high rainfall event periods (V. Léonardi and P. Gavrilenko, 2004).

We interpreted the residual movements as the result of a slip surface in purple limestone. Their weathering degree (observed on the field and on outcrops) explains the restriction of the movement at this geological unit. It should be noted that target 3 is located on this unit, but any residual movement is recorded. This target is fixed on a deep concrete structure, so is not directly connected to the ground surface. The absence of movement on this target indicates low depth of the slip surface.

The residual movement ("short term" movement) is related to a superficial sliding movement. The slide surface is created by the stratigraphic contact between grey and purple calcareous and emerges at the surface along the normal fault (elevation: 200 m). The mobile mass is circular overall, with a thickness of approximately 30 m.

The quantification of the "long term" kinematic movement is limited by the survey period, but is estimated at around 10 mm y-1. It is observed in the structurally complex zones: intersection between the stratigraphic contacts, E-W and N150 faults. Target 3 presents one of the largest movements, indicating a deep movement (it is sealed on the entry of the drainage tunnel). This movement consists of a deformation at a large scale, and is also related to morphological signs such as tension cracks at the top of the slope and deformations observed close to the national road.

3.2. Tectonic Component

The instrumental survey of displacements shows non-gravitational movement records on the target S3. These observations are located above the highly faulted area mapped in the field, that we associated with the Psathopyrgos fault zone. Previous works (D. Latorre et al, 2004 ; N. Flotté et al, 2005 ; P. Bernard et al, 2006) described it like one individualized lineament, with a N100 50-60 N. The fault was geometrically characterized with a 15+/-2 km length and a 9+/-2 km depth by P. Bernard et al, 2006. It is connected at 6 km depth on a horizontal ramp (A. Rigo et al, 1996 ; D. Latorre et al, 2004 ; P. Bernard et al, 2006). Based on a GPS survey and seismological data, P. Bernard et al, 2006 proposed that the Psathopyrgos fault is connected to a part of this ramp subject to brittle creep with a high seismicity rate. In their study, they underlined a minimum slip rate of 15 mm y-1 showing a kinematic characterizing the end of seismic cycle. This displacement rate is close to the value that we measured for the target S3. Moreover, directions can be considered as globally linked considering the accuracy of measurement for target S3. It should indicate that S3 measurement is dominated by a tectonic component related to the Psathopyrgos fault zone.

Furthermore, the fact that other targets do not show this component indicates that the area could be considered in two blocks bounded by the upper normal fault. Its supposed activity is correlated with the thick and recent colluvium deposits at the south eastern part of the slope. Considering this boundary, the measurement station is embedded in the same block with other targets (except S3). The tectonic component is therefore only expressed by the relative positioning of S3 versus the measurement station. Other target positions do not allow the quantification of the tectonic component and movements recorded are associated with the gravitational component (assumed by the fact that movements and slope direction are equivalent).

The tectonic component is divided on the two slope units and Figure 3 shows resulting absolute displacements on the slope. However, considering the accuracy of S3 displacement, it is not currently possible to relate the supposed tectonic slip and the long term gravitational movement characterized here.

Research Studies by S. El Bedoui, Geoazur, UNS, Nice, France.

 Figure 3: Gravitational and Tectonic Components