1 Introduction

Mass movements are geohazards that can directly influence human existence in mountainous terrains. The Indian Himalayan belt, which supports a wide variety of ecosystems, is prone to various types of natural disasters due to its geological conditions and typical climatic conditions. Landslides occur both in the Lesser and in the Higher Himalayas and also in the barren cold desert regions of Ladakh.

In areas prone to landslides, communication infrastructure often has to be repaired. Many repairs are made without getting any substantial benefit because of the lack of a full understanding of the geometry and hydrologic regime of the affected sites and the landslide processes (Bogoslovsky and Ogilvy 1977). In addition, the blind implementation of a traditional engineering repair schemes, for example recompaction, may not serve to mitigate adequately all types of future slope stability (McCann and Forster 1990). Scores of different methods (Pachauri and Pant 1992; Rautela and Lakhera 2000; Sarkar and Kanungo 2002; Saha et al. 2005) for assessing landslide hazards have been proposed. However, a common ground is lacking in the preparation of landslide hazard zonation (LHZ) maps by different specialists such as geo-environmentalists, engineers, policy-makers, or developers in hilly regions. Besides, these zonation maps in general do not consider the subsurface conditions or parameters, which play a major role in landslide processes.

Geophysical methods are non-invasive and in situ, and the results of many of them can be translated into relevant geotechnical information on the subsoil (Cosenza et al. 2006). Geophysical methods can provide highly resolved two-dimensional distributed data (Mauritsch et al. 2000). A particular geophysical method may not be suitable to study all types of mass movement processes. A careful selection of one method or a combination of several methods is necessary, by considering the local geological and structural setting and the type of mass movement. By applying appropriate geophysical interpretation, three-dimensional models of the investigated slope may be developed (McCann and Forster 1990; Mauritsch et al. 2000, Lebourg et al. 2003). Also, geophysical methods provide new means for the rapid investigation of vast areas at a relatively low cost (Bogoslovsky and Ogilvy 1977).

Electrical resistivity is sensitive to four parameters, viz., the water content, the mineralization of pore water, the cation exchange capacity of clay minerals, and temperature (Revil et al. 1998, 2002; Niwas et al. 2007; Jin et al. 2007; Shevnin et al. 2007; Hayley et al. 2007). Some of these parameters can play an important role in landslide processes, so this is the reason why ERT finds a natural application in landslide investigations. The role of micro-gravity in near-surface investigations is well recognized in the geophysical literature (Loj 2010; Rybakov et al. 2001).

Lebourg et al. (2003) studied deep-seated landslides in the French Alps through 2D and 3D geophysical imaging methods. Scores of authors (Erginal et al. 2009; Piegari et al. 2009; Havenith et al. 2000; Bichler et al. 2004; Lapenna et al. 2003, 2005; Colangelo et al. 2006; Friedel et al. 2006; Jomard et al. 2007; Sastry et al. 2006, 2007, 2008; Mondal et al. 2007, 2008; Park 1998) have utilized and discussed various geophysical methods in order to delineate the plane of failure, the hydrogeological regime, and to monitor the activity of the landslide. 3D and 4D ERT (Chambers et al. 2011; Wilkinson et al. 2010; Jongmans and Garambois 2007; Schmutz et al. 2009; Friedel et al. 2006; Lebourg et al. 2005) and fast 2D (Bichler et al. 2004; Perrone et al. 2004; Yang et al. 2004; Lapenna et al. 2003, 2005; Lebourg et al. 2005; Drahor et al. 2006; Sastry et al. 2006, 2007, 2008; Mondal et al. 2007) have also been used for landslide studies. The utility of ERT to investigate roto-translational slides and a translational earth flow in southern Italy are discussed by Lapenna et al. (2005). Israil and Pachauri (2003) have utilized a joint application of vertical electrical sounding (VES), seismic refraction, and the spectral analysis of surface waves (SASW) to characterize a landslide in Himalayan foothills region. Active landslide characterization using the 2D ERT method (Sastry et al. 2006; Mondal et al. 2007, 2008) and a combined application of 2D ERT and the gravity method have been discussed by Sastry et al. (2007, 2008).

Gravity surveys could reveal the subsurface slope morphology (Sastry et al. 2008), while lithological variations leading to bulk density variations are used for quantitative slope stability analysis (Del Gaudio et al. 2000). Del Gaudio et al. (2000) have also discussed the use of a micro-gravity survey on slump earth flows in Italy. Bláha et al. (1998) claimed that gravimetric surveys provided an effective contribution to the description of the structures and their dynamic control over time. Gravity measurements can detect local faults and subsurface failure surfaces both within the slide mass and in its vicinity, and they can contribute to a better characterization of active landslides (Sastry et al. 2007).

As per Jongmans and Garambois (2007), induced polarization methods have not been applied so far to landslide problems even though IP method can distinguish clayey zones from water-saturated zones, which exhibit almost the same resistivities. Recently, Zanetell (2011) has used both IP and seismic methods for landslide investigations. Here, we employ ERT, IPI, and gravity method in addressing the active landslide at the Salna sinking zone. The geophysical investigations included gravity, ERT, and IPI methods. Gravity has helped decipher several faults, which seem to control the sinking zone. ERT and IPI profiles infer the shallow geological structures and lithologies with fault signatures. Geotechnical soil classification has been carried out at two test sites, and slope stability studies were carried out along four slope profiles. These integrated studies have led a new understanding of the reasons for the sinking region at Salna, India.

2 Study Area and Physiography

The Himalayan mountain chains are seismically very active. Many small earthquakes occur almost every day along some of the neo-tectonically active fault zones (Rao et al. 2006; Bilham et al. 2001; Gaur et al. 1984; Kayal 2001; Khattri and Tyagi 1983; Khattri et al. 1989; Ni and Barazangi 1984; Sarkar et al. 1993). The tectonic activities due to earthquakes along numerous faults in the Himalaya have resulted in contemporary morphological adjustments, inducing a variety of mass-wasting processes (Saraf 2000). The Chamoli earthquake not only triggered many new landslides, but also reactivated old ones (Ravindran and Philip 1999). One reactivated mass movement site is the sinking zone developed near the village of Salna in the Chamoli district of Uttarakhand. A portion of almost 200 m of the Pokhri–Gopeshwar motor road has sunk 20 m vertically (Fig. 1a) since 1999 and a landslide buried several residential buildings and acres of cultivated land (Fig. 1b).

Fig. 1
figure 1

a Sinking of the Pokhri—Gopeshwar road. The arrow compensates the earlier elevation of the road to the present one. b A scenic view of the Salna village sinking zone showing the destruction of cultivated land due to sinking

Full size image

Salna village is a small hamlet of about 200 people in Pokhri tehsil of the Chamoli district of Uttarakhand, India. The Salna village sinking zone (30°23′44′′N, 79°12′13′′E) is located at the right side of Nagol gad (river) immediately to the northeast corner of the village Salna, at 11 km from Pokhri along Pokhri—Gopeshwar road (Fig. 2), a distance of about 220 km from the state capital Dehradun. This road serves as an alternate for National Highway No. 58 and connects two major tourist attraction sites, Badrinath and Kedarnath. Further, this road is the only means of communication for the inhabitants of the northwest Alkananda valley to access major towns like Gopeshwar (at 55 km) in the North, Rudraprayag (70 km) in the West, and Karnaprayag (40 km) to the South of the landslide site.

Fig. 2
figure 2

Map showing the location of the Salna village sinking zone, Uttarakhand, India. The Digital Elevation Model (DEM) of the study region is prepared using ASTER GDEM (ArcGIS 9.3), which has a ground resolution of 30 m. In the image, the height values are grouped according to their increasing values (Aster GDEM is freely downloadable from http://www.gdem.aster.ersdac.or.jp)

Full size image

A major part of this landslide slope is covered with cultivated land as terrace farming. The slope has a gradient of 50–60º in the vicinity of the landslide. The elevation difference between the toe (1,380 m above msl) and the crown (1,500 m above msl) of the landslide is around 120 m. The annual temperature varies from 5 °C in January to 35 °C in June or July. Nearly 70 % of the rainfall occurs during the period of June to September and the average annual rainfall varies from 1,200 to 1,500 mm. The relative humidity in the area is high during the monsoon season, generally exceeding 70 % on the average. The driest part of the year is the pre-monsoon period when the humidity drops to 35 % in the afternoon.

3 Geological Setting

The study area (Fig. 2) was marked by a complex geological setting (Fig. 3) with a large number of litho-units, which comprise of quartzite, phyllite, schists, gneisses, and metavolcanics of various genres (Kumar and Agarwal 1975). Regionally, the geological setup comprises the Garhwal group of rocks, which were sandwiched between the northern bound of Central Crystallines and the southern bound of Dudatoli group. Both boundaries were of tectonic origin and were thrust faults. The Main Central Thrust (MCT) marks the boundary between the Central Crystallines (Fig. 3) and the Garhwal group of rocks, while the North Almora Thrust (not shown on the present map) marks the boundary between the Garhwal group and the Dudatoli group. The present study site was a part of the Patroli formation of the Garhwal group (Kumar and Agarwal 1975), which was bounded by these two major regional thrust faults. This group was broadly divided into five formations, namely Rudraprayag, Lameri, Chamoli, Gwanagarh, and Patroli.

Fig. 3
figure 3

Geological map of the study area (after Kumar and Agarwal 1975). Notation RSG Ragsi Schist and Granite, PQ Patroli Quartzites, PBS Patroli Biotite Schist, PCP Patroli Chlorite Phylite, DOB P Dobri Phylite, BM Bhekuna Metavolcanics, NQ Nagthat Quartzite, DP Deothan Phylite, HQ Hariali Quartzite, SQS Sericite Quartz Schist, MAG Magnesite, CF Chamethi Fault, PF Pokhri Fault, KF Kande Fault, TG Tangi Anticline, MS Maithana Syncline, DS Dudhbhanga Syncline, and Ant/Syn Anticline/Syncline Axis

Full size image

The Garhwal group consists of quartzite, phyllite, slate, and limestone. Acidic and basic igneous rocks intruded the Garhwal group. The structures of the Garhwal group of rocks around the study area were complicated due to the effects of various tectonic episodes. There were at least three phases of tectonic activity, which are clearly visible in the present day regional structural patterns (Kumar and Agarwal 1975). The second generation NE-SW trending broad and open plunging folds and faults were superimposed on the earlier NW-SE doubly plunging folds and faults, such as the Maithana syncline. Toward the end of this period, a large number of NW-SE trending faults, such as the Kande and Chamethi faults were developed resulting in the mylonitisation of granite giving rise to sericite quartz schist. Due to the Main Central Thrust, the closer part of the southeasterly plunging Maithana syncline was terminated near Kalsir. The Chamethi and Pokhri faults are two parallel faults, which die out eastward and continue westward beyond Mohankhal. The other two major structures are the Tangi anticline and the Dudhbhanga syncline (Kumar and Agarwal 1975). The rocks exposed in the vicinity of the landslide site are biotite schist and sericite quartz schist of the Patroli formation of the Garhwal group. The landslide is located less than 300 m from the Kande Fault in the North–East and the Tangi Anticline in the North–West directions.

4 Geophysical Investigations

A SYSCAL Jr Switch—72 DC electrical resistivity imaging system of IRIS Instruments was used for ERT and IPI studies. This is a multi-node resistivity imaging system (www.iris-instruments.com) with an internal switching board for 72 electrodes and an internal 100 W power source. For sequence preparation, Electre II was used. The locations of the geophysical profiles are shown in Fig. 4. A total of eight combined ERT and IPI profiles along profiles AA′, BB′, CC′, DD′, EE′, FF′, GG′ & HH′ (Fig. 4) were undertaken during June 2006 using a Wenner–Schlumberger array configuration with different inter-electrode spacings. Relevant data acquisition details are included in Table 1.

Fig. 4
figure 4

Map showing the location of the ERT (AA′ to HH′), IP (AA′ to HH′) and gravity profiles (Profiles 1, 2, 3, 4 and 5). The landslide boundary is marked by the green dashed line

Full size image
Table 1 ERT and IPI data acquisition details
Full size table

The CG-5 Micro-gravimeter (M/S Scintrex make) has been employed for gravity data acquisition (Fig. 4) with a station interval of 10–30 m at 68 stations, in five different profiles (Gravity profiles 1, 2, 3, 4, and 5). Gravity profile 1 was conducted along the road, profiles 2, 3, and 4 were carried out in parallel to profile 1 in the central part of landslide, at the crown and near the toe regions, respectively. Profile 5 was conducted parallel to the landslide axis. The positioning of the electrodes of multi-electrode resistivity system and the gravity stations was completed with a Leica CS 5 GPS and a Topcon GTS 710 electronic total station.

For assessing the average density of hard rock, both laboratory density measurements of collected hard rock samples from landslide and adjacent areas and in situ overburden density estimation using a core cutter (Ranjan and Rao 2005) were carried out.

4.1 Geophysical Data Processing and Interpretation

The geophysical data sets include ERT, IPI, and micro-gravity. Interpretation of geophysical data is made through knowledge gained from extensive fieldwork done at the site and through the available literature. The processing and interpretation details are mentioned below.

4.1.1 ERT & IPI Data Processing and Interpretation

Both ERT and IPI data were processed using IRIS software and inverted respectively to 2-D true resistivity (RES2DINV software of Loke 2006) and chargeability sections using the Loke and Barker (1995a, b, 1996a, b) and Oldenburg and Li (1994) algorithms. Based on outcrops and the geoelectrical literature (Telford et al. 1990), resistivity–lithology and chargeability–lithology conversions (Tables 3, 4) were prepared, which served as the basis for the preparation of litho-sections along profiles AA′, BB′, CC′, DD′, EE′, FF′, GG′ & HH′ (Fig. 4), and also for their common color code presentation. The IPI sections were performed along with the ERT data along profiles AA′, BB′, CC′, DD′, EE′, FF′, GG′ & HH′ (Fig. 4). A common color code is developed and adopted for the presentation of true chargeability sections along all the profiles. The true chargeability variation ranged from 0.01 to 153 mV/V.

4.1.2 Shaliness Estimation

Here, we estimated the relative shaliness (shaleyness) of the subsurface materials from true chargeability data using the following formula similar to the one proposed in the well-logging literature (Serra 1984).

$$ {\text{SHP}} = \frac{{{\text{CHT}} - {\text{CHT}}_{\text{Min}} }}{{{\text{CHT}}_{\text{Max}} - {\text{CHT}}_{\text{Min}} }}\% $$
(1)

where SHP is the relative shaliness percentage, CHT the present true chargeability within the subsurface, CHTMin the minimum true chargeability, and CHTMax the maximum true chargeability.

Table 2 was adopted for the assigning the shaliness to the lithologies. The IPI (True Chargeability) section can be converted to a relative shaliness plot, which in turn can infer broad lithologies. Relative shaliness may not literally mean the presence of clay/shale. However, such plots can serve as additional constraints to ERT plots for lithological information on the subsurface. Also, these plots need to be analyzed with respect to the local geology of a site.

Table 2 Percentage of shaliness and related lithology
Full size table

4.1.3 Gravity Data Processing and Interpretation

Interpretation of ERT and IPI sections was first needed to be able to process the gravity data. By considering the resistivity images, a Bouguer datum was selected which led to the Bouguer anomaly map of the study region. The Bouguer gravity anomaly map was subjected to both qualitative and quantitative interpretation. The qualitative interpretation involved identification of fault signatures and their role on the structurally controlled landslides. Then, several gravity profiles were drawn on the Bouguer anomaly map across the strike of inferred faults, and they were subjected to quantitative interpretation for fault parameters.

4.1.4 Density Contrast Estimation

Laboratory density measurements of collected hard rock samples from landslide and adjacent areas were carried out to assess the average density of hard rock. The in situ overburden density was measured using a core cutter (Ranjan and Rao 2005). The lateral density contrast was estimated using the above density values.

5 Results and Discussion

As per the procedure outline earlier, ERT, IPI, and gravity data analysis results are considered here with relevant discussions in the relevant subsections. Some illustrative examples are considered.

5.1 Development of Resistivity–Lithology and Chargeability–Lithology Conversion Tables

As mentioned earlier, based on the outcrops and geoelectrical literature (Telford et al. 1990), resistivity–lithology and chargeability–lithology conversion tables (Tables 3, 4) were prepared. From the current experience of geophysicists the world over, by no means can one claim uniqueness in developing such conversion tables. However, special care was taken in designing these conversion tables placing due emphasis on the local geological conditions.

Table 3 Resistivity–lithology conversion
Full size table
Table 4 Chargeability–lithology conversion
Full size table

5.2 Electrical Resistivity (ERT) and Induced Polarization (IPI) Tomography Sections

Given the tectonic setting of the study region, a high-resolution geological picture of the subsurface is needed to meet the basic objectives of the study. At shallower depth ranges combined ERT and IPI studies along eight profiles (Fig. 4) were implemented, which provided detailed subsurface geology up to 150 m depth. Usually, the presence of moist clay in the subsurface is an important factor facilitating landslides. Jongmans and Garambois (2007) remarked that very little attention is paid to IPI in landslide investigations even though it can distinguish moist clays from water wet sands. So, our choice of combined ERT and IPI studies in landslide investigations is unique in inferring the presence of clay in the subsurface, which also helped in the indirect mapping of subsurface faults.

In our study region, the subsurface lithology includes wet sand and schist in different stages of weathering. Figure 4 along with Table 1 provides all the relevant details of eight different ERT and IPI profiles. Tables 4 and 5 respectively outline the adopted resistivity–lithology and chargeability–lithology conversions. For the sake of brevity, we report here only the results along three profiles, AA’, BB’, and GG’ (Fig. 4).

Table 5 Summary of fault analysis of Salna village sinking zone
Full size table

ERT, IPI, and shaliness percentage sections (Figs. 5, 6, 7) are correlated for the confirmation of various lithologies inferred from the resistivity sections. The IPI sections revealed the absence of clay horizons and the dominance of fresh schist/dry sand/wet sandy zones. The correlation between ERT, IPI, and relative shaliness percentage along three profiles are described below:

Fig. 5
figure 5

Display of a resistivity, b chargeability and c relative shaliness percentage along profile AA′ (Fig. 4)

Full size image
Fig. 6
figure 6

Display of a resistivity, b chargeability and c relative shaliness percentage along profile BB′ (Fig. 4)

Full size image
Fig. 7
figure 7

Display of a resistivity, b chargeability and c relative shaliness percentage along profile GG′ (Fig. 4)

Full size image
  1. (a)

    Figure 5a depicts the ERT section along profile AA′. This profile was conducted across the limits of the landslide (Fig. 4) to ascertain the depth of the slip surface. A varied lithology is observed throughout the ERT section up to a depth of 90 m, and no continuous slip surfaces was noticed until that depth. The fresh schist (>8,000 Ωm) is observed at a shallow depth of 10–15 m range on the right side of the ERT section at surface distances of 420–520 m where, in the central part of the profile, hard rock (fresh schist) is observed at much greater depth (80 m). A zone of low resistivity of 500–1,000 Ωm (highly weathered schist) overlain by a high resistivity zone of 1,000–3,000 Ωm (semi-weathered schist) was found 10–15 m below the surface, in between electrode numbers 35 and 43. A heterogeneity in the resistivity values up to a great depth produced perplexity about the internal structure of the landslide.

The IPI section along profile AA′ (Fig. 5b) shows the dominance of very low chargeability (<30 mV/V) material, presumably fresh schist/dry sand/wet sand. A relative higher chargeability zone is observed in between electrode numbers 23–35 at a depth of 80 m below the surface. The corresponding zone in the ERT section (Fig. 5a) shows a higher resistivity value. The high chargeability at such great depth may be due to the low resolving power of the IRIS equipment for IP data acquisition. As the relative shaliness percentage section (Fig. 5c) is derived from the true chargeability data, it provides similar information.

  1. (b)

    The ERT along BB′ (Fig. 6a) was conducted in the central portion of the landslide parallel to the landslide axis (Fig. 4). The profile is of limited length due to the presence of buildings in Salna village and a busy road with heavy traffic. Except for the presence of two zones of high resistivity (>3,000 Ωm), most of the ERT exhibits intermediate resistivity zones representing wet sand (150–500 Ωm) and highly weathered schist (500–1,000 Ωm). In the IPI section along profile BB′ (Fig. 6b), the dominance of very low chargeability (<30 mV/V) representing fresh schist/dry sand/wet sand is observed, followed by a relatively higher chargeability of 30–60 mV/V (highly weathered schist) and occasional patches of sandy clay material (70–130 mV/V). An increase in thickness of the low chargeability zone (<30 mV/V) is observed from crown to toe. This may be due to debris material deposited during the landslide process. The corresponding relative shaliness plot (Fig. 6c) also provides a similar result. Relatively higher chargeability zones (30–60 mV/V) with respect to corresponding higher resistivity zones (Fig. 6a) may be due to low-resistive highly weathered schist zones surrounding the high resistivity fresh schist regions.

  2. (c)

    The ERT section along profile GG′ was conducted below the road in the toe region in continuation of ERT BB′ (Fig. 4). The length of the profile was limited to 144 m due to the steeply sloping ground. The first half of the tomogram lying in the lateral distance range of 0–72 m shows a dominance of highly weathered schist (500–1,000 Ωm), with occasional pockets of wet sandy (150–500 Ωm) material. Zones of highly resistive semi-weathered schist (1,000–3,000 Ωm) material are noticed in a depth range of 5–10 m. Similarly, the second half of the ERT (72–144 m.) shows the presence of wet sand (150–500 Ωm) up to a depth of 8 m along with few tiny pockets of wet sandy clay (40–150 Ωm). The high resistive zones (>1,000 Ωm) at the surface level are due to dislodged boulders. Thus, the ERT section along profile GG′ (Fig. 7a) shows a distribution of three major resistivity zones, namely wet sand (150–500 Ωm), highly weathered schist (500–1,000 Ωm), and semi-weathered schist (1,000–3,000 Ωm). The highly weathered schist is present mostly in the first half of the section between the surface distances 0–72 m. The IPI section (Fig. 7b) exhibits a very low chargeability (<30 mV/V) which may be due to low or nil clay content within the highly weathered schist material. The respective relative shaliness plot (Fig. 7c) explains the presence of sandy material in the subsurface as wet sand and semi-weathered schist will have very low chargeability values.

5.3 Gravity Data Analysis and Delineation of Fine Structure of Hard Rock

ERT sections along profiles AA′–HH′ revealed the presence of hard rock at a deeper level. So, the chosen common elevation datum was 1,385 m, and, accordingly, a common Bouguer datum for all the profiles was fixed. Station-wise surface elevations and relative Bouguer anomaly variation for profiles 1, 2, 3, 4, and 5 are shown in Fig. 8. The relative Bouguer anomaly contour map is depicted in Fig. 9. The qualitative interpretation of the Bouguer anomaly map has indicated the presence of several faults. Eleven profiles (AA′–KK′) were drawn across the strike of those inferred faults (Fig. 8). Those profiles yielded the fault parameters of 57 inferred faults (F1–F57), which are shown in Fig. 9. Table 5 summarizes all the details of the inferred faults (F1–F57).

Fig. 8
figure 8

Plots showing gravity station elevation versus relative Bouguer gravity anomaly along profiles 1 to 5 (Fig. 4)

Full size image
Fig. 9
figure 9

Bouguer anomaly contour map with eleven profiles (AA′ to KK′) for interpretation

Full size image

The gravity-inferred faults were extended laterally by considering their up-thrown and down-thrown signatures. In such a process, enough care is taken to weigh the strike of the fault and its depth of occurrence. As a result, Fig. 11 illustrates the location of gravity-inferred faults and their lateral extensions. Further, as per Fig. 9, the inferred faults are oriented mainly in two directions, namely the E-W trending faults, parallel to the regional Kande fault (Fig. 3) and NNE-SSW trending faults parallel to the Tangi anticline axis (Fig. 3). In addition, the faults parallel to the Kande fault have been dissected by later formed ones parallel to the Tangi anticline axis, thereby dividing the entire region into several blocks (Fig. 11).

The role of initial micro-gravity observations along five profiles (Fig. 4) ended up in generating the Bouguer anomaly map (Fig. 9) of the study region. Given the specific geological features of the study region, the choice of eleven gravity profiles (AA′–KK′) for the study region (Fig. 10) seemed to be a better strategy in deciphering the fine structure of the basement. At the outset, the number of inferred faults (57No.) seemed to be very much on the high side, but a close inspection of Figs. 10 and 11 reveals that they are indeed part of 13 curvilinear faults in total, with 8 of them parallel to the Tangi anticline and the other 5 parallel to the Kande fault. In our study region, micro-gravity played an important role in explaining the possible source(s) for the sinking and related landslides. As gravity anomaly sources are deep and extensive, it is natural to correlate the gravity-inferred faults with relevant signatures in the ERT and IPI sections.

Fig. 10
figure 10

Gravity inferred fault parameters are shown at respective locations on the profiles AA′–KK′ (Fig. 9). The depths and relative positions of up-thrown and down-thrown blocks of different inferred faults (F1–F57) are included in Table 5

Full size image
Fig. 11
figure 11

Illustration showing gravity inferred faults and their lateral extensions in the Salna village sinking zone. Black and blue dashed lines indicate faults parallel to the regional Kande Fault and the Tangi anticline axis, respectively

Full size image

5.4 Fault Signatures on ERT Sections and Their Correlation with Gravity-Inferred Faults

The positions of ERT profiles AA′–HH′ (Fig. 4) are overlaid on the map of gravity-inferred faults and their lateral extensions (Fig. 11) for checking possible fault signatures on the respective ERT sections. In such an effort, we have looked for independent analysis of ERT sections (Fig. 12) with gravity-inferred faults serving as the starting point. For that, the position of the gravity derived faults was projected onto the corresponding ERT sections, and the resulting fault signatures on ERT profiles AA′–HH′ are verified. For illustration purposes, we include here the ERT sections AA′ (Fig. 13) and GG′ (Fig. 14), respectively. Figures 13 and 14 reveal that gravity-inferred faults are almost vertical to subvertical in nature. We use the fault numbering adopted in Figs. 10 and 11. The resistivity values in all the ERT sections are expressed on a logarithmic scale.

Fig. 12
figure 12

A combined plot showing gravity inferred fault locations and ERT profiles

Full size image
Fig. 13
figure 13

ERT section along profile AA′ (Fig. 4). Resistivity values are expressed using a logarithmic scale. Black dashed lines are the inferred faults along with arrows showing up-thrown and down-thrown blocks. Short horizontal segments in red color refer to the depth of up-thrown blocks derived from gravity studies

Full size image
Fig. 14
figure 14

ERT section along profile GG′ (Fig. 4). Resistivity values are expressed using a logarithmic scale. Black dashed lines are the inferred faults along with arrows showing up-thrown and down-thrown blocks. Short horizontal segments in red refer to the depth of up-thrown blocks derived from gravity studies

Full size image

(a) The ERT section along profile AA′ (Fig. 13) shows the presence of three faults, which coincide with the gravity-inferred faults, F14, F45, and F46 (Figs. 10, 11). Extension of the gravity-inferred fault F 14 is located in between electrode number 19 and 20. The left side of the fault is the down-thrown side and the right side is the up-thrown side. Similarly, the extended line of fault F 45 crosses the ERT section AA′ in between electrode number 32 and 33 at a depth of 1,365 m, and here, the left side of the fault is the up-thrown side and the right the down-thrown side. Fault F46 crosses the ERT in between electrode 43 and 44 at a depth of 1,371 m. The left side of the fault is up-thrown in relation to the right side.

(b) The gravity-inferred fault F19, which is located at depth of 1,374 m, just below the ERT section in between electrodes 22 and 23, has a clear signature on the ERT section GG′ (Fig. 14). The high resistive right block is not continuing on the left side of the inferred fault (Fig. 14). Further, the right block forms the up-thrown side of the fault F40. In view of the above, it is clear that gravity-inferred faults divide the hard rock at depth into a number of segments creating a blocky appearance. One can look for their signatures on the ERT sections. Accordingly, we have carried out an independent structural interpretation of the ERT/IPI section(s) using gravity analysis results in the background. At first sight, it seems that our interpretation approach is based heavily on gravity with ERT/IPI sections as support at shallow depth levels. However, we have shown that our analyses of both geophysical data sets are independent of each other. It is also true that not all gravity-inferred faults show up in the ERT sections. As the region happens to be a sinking zone, gravity played a key role in our investigations.

We presume that a relative movement of gravity-inferred blocks of hard rock, triggered by either neo-tectonic activity of region or monsoon rains, could serve as a source for landslides along with gradual sinking of the study region. Some of these faults could be traced in the ERT sections at shallow depth levels. Probably, the Chamoli earthquake in March 1999 could have triggered seismically induced ground movements in this region. The possible gravity-inferred faults that do not show up in ERT sections are like blind faults in the seismological literature and such sites could be the source(s) of future neo-tectonic activity including sinking and landslide triggering episodes.

6 Geotechnical Investigations

Limited geotechnical investigations were carried out in the field and in the laboratory. The results of geotechnical investigations are mentioned below.

6.1 Classification of Landslide Material

Soil/debris samples were collected from different locations of the landslide region (Fig. 15). Laboratory analysis for grain size distribution and Atterberg limits were carried out, and the soil material was classified using the Indian Standard Soil Classification System (ISSCS). The grain size distribution plots of samples are depicted in Fig. 16. The classification of soil samples is given in Table 6.

Fig. 15
figure 15

Map showing the location of slope cross sections (AA′–DD′) and soil samples (Salna 1 and Salna 2). The landslide boundary is demarcated by the green dashed line

Full size image
Fig. 16
figure 16

Plot of grain size distributions for soil/debris samples of the Salna village sinking zone

Full size image
Table 6 Classification of soil/debris sample
Full size table

6.2 Slope Stability Analysis

Detailed stability analyses were carried out for four representative slope cross sections (AA′–DD′ in Fig. 15) using the geological map and topographical data. The shear strength parameters (c and \( \varphi \)) were determined through direct shear tests of soil/debris samples and back analysis of critical slopes. The general slope condition is dry at the surface and moist at depth throughout the year. However, it becomes saturated during the monsoon. Considering the hydro-geological condition of the slope, the factor of safety (F) for all the sections was calculated both for dry and saturated conditions.

For illustration, we include a centrally located profile CC′ (Fig. 17), and the relevant stability analysis results in the form of output computer listing is included in section Appendix. The shear strength parameters c and \( \varphi \) for this slope profile are 2 and 16, respectively. The stability analysis is undertaken for rotational circular failure at depth. Table 7 outlines the minimum factor of safety for the slope cross sections (AA′–DD′) both for dry and for saturated conditions along with the coordinates of the critical slip surfaces.

Fig. 17
figure 17

Slope section CC′ (Fig. 15) for the geotechnical stability study

Full size image
Table 7 Factor of safety of slope cross sections in dry and saturated condition along with values for cohesion and angle of internal friction
Full size table

6.2.1 Factor of Safety for Sections AA′–DD′

For profile AA′, the slide material is composed of an almost equal mixture of gravel, sand, and plastic fines and belongs to silty sand class (SM). The factor of safety (F) value varies from 1.013 in dry conditions to 0.0197 in saturation conditions, inferring thereby that this part of the slope is in the critical stage for both conditions. For profile BB′, the slide material is similar to that of section AA′. The factors of safety (F) are 1.153 and 1.078 for dry and saturation conditions, respectively. This shows that this part of slope may fail during the rainy season.

For profile CC′, the slide material is composed of sand with gravel and non-plastic fines and belongs to the poorly graded sand (SP) class. The factors of safety (F) are 1.197 and 1.176, respectively, thereby making the slope stable under both dry and wet conditions. For Profile DD′, the slide material is similar to that of section CC′. The factors of safety (F) are 1.116 and 1.031 for dry and saturated conditions, respectively. The slope along this section may fail during the rainy season.

7 Summary and Conclusions

  1. (a)

    The Salna village sinking zone belongs to a completely different landslide mechanism, where the land subsidence is occurring along neo-tectonically activated faults, and this subsidence seems to be playing a crucial role in associated landslides periodically. For addressing the Salna sinking zone and associated landslide problem, we have carried out integrated geophysical and geotechnical studies. The geophysical studies include ERT, IPI, and gravity methods. We have carried out eight ERT and IPI profiles and five gravity profiles spanning the study region. Additionally, we have collected soil samples at two locations and carried out relevant soil classification studies in a geotechnical laboratory. We have also carried out both in situ and geotechnical laboratory density measurements, which were used both in gravity data processing and slope stability studies.

  2. (b)

    In view of the fact that gravity and tectonics go together, the more so for characterizing a sinking zone, our close micro-gravity data acquisition and its analysis have inferred a criss-cross network of faults in the subsurface. Unlike traditional landslide investigations, in our case study micro-gravity has played a key role in deciphering a fine subsurface structure at deeper depth level with respective fault signatures on ERT sections at shallow depths. Further, judging by the fault signatures on several ERT sections, it is inferred that these faults are vertical to subvertical in nature. The gravity-inferred faults seem to have divided the entire subsurface into several blocks, and the relative movement of these blocks is possibly leading to the sinking of the site.

  3. (c)

    At first sight, the faults deciphered from gravity in Fig. 10 seem to be on the higher side in comparison to the limited number of micro-gravity measurements carried out along five profiles (Fig. 4). We feel that the role of gravity measurements must end with the preparation of a reliable Bouguer gravity map (Fig. 8) and its analysis needs to be tackled separately. Given the complex geological situation (Fig. 3) of the study region, we have considered 11 gravity profiles (Fig. 9) on our Bouguer anomaly map and carried out their analysis. This yielded 57 fault signatures along these profiles arising from a total of 13 faults, which run parallel to the major structural elements of the study region (Fig. 3).

  4. (d)

    Although most of the surface geophysical methods are routinely being used for landslide investigations worldwide, our utilization of gravity along with ERT and IPI is novel. It has helped us to infer deep-seated subsurface fault networks. The absence of few gravity-inferred faults in shallow ERT sections may hint at blind faults, which could serve as future source(s) for geohazards in the study region. Thus, these integrated studies have yielded a better understanding of the mass-wasting mechanism for the study region.

  5. (e)

    The slope stability investigation was conducted for the debris material at the site as per geotechnical norms. The results of slope stability analyses corroborate well with the geophysical inferences.

We have carried out geotechnical investigations along with field investigations as per the standard practice of geotechnical engineering. Soil sample sites and location of slope cross sections are indicated in Fig. 15. The grain size distribution plots are included in Fig. 16. Factors of safety are evaluated all along four sections (Fig. 15) and with critical values of cohesion and angle of internal friction, φ through standard back analysis. For illustration sake, we included slope stability analysis of profile CC′ (Fig. 17) and computation details in section Appendix.