Review Article
A Review of the Advantages and Limitations of
Geophysical Investigations in Landslide Studies
Veronica Pazzi
, Stefano Morelli
, and Riccardo Fanti
Department of Earth Sciences, University of Firenze, Via G. La Pira 4, 50121 Firenze, Italy Correspondence should be addressed to Veronica Pazzi; [email protected] Received 30 November 2018; Accepted 27 June 2019; Published 14 July 2019 Academic Editor: Pantelis Soupios
Copyright © 2019 Veronica Pazzi et al. This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited. Landslide deformations involve approximately all geological materials (natural rocks, soil, artificial fill, or combinations of these materials) and can occur and develop in a large variety of volumes and shapes. The characterization of the material inhomogeneities and their properties, the study of the deformation processes, and the delimitation of boundaries and potential slip surfaces are not simple goals. Since the ‘70s, the international community (mainly geophysicists and lower geologists and geological engineers) has begun to employ, together with other techniques, geophysical methods to characterize and monitor landslides. Both the associated advantages and limitations have been highlighted over the years, and some drawbacks are still open. This review is focused on works of the last twelve years (2007-2018), and the main goal is to analyse the geophysical community efforts toward overcoming the geophysical technique limitations highlighted in the 2007 geophysics and landslide review. To achieve this aim, contrary to previous reviews that analysed the advantages and limitations of each technique using a “technique approach,” the analysis was carried out using a “material landslide approach” on the basis of the more recent landslides classification.
1. Introduction
Large landslides and smaller-scale mass movements are natural widespread processes that result in the downward and outward movement of slope-forming materials, significantly sculpting the landscape and redistributing sediment and debris to gentler terrain. The rapid population growth and the pressure from human activities have strongly influenced their extension and occurrence so that they have become disasters causing vast direct and indirect socioeconomic consequences [1]. These deformations involve approximately all geological materials (natural rocks, soil, artificial fill, or combinations of these materials) and can occur and develop in a large variety of volumes and shapes [2]. Artificial fills are usually composed of excavated, transported, and placed soil or rock, but they can also contain demolition debris, ash, slag, and solid trash. The term rock refers to hard or firm bedrock that was intact and in place prior to slope movement. Soil, either residual or transported material, is used for unconsolidated particles or poorly cemented rock or aggregates. Soil is usually further distinguished on the basis of texture as debris (coarse fragments) or earth (fine fragments) according to the well-established Varnes Classification [3]. Following the
recent updating of [4], more reasonable use of geotechnical material terminology (clay, silt, sand, gravel, and boulders) is starting to spread, although some classical terminologies (mud, debris, earthflow, peat, and ice) are maintained after a recalibration of their definitions, because they have acquired a recognized status in landslide science by now. The Hungr classification includes aggregations of different materials that have been mixed by geomorphic processes such as weather-ing, mass wastweather-ing, glacier transport, explosive volcanism, or human activity. The use of geotechnical terminology is indeed most useful, as it relates best to the mechanical behaviour of the landslide as stated by [4] and even to most common investigation methods. In any case, the distinction between different materials is usually based on interpretation of the main geomorphic characteristics within landslide deposits but can also be inferred from the geological attributes of the involved parent material. The type of material is one of the most important factors influencing the movement of landslides, which can be categorized as falls, topples, spreads, slides, or flows according to their behaviour from the source area to the final deposit through distinctive kinematics [2, 3, 5]. Actually, the most common criterion used in landslides classification is based on the combination of the materials
with the type of movement, but it is possible to find many other classification criteria, including velocities, volumes, water content, geotechnical parameters, and processes related to the formation of the mobilized material, among others. This is because, as stated by [5], engineering geology literature on landslides is affected by inconsistent terminology and ambiguous definitions from older classifications and current key terms for both specialists and the public. Currently, the most widely accepted and used classification is that of [2], which enhances the previous system devised by D.J. Varnes [3, 6]. Since then, only small improvements for specific cate-gories have occurred, such as that for flow-like landslides by [5]. In 2014 Hungr et al. [4], by maintaining the consolidated concepts introduced by [2], redefined some basic elements (basically typology and material) that still refer to the original characterization of [3] and, consequently, updated the total amount of categories (from 29 to 32), along with revisiting some of their descriptions. This new landslides classification version (Table 1), which was proposed to simplify landslides studies, is increasingly circulating in the academic world, and for this reason, it is used as the reference in the present paper. Characterizing landslide material inhomogeneities and their properties, studying the deformation processes, and delimiting boundaries and potential slip surfaces are not sim-ple goals. They require the availability of a wide range of data, observations, and measurements (e.g., kinematic, geomor-phologic, geological, geotechnical, and petro-physical data [7]) and the evaluation of geologic and hydrologic conditions related to phenomena occurrences [8]. To obtain the needed information, many techniques including both traditional methods (detailed geomorphological surveys, geotechnical investigations, local instrumentation, and meteorological parameters analyses) and more recent methods (remote-sensing satellite data, aerial techniques, and synthetic aper-ture radar interferometry) can be employed [[9, 10] and references within]. Among the latter, geophysical techniques are also included, since they are very useful in detecting the petro-physical properties of the subsoil (e.g., seismic wave velocity, electrical resistivity, dielectric permittivity, and gravitational acceleration [7]). Even though linking geophysi-cal parameters and geologigeophysi-cal/geotechnigeophysi-cal properties should always be supported with direct information (e.g., data from drillings), geophysical methods can provide the layered structure of the soil and certain mechanical parameters [11]. Therefore, because almost all of the advantages of geophysical methods correspond to disadvantages of geotechnical tech-niques and vice versa, the two investigation techtech-niques can be considered complementary. Finally, the geophysical inversion data, and, therefore, the creation of a reliable subsoil model, is a complex and nonlinear problem that must be evaluated by taking into account all the available data on the site [11].
It is to be noted that the success of geophysical methods is mostly dependent on the presence of a significant and detectable contrast in the physical properties of different lithological units. However, in landslide characterization, geophysical contrast (i.e., differences in mechanical and phys-ical properties) cannot be associated only with a boundary in mechanical properties (i.e., landslide boundaries) and therefore be of interest relative to the slope stability. These
measured variations, in fact, could be local anomalies within the landslide or caused by the rough topography, and as a result, they could be of no or little interest [12]. This is why according to [11], the references for landslide investigation purposes are relatively few, and according to [13], there have been few landslides in which geophysical techniques were very useful. Nevertheless, the application of these techniques has changed over the years thanks to technological progress, the availability of cheaper computer electronic parts, and the development of more portable and faster equipment and new software for data processing [12], allowing the adequate investigation of 3D structures, which addresses one of the most ancient geophysical method limitations according to [11].
This review work, which starts from [11], is focused on the last twelve years of works (2007-2018) published in international journals and available online. The main goal was to analyse the geophysical community efforts in over-coming the geophysical technique limitations highlighted in the conclusion section of [11]. The drawbacks pointed out were as follows: (i) geophysicists have to make an effort in the presentation of their results; (ii) the resolution and penetration depth of each method are not systematically discussed in an understandable way; (iii) the geological interpretation of geophysical data should be more clearly and critically explained; (iv) the challenge for geophysicists is to convince geologists and engineers that 3D and 4D geophysi-cal imaging techniques can be valuable tools for investigating and monitoring landslides; and finally, (v) efforts should also be made towards achieving quantitative information from geophysics in terms of geotechnical parameters and hydrological properties. To reach the aim, contrary to the four geophysics and landslide reviews discussed in section number 2 [8, 11, 12, 14] that analysed the advantages and limitations of each technique using a “technique approach,” the analysis in this paper was carried out on the basis of a “material landslide approach” according to the recent landslide classification discussed above [4]. Finally, since it is beyond the aim of the work, we do not discuss the theoretical principles of the different geophysical techniques nor how to perform field surveys in this paper.
2. Geophysical Techniques and Landslides:
The State of the Art of Review Papers
One of the first papers related to the application of geophys-ical techniques for the investigation of landslides, defined as a pioneering work by [11], is [8]. Herein, “landslides” are defined as a sudden or gradual rupture of rocks and their movement downslope by the force of gravity. In this paper, the main advantages of applying geophysical methods are as follows: (a) the rapid investigation of vast areas, collecting a larger number of sample points than those acquired by geologic engineering techniques; (b) the determination of the mechanical properties of wet and dry soils based on the measurements of large rock volumes directly involved in the processes; (c) the measured parameters reflect the combined geological and hydrological characteristics, which sometimes cannot be identified separately; and (d) the measurements
Table 1: Nomenclature of the newly proposed landslide classification version according to [4] based on the Varnes classification system. Words divided by / (slash symbol) have to be used alternatively. In italic movement types that usually reach extremely rapid velocities as defined by [2], while for the others, the velocity varies between extremely slow to very rapid (for details, refer to [4]).
TYPE OF
MOVEMENT ROCK SOIL
Fall Rock/ice fall Boulder/debris/silt fall
Topple Rock block topple Gravel/sand/silt topple
Rock flexural topple
Slide
Rock rotational slide Clay/silt rotational slide
Rock planar slide Clay/silt planar slide
Rock wedge slide Gravel/sand/debris slide
Rock compound slide
Clay/silt compound slide Rock irregular slide
Spread Rock slope spread Sand/silt liquefaction spread
Sensitive clay spread
Flow Rock/ice avalanche
Sand/silt/debris dry flow Sand/silt/debris flowslide Sensitive clay flowslide Debris flow Mud flow Debris flood Debris avalanche Earthflow Peat flow Slope Deformation Mountain slope
deformation Soil slope deformation
Rock slope deformation Soil creep
Solifluction
can be repeated any number of times without disturbing the environment. Four main goals can be reached by applying vertical electric sounding (VES), seismic refraction (SR), self-potential (SP), and electromagnetic measurements (EM), listed as follows: (i) the investigation of the landslide geo-logic configuration, (ii) the investigation of the groundwater (determining the level and its fluctuation with time) as a land-slide formation factor, (iii) the study of the physical properties and status of the landslide deposits and their changes with time, and (iv) the investigation of the landslide displacement process. Reference [8] also showed how electrical resistivity values and seismic waves velocities decrease between the bedrock and the rocks in the landslide body. Finally, in the conclusion section of [8], microseismic noise (SN) analysis is mentioned as a valuable method by which to characterize the slope soil strata.
Reference [14] conducted a review of the geophysical methods employed in landslide investigations. They high-lighted that the selection of the method/s to be applied depends on its/their suitability for solving the problem. To estimate this adequacy, there are four main control factors: (i) the definition/understanding of the geophysical contrasts that have to be investigated, (ii) the evaluation of the charac-teristics (penetration depth and resolution) of the geophysical
methods, (iii) the calibration of the acquired data by means of geological/geotechnical data, and finally, (iv) the signal-to-noise ratio. In the paper, several case studies are shown wherein the SR was successfully employed to determine the lower landslide boundary.
Ten years later, the SR, seismic reflection (SRe), electrical resistivity (ER), SP, EM, and gravimetry were discussed by [12] as the most frequently used methods in landslide charac-terization. For each method, the author gives (i) the theoreti-cal principles, (ii) how to perform the measurements, (iii) the sources for those which are active techniques, and, finally, (iv) some expected results. Moreover, he presents some summary tables with the physical property ranges (e.g., those of the P-wave velocity, density, and electrical resistivity) of the most common soil and rock masses in their crude form (without taking into account variations caused by different clay con-tents, weathering, saturation, etc.). Finally, for each discussed method, [12] synthesizes in one table its suitability for use in landslide characterization, human artefact (like pipes and foundations) identification, and physical properties determi-nation for geotechnical purposes. Overall, the SP method results are not or only marginally suitable in all fields. Never-theless, in the same year, [15] and, later, [16–18] showed how the SP method could be helpfully employed. From the table
in [12], the seismic tomography and 2D and 3D geo-electric results correspond to the best methods for use in landslide characterization.
Reference [11] presents the state of the art of the geophys-ical techniques applied in landslide characterization based on papers after 1990. According to this review, the methods could be divided into seldom, widely, and increasingly used categories. Among the first methods they enumerate are SRe, ground penetrating radar (GPR), and gravimetry, while among the second group are SR, ER VES, or tomographies (ERT), and SP, and, finally, among the third group are SN, surface waves (SW), and EM. Moreover, they indicate seismic tomography (ST) as method useful only for limited site con-ditions (rock slides). They synthetize in a table (a) the main geophysical methods used, (b) the measured geophysical parameters and information type, (c) the geological context, (d) the landslide classification following [2], (e) the geomor-phology, and (f) the applications (targets). According to the review in [11], there are three main advantages and three main limitations in employing geophysics for the subsurface map-ping of landslides. As benefits of the geophysical methods, the author enumerates (i) the flexibility and the relative efficiency on slopes; (ii) the noninvasiveness and the generation of information on the internal structures of soil or rock masses; and (iii) the allowance of examining large volumes of soil. As drawbacks, he highlights that (i) the resolution, which is dependent on the signal-to-noise ratio, decreases with depth; (ii) the solution for a set of data is nonunique, and the results must be calibrated; and (iii) these methods yield indirect information on the subsoil, such as physical parameters rather than geological or geotechnical properties. One of the main conclusions of the review is that in landslide char-acterization, the geophysical survey design is still a much-debated question, and no unique strategy has arisen from the literature.
Reference [11] is the last review published in an interna-tional journal and available online that focused on the advan-tages and limitations of the geophysical methods applied in landslides characterization. Reference [19], in fact, discusses, by means of case studies, benefits and drawbacks of the most common geophysical techniques (GPR, ER, and SR) in geomorphological applications. Therefore, in this paper landslides are just one of the possible fields of application. Two more recent reviews about geophysics and landslides are [20, 21]. The first is focused only on the ERT tech-nique applied in landslide investigations and analyses the advantages and limitations of 2D-, 3D-, and 4D-ERT (or time-lapse ERT: tl-ERT) surveys based on papers of the period from 2000 to 2013. The second is a review of the current state of the art and the future prospects of the near surface geophysical characterization of areas prone to natural hazards (e.g., landslides, rockfalls, avalanches and rock glaciers, floods, sinkholes and subsidences, earthquakes, and volcanos) published in a book series (and, therefore, not freely available online for download), wherein the anal-ysis of the geophysical techniques applied in landslides characterization is limited to subsections of the case study section.
3. Geophysical Techniques and Landslides:
A
(Landslide Approach) Analysis
As mentioned in Introduction, this review work is based on a “material landslide approach” analysis on the basis of the more recent landslide classification presented by [4] and discussed in Introduction. Even though this classification is not widely employed (only 20% of the analysed papers from the years 2015-2018 adopted it, and these papers are marked with # in Tables 2 and 3), we decided to use it considering that the same landslide could assume different names from paper to paper, though the authors could be more or less the same. Among the analysed papers, examples are the Super Sauze landslide and the La Vallette landslides (marked in
Table 2 with (∘) and (∘∘), respectively) or the Randa landslide
(marked with (∘) in Table 3). This means that the analysed
works are clustered and discussed in two groups, “soil” and “rock,” respectively, on the basis of the material landslide type (columns 2 and 3 of Table 1).
Moreover, we decided to analyse the works starting from 2007 because the review in [20] is focused only on the ERT technique application; nevertheless, we do not analyse in detail all references already discussed therein, but we synthetize the results. The results of the review analysis are summarized in Tables 2 and 3, where for each work, we specify: (a) the landslide typology according the authors of the paper (i.e., how they refer to the landslide in the text) and (b) according to the classification from [4] (where possible, since sometimes it is not easy to identify the landslide classes from [4] on the basis of only the text); (c) the materials involved in the landslides; (d) which geophysical methods and (e) which other traditional techniques were employed; and (f)-(l) how many efforts were performed to overcome the five drawbacks highlighted by [11] and listed in Introduction. To quantify these efforts, a three-level scale was employed, where +, -, and n.d. mean, respectively, that many/some, insufficient, and nondiscussed efforts were made to overcome the limitations. Unfortunately, we know that the evaluation of how many efforts were performed could seem subjective. Therefore, in Table 4, for each drawback, we summarize how we evaluated the efforts.
3.1. “Soil” Landslides. “Soil” landslides, with respect to “rock”
landslides, are the typology most studied with geophysical techniques. Among the 120 analysed papers, more than half (e.g., 66 papers, which means 75 landslides analysed without considering those reported in [20]) were about “soil” landslides, and among them, more than half were on the flow type. As summarized in Table 5, in fact, no one was focused on falls, topples, or spreads, while 28 landslides (the 37.3%) were analysed focused on the slide (6 clay/silt rotational slides, 8 clay/silt planar slides, 11 rotational and planar slides, 1 debris slide, and 2 clay/silt compound slides), 41 (the 54.6%) on the flows (5 sensitive clay flowslides, 9 debris flows, 5 mud flows, and 22 earthflows), and 6 (the 8.1%) on the slope deformations (soil slope deformation). Only two of the analysed landslides were marine landslides [33, 35], indicating that it is not easy to conduct geophysical surveys to characterize landslides that dive into the sea. It is also
T able 2: This ta b le summa rizes th e anal ys ed scien tific pa p er s fr o m th e last tw el ve ye ar s (20 07 -2018) fo cus ed o n “s o il la n dslide”. The la ndslide typ o log y and ma ter ials ar e defined as in th e p ap er s them sel ves. M o re o ver , w her e p o ssib le , w e added the la ndslide cl assifica tio n acco rdin g to [4]. P ap er s m ar ke d w it h # al re ad y ado p ted th is cl assifica tio n. P ap er s ma rk ed wi th ( ∘) an d ( ∘∘ ) fo cus o n th e Su p er Sa uze and L a V allet te la ndslides, resp ec ti ve ly . Dra w b ac ks 1 to 5 ar e the limi ta tio n s o f the ge o p h ysical tec hniq ues ap p lied to la nd slide ch arac te riza tio n s p o in ted o u t by [11]. The y ar e, resp ec ti vel y, as fo llo w s: Dr aw ba ck 1: geo p h ysicists ha ve to ma ke an eff o rt in the p res en ta ti o n o f th eir resu lts; Dr aw ba ck 2: th e re so lut ion an d th e p en et ra ti on d ept h o f ea ch met h o d ar e n o t syst ema ticall y dis cuss ed in an u nder st an da b le wa y; Dr aw ba ck 3: the ge o logical in te rp re ta ti o n o f ge o p h ysical d at a sho uld b e m o re cle ar ly an d cr it icall y exp la ined; Dr aw ba ck 4: the ch allen ge fo r ge o p h ysicists is to co n vince geo logists an d en gineer s th at 3D an d 4 D geo p h ysical ima gin g tec hniq ues ca n b e val u ab le to o ls fo r in ve st iga tin g and mo ni to ri n g la ndslides; an d Dr aw ba ck 5: eff o rt s sho uld als o b e m ade to w ar ds o b ta inin g q ua n ti ta ti ve inf o rma ti o n fr o m ge o p h ysics in te rm s o f geo te chnical pa ra met er s an d h yd ro logical p ro p ert ie s. + ,-, an d n .d .m ea n, re sp ec ti ve ly ,t ha t m an y/s o me ,i n sufficien t, and no n-dis cuss ed eff o rt s w er e m ade to o ve rc o m e the limi ta ti o n s (s ee T ab le 4), w hile + ∗ me an s tha t the in ter p re ta ti o n is lin ke d to a n umer ical m o del o r la ndslide fe at ur e iden tifica tio n. VES: ve rt ical elec tr ical so undin g ,ER T :elec tr ical re sist iv it y to m ogra p h y, IP/S IP :ind uced po la ri za ti o n /s pectr al in d u ced po la ri za ti o n ,S P : se lf-p o ten ti al ,S PT :s elf-p o ten tia l to m og ra p h y, F/TD EM: fr eq uenc y/t ime do m ain elec tr o mag n et ism, VLF -EM: ver y lo w-f re q uenc y elec tr o m ag net is m, EM: elec tr o ma gnet ism, RMT :radio ma gnet o tell ur ic, AE: aco ust ic emissio n, SN :s eismic no is e, SR :s eismic ref rac tio n, SRe: seismic reflec tio n, SW :s ur face wa ves, D H: do w nho le ,CH: cr osshole, GPR: gr o u nd p enet ra ting rad ar , M G :micr o gra vi ty ,D T M: digi tal te rra in m o del ,S AR: syn th et ic ap er tur e rada r, GB-I nSAR: gr o und b as ed in te rf er o m et ric SAR ,G PS: glo bal p osi tio n sys te m, T S: to tal st at io n, TLS: te rr est rial las er sca nner s, CPT :co ne p enet ra tio n test, TD R: ti me do ma in reflec to m et ry . Ye ar R ef er en ce L andslide typ o log y (as d efine d by the au th o r/ s) L andslide typ o log y (acco rdin g [4]) Mat er ia l Geo p h ys ic al T ec hniq ue/s Oth er av aila b le da ta Dra w b ac k 1 Dra w b ac k 2 Dra w b ac k 3 Dra w b ac k 4 Dra w b ac k 5 2007 [22] ( ∘ ) m udslide ea rt hflo w cla y fo rma tio n SN (lo cal) GPS, ge o det ic an d st ra in in st ru men t, pi ez om et er -n.d . n.d . n.d . n.d . 2007 [23] ea rt hflo w (t h e o ldest mo vemen t), cl ay pl an ar sl id e (r ea ct iv at io n s) ea rt hflo w cl ay bod y o ve r mu d st o n e-sh al e ba se m en t ER T ,S R ,SW , SN (lo cal), D H SAR d at a, inc lino m et er s, st ra ti gr ap h y + + -n.d . n.d . 2007 [2 4] / / mu d st o n e, co ll u vi um, limesto n e and ca rb o n at e b re ccia SN (r egio nal) / -n.d . -n.d . + 2007 [2 5] ( ∘ ) in tra-ma te ri al m u dslide ea rt h fl o w cl ay for m at ion SR ,E R T ,S W , SN (l o ca l) b o re ho les + -+ n.d . + 2007 [26] ro ck an d d eb ri s flo ws ro ck av al anche and deb ris flo ws / SN (lo cal) / -n.d . + ∗ n.d . n.d . 2007 [27] ( ∘ ) (a) so ft-r o ck (m udslide o r cl ay ey flo w-lik e la ndslide) (b) tra n sla tio n al la ndslide (a) ea rt hflo w (b) cla y p la na r la ndslide cl ay fo rm at io n SN (lo cal), ER T (a) geo log y, ge om or ph ol o gy , ge o tec hnics and h ydr o log y (b) ge o mo rp holog y an d geo te chnics ++ -n .d . 20 08 [2 8] ea rt hflo w ea rt hflo w lo ess ER T ,S R ,S N (lo cal) DE M -n .d. + n .d . +
Ta b le 2: C o n ti n u ed . Ye ar R ef er en ce L andslide typ o log y (as d efine d by the au th o r/ s) L andslide typ o log y (acco rdin g [4]) Mat er ia l Geo p h ys ic al T ec hniq ue/s Oth er av aila b le da ta Dra w b ac k 1 Dra w b ac k 2 Dra w b ac k 3 Dra w b ac k 4 Dra w b ac k 5 2008 [29] / / mu d st o n e, co ll u vi um, limesto n e and ca rb o n at e b re ccia SN (lo cal) / -n .d . -n.d . + 20 0 9 [3 0] deb ris flo w deb ri s flo w met as edimen ts, gn eiss es, q ua rt zi tes SN (lo cal) m et eo ro logical da ta + n .d . +∗ n.d . n.d . 20 0 9 [3 1] ea rt hflo w ea rt hflo w fly sch ,c al car eo us Al ps SR e b o reho les, D EM -n .d . + n.d . n.d . 20 0 9 [3 2] ea rt hflo w ea rt hflo w co ar se co m p o n en ts in sil ty-c la ye y ma tr ix SR ,S Re b o re ho les -n.d . n.d . n.d . n.d . 2009 [3 3] / cl ay ro ta ti o n al and pl an ar sl id e / SR e / ---n .d n .d . 20 10 [3 4] sa n d y-cl ay tra n sla tio nal la ndslide cl ay p la n ar slide ma rl y limest o ne ER T pi ez om et er ,fi el d o bs er va tio n s, bo re h o le s -n.d . + n.d ./+ + 20 10 [3 5] tra n sl at io n al sl o pe la ndslide cl ay ro ta ti o n al slide , se n si ti ve cl ay flo wslide ma ri ne an d gl acio m ar ine cla y (F jo rd-del ta ic se dimen ts) seismic sub-b o tt o m profi le bo re h o le s, C P T , ge o tec hnical te st, tr enc h es, ba th ym et ric d at a + n.d . + n.d . n.d . 2010 [3 6] deep la ndslide cl ay /sil t ro ta tio nal slide ,m ud flo w gl acio lac ust rine cl ay s SN (l o ca l) inc lino m et er s, bo re h o le s, G P S + n.d . + +/n.d . n.d . 2010 [3 7] deb ris slide deb ris slide sc hist an d gneiss V LF -EM, VES b o reho les, T S -n.d . -n .d . n .d . 20 10 [3 8] co m p osi te m u lt ip le ea rt h slide – ea rt h flo w ea rth fl o w mu d st o n e, sa n d st o n e ER T GPS + + -n.d ./+ n.d . 2011 [3 9] deep-s ea te d la ndslide so il sl o p e def o rm at io n co ar se co ll u vi um ov er mu d st o n e SN (lo cal) / -n .d . + n.d . +
Ta b le 2: C o n ti n u ed . Ye ar R ef er en ce L andslide typ o log y (as d efine d by the au th o r/ s) L andslide typ o log y (acco rdin g [4]) Mat er ia l Geo p h ys ic al T ec hniq ue/s Oth er av aila b le da ta Dra w b ac k 1 Dra w b ac k 2 Dra w b ac k 3 Dra w b ac k 4 Dra w b ac k 5 2011 [4 0] / cla y co m p o und slide sa nd y, si lt y cla y SN (lo cal) met eo ro logical da ta , ext en so m et er s, SR , ro b o ti c T S -n.d . + n.d . -2012 [41] cl ay ey la ndslide ea rt hflo w b lac k ma rls SR ge ome tr ic al m o d el, DE M + -+ n.d . n.d . 20 12 [4 2] ma n y / b la ck ma rl ER T ,S R ,S W / -n .d. + n.d. + 20 12 [4 3] ( ∘∘ ) in tra-ma te ri al la ndslide rot at ion al an d p la n ar slide ,m ud flo w cl ay-shale dep osi ts ER T ,S R ,SW / -+ n.d . + 2012 [4 4] deb ris flo w deb ri s flo w / SN (r egio n al) / -n.d . + ∗ n.d ./-n.d . 20 12 [4 5] co m p lex re tro gre ss iv e ro t-t ra n sla tio nal slide rot at ion al an d pl an ar sl id e sc hist, fl ys ch ER T TD R ,w ea ther st at io n + n.d . -n.d ./+ -20 12 [4 6] ( ∘∘ ) flo w -lik e la n dslide rot at ion al an d p la n ar slide ,m ud flo w b la ck m ar ls 3D -SR b ore h o le s, SR ,E R T + n .d . + + /n .d . n .d . 20 12 [7] ( ∘) flo w-lik e la ndslides (m udslides) ea rt hflo w / SR ,E R T ,G P R , EM, M G m an y +++ n .d n .d . 20 12 [4 7] deep-s ea te d la ndslide so il sl o p e def o rm at io n /S N (r eg io n al )/ -n .d . +∗ n.d . n.d . 2013 [4 8] q uic k cla y la n dslide se n si ti ve cl ay flo w-lik e cl ay E R T ,I P re si st iv it y C P T , ge o tec hnical te st -n.d . + n.d n.d . 2013 [4 9] deep-s ea te d la ndslide so il sl o p e def o rm at io n cl ay -r ic h for m at ion an d col luv ia l dep osi ts SN (lo cal) / -+ n .d . n .d . 2013 [5 0] shallo w la n dslide pl an ar slide-e ar thflo w sil ty sa n d , ma rlsto ne, sa n d st o n e ER T TRD ,t en sio m et er -+ n.d ./+ +
Ta b le 2: C o n ti n u ed . Ye ar R ef er en ce L andslide typ o log y (as d efine d by the au th o r/ s) L andslide typ o log y (acco rdin g [4]) Mat er ia l Geo p h ys ic al T ec hniq ue/s Oth er av aila b le da ta Dra w b ac k 1 Dra w b ac k 2 Dra w b ac k 3 Dra w b ac k 4 Dra w b ac k 5 2013 [5 1] ea rt h flo w (q uic k cl ay la ndslide) se n si ti ve cl ay flo wslide q uic k cla y 2D an d 3D SR , SR e, SN (lo cal), 2D an d 3D ER T ,EM, GP R ,RMT , MG bo re h o le s, C P T , L iD A R ,g eo tec hnical te st -n.d . + -/n.d . n.d . 2013 [5 2] ea rt h flo w (q uic k cl ay la ndslide) se n si ti ve cl ay flo wslide q uic k cla y SRe bo re h o le s, C P T , L iD A R ,g eo tec hnical te st ,E R T ,I P -n.d . + -/n.d . n.d . 2013 [5 3] / cl ay /sil t ro ta tio nal slide gr av el, san d, an d pl ast ic and no n-pl ast ic fi nes fr om qu ar tz it e, ph yl li te, sl ate, and limest o n e ER T ,IP ,M G geo te chnical an al ysis -n .d . -n.d . n.d . 2013 [5 4] sur ficia lfa il ur es al ong a pl an ar so il/r o ck in ter face cl ay /sil t p la na r slide cl ay ey sa nd, si lt y sa nd, sa n dsto ne ER T bo re h o le s, ge o tec hnical an al ys is , met eo ro logical st at io n, TD R . pi ez om et er -n.d . -n.d ./-+ 2013 [5 5] m udd y la n dslide deb ris flo w ,r o ck fall (a) b lac k ma rls, (b) fly sc h and cl ay -s ha le SN (l o ca l) / -n .d . +∗ n.d . n.d . 2013 [9] m u dslide an d rot at ion al /p la n ar slide ( ∘∘ ) pl an ar sl id e bl ac k m ar ls ,fl ys ch SR Li D A R ,T L S, fi el d in ve st iga tio n s, G PS, DE M + n.d . + n.d . n.d . 2013 [5 6] deep-s ea te d la ndslide so il sl o p e def o rm at io n / SN (lo cal) L iD A R -n.d . + ∗ n.d . n.d . 20 14 [5 7] tr ans it ion al sl id e cl ay pl an ar sl id e cl ay sa ndy calca reni tic successio n 2D an d 3D ER T ,S R ,S N (lo cal), SW b o re ho les + n.d . + +/n.d . n .d . 201 4 [58] / / / SN (lo cal) / -n.d . n.d . n.d . + 20 14 [5 9] co m p osi te ro ta tio nal la ndslide (co m p lex la ndslide) cl ay ro ta ti o n al la ndslide ma rls, cha lks ER T ,S R field in ve st iga tio n, aer ial or th oph o to gr ap hs , Li D A R ,bo re h o le s, DE M + n.d . + n.d . n.d .
Ta b le 2: C o n ti n u ed . Ye ar R ef er en ce L andslide typ o log y (as d efine d by the au th o r/ s) L andslide typ o log y (acco rdin g [4]) Mat er ia l Geo p h ys ic al T ec hniq ue/s Oth er av aila b le da ta Dra w b ac k 1 Dra w b ac k 2 Dra w b ac k 3 Dra w b ac k 4 Dra w b ac k 5 20 14 [60] co m p lex, co m p osi te , successi ve ea rt h-slide ea rt h flo w cl ay /sil t ro ta tio nal slide ,e ar thflo w mu d st o n e E R T bo re h o le s, Li D A R , aer ia lp ho tos, D EM, GPS, inc lino m et er -n.d . + +/n.d . -20 14 [20] m an y / m an y ER T (ER T te chniq ue re vi ew p ap er ) ma n y + -+/+ n.d . 20 14 [6 1] deep-s ea te d gra vi ta tio nal def o rm at io n disma n tl ed in to slides and flo ws cl ay /sil t ro ta tio nal an d p la na r slide , ea rt hflo w ,cr eep bo ul d er s, gr av el an d co ar se sa nd o ver arg il lite s an d limest o n e SR e, E R T inc lino m et er s, bo re h o le s, SR + n.d . -n.d . n.d . 20 14 [6 2] (a) and (b) rot at ion al -tra n sla tio nal la ndslide (a) and (b) cla y rot at ion al /p la n ar slide (a) co ll u vi um, (b) co ll u vi u m and am ph ib ol ite ER T inc lino m et er ,3 D disp lacemen t me asur es, pi ez om et ri c w at er le ve l, so il tem p era tur e + + + n.d ./+ n.d . 2015 [6 3] (#) d eb ri s fl o w deb ris flo w gl acia lt il l, o u twash se dimen ts SN (re gi on al ) Li D A R ,G IS , nu m er ic al mo dellin g +n .d . +∗ n.d . n.d . 2015 [6 4] deb ris flo w deb ri s flo w SN (r egio nal) / -n.d . + ∗ n.d . n.d . 2015 [6 5] ea rt h-flo w ea rt hflo w sa ndsto ne and limest o n e ER T GPS + n.d . n.d . + n.d . 2015 [6 6] ea rt hflo w ea rt hflo w ma rl, cl aysto ne, m u dst o ne ,a ll u vi um, limest o n e ER T ,S P, / -n.d . + n.d . n.d . 201 6 [67] (#) sh allo w cla y p la n ar slide cl ay pl an ar sl id e p la st icb lu em ar ls , limest o n e, b lue cl ay s ER T GPS, aer ia lp ho to , b o re holes, au ger and p enet ra tio n, ge o tec hnical an al ysi s, ra infall an d gr o u n d w at er lev el da ta + -+ n.d . n.d . 20 16 [68] ( ∘ ) cla ye y la n dslide ea rt hflo w b lac k ma rls E R T / + + + n.d ./+
-Ta b le 2: C o n ti n u ed . Ye ar R ef er en ce L andslide typ o log y (as d efine d by the au th o r/ s) L andslide typ o log y (acco rdin g [4]) Mat er ia l Geo p h ys ic al T ec hniq ue/s Oth er av aila b le da ta Dra w b ac k 1 Dra w b ac k 2 Dra w b ac k 3 Dra w b ac k 4 Dra w b ac k 5 20 16 [6 9] shallo w cr eep in g la ndslide so il sl o p e def o rm at io n cl ay ey si lt ,m il d cl ay an d gr avel ly san d, co ve re d by grass pea t an d tu rf . (b edr o ck :m udst o n e, sa n d st o n e) ER T ,GPR b o reho les -n .d . -n.d . n.d . 2016 [7 0] (#) d eb ri s fl o w deb ris flo w sa ndsto n e, cl ay sto n e, si lt st on e, vol can ic ro ck s, Q u at er n ar y se dimen ts ER T ,EM, SP , MG ga mma-ra y, so il R ado n, b o re ho les, GPS -n.d . -n.d . n.d . 20 16 [7 1] tra n sla tio nal la ndslide cl ay /sil t p la na r slide si lt y cl ay w it h gra ve ls o ve r sa ndsto ne and mu d st o n e ER T bo re h o le s, inc lino m et er s, pi ez om et er s, pluv io me te r, osmo meter + n.d . + n.d ./-n.d . 2016 [7 2] q uic k-c la y la ndslide se n si ti ve cl ay flo wslide q uic k cla y FD EM, ER T , SR ,S W L iD A R ,g eo tec hnical te st ,r es ist iv it y-C P T , bo re h o le s + n.d . + n.d . n.d . 2016 [7 3] lo ess la n dslide cl ay /sil t ro ta tio nal slide sa ndy and cl ay ey lo ess ER T / -+ n.d . + 201 6 [7 4] (#) mu lt ip le ea rt h slide-e ar th flo w ea rth fl o w mu d st o n e, sa n d st o n e SR 3D-ER T , ge o tec hnical an al ysis +-+ n .d . -2016 [7 5] co m p lex la n dslide cl ay co m p o u nd slide cl ay s, sa nds, gra vels SW ,S N (lo cal), VES b o re ho le + -+ n .d . n .d . 20 17 [1 0] (a), (b) co m p lex rot o -t ra ns la ti on al la ndslide (a) cla y/sil t rot at ion al an d pl an ar sl id e (b) cla y/sil t rot at ion al an d pl an ar sl id e, ea rt hflo w limest o n e SN (lo cal) bo re h o le s, inc lino m et er s, + n.d . -+/n.d . n.d . 201 7 [7 6] deb ris flo w d eb ri s fl o w sc hists, do lo b reccias, qu ar tz ite s SN (l o ca l) fl ow st ag e se n so rs -n .d . +∗ n.d . n.d . 20 18 [7 7] Sac kun g-li ke mo vemen t so il sl o p e def o rm at io n sa n d st o n e, co ll u vi um Qua te rn ar y de posi ts ER T ,S R ,S N (lo cal) bo re h o le s, ge o tec hnical an al ys is ,DE M , re m o te im ag es ,G P S + n.d . + n.d . n.d .
Ta b le 2: C o n ti n u ed . Ye ar R ef er en ce L andslide typ o log y (as d efine d by the au th o r/ s) L andslide typ o log y (acco rdin g [4]) Mat er ia l Geo p h ys ic al T ec hniq ue/s Oth er av aila b le da ta Dra w b ac k 1 Dra w b ac k 2 Dra w b ac k 3 Dra w b ac k 4 Dra w b ac k 5 2018 [7 8] flo w-lik e la ndslide rot at ion al an d p la n ar slide ,m ud flo w sa ndsto ne, limesto ne IP ,S IP Li D A R ,G P S, bo re h o le s ge o tec hnical an al ys is ,E R T ,S R + -+ n.d . n.d . 20 18 [7 9] rot o -t ra ns la ti on al slide ,e ar th slide an d flo ws cl ay /sil t ro ta tio nal an d p la na r slide , ea rt hflo w fl ys ch, met amo rp hic ro ck s SN (lo cal), ER T D T M, aer ial ima ges -n.d . + n.d . n.d . 20 18 [80] / ea rth flo w (?) sa ndsto n e, sha le, co al sh ale ,m ar l, cla y, sil t ER T bo re h o le s, ge o tec hnical an aly si s, SP T ,fi el d in ve st iga tio n s + -+ n.d ./+ + 201 8 [8 1] (a) m udflo w ,(b) deb ris flo w (a) m ud flo w ,(b) deb ris flo w / SN (lo cal), A E ul tras o nic ga ug e, video ca m eras +n .d . +∗ n.d ./-n.d . 2018 [8 2] lo ess la n dslide cl ay /sil t ro ta tio nal slide cl aye y an d san dy lo ess E R T p re ss u re p o re -n .d . + n .d . n .d . 201 8 [8 3] (#) (a), (b) cla y-r ic h deb ris slide (c la ye y la ndslide) (a), (b) ea rt hflo w cl ay -r ich m at ri x, ma rls and limesto n e SN (l o ca l) / + n .d . +∗ n.d . n.d .
T able 3: This ta b le summa rizes th e anal ys ed scien tific pa p er s fr o m th e last tw el ve ye ar s (20 07 -2018) fo cus ed o n “r o ck la n dslide”. L andslide typ o log y and ma ter ials ar e defined as in th e p ap er s th em se lv es. M o reo ver ,w h er e p ossib le ,w e added the la ndslide cl assifica tio n acco rdin g to [4]. P ap er s m ar ke d w it h # alr ead y ado p ted this cl ass ifica tio n. P ap er s ma rk ed wi th ( ∘ )f o cu s o n th e R an da la ndslides. Dra w bac k s 1 to 5 ar e th e limi ta ti o n s o f the ge o p h ysical tec hniq ues ap p lied to la ndslide ch arac te ri za ti o n s p o in ted o u t by [11 ]. The y ar e, resp ec ti vel y, as fo llo w s Dr aw ba ck 1: geo p h ysicists ha ve to ma ke an eff o rt in the p res en ta ti o n o f th eir resu lts; Dr aw ba ck 2: the re so lu tio n and p enet ra tio n d ep th o f ea ch met h o d ar e n o t syst ema ticall y dis cuss ed in an under st anda b le wa y; Dr aw ba ck 3: the ge o logical in te rp re ta ti o n o f ge o p h ysical d at a sho uld b e m o re cle ar ly an d cr it icall y exp la ined; Dr aw ba ck 4:t h e ch al le n gef o r ge o p h ys ic is tsi st oc o n vi n ce ge o logists an d en gineer s th at 3D an d 4D geo p h ysical ima gin g tec hniq ues ca n b e val u ab le to o ls fo r in vest iga tin g and mo ni to rin g la ndslides; an d Dr aw ba ck 5:e ff o rt ss h o u lda ls ob em ad e to wa rd s o b ta inin g q ua n ti ta ti ve inf o rma ti o n fr o m ge o p h ysics in te rm s o f geo te chnical pa ra met er s an d h yd ro logical p ro p er ti es. +, -, and n.d .me an ,r esp ec ti vel y, th at ma n y/s o m e, in sufficien t, an d n o n-dis cu ss ed eff o rts w er e made to o ver co me th e limi ta tio n s (s ee T ab le 4), w hile + ∗ me an s tha t the in ter p ret at io n is lin ked to a n umer ical mo del o r la n dslide fe at ur e iden tifica tio n. VES: ve rt ical elec tr ical so undin g ,ER T :elec tr ical re sist iv it y to m ogra p h y, IP/S IP :i nd uced p o la ri za ti o n /sp ec tral ind uced p o la ri za ti o n ,S P :s elf-p o ten tial ,S PT :s elf-p o te n tial to m ogra p h y, F/TD EM: fr eq uenc y/t ime do ma in elec tr o m ag net ism, VLF -EM: ver y lo w-f req uenc y elec tr o ma gnet ism, EM: elec tr o ma gnet ism, RMT :r adio ma gnet o tell ur ic, AE: aco ust ic emissio n, SN :s eismic no is e, SR : seismic ref rac tio n, SRe: seismic reflec tio n, SW : sur face wa ves, D H: do w n h o le , CH: cr ossho le , GPR: gr o und p enet ra tin g rada r, M G: micr o gra vi ty , D T M: digi tal te rra in m o d el , SAR: syn thet ic ap er tur e rada r, GB-I nSAR: gr o und b as ed in ter fer o met ric SAR ,GPS: glob al p osi tio n syst em, T S: to ta ls ta tio n, TLS: ter rest ri al las er sca n ner s, C PT :c o n e p enet ra ti o n te st, T D R :t ime do ma in reflec to m et ry . Ye ar R ef er en ce L andslide typ o log y (as defined by th e au th o r/ s) L andslide typo log y (acco rdin g [4]) Mat er ia l Geo p h ys ic al T ec hniq ue/s Oth er av aila b le da ta Dra w b ac k 1 Dra w b ac k 2 Dra w b ac k 3 Dra w b ac k 4 Dra w b ac k 5 2007 [84] Ro ckfall/ d eb ri s flo ws ro ck fa ll ,d ebr is flo w ba sal tic SN (local) / -n .d . + ∗ n.d . n.d . 20 07 [8 5] ro ck fall ro ck fall limest o n e G PR b o re ho les, minin g -+ n.d . n.d . 2007 [26] ro ck an d d eb ri s flo ws ro ck av al anche an d d eb ri s fl o w s / SN (lo cal) / -n.d . + ∗ n.d . n.d . 2007 [86] ( ∘ ) roc k fall roc k fall heter o gene o us gn eiss es SN (lo cal), 3D-S R T bo re h o le s, G P R , ge o det ic, ge o tec hnical , met eo ro logical mo ni to ri ng system + n.d . -+/n.d . n.d . 20 0 8 [8 7] ro ck fa ll ro ck fa ll li m es ton e E R T ,G P R TLS, photo gr am m et ry +-+∗ n.d . n.d . 20 08 [88] ro ck fa ll an d ro ck -f al l av al an ch es ro ck fa ll an d ro ck av al anche limest o n e, am ph ib ol ite, gr an ite SN (r egio nal) / -n.d . + ∗ n.d . n.d . 20 08 [8 9] ro ck fall ro ck fall la va do me SN (r egio nal) / -n.d . + ∗ n.d n.d . 2008 [90] rot at ion al slidin g ro ckslide (a n y sp ecific typ o log y emer ge d) ma rl an d schist ai rb or n e E M ,E R T , ge o p h ysical logs, SP bo re h o le s, ge o tec hnical te st, hy d ro p hy si ca ll o gs , DE M ,g am m a ra y sp ec tr o m et er , inc lino m et er s --+ n .d /-n .d .
Ta b le 3: C o n ti n u ed . Ye ar R ef er en ce L andslide typ o log y (as defined by th e au th o r/ s) L andslide typo log y (acco rdin g [4]) Mat er ia l Geo p h ys ic al T ec hniq ue/s Oth er av aila b le da ta Dra w b ac k 1 Dra w b ac k 2 Dra w b ac k 3 Dra w b ac k 4 Dra w b ac k 5 20 08 [9 1] ro ck fa ll (a rt ificial) ro ck fal l / SN (lo cal) videos, p ho to s + n.d . + ∗ n.d . -2008 [9 2] ( ∘ ) ro ckslide ro ck w edg e slide heter o gene o us gn eiss es SR ,G P R , b o re ho les, field ma p p in g +-+ -/ n .d . n .d . 20 0 9 [9 3] ro ck fa ll ro ck fa ll met allif er o u s limest o n e SN (lo cal) T S, TLS + + -∗ n.d . n.d . 2009 [9 4] ro ck -fall roc k fall ch alk SN (local) + n.d . -∗ n.d . n.d . 20 10 [95] ro ck fa ll ro ck fa ll o thog n eiss es SN (lo ca l) ther m o m et er -n .d . -∗ n.d . n.d . 2010 [9 6] ro ck slide ro ck b lo ck toppl ing heter o gene o us gn eiss es SN (l o ca l) G P S + n .d . +∗ n.d . n.d . 2010 [97] ro ck slide ro ck com p ou n d slide met a-g ra no dio ri te an d tw o -mica gneiss SN (l o ca l) GPS, met eo ro logical da ta -n.d . + n.d . n.d . 2010 [98] ro ck slice co ll aps e ro ck rot at ion al slide limest o n e SN (lo cal) ext en so m et er -n.d . + ∗ n.d . n.d . 2010 [9 9] ro ckslide ro ck fall micas chists SN (lo cal) met eo ro logical st at io n, GPS +n .d . +∗ n.d . n.d . 20 10 [1 00 ] roc k fall roc k fall limesto n e and ma rly limest o n e SN (lo ca l), SR ext en so m et er + n.d . +∗ n.d . n.d . 20 10 [1 01 ] ro ck -i ce av al an ch e ro ck /i ce av al an ch e plut on ic ro ck s SN (re gi on al ) D E M -n .d . + ∗ n.d . n.d . 20 10 [1 02 ] deep-s ea te d wi th se ve ral deb ris flo w ro ck sl op e def o rm at io n fl ysh an d eva p o ri tes SR ,S R e / + n .d + n .d . n .d . 2011 [10 3] ro ckslide ma n y ma n y SN (r egio nal) / -n.d . + ∗ n.d . n.d . 20 11 [1 0 4 ] roc kfall roc k fall vo lca n o s SN (local) / -n.d . +∗ n.d . n.d . 2011 [105] ro ckslide ro ck fall micas chists SN (lo cal) GPS, b o reho les + + +∗ n.d . n.d . 2011 [1 0 6 ] ro ckfall ro ck fall limest o ne SN (lo cal), SR -n .d . + ∗ n.d n.d .
Ta b le 3: C o n ti n u ed . Ye ar R ef er en ce L andslide typ o log y (as defined by th e au th o r/ s) L andslide typo log y (acco rdin g [4]) Mat er ia l Geo p h ys ic al T ec hniq ue/s Oth er av aila b le da ta Dra w b ac k 1 Dra w b ac k 2 Dra w b ac k 3 Dra w b ac k 4 Dra w b ac k 5 2011 [1 07] ro ck slo p e ro ck b lo ck toppl ing pa ra gn ei ss an d sc hists SN (l o ca l) re gi on al ea rt hq u ak es, fib re o p tic st ra in se n so rs +n .d . +∗ n.d . n.d . 2011 [1 08] ro ckslide (l ong -r u n out ) ro ck pl an ar sl id e fly sch (cl ay -sto ne/m u dsto ne se q u ence) ER T radio ca rb o n an d d en d ro ge om or ph o -logical an al ysi s, ge om or ph ic ma p p ing ,GPS, ge o tec hnical an al ysis + n.d . + n.d . n.d . 2011 [1 0 9] deep-s ea te d gra vi ta tio nal slo p e def o rm at io n mo un ta in slo p e def o rm at io n fly sch E R T field sur ve y, k inema ti c ana lysis, tr enc h es + -+ n.d . n.d . 2012 [11 0] ro ck fa ll (p re co ndi tio n for ) ro ck fall gra n it ic gneiss AE met eo ro logical da ta -+∗ n.d . n.d . 2012 [111] / ro ck fa ll /ro ck bl o ck toppl e (? ) o rt h ogneiss SN (lo cal) / + n.d . +∗ n.d . n.d . 20 12 [112 ] ro ck -fall roc k fall gn ei ss an d ga b b ro SN (local) th erm o m et er s, GPS -n.d . +∗ n.d ./-n.d . 2012 [113] ro ck fa ll an d la te ral sp re adin g ro ck fa ll limest o n e, cl ay fo rm at io n SN (lo cal) + n.d . + n.d . n.d . 2012 [11 4] ro ck fall in th e so ur ce ar ea o f a m u dslide ro ck fa ll b lac k m ar ls SN (r eg io na l) ext en so m et er s, DE M +n .d . +∗ n.d . n.d . 2012 [115] ro ckfall ro ck fall fly sc h ,marls and limest o n e SN (l o ca l) met eo ro logical an d h ydr o logical da ta , lo cal seismici ty +n .d . +∗ n.d ./-n.d .
Ta b le 3: C o n ti n u ed . Ye ar R ef er en ce L andslide typ o log y (as defined by th e au th o r/ s) L andslide typo log y (acco rdin g [4]) Mat er ia l Geo p h ys ic al T ec hniq ue/s Oth er av aila b le da ta Dra w b ac k 1 Dra w b ac k 2 Dra w b ac k 3 Dra w b ac k 4 Dra w b ac k 5 2013 [11 6] ro ck slide-deb ri s flo w ro ck ir re gu la r slide ,d eb ri s fl o w rh yo daci te b reccias, tu ff SN (r egio nal) / -n.d . + ∗ n.d . -2013 [11 7] (a), (b), an d (d) toppl ing /b as al slidin g ,(c) ro ck co m p o u nd slide (a), (b), an d (d) ro ck b lo ck toppl e, (c ) ro ck co m p o u nd slide (a ) ar gill it es (b) and (d) limest o n e (c) sha le-sa n dsto ne se ri es SN (l o ca l) (a) SR ,(c) ST met eo ro logical st at io n (a) d isp lacemen t me asur es, (b) an d (c) ext en so met er s, -n.d . + n.d . n.d . 2013 [11 8] ma n y: ro ck fa ll, deb ris av ala n ch e, slides ma n y: ro ck fa ll, deb ris avala n ch e (a n y sp ecific so il or ro ck emer ge d) ma n y (no t sp ecified) SN (r egio nal/ca tc hmen t) met eo ro logical da ta , sa te ll it e im ag es , aer ia lp ho tos +n .d . +∗ n.d . n.d . 2013 [11 9] pl an an d toppl ing fai lu re s / ma in ly cl ay st o n e an d si lt ston e VES D EM -n.d . -n.d . n.d . 2013 [5 5] m udd y la n dslide deb ris flo w ,r o ck fall (a) b lac k ma rls, (b) fly sc h and cl ay -s ha le SN (l o ca l) / -n .d . +∗ n.d . n.d . 20 14 [120] roc kfall / lim es to n e SN (local) Li D A R , p h o togra mmet ry , video ca m eras, ext en so m et er , til tm et er +n .d . +∗ n.d . n.d . 20 14 [121 ] roc kfall roc k fall vo lca n o s SN (local) / -n.d . +∗ n.d . n.d . 20 14 [122] deep-s ea te d la ndslide mo un ta in slo p e def o rm at io n fl ys ch, sa ndst o n e E R T field sur ve y + n .d . + n.d . n.d . 2015 [12 3] ro ckslide / ar gilla ceo u s ma ter ia ls in te rcala ted wi th m u dsto ne /l im esto ne ER T ,VES, SR -n.d . + n.d . n.d . 2015 [12 4 ] (#) til ti n g an d ro ck fall ro ck sl op e sp re ad do lo mi te ER T TLS, GPS + n.d . + +/n.d n.d .
Ta b le 3: C o n ti n u ed . Ye ar R ef er en ce L andslide typ o log y (as defined by th e au th o r/ s) L andslide typo log y (acco rdin g [4]) Mat er ia l Geo p h ys ic al T ec hniq ue/s Oth er av aila b le da ta Dra w b ac k 1 Dra w b ac k 2 Dra w b ac k 3 Dra w b ac k 4 Dra w b ac k 5 20 16 [12 5] roc kfall roc k fall gra ni te CH, SR ge o tec hnical an al ys is , crac kmet er s, te m p er at u re p ro b es , inc lino m et er s, SN -n.d . + n.d . + 20 16 [126] (#) 21 la ndslides: 12 ro ck fa ll s, 8 ro ck slides and 1 ro ck av al an ch e 21 la ndslides: 12 ro ck fa ll s, 8 ro ck slides (a n y sp ecific typ o log y emer ge d) an d 1 ro ck ma n y SN (lo ca l) / -n .d. +∗ n.d . n.d . 2016 [12 7] ro ckslide pl an ar ro ck sl id e, ro ck fa ll ,d ebr is av al an ch e do lo mi te SN (r egio nal) / -n.d . + ∗ n.d ./-n.d . 20 17 [128] roc kfall roc k fall lim es to n e SN (local) T LS + + +∗ n.d . n.d . 20 17 [129] roc kfall roc k fall lim es to n e SN (local) G PS, p h o to ca m era + n .d . +∗ n.d . n.d . 201 7 [13 0] ro ck slide ro ckslide (a n y sp ecific typ o log y emer ge d) p ara gneiss SW ,S N / -n .d . + n.d . n.d . 201 7 [13 1] ro ck fall ro ck fall b lac k m ar ls SN (lo cal) / -n .d . +∗ n.d . n.d . 201 7 [13 2] deep-s ea te d la ndslide mo un ta in slo p e def o rm at io n se dimen ta ry ro ck s an d fl ys ch ER T ,GPR ,S R ,M G fi eld sur ve y, GPS + -+ n.d . n.d . 2018 [13 3] ro ck fal l ro ck fal l limest o n e SN (lo cal) we at he r st at ion , DE M -n .d . +∗ n.d . n.d . 201 8 [13 4] ro ck slide ro ck com p ou n d slide ,r o ck fall , deb ris avala n ch e am ph ib ol it ic and au ge n gn ei ss SN (l o ca l) ext en so m et er s, met eo ro logical st at io n, GPS +n .d . +∗ n.d . n.d . 2018 [13 5] ro ck fal l ro ck fal l limest o n e SN (lo cal) ext en so m et er , met eo ro logical da ta + n.d . -n.d ./-n.d .
Table 4: For each drawback, this table explains how the three-level scale (+, -, and n.d., which mean that many/some, insufficient, and non-discussed efforts were made to overcome the limitations) was applied.
+ - n.d.
Drawback 1
(i) Coloured figures (ii) 3D figures (iii) Figures with
interpretations
(i) B&W figures (ii) Non-interpreted
figures (iii) Figures too small
(iv) Only raw data
/
Drawback 2
There is wide discussion about the technique/s penetration depth and/or
resolution
There are only some mentions of the technique/s
penetration depth and/or resolution
There are no mentions of the technique/s penetration
depth and/or resolution
Drawback 3
There is wide discussion about the geological interpretation of the geophysical data
There are only some mentions of the geological
interpretation of the geophysical data
There are no mentions of the geological interpretation of the
geophysical data Drawback 4 3D/4D data are presentedand discussed
3D/4D data are presented but they are not discussed
in depth
No 3D/4D data are presented or discussed
Drawback 5
There is wide discussion on how to link geophysical
data with geotechnical and/or hydrological
properties
There are only some mentions of how to link
geophysical data with geotechnical and/or hydrological properties
There are no mentions of how to link geophysical
data with geotechnical and/or hydrological
properties Table 5: For each type of movement and “soil” landslide typology, the table summarizes how many papers are focused on it. In italic movement types that usually reach extremely rapid velocities as defined by [2], while for the others, the velocity varies between extremely slow to very rapid (for details, refer to [4]).
TYPE OF MOVEMENT Number of papers SOIL Number of papers
Fall / Boulder/debris/silt fall /
Topple / Gravel/sand/silt topple /
Slide 28
Clay/silt rotational slide 6
11 Clay/silt planar slide 8
Gravel/sand/debris slide 1
Clay/silt compound slide 2
Spread / Sand/silt liquefaction spread /
Sensitive clay spread /
Flow 41
Sand/silt/debris dry flow /
Sand/silt/debris flowslide /
Sensitive clay flowslide 5
Debris flow 9 Mud flow 5 Debris flood / Debris avalanche / Earthflow 22 Peat flow / Slope Deformation 6
Soil slope deformation 6
Soil creep /
Solifluction /
important to point out that in our analysis, we do not consider papers focused on the geophysical characterization of quick-clay that could evolve into a sensitive quick-clay flowslide but only papers focused on those that already occurred [35, 51, 52, 72]. In only 8 works (12.1% of the analysed “soil” landslide works), it is possible to find a detailed discussion of the theory
applied to landslides, concerning either how to formulate the inversion problem [41, 46, 52, 55, 68, 83] or how to combine data from different surveys [7, 42]. All the other papers deal with the discussion of a case study.
A detailed analysis of the applied techniques is discussed in Section 4. Below, we present only the main considerations
from some papers. ERT is an active geophysical method that can provide both 2D and 3D images of the subsoil. A wide review of this technique applied to landslides is provided in [20]. Therefore, here, we limit discussion to saying that in most papers (29 of 33 that present ERT applications, i.e., 88.0%), 2D ERTs are shown, while only in 6.0% (2 papers of 33), 3D ERTs are shown, and in the remaining 6.0% (2 papers of 33), both 3D and 2D applications are presented.
Since the ‘60s, passive seismic techniques have been developed to monitor and characterize signals triggered by landslide dynamics and related changes in the material mechanical properties (i.e., (i) material bending, shearing, or compression; (ii) fissure opening; (iii) slipping at the bedrock interface; and (iv) debris flows or mudslides) [22, 55]. They are of great interest in (a) detecting debris flows [30], (b) assessing site effects [24, 29], (c) detecting landslide slip sur-faces [10], and (d) estimating the thickness of a material that could be mobilized by a landslide [136]. Another advantage of this method is its ability to detect remote events that might otherwise go unnoticed for weeks or months. The main difficulties arise from two issues: (i) the seismic signatures of landslides and mud/debris flows are very complex and cannot be effectively identified without a detailed waveform analysis and (ii) the epicentres of landslides and mud/debris flows cannot be confidently determined by conventional earthquake-locating methods, mainly due to the lack of clear arrivals of P and S phases [44].
3.2. “Rock” Landslides. Among the 120 analysed papers, less
than half (e.g., 54) were about “rock” landslides, and the majority discussed were of the rock fall type. As summarized in Table 6 the landslide typology is divided as follows: 41 (the 54.6%) falls, 5 (the 6.7%) topples (5 block topples), 18 (the 24.0%) slides (1 rotational, 2 planar, 1 wedge, 3 compound, 1 irregular), 1 (the 1.3%) spread (rock slope spread), 6 (the 8.0%) flows (avalanches), and 4 (the 5.4%) slope deformations (3 mountain slope deformations and 1 rock slope deformation). In all the works that discuss the application of seismic techniques [26, 55, 84, 86, 87, 89, 91, 93–101, 103–107, 111– 118, 120, 121, 126–128, 130, 131, 133, 134], it is possible to find a more- or less-detailed discussion on the theory of the seismic wave analysis carried out to find the “rock” landslide features. “Rock” landslides are well-known phenomena but are poorly understood. Contrary to other landslide types, rock-falls are usually sudden phenomena with few apparent pre-cursory patterns observed prior to the collapse. A key point in the prediction of rock slope failure is better knowledge of the internal structure (e.g., the persistence of joints), which requires an interdisciplinary research field among rock mechanics, rock engineering, and mining [98]. This is why in 64.8% of the analysed papers, the geophysical technique is carried out along with more traditional methods (i.e., boreholes, mining, extensometers, and inclinometers). Moreover, there are at least two limitations in applying geophysical methods for rock deposits: (a) the difficulty of deploying sensors (i.e., ER electrodes, geophones, or GPR antennas) on sharp and blocky ground with a high void ratio and (b) the low geophysical contrast between the rock deposit and the underlying layers with comparable properties [[137],
not listed in Table 3 because it was already analysed by [20]]. In [137], there is another limitation in applying geophysics for rock deposits: the presence of a shallow geophysical contrast caused by the subsoil water table that could mask deeper interfaces. Nevertheless, this limitation also has to be considered for “soil” landslides.
More recently, to overcome these limitations, rock slope stability characterization and monitoring has been carried out using passive seismic techniques (see also the discussion session), implemented initially in open-mine monitoring [98]. These techniques, in fact, could help in (i) understand-ing the seismic responses of rock to slope deformation (e.g., the release of stored elastic energy under particular condi-tions) [135, 138], (ii) detecting and locating microearthquakes generated by fracturing within unstable rock masses (major effort is required for classifying seismic signals and extracting those related to landslides [86, 99, 129]), and (iii) identifying remote events that could otherwise go unnoticed for weeks or months. Therefore, these methods are applied to avalanches [26, 84, 101, 126], rock topplings [107, 111, 117, 134], rockslides [55, 96–99, 103, 116, 126, 127, 130], and rock falls or cliff failures [86, 88, 89, 91, 93–95, 100, 104–106, 112–115, 118, 120, 121, 126, 128, 131, 133]. Finally, some works are focused on finding the relation among “rock” landslides, displacement rate mea-surements, and meteorological (i.e., rain and temperature) parameters [95, 99, 100].
4. Discussion
Most studies focused on geophysical surveys are applied (a) to explore the subsoil for mineral deposits or fossil fuels, (b) to find underground water supplies, (c) for engi-neering purposes, and (d) for archaeological investigations [19]. Technological progress and the availability of cheaper computer electronic parts has allowed the improvement of more portable equipment and the development of 2D and 3D geophysical techniques [11, 12]. Therefore, the applicability of geophysical methods in landslide characterization has grown over the years. Starting from the state of the art of the geophysical techniques applied in landslide characterizations pointed out in [12], this review focused on the papers from the last twelve (2007-2018) years and tried to understand how many efforts have been made by the international scientific community to overcome the drawbacks. These geophysical techniques limitations are listed in Introduction. To reach the goal of this paper, contrary to the four reviews discussed in Section 2 [8, 11, 12, 14], the analyses of the geophysical method advantages and limitation were carried out on the basis of the latest landslide classification, which is mainly based on the involved materials and geotechnical properties [4]. Therefore, the 120 analysed papers were divided into two classes: “soil” (in red in the following figures) and “rock” (in green in the following figures), which account for 66 and 54 works, respectively.
Even though it is well known that it is better to integrate more than one geophysical technique because of the intrinsic limitations of each approach, in 68.3% of the analysed papers (Figure 1), only one geophysical method is presented and discussed. However, in 64.6% of these works (which
Table 6: For each movement type and “rock” landslide typology, the table summarizes how many papers are focused on it. In italic movement types that usually reach extremely rapid velocities as defined by [2], while for the others, the velocity varies between extremely slow to very rapid (for details, refer to [4]).
TYPE OF MOVEMENT Number of papers ROCK Number of papers
Fall 41 Rock/ice fall 40
Topple 5 Rock block topple 5
Rock flexural topple /
Slide 18
Rock rotational slide 1
Rock planar slide 2
Rock wedge slide 1
Rock compound slide 3
Rock irregular slide 1
Spread 1 Rock slope spread 1
Flow 6 Rock/ice avalanche 6
Slope Deformation 4 Mountain slope deformation 3
Rock slope deformation 1
correspond to 44.1% of the total analysed papers, as indicated by the bottom/darker part of the blue bar in Figure 1), the geophysical results are interpreted on the basis of other techniques. This means that only in 24.2% of the analysed works (the top/lighter part of the blue bar in Figure 1) is just one technique presented, and in 80% of these 24.2% (which means four works out of five), the employed method is a passive seismic technique. This is probably because these techniques (a) require quite light equipment, (b) can be employed to both monitor and characterize seismic signals triggered by landslide dynamics [55, 133, 134], and (c) can be useful for overcoming the unpredictable occurrence of rockfalls [128], even though it is not easy to correlate seismic signal features with landslide geological properties [120, 134]. In general, active and passive seismic methods are the most employed in landslide characterization and monitoring (Figure 2). In “soil” landslides, the three most employed techniques are ERT, SN (at local and regional scales), and SR. The last, together with SRe and SW, is largely used in this kind of landslide typology, and in general, it is easier to find papers focused on “soil” landslides that integrate the abovementioned seismic techniques with other less-common techniques (e.g., MG, IP, SP, and EM). Our analysis of “soil” landslides confirms the conclusions of [20]; i.e., (a) ERT and SR integration proves to be the most effective, (b) the joint application of ERT, SR, and GPR seems to solve and overcome the resolution problems of each single method, and (c) in the literature, there are very few examples of ERT combined with IP to distinguish clayey material or to better interpret ERT. In “rock” landslides, the three most employed techniques are SN (at local and regional scales), ERT, and SR, indicating that passive seismic techniques are preferred over electrical ones. As mentioned above, this is probably because they can be employed to both monitor and characterize seismic signals triggered by landslide dynamics [55, 133, 134]. At the fourth position is GPR, although the authors highlight both the difficulty of deployment on cliffs and the limitation of its applicability to only highly resistive rock slopes [87, 88, 92, 132]. 60.6%77.8% 68.3% 39.4% 22.2% 31.7%
One technique More than one technique
0 10 20 30 40 50 60 70 80 90
SOIL ROCK Total
Figure 1: For each landslide typology (“soil” in red, “rock” in green, and total in blue), the bar graph shows the number of papers focused on just one technique or on more than one. Numbers on the top of the bars are the percentage values with respect to the total number of analysed papers. The darker colours of the “soil” and “rock” bars of the “one-technique” group indicate in how many works the passive seismic technique was employed alone. The dark blue portion of the “one-technique total bar” indicates in how many works other nongeophysical techniques were employed.
In Figure 3, for each drawback, the percentages and the numbers of papers (numbers on the top of the bars) that fall into each level of the three-level scale (+, -, and n.d., which mean that many/some, insufficient, and nondiscussed efforts were made to overcome the limitations, as shown in Table 4) are summarized. In general, it is possible to observe that great efforts were made (95 papers out of the 120 analysed, which is 79.1%, are on the + level of the scale) to improve the geological interpretation of the geophysical data and to explain it more clearly and critically (drawback 3). In contrast, very few efforts were made to (a) systematically discuss, in an understandable way, the resolution and penetration depth of each method (drawback 2: 91 papers out of the 120 analysed, which are 75.8%, are on the n.d. level of the scale), (b) to convince geologists and engineers that 3D and 4D geophysical imaging techniques can be valuable tools for investigating and monitoring landslides (drawback
VES ERT IP SIP SP LOGS GPR FDEM VLF-EM EM RMT AE SN (local) SN (regional) SR SRe SW DH CH MG 0 10 20 30 40 50 60
SOIL ROCK Total
Figure 2: For each landslide typology (“soil” in red, “rock” in green, and total in blue), the bar graph shows the number of papers focused on each geophysical method.
+ - + - n.d. + - n.d. + - n.d. + - n.d. + - n.d. 0.0 10.0 20.0 30.0 40.0 50.0 60.0 70.0 80.0 90.0 100.0
SOIL ROCK Total
Drawback 1 Drawback 2 Drawback 3 Drawback 4 (3D) Drawback 4 (4D) Drawback 5
302757 36 2763 6 3 9 13 720 47 44 91 47 48 95 15 6 21 4 4 7 29 3 1 4 56 51 107 9 9 4 5 9 53 49 102 13 1 14 5 2 7 49 51 99
Figure 3: The bar graph indicates the percentage of efforts made (+ means many/some, - means insufficient, and n.d. means nondiscussed) to overcome each drawback. The percentages of papers focused on “soil” landslides are in red, those of papers focused on “rock” landslides are in green, while in blue are the total percentages. The numbers on the top of each bar indicate the numbers of papers.
4: 107 papers for 3D applications and 102 papers for 4D applications out of the 120 analysed, which are 89.2% and 85.0%, respectively, are on the + level of the scale), and (c) to obtain quantitative information in terms of geotechnical parameters and hydrological properties from geophysical data (drawback 5: 99 papers out of the 120 analysed, which are 82.5%, are on the n.d. level of the scale). Finally, thanks to the development of new 2D and 3D imaging software, some efforts, but still not enough (57 papers out of the 120 analysed, which is 47.5%, are on the + level of the scale), were made to show the geophysical results more clearly (drawback 1).
In the following discussion, we analyse point-by-point the efforts made to overcome each drawback highlighted by [11].
Drawback 1: Geophysicists Have to Make an Effort in the Presentation of Their Results. According to our analysis
(Fig-ure 3), the efforts to overcome this drawback were performed more or less in the same way for both “soil” and “rock” landslides. This means that a tendency to show and present the results more objectively is beginning to emerge. This could be possible thanks to the development of new 2D and 3D software that allow the integration of data from
different sources and surveys (e.g., geophysical, geotechnical, and borehole data). Nevertheless, the presentation of seismic data is sometimes still hard, since authors often show the rough traces or spectra (e.g., [22, 24, 26, 29, 39, 40, 44, 47, 49, 55, 56, 58, 64, 76, 84, 89, 95, 97, 98, 101, 103, 104, 106, 112, 116, 117, 121, 126, 127, 131, 133]) that could be difficult to read for a nonexpert audience.
Drawback 2: The Spatial Resolution and Penetration Depth of Each Method Are Not Systematically Discussed in an Under-standable Way. Each technique has a different resolution and
penetration depth that contribute to the final quality of a geo-metrical model. According to [7], several preprocessing steps are needed to carefully check the data quality and, therefore, the resolution and penetration depth before incorporation into a 3D model. In total, 75.8% of the analysed papers (47 of those on “soil” landslides and 44 of those on the “rock” type) do not discuss either the resolution or the penetration depth of the presented methodology (Figure 3). Additionally, in the review in [20], none of the cited papers within the year range (2007-2013) examine these two points. In contrast, in the remaining 24.2% (Figure 3) of the examined works,
