Next Article in Journal
A Study on the Object-Based High-Resolution Remote Sensing Image Classification of Crop Planting Structures in the Loess Plateau of Eastern Gansu Province
Previous Article in Journal
An Assessment of Changes in the Thermal Environment during the COVID-19 Lockdown: Case Studies from the Greenland and Norwegian Seas
Previous Article in Special Issue
Geometrical Variation Analysis of Landslides in Different Geological Settings Using Satellite Images: Case Studies in Japan and Sri Lanka
 
 
Search for Articles:
Title / Keyword
Author / Affiliation / Email
Journal
Article Type
 
 
Advanced Search
 
Section
Special Issue
Volume
Issue
Number
Page
 
Journals
Remote Sensing
Volume 16
Issue 13
10.3390/rs16132478
remotesensing-logo
Submit to this Journal Review for this Journal Propose a Special Issue
► Article Menu

Article Menu

share Share announcement Help format_quote Cite question_answer Discuss in SciProfiles thumb_up
...
Endorse textsms
...
Comment
first_page
settings
Order Article Reprints
Font Type:
Arial Georgia Verdana
Font Size:
Aa Aa Aa
Line Spacing:
Column Width:
Background:
Article

Evaluation of Mass Movement Hazard in the Shoreline of the Intertidal Complex of El Grove (Pontevedra, Galicia)

by
Joaquín Andrés Valencia Ortiz
*,
Carlos Enrique Nieto
and
Antonio Miguel Martínez-Graña
Department of Geology, Faculty of Sciences, University of Salamanca, Plaza de los Caídos s/n, 37008 Salamanca, Spain
*
Author to whom correspondence should be addressed.
Remote Sens. 2024, 16(13), 2478; https://doi.org/10.3390/rs16132478
Submission received: 17 May 2024 / Revised: 3 July 2024 / Accepted: 5 July 2024 / Published: 6 July 2024
(This article belongs to the Special Issue Geomatics and Natural Hazards)
Browse Figures
Versions Notes

Abstract

:
Knowledge of hazard conditions due to mass movements is one of the non-structural measures for risk management, urban planning, and protection of natural resources. To obtain this type of mapping, a spatial construction was started by correlating the historical movements with the inherent variables of the terrain by means of the bivariate statistical method, which assigns densities or weights of evidence to estimate the degree of susceptibility. This model was combined with the triggering factors (rainfall and earthquake) to determine the spatiotemporal conditions (hazard). From this procedure, it was obtained that the susceptibility model presents 34% (32.33 km2) of the total area in the high and very high categories, especially in the regions of Mount Siradella and Mount Faro. The validation of the present model obtained a value of 0.945 with the ROC curve. For the hazard condition, 34.1% (32.06 km2) of the study area was found to be in the high and very high category, especially in the municipalities of El Grove, Sanxenxo, and A Illa de Arousa, which have the greatest extension. The present evaluation is an advance in the knowledge of the risk and the actions that can be derived, as in turn, this type of study is an easy tool to obtain due to its low cost and information processing.

1. Introduction

The evaluation of potentially unstable conditions from the point of view of mass movement processes is a relevant factor in government policies aimed at risk management, the evaluation of disaster management plans, the development of infrastructure, and the protection of natural resources. This aspect is one of the areas that has been developing more and more in recent decades, due to changes in the environment, especially climate change. The latter is an important point to describe since climatic conditions have presented a significant change due to an increase in average, maximum, and minimum air temperatures that provide rainfall scenarios with greater intensity and recurrence, especially in mountain areas where they can easily activate this type of processes [1,2,3,4,5].
To determine these types of processes and how they affect anthropic environments, studies are initially contemplated focused on the characterization of the inherent conditions of the terrain, where aspects concerning geological, geomorphological, land cover, and land use attributes that are relevant factors in the generation of mass movements are evaluated [6,7,8]. These inherent elements of the terrain are part of a geoenvironmental characterization that determines the influence or not of unstable zones [9], as well as assessments directly associated with mass movements, defining the type of movement and its characterization [10]. The evaluation of these two aspects (inherent factors and processes) determines the probability in terms of susceptibility of occurrence of a spatially shaped movement. To recreate these probabilities, specific algorithms or simulations have been employed where these two aspects are correlated resulting in a non-temporal spatial vector mapping element that considers the classification, volume (or area), and spatial distribution of mass movements [9,11,12,13]. This result approximates the spatial type of probability to generate a terrain instability process, which is manifested by the given displacement over a surface that will depend directly on the method employed and the scale of work.
For the construction of this type of correlation, different methods are used that are grouped according to their scale, availability, quality, data accuracy, susceptibility resolution and the type of subtraction to be obtained, including heuristic, stochastic, statistical, and deterministic methods [14,15]. As Cascini [16] states, in order to obtain susceptibility cartographic models, a suitable working scale must be adopted according to the method proposed, and within the different scales described, to obtain cartography at an intermediate level of detail (scale 1:25,000 or 1:10,000), stochastic or statistical methods must be considered. This is the statistical method adopted for the present study since its basis focuses on establishing combinations between inherent factors and mass movements in a statistical way [14]. To perform this correlation, the present study will take as a basis the Bivariate algorithm, which estimates the densities of “Weights of evidence” between each of the evaluated attributes of the inherent factors established and the mass movements [9,17,18].
With the aspects described above, determining spatiotemporally the occurrence of a terrain instability process involves taking the concept of susceptibility to hazard, which encompasses temporary conditions associated with an external stimulus (trigger) due to factors such as heavy rainfalls, rapid thawing, changes in water level, volcanic eruptions or strong terrain shaking by the action of an earthquake [19,20]. In the specific case of the present study, rainfall and seismic activity will be the reference points for the estimation of hazard conditions in the study area. Rainfall has been described as one of the most influential factors in the generation of mass movements [21,22,23,24,25]. This description refers especially to the action of intense precipitation episodes, of long or short duration, which generate an increase in interstitial pressure or loss of cohesion between soil particles [26,27]. But this aspect is affected by local factors such as slope, soil type, vegetation, lithology, morphology, kinematics, and material involved, among others, which increase or decrease the action of rainfall on the soil [21,28,29]. The second trigger that is related to seismic activity is a factor that has been considered worldwide as one of the precursors of unstable terrain points, which has caused considerable human and infrastructure losses created by the same seismic activity and the activation of mass movement [30,31]. Its evaluation for the present study is limited to the values of the Peak Ground Acceleration (PGA), however, seismic activity as a trigger of mass movements is a topic under study at present, because it has a local mode that can define the phenomenon that triggers the collapse of the slope, but at a more regional level this perspective changes due to inherent factors of the terrain and the physical properties given at the time of seismic activity [20,32].
With the present information, the evaluation of mass movement hazard conditions is an important point in the knowledge of risk from a non-structural perspective. The creation of a susceptibility model, and subsequently a hazard model, can provide useful information for local or regional decision-makers, who can use this study as a reference in land use planning and the preservation of natural resources. For this reason, the development of this type of study is in line with the objectives set out in the Sendai Framework for Disaster Risk Reduction, 2015–2030 [33].

2. The Test Area

The study region is located in Northwestern Spain, in the Pontevedra region, and includes the municipalities of A Illa de Arousa, Cambados, Meaño, O Grove, Ribadumia, Sanxenxo, Vilagarcía de Arousa and Vilanova de Arousa (Figure 1). The region has an area of 95.3 km2, an altitudinal difference of 196 m (0–196 AMSL), with a slope between 0° and 84°, an average of 6° and a dominance between the range of 0° to 5°, with surfaces that are mostly oriented towards the west (240° to 300°). Geologically (Figure 2A), the region presents rocks of origin between the Precambrian and Silurian periods corresponding to alkali feldspar granite (PcSGf), biotitic granodiorite with amphibole (PcSGd), granodiorite with feldspar megacrystals (PcSGdf), coarse-grained biotite–amphibole late granodiorite (PcSGdt), paragneises with plagioclase and biotite, micaschists (PcSPG), micaschists, quartz schists and paragneises (PcSM), quartzites (PcSc), feldspar granite (PcSG), calcsilicate rocks and para-amphibolites (PcSC) and quartz dykes, aplites and pegmatites (Sd) [34,35,36,37].
In turn, there are deposits from the Quaternary period such as the dejection cone (Qcd), old beaches and coastal rasa (Qpa), Pleistocene terrace deposits (Qtp), colluvial deposits (Qc), beach sands (Qap), dune sands (Qad), alluvial–colluvial deposits (Qac) and alluvial deposits (Qa) [34,35,36,37]. From the geomorphological aspect (Figure 2B), the region is characterized by the development of marine environment units such as aeolian dunes with vegetation (Ad), abrasion platform (Ap), beach (B), sand bars (Sb) and sandstone (S); of the denudational environment presents the residual relief (Rr), colluvial (C), glacis (G), Piedemonte (F); and of the fluvial and eolian environment we find the alluvial (A), alluvial fan (Af) and current aeolian dunes (Cad), respectively [38].

3. Methodology and Materials

The evaluation of mass movement hazard as a non-structural measure is a fundamental element in disaster risk management, especially in land use planning and natural resource protection. These mass movements are considered as a displacement of a mass of rock, debris, or soil downslope that generates serious impacts on anthropic and environmental elements [39]. In order to establish the evaluation of the hazard condition due to mass movements in the study area, two procedures were carried out following the guidelines established by the United Nations Office for Disaster Risk Reduction [40]. On the one hand, we began with the calculation of the susceptibility to mass movements (spatial condition) where, by means of an algorithm (statistical method), the inherent or conditioning variables were correlated against the dependent variable. On the other hand, the result obtained from the mass movement susceptibility calculation was used as a basis and this was crossed by means of a combination matrix with the triggering factors (rainfall and earthquake—temporal condition), to take the expression to terms of mass movement hazard (spatial–temporal relationship). For the evaluation of each of these conditions, the present study followed the flow diagram described in Figure 3.

3.1. Calculation of Susceptibility to Mass Movement

Based on the flow diagram designed (Figure 3), we begin by calculating the susceptibility to mass movements. This calculation was made based on the bivariate statistical method. This method is part of the statistical methods that make it possible to establish combinations between the inherent elements (conditioning factors) of the surfaces and the mass movements (dependent) that have historically been generated in a region, resulting in a quantitative predictive model [14]. This model discriminates the regions with a spatial probability (susceptibility) in the generation of an event, a situation that shares similar conditions to the inherent elements that were entered for the calculation performed with the bivariate statistical method [14]. In order to establish these conditions between the inherent elements and the mass movements, an evaluation of the surfaces where mass movements occur was carried out. This evaluation considers all the variables associated with the process that are especially related to geology, geomorphology, land covers, and land uses, which are relevant factors in the generation of mass movements [6,7,9,12].
The bivariate statistical method is based on the Bayesian theorem that indicates the probability that an event A occurred because an event B occurred [41,42]. From this relationship, an association is sought by means of occurrence densities between inherent elements and mass movements, by means of a spatial correlation, this correlation generates numerical values that vary depending on the degree of association, values also called “weights of evidence” [9,17,18]. In other words, the aim is the catheterization and subsequent correlation of the geoenvironmental variables that are directly associated with ground stability problems, and thus determine the spatial probability of the regions susceptible to a displacement of a mass of rock, debris, or soil. For this calculation of the weights of evidence, a database with the historical record of mass movements was elaborated. These movements were obtained by a process of photointerpretation on images arranged in the Google Earth® platform, multitemporal images with an optimal resolution where it was possible to determine the characteristics of type, geometry, distribution, and material of the mass movements [43,44]. For the characterization of the movements, the descriptions made by Varnes [10] were considered, and also the new aspects described by Hungr et al. [45], and Skempton and Hutchinson [46] for the movements of landslide and flow type, which can be classified into superficial and deep. In turn, for the construction of the geometric elements of the mass movements, as well as the inherent variables, cartographic inputs were acquired from the National Geographic Institute of Spain (IGN) at a scale of 1:25,000, and the Digital Elevation Model (DEM) with a resolution of 1 m [47]. These inputs provide within each model a good resolution for a scale of 1:25,000 on which the present work is based. The construction of the layers and the calculations will be developed under the ArcGIS V10.8 software platform.

Construction of Geoenvironmental Variables

In the construction of the inherent variables (9 variables), such as lithology, distance to faults, morphogenesis, slope, orientation, rugosity, relative relief, land cover, and land use were correlated with the mass movement inventory to generate the weights of evidence. The first two variables (lithology and distance to faults) correspond to the geological attribute of the study region that was acquired from the Geological Institute of Spain at a scale of 1:50,000 [34,35,36,37]. From the geomorphological attribute, the morphogenetic variable was acquired [38], which describes the endogenous and exogenous factors characteristic of terrestrial dynamics, added to the degradation and aggradation of the terrain resulting from weathering, erosion, and transport processes over time [48]. In turn, for the geomorphological attribute, the morphometric variables of slope, orientation, rugosity, and relative relief were constructed through DEM processing. For the land cover attribute, in the study of susceptibility to mass movements, the variable of the current state of land cover was taken, which is developed with the CORINE land cover methodology [49]. This variable is an important factor, since, depending on the type of cover, it offers better soil protection and contributes to the dissipation of rainfall energy [50]. This input was acquired from the IGN and corresponds to the European CORINE Land Cover (CLC) project, which is composed of 44 classes and has a 2018 reference version [51]. For the present study and based on the reference levels created by the CORINE Land Cover methodology, the present study took level 3 and below as a variable in the calculation of susceptibility, since it is in accordance with the scale of work of this research. For the land use attribute, the cartography generated by the Spanish Land Use Information System (SIOSE), which has a high-resolution input based on the integration of high-detail geospatial sources (SIOSE AR), was used. This input corresponds to the characteristics of land occupation for the year 2017 (most current version) [52] and was taken as a variable in the calculation of susceptibility.
With the acquisition of the inputs and the construction of each variable, the calculation of susceptibility to mass movements was carried out. The final susceptibility model adopted five categories (very low, low, medium, high, and very high) for the classification of areas with a natural predisposition to generate an instability process according to the inherent variables that were evaluated. For the validation of the susceptibility calculation, two different tests were performed. The first test was performed by means of the receiver operating characteristic (ROC) curve where the area under the curve (AUC) was estimated. This test consists of the calculation and its graphical representation of sensitivity versus specificity for a binary classifier system as the discrimination threshold is varied, and from this, its area under the curve (ABC) is estimated as a statistical measure of the success and prediction rates of each model [53,54]. The second test consisted of calculating accuracy, precision, recall, and harmonic mean (F1-score) values that are based on the degree of classification generated from a confusion matrix. The confusion matrix basically compares the modeled data against the actual data, in this sense it seeks to understand the model performance by combining by means of a table the data generated from the model prediction against the actual data [55,56].

3.2. Mass Movement Hazard Calculation

Based on the susceptibility model, the hazard map was constructed, which incorporates the triggering factors of mass movements. This condition is established since the susceptibility model determines the spatial aspects of the movements, but the triggering action is a phenomenon associated with rainfall, earthquakes, and anthropic activity, which takes the expression to spatiotemporal terms [20,57,58]. For the present study, only rainfall and earthquakes were taken as mass movement triggers. For the rainfall trigger, this input was acquired from the Centro de Estudios y Experimentación de Obras Públicas (CEDEX), with its hydrographic studies center attached to the Spanish Ministry of Transport and Sustainable Mobility, where a model of the impact of climate change on maximum precipitation in Spain (2021, 2022) was generated. The respective result layers of this study present the rates of change in maximum annual daily precipitation generated with an SQRT-R model for Tr return periods of 10, 100, and 500 years [59]. For the seismic trigger, the Seismic Hazard Map of Spain 2015 was acquired, with Peak Ground Acceleration values (PGA), with a return period of 475 years, found in the portal of the National Geographic Institute of Spain [60]. For the final part of this study, the mass movement hazard map was constructed by means of a combination matrix between the susceptibility model and the rainfall and seismic triggers. The combination of these three elements describes the spatiotemporal instability conditions for the study region, which were classified into four hazard categories (low, medium, high, and very high).

4. Results

For the evaluation of the mass movement hazard, the starting point was the construction of a database of historical mass movements. For its construction, we first consulted the mass movements registered in the database of the Geological and Mining Institute of Spain (BDMOVES), in which, for the study region, there are no reported events. Based on this, the second step was a cartographic construction of the mass movements by means of a process of photointerpretation of satellite images contained in the Google Earth® platform. For the construction of each of these geometries where mass movement was evidenced, a total of 22 satellite images from 1985 to 2023 were analyzed. This sequence of images presented a good spatial resolution for the identification of the respective movements, but in some sectors due to atmospheric conditions, especially cloudiness, it was not possible to clearly appreciate the surfaces on the analyzed image. Another limitation found in the photointerpretation process is the size of the image, which for some sequences did not cover the study region, presenting an overlap with another image with a different date of acquisition, generating a bias on the region that does not present a coherent date. Taking into account these limitations, this process resulted in a record of 60 mass movements with a total area of 0.055 km2, less than 1% of the total study area (Figure 4). From this inventory, a total of 35 surface landslide-type movements were described, with a total area of 0.024 km2 (44% of the total area of movements), and a total of 25 surface flow-type movements with a total area of 0.031 km2 (56% of the total area of movements). These movements are concentrated towards the region of Monte de Siradella (east zone—El Grove), which presents a total of 18 movements (8 landslides and 10 flows), and for the region of Monte Faro (south zone—Sanxenxo), which presents a total of 35 movements (24 landslides and 11 flows).
Based on the mass movement inventory, the calculation of the weights of evidence for each of the constructed variables was carried out. From the geological attribute and its lithological variable, it is observed that the rocks that contribute to the generation of mass movements are especially the granodiorite with feldspar megacrystals (PcSGdf), coarse-grained biotite–amphibolic late granodiorite (PcSGdt) and the micaschists, quartz schists and para-gneisses (PcSM). This is largely due to the weathering of these rocks by physical-mechanical processes associated with the trace of the surrounding faults, as well as to the changes produced by the weathering generated by the conditions of seasonality change. The lithological units that do not contribute to the generation of mass movements are the alluvial deposits (Qa), alluvial-colluvial deposits (Qac) and the deposits of old beaches and littoral rasa (Qpa). This is largely due to the fact that these deposits (90% of the area of these units) are located on areas with slopes < 10°, which generates a certain degree of stability to the deposit in the face of mass movement processes, but it is susceptible to surface or laminar erosion processes (Figure 5A). For the variable of distance to faults, processing was first performed on the fault trace where this distance was calculated by means of an algorithm (Euclidean distance) in ArcMap V10.8, with a separation radius of 500 m, obtaining 7 ranges of separation. These ranges were correlated with the inventory of mass movements to obtain the respective weights of evidence. From this result it is observed that the range of 0–500 m contributes to the generation of movements and the range between 1500 and 2000 does not contribute to the generation of movements (Figure 5B).
For the geomorphological attribute and its morphogenetic variable, it is observed that the units of the denudational environment associated with plutonic and metamorphic rocks contribute to the generation of mass movements. The combined effect of seasonal change, physical–mechanical weathering produced by fault stroke, slope, and abrupt changes in the surface planes (rugosity) have an impact on the denudational processes of the surfaces, creating situations that are exploited for the development of degradational processes within the morphological aspect on this type of rocks. On the contrary, the units of the aeolian, anthropic, and marine environments, such as aeolian dunes with vegetation, roads, and sanding, respectively, do not contribute to the generation of mass movements. This is largely due to the location of these units since they are located in areas with a slope < 10° (92% of the area of these units), and the processes that prevail in these regions, especially in the marine environment units, are due to accentuated erosion caused by wave action (Figure 6).
For the morphometric variables, the correlation was first made with the slope, which was obtained from the DEM and was reclassified into 10° ranges (nine ranges), from this reclassification it was correlated with the mass movement inventory to determine the weights of evidence. From this correlation, it was obtained that slopes between 60°–70° and 50°–60° contribute to the generation of mass movements (Figure 7A).
This characteristic would be somewhat in contradiction with respect to the descriptions that have been made at the international level, where it is described that the most susceptible regions in the generation of mass movements would be between 25° and 35° [61,62,63]. This condition would be more associated with the aspects described by Valencia Ortiz et al. [9] where he states that the lack of relationship is due to the way the bivariate method works, which is more a function of the relationship in areas. Performing a parallel exercise only correlating the slope with the mass movements, disregarding the area function (histogram), it is observed that 41% of the events occur on slopes between 20° and 40°, which is consistent with what is described at the international level. On the other hand, slopes between 20°–30° and 0°–10° do not contribute to the generation of mass movements.
For the morphometric variable of orientation, it is observed that surfaces oriented between 45°–90° and 315°–360° contribute to the generation of mass movements, and surfaces oriented between 225°–270° and 270°–315° do not contribute to the generation of mass movements (Figure 7B). The relative relief variable was based on the descriptions made by van Zuidam [64] and was constructed from the processing of the DEM that was divided into 500 m squares, where a neighborhood operation was performed by ranges that were later reclassified into 30 m ranges, obtaining five ranges. The correlation of this variable with the inventory of movements to obtain the weights of evidence describes that the ranges between 90–120 and >120 m contribute to the generation of mass movements. On the contrary, the ranges between 0–30 and 30–60 do not contribute to the generation of mass movements (Figure 7C). The last morphometric variable calculated from the DEM was rugosities, which was constructed based on the aspects described by Felicisimo [65] and was classified into zones with very high rugosities to very low rugosities. With this classification and correlating it with the inventory of movements for the weights of evidence, it is observed that the zones with very high rugosities contribute to the generation of mass movements, and the zones with very low rugosities do not contribute to the generation of movements (Figure 7D).
For the land cover attribute and its variable of current land cover status, a layer was obtained that presents a total of 37 level 3 cover classes. From this layer, correlating it with the inventory of mass movements to obtain the weights of evidence, it is observed that coverages such as mixed woodland, citrus fruit trees, evergreen hardwoods, and conifers are regions that contribute to the generation of mass movements. On the other hand, coverages such as paved or sealed areas, buildings, artificial green areas urban trees, and vineyards are regions that do not contribute to the generation of mass movements (Figure 8A). Finally, the layer of current land uses was obtained, which has a total of 48 classes of uses. This layer was correlated with mass movements, obtaining a result where uses such as natural land areas, abandoned areas, electric power, gas and thermal energy distribution services and textile manufacturing regions contribute to the generation of mass movements. The regions that do not contribute to the generation of mass movements are associated especially with sports facilities, secondary production, transitory areas, and commercial agricultural production (Figure 8B).
With each of the variables constructed, and calculated the weights of evidence of each class, the sum of the weights calculated for the class to which the pixel of each of the selected factors belongs is made, thus obtaining the final susceptibility function or LSI (Landslide Susceptibility Index) [9,17,41]. Once the LSI is defined, a reclassification of the data into five categories (very low, low, medium, high, and very high) is performed using the geometric interval algorithm. This algorithm creates breaks of the classes into class intervals that have a geometric series, which ensures that each class range has approximately the same number of values in each class and that the change between intervals is fairly consistent, generating a visually appealing and cartographically understandable result [66]. From the present reclassification into five categories, it is observed that the very high susceptibility category has an area of 17.9 km2 (18.8%) of the study area, and a total of 43 mass movements were grouped on this category with an affected area of 53943 m2 (98.21%) of the total area of movements. The high susceptibility category presents an area of 14.4 km2 (15.2%), with a total of 16 mass movements, with an area of 966 m2 (1.76%); the medium susceptibility category presents an area of 13.9 km2 (14.6%), with a total of 1 (1) mass movement with a total area of 17 m2 (0.03%). Finally, low, and very low susceptibility present an area of 48.9 km2 (51.4%), these categories do not include any mass movement (Table 1). This mass movement susceptibility map shows the largest area in the very high category, which is especially sectored for the Siradella and Faro Mountains, and to a lesser extent in the northeastern region of the municipality of A Illa de Arousa and the southeastern region of the municipality of El Grove (Figure 9).
Based on the final susceptibility model, the different validation tests were carried out using the methods proposed. For the ROC curve method, a value equal to 0.945 was obtained, and for the confusion matrix, the values of accuracy = 0.9821, precision = 0.9824, recall = 0.9997, and F1-score = 0.991 were obtained. With the construction of the AUC curve and the analysis based on the confusion matrix and its results, it is observed that the calculation of susceptibility by means of bivariate statistical analysis presents a good performance in the prediction of regions that have historically been unstable, giving a very high level of certainty about the behavior of susceptibility in the entire study area. All these results describe a good behavior of the susceptibility model (Figure 10).
With the validated mass movement susceptibility model, the mass movement hazard map was constructed, which integrates the triggering factors (rainfall and earthquake) of the study region. For the rainfall trigger obtained from CEDEX, two scenarios are described, RCP 4.5 and RCP 8.5 defined by the Intergovernmental Panel on Climate Change—IPCC [67]. The models generated by CEDEX describe that, for RCP 4.5 with a mid-century projection, rainfall can reach values between 11 and 17 mm (maximum annual daily precipitation) for a Tr of 10 years, values of 15–22 mm for a Tr of 100 years and 17–25 mm for a Tr of 500 years [59]. For the scenario with an RCP 8.5, values between 12 and 18 mm for a Tr of 10 years, between 10 and 24 mm for a Tr of 100 years, and between 13 and 28 mm for a Tr of 500 years are described [59]. From the respective models generated, the mid-century projection is taken, with data from the RCP 4.5 model and a Tr of 100 years, which would be more discrete data as a mass movement triggering factor for the study area (Figure 11A). For the triggering earthquake, the model generated in 2015 from the Peak Ground Acceleration (PGA) map present in the IGN portal [60] was acquired. This map describes acceleration values for the study region between 0.04 g ≤ PGA < 0.08 g (39.2 cm/s2 ≤ PGA < 78.4 cm/s2) (Figure 11B).
With the present rainfall and acceleration map, which are qualified in five ranges, they were assigned a numerical value (1 to 5) in order to correlate the present values of the study area with the data obtained from the susceptibility model by means of a combination matrix (Table 2). This assignment of values is established to perform a concordant correlation between the values and results of the triggers and the mass movement susceptibility [57,58]. For the correlation by means of the combination matrix, the susceptibility classification is taken as the abscissa axis and the sum by independent of the triggers with the susceptibility values on the ordinate axis, the crossing between axes describes the mass movement hazard [58]. This hazard map represents the spatiotemporal conditions of the study area in four categories (low, medium, high, and very high) (Figure 12).
The present mass movement hazard map shows a total area of 17.76 km2 (18.89%) of the study area in the very high category, 14.30 km2 (15.21%) in the high category, 41.50 km2 (44.14%) in the medium category and 20.45 km2 (21.75%) in the low category. These hazard conditions are associated with maximum daily rainfall between 15 and 22 mm, a condition that may change due to climate change (this aspect will be explained in the following chapter), and acceleration values with a PGA between 0.064 and 0.108 for the study region. On the other hand, indirectly, the seasonality of the region must be considered, in addition to the snow conditions that can create weathering and erosion conditions of a superficial type, contributing to a certain extent to the creation of instability planes. The municipalities with the largest area in the very high category of mass movement hazard are El Grove, A Illa de Arousa, and Sanxenxo (Table 3).

5. Discussion

The development of this type of cartography, which is built on the basis of integrating in a mathematical function the inherent elements of the terrain correlated with historical processes to establish a probability that is subsequently crossed with the triggering factors to describe the spatiotemporal relationships, are key points to consider in the evaluation of risk management plans and especially in the planning of urban development and protection of natural resources. This aspect is discussed by Bathrellos et al. [68], where he describes the correct application of a methodology that is appropriate for a specific environment where the use of a coherent scale of work will determine the potentially susceptible sites so that people such as planners, engineers, and policy-makers can improve the quality of life of its inhabitants. In turn, this type of information generated is useful for governmental and non-governmental bodies to define an appropriate level of management and preparation of necessary measures based on the information obtained in combination with risk analysis, thus estimating the possible average costs associated with the damage generated by mass movement [69].
This type of evaluation is relevant in the aspects described above, but also, in the construction of a susceptibility model, the intrinsic value associated with the appraisals and descriptions made by the expert criteria must be taken into account, according to the scale of work, since the mathematical methods adopted work on the basis of the data entered and their corresponding correlation of the information [9]. This is largely due to the cause–effect relationship where the real conditions of an environment can be overestimated or underestimated, which, in this specific case, would be given by that spatial relationship (susceptibility) and its triggering factors. Conditions that are very focused on the aspects of a low complexity system of unidirectional nonlinear type, where the relationship of triggers, such as rainfall and earthquakes (for this study) has an increasing effect on mass movements, but the movements have no effect on the triggers [20,70]. But it is also important to have designed inputs on the environmental conditions present in the evaluated region (site effect) since regional models can bias or homogenize their behavior, as is the case of the spatio-temporal relationship, where susceptibility is integrated with the triggers of a region [9,20,57]. In turn, within the cartographic conditions for the creation of a suitable susceptibility model that is in accordance with the level of interpretation and detail required within an urban planning and regional zoning scheme, it is required that the cartographic inputs be in accordance with the present level since it can lead to problems on anthropic elements [16].
A relevant aspect within these cartographic constructions, especially in the creation of the mass movement inventory, is the limitation of the record obtained, both by national or international databases and by the photointerpretation process, where the record, on some occasions, is somewhat limited. If we look only at the photo interpretation process, from aerial photographs, we only have very specific dates on which these photographs were acquired, or from satellite images, in which the acquisition is limited to the last decades. This means that we do not have a complete picture of the region under evaluation or the regions that could potentially be unstable. From these aspects arises the importance of acquiring first-hand information sources for a correct interpretation and evaluation of a region [9,11].
In a parallel example to this study but involving somewhat similar conditions in terms of the scale of work, expert criteria, and level of detail, are the descriptions made by De Moel et al. [71], where he mentions that there are notable contrasts in the different levels of information described for the knowledge of irrigation associated with floods. From this perspective and only taking as a reference a local scale, shortcomings and urgent needs are described in the study and analysis of the effects generated by floods on critical infrastructures, given their importance for society, the economy, emergency management, and reconstruction [71]. These factors that imply further progress in the knowledge of the elements associated with risk can be extrapolated to the processes of mass movements since the conditions of generation of an event to a large extent are arranged by hydrometeorological effects. On the other hand, conditions at the level of mass movement processes, such as floods, are factors that should be given priority in studies at larger scales, due to the implications they have on anthropic environments. These aspects are becoming more and more relevant due to the changing climatic conditions we are currently experiencing.
An overview of these changes can be seen in the reports generated by the Research on the Epidemiology of Disasters (CREDS), where in its 2023 report it analyzes the number of disasters generated for the year 2022 and compared with the annual average between 2002 and 2021. This report describes that for the year 2022, a total of 387 different events were generated, and for the period from 2002 to 2021 370 were generated, of which 16 droughts, 27 earthquakes, 168 floods, 18 landslides, 104 storms, 6 volcanic activities, and 11 forest fires were recorded, and only for the year 2022 there were 22 droughts, 31 earthquakes, 176 floods, 17 landslides, 108 storms, 5 volcanic activities, and 15 forest fires [72]. The above describes a considerable increase in just one year in the number of events recorded worldwide. These conditions in the increase in natural disasters due to climate change in recent years are factors to be considered indirectly in the way of establishing relationships with future climate projections in the evaluated environment. If we look at the projections for the study region and only take the climatic factors of temperature (maximum in the year) and rainfall (maximum in 24 h and number of days with rainfall) described in the climate scenario viewer of the Ministry for Ecological Transition and the Demographic Challenge of Spain [73], we observe three scenarios for an RCP of 4.5. A near future scenario would have temperature values between 19° to 21° C, maximum 24 h rainfall between 51 to 67 mm, and a number of rainy days between 122 to 135 [73]. For a medium future scenario, temperature values would be between 20° to 21° C, maximum 24 h rainfall between 54 to 70 mm, and number of rainy days between 115 to 135 [73]. For a distant future scenario, temperature values would be between 20° to 22° C, maximum 24 h rainfall between 55 to 71 mm, and number of rainy days between 116 to 135 [73].
In order to better understand whether these scenarios present an increase or decrease over the base values for the study region, a comparison should be made with the historical data reported. In this sense, the information reported by the Spanish State Meteorological Agency (Aemet) is used, where average maximum temperature values are between 17.5° to 20° C, average maximum daily precipitation between 70 to 80 mm, and average number of days with rainfall greater than 1 mm between 100 to 125 days [74]. With this information, a variation of the data is observed, especially for the number of days with rain that present an increase. The effect of this type of change, where there is an increase in the number of days with rain, or the maximum rainfall of 24 h, can intensify the action as a trigger for mass movements. But it is also important to consider that, in order to define a complete scenario, detailed studies of rainfall as a trigger of mass movements must be carried out, where the threshold at which an event can be triggered, the probability of exceeding this threshold and its recurrence are estimated [20]. These aspects greatly improve the understanding of the phenomenon and its causality. This contrast indicates conditions of increase in each of the observed parameters, indicating a greater possibility of generating surface processes, especially mass movements. Hence the importance of expanding the knowledge of the risk associated with these types of hazards to improve from a structural and non-structural condition at a local level of detail, the actions, and measures to mitigate the impacts that may be generated. Thus, the characterization of an environment that is disturbed by climatic or internal land conditions is necessary to estimate the degree of impact that may occur on these natural and anthropic environments [75,76].

6. Conclusions

The evaluation of the mass movement hazard constructed from the susceptibility model generated by the bivariate method, and combined with the trigger’s rainfall and earthquake, describes in four categories (low, medium, high, and very high) the regions in danger condition. The susceptibility model is presented in five categories of which high and very high susceptibility represent 34% (32.33 km2) of the total area, with a greater extension on the areas of Mount Siradella and Mount Faro. This susceptibility model was created based on nine variables (lithology, distance to faults, morphogenesis, slope, orientation, rugosities, relative relief, land cover, and land use) that were correlated with the mass movement inventory, which has 60 records, thus obtaining the respective weights of evidence. The final susceptibility model according to the methods proposed for its validation obtained a value of 0.945 for the ROC curve, and for the confusion matrix the values of accuracy = 0.9821, precision = 0.9824, recall = 0.9997, and F1-score = 0.991 were obtained. This reclassified susceptibility model was combined in a matrix with the rainfall and earthquake triggers to obtain the hazard. The rainfall trigger for the study region presents maximum annual daily precipitation values between 15 and 22 mm for a 100-year Tr according to RCP 4.5 with projection to mid-century. For the earthquake trigger, it is described that the Peak Ground Acceleration (PGA) values in a 475-year return period are between 0.04 g ≤ PGA < 0.08 g (39.2 cm/s2 ≤ PGA < 78.4 cm/s2). The mass movement hazard map in its high and very high category has an area of 32.06 km2 (34.1%) of the study area, with the municipalities of El Grove, Sanxenxo, and A Illa de Arousa, respectively, having the greatest extent of hazard in these two categories. The study carried out here is an advance in the non-structural measures of the knowledge of the risk associated with these processes. In addition to this, the evaluation carried out and part of the inputs used come from public entities such as the Geological and Mining Institute of Spain (IGME), the National Geographic Institute (IGN), and the Center for Studies and Experimentation of Public Works (CEDEX), which present the thematic information freely available and in different formats so that they can be processed according to the study to be developed. Also, this type of processing to obtain the susceptibility and hazard map can be completed with spreadsheets and free format geographic information systems, which means this type of study can be generated at a low cost, but with products that are relevant at the time of concrete actions in urban planning and preservation of natural resources. At the same time, it is a tool that can be easily involved in actions concerning disaster risk management, due to the sectorization and characterization of regions in hazardous conditions. However, this type of study should be carried out continuously, since a periodic update with respect to new records of mass movements, inherent factors and analysis of triggers will improve the prediction and quality of the products obtained, as well as improve the aspects described above in a process of positive feedback of the system.

Author Contributions

Conceptualization, data collection, data analysis, writing draft and final manuscript, calculation of susceptibility and hazard due to mass movements, J.A.V.O. Concept, writing draft manuscript, C.E.N. and A.M.M.-G. All authors have read and agreed to the published version of the manuscript.

Funding

This research received no external funding.

Data Availability Statement

The original contributions presented in the study are included in the article, further inquiries can be directed to the corresponding author.

Acknowledgments

This research was assisted for Grant 131874B-I00 funded by MCIN/AEI/10.13039/501100011033 and the GEAPAGE research group (Environmental Geomorphology and Geological Heritage) of the University of Salamanca.

Conflicts of Interest

The authors declare no conflicts of interest.

References

  1. Intergovernmental Panel on Climate Change (IPCC). Climate Change 2007: The Physical Science Basis Contribution of Working Group I to the Fourth Assessment Report of the Intergovernmental Panelon Climate Change; Solomon, S., Qin, D., Manning, M., Chen, Z., Marquis, M., Averyt, K.B., Tignor, M., Miller, H.L., Eds.; Cambridge University Press: Cambridge, UK, 2007.
  2. Gariano, S.L.; Guzzetti, F. Landslides in a changing climate. Earth Sci. Rev. 2016, 162, 227–252. [Google Scholar] [CrossRef]
  3. Alvioli, M.; Melillo, M.; Guzzetti, F.; Rossi, M.; Palazzi, E.; von Hardenberg, J.; Brunetti, M.T.; Peruccacci, S. Implications of climate change on landslide hazard in Central Italy. Sci. Total. Environ. 2018, 630, 1528–1543. [Google Scholar] [CrossRef]
  4. Patton, A.I.; Rathburn, S.L.; Capps, D.M. Landslide response to climate change in permafrost regions. Geomorphology 2019, 340, 116–128. [Google Scholar] [CrossRef]
  5. Wood, J.; Harrison, S.; Reinhardt, L.; Taylor, F. Landslide databases for climate change detection and attribution. Geomorphology 2020, 355, 107061. [Google Scholar] [CrossRef]
  6. Popescu, M. Landslide Causal Factors and Landslide Remedial Options. In Proceedings of the 3rd International Conference on Landslides, Slope Stability and Safety of Infra-Structures, Singapore, 11–12 July 2002; pp. 61–81. [Google Scholar]
  7. Van Westen, C.J.; Castellanos, E.; Kuriakose, S.L. Spatial data for landslide susceptibility, hazard, and vulnerability assessment: An overview. Eng. Geol. 2008, 102, 112–131. [Google Scholar] [CrossRef]
  8. Lv, L.; Chen, T.; Dou, J.; Plaza, A. A hybrid ensemble-based deep-learning framework for landslide susceptibility mapping. Int. J. Appl. Earth Obs. Geoinf. 2022, 108, 102713. [Google Scholar] [CrossRef]
  9. Ortiz, J.A.V.; Martínez-Graña, A.M.; Méndez, L.M. Evaluation of Susceptibility by Mass Movements through Stochastic and Statistical Methods for a Region of Bucaramanga, Colombia. Remote Sens. 2023, 15, 4567. [Google Scholar] [CrossRef]
  10. Varnes, D.J. Slope movement types and processes. In Landslides: Analysis and Control; Schuster, R.L., Krizek, R.J., Eds.; Transportation and Road Research Board, National Academy of Science: Washington, DC, USA, 1978; pp. 11–13. [Google Scholar]
  11. Fell, R.; Corominas, J.; Bonnard, C.; Cascini, L.; Leroi, E.; Savage, W.Z. Guidelines for landslide susceptibility, hazard and risk zoning for land use planning. Eng. Geol. 2008, 102, 85–98. [Google Scholar] [CrossRef]
  12. Ortiz, J.A.V.; Martínez-Graña, A.M. A neural network model applied to landslide susceptibility analysis (Capitanejo, Colombia). Geomatics Nat. Hazards Risk 2018, 9, 1106–1128. [Google Scholar] [CrossRef]
  13. Liu, R.; Ding, Y.; Sun, D.; Wen, H.; Gu, Q.; Shi, S.; Liao, M. Insights into spatial differential characteristics of landslide susceptibility from sub-region to whole-region cased by northeast Chongqing, China. Geomat. Nat. Hazards Risk 2023, 14, 2190858. [Google Scholar] [CrossRef]
  14. Soeters, R.; van Westen, C.J. Slope instability recognition, analysis and zonation. In Landslide Types and Processes; Turner, A.K., Schuster, R.L., Eds.; Special Report National Research Council 247; National Academy of Sciences: Washington, DC, USA, 1996; Volume 247, pp. 129–177. [Google Scholar]
  15. Huang, F.; Cao, Z.; Guo, J.; Jiang, S.-H.; Li, S.; Guo, Z. Comparisons of heuristic, general statistical and machine learning models for landslide susceptibility prediction and mapping. CATENA 2020, 191, 104580. [Google Scholar] [CrossRef]
  16. Cascini, L. Applicability of landslide susceptibility and hazard zoning at different scales. Eng. Geol. 2008, 102, 164–177. [Google Scholar] [CrossRef]
  17. van Westen, C. Guidelines for the Generation of 1:50.000 Scale Landslide Inventory, Susceptibility Maps, and Qualitative Risk Maps, Illustrated with Case Studies of the Provinces Thanh Hoa and Nghe An; University of Twente: Enschede, The Netherlands, 2013. [Google Scholar]
  18. Chen, W.; Chai, H.; Sun, X.; Wang, Q.; Ding, X.; Hong, H. A GIS-based comparative study of frequency ratio, statistical index and weights-of-evidence models in landslide susceptibility mapping. Arab. J. Geosci. 2016, 9, 204. [Google Scholar] [CrossRef]
  19. Wieczorek, G.F. Landslides: Investigation and mitigation. In Chapter 4-Landslide Triggering Mechanisms; Transportation Research Board Special Report: Washington, DC, USA, 1996; pp. 76–90. [Google Scholar]
  20. Ortiz, J.A.V.; Martínez-Graña, A.M. Calculation of precipitation and seismicity thresholds as triggers for mass movements in the region of Bucaramanga, Colombia. Ecol. Indic. 2023, 152, 110355. [Google Scholar] [CrossRef]
  21. Corominas, J.; Moya, J. Reconstructing recent landslide activity in relation to rainfall in the Llobregat River basin, Eastern Pyrenees, Spain. Geomorphology 1999, 30, 79–93. [Google Scholar] [CrossRef]
  22. Iverson, R.M. Landslide triggering by rain infiltration. Water Resour. Res. 2000, 36, 1897–1910. [Google Scholar] [CrossRef]
  23. Dai, F.; Lee, C. Frequency–volume relation and prediction of rainfall-induced landslides. Eng. Geol. 2001, 59, 253–266. [Google Scholar] [CrossRef]
  24. Rosi, A.; Peternel, T.; Jemec-Auflič, M.; Komac, M.; Segoni, S.; Casagli, N. Rainfall thresholds for rainfall-induced landslides in Slovenia. Landslides 2016, 13, 1571–1577. [Google Scholar] [CrossRef]
  25. Dikshit, A.; Sarkar, R.; Pradhan, B.; Segoni, S.; Alamri, A.M. Rainfall Induced Landslide Studies in Indian Himalayan Region: A Critical Review. Appl. Sci. 2020, 10, 2466. [Google Scholar] [CrossRef]
  26. Sidle, R.C.; Swanston, D.N. Analysis of a small debris slide in coastal Alaska. Can. Geotech. J. 1982, 19, 167–174. [Google Scholar] [CrossRef]
  27. Fredlund, D.C. Slope stability analysis incorporating the effect of soil suction. In Slope Stability; Anderson, M.G., Ed.; John Wiley & Sons Ltd.: Saskatchewan, SK, Canada, 1987; Chapter 4. [Google Scholar]
  28. Crosta, G. Regionalization of rainfall thresholds: An aid to landslide hazard evaluation. Environ. Geol. 1998, 35, 131–145. [Google Scholar] [CrossRef]
  29. Aleotti, P. A warning system for rainfall-induced shallow failures. Eng. Geol. 2004, 73, 247–265. [Google Scholar] [CrossRef]
  30. Rodríguez, C.; Bommer, J.; Chandler, R. Earthquake-induced landslides: 1980–1997. Soil Dyn. Earthq. Eng. 1999, 18, 325–346. [Google Scholar] [CrossRef]
  31. Bird, J.F.; Bommer, J.J. Earthquake losses due to ground failure. Eng. Geol. 2004, 75, 147–179. [Google Scholar] [CrossRef]
  32. Corominas, J.; Van Westen, C.; Frattini, P.; Cascini, L.; Malet, J.-P.; Fotopoulou, S.; Catani, F.; Van Den Eeckhaut, M.; Mavrouli, O.; Agliardi, F.; et al. Recommendations for the quantitative analysis of landslide risk. Bull. Eng. Geol. Environ. 2014, 73, 209–263. [Google Scholar] [CrossRef]
  33. UNDRR. Sendai Framework for Disaster Risk Reduction 2015–2030; United Nations (ONU), United Nations Office for Disaster Risk Reduction: Geneva, Switzerland, 2015. [Google Scholar]
  34. IGME. Mapa Geológico de España E. 1:50.000, Puebla de Caramiñal, Hoja 151 (3–9), Segunda Serie—Primera Edición; Instituto Geológico y Minero de España, Servicio de Publicaciones Ministerio de Industria y Energía: Madrid, Spain, 1981. [Google Scholar]
  35. IGME. Mapa Geológico de España E. 1:50.000, Grove, Hoja 184 (3–10), Segunda Serie—Primera Edición; Instituto Geológico y Minero de España, Servicio de Publicaciones Ministerio de Industria y Energía: Madrid, Spain, 1981. [Google Scholar]
  36. IGME. Mapa Geológico de España E. 1:50.000, Pontevedra, Hoja 185 (4–10), Segunda Serie—Primera Edición; Instituto Geológico y Minero de España, Servicio de Publicaciones Ministerio de Industria y Energía: Madrid, Spain, 1981. [Google Scholar]
  37. IGME. Mapa Geológico de España E. 1:50.000, Villagarcía De Arosa, Hoja 152 (4–9), Segunda Serie—Primera Edición; Instituto Geológico y Minero de España, Servicio de Publicaciones Ministerio de Industria y Energía: Madrid, Spain, 1982. [Google Scholar]
  38. Martinez-Graña, A.M.; Arias, L.; Goy, J.L.; Zazo, C.; Silva, P. Geomorphology of the mouth of the Arosa estuary (Coruña-Pontevedra, Spain). J. Maps 2017, 13, 554–562. [Google Scholar] [CrossRef]
  39. Cruden, D.M. A simple definition of a landslide. Bull. Eng. Geol. Environ. 1991, 43, 27–29. [Google Scholar] [CrossRef]
  40. UNDRR. Natural Disasters and Vulnerability Analysis: Report of Expert Group Meeting, 9–12 July 1979. Office of the United Nations Disaster Relief Coordinator. Available online: https://digitallibrary.un.org/record/95986?ln=en&v=pdf (accessed on 11 April 2024).
  41. Bonham-Carter, G.F.; Agterberg, F.P.; Wright, D.F. Integration of geological datasets for gold exploration in Nova Scotia. Photogramm. Eng. Remote Sens. 1988, 54, 1585–1592. [Google Scholar] [CrossRef]
  42. Bonham, G. Geographic Information Systems for Geoscientists: Modelling with GIS (No. 13); Pergamon, Ed.; Elsevier: Ámsterdam, The Netherlands, 1994; Volume 13. [Google Scholar]
  43. Sato, H.P.; Harp, E.L. Interpretation of earthquake-induced landslides triggered by the 12 May 2008, M7.9 Wenchuan earthquake in the Beichuan area, Sichuan Province, China using satellite imagery and Google Earth. Landslides 2009, 6, 153–159. [Google Scholar] [CrossRef]
  44. Guzzetti, F.; Mondini, A.C.; Cardinali, M.; Fiorucci, F.; Santangelo, M.; Chang, K.-T. Landslide inventory maps: New tools for an old problem. Earth-Sci. Rev. 2012, 112, 42–66. [Google Scholar] [CrossRef]
  45. Hungr, O.; Leroueil, S.; Picarelli, L. The Varnes classification of landslide types, an update. Landslides 2014, 11, 167–194. [Google Scholar] [CrossRef]
  46. Skemton, A.W.; Hutchinson, J.N. Stability of natural slopes and embankment foundations. In Proceedings of the Seventh lntemational Conference on Soil Mechanics and Foundation Engineering 4, Mexico City, Mexico, 25–29 August 1969; Sociedad Mexicana de Mecánica de de Suelos: Mexico City, México, 1969; pp. 291–340. Available online: https://trid.trb.org/view/125702 (accessed on 11 April 2024).
  47. IGN; Instituto Geográfico Nacional. Centro de Descargas—CartoBase ANE. Available online: https://centrodedescargas.cnig.es/CentroDescargas/index.jsp# (accessed on 11 April 2024).
  48. Zinck, J.A. Geopedología: Elementos de geomorfología para estudios de suelos y de riesgos naturales: Enschede. In International Institute for Geo-Information Science and Earth Observation (ITC); ITC Special Lecture Notes Series; University of Twente: Enschede, The Netherlands, 2012. [Google Scholar]
  49. Bossard, M.; Feranec, J.; Otahel, J. CORINE Land Cover Technical Guide: Addendum 2000; European Environment Agency: Copenhagen, Denmark, 2000; Volume 40.
  50. Charman, P.V.; Murphy, B.W. Soils: Their Properties and Management, 2nd ed.; Oxford University Press: Melbourne, Australia; Oxford, UK, 2000. [Google Scholar]
  51. IGN; Instituto Geográfico Nacional. Centro de Descargas—CORINE Land Cover. Available online: https://qrcd.org/5bHm (accessed on 12 April 2024).
  52. IGN; Instituto Geográfico Nacional. Centro de Descargas—SIOSE AR. Available online: https://qrcd.org/5rAl# (accessed on 12 April 2024).
  53. Dahal, R.K.; Hasegawa, S.; Nonomura, A.; Yamanaka, M.; Dhakal, S.; Paudyal, P. Predictive modelling of rainfall-induced landslide hazard in the Lesser Himalaya of Nepal based on weights-of-evidence. Geomorphology 2008, 102, 496–510. [Google Scholar] [CrossRef]
  54. Vakhshoori, V.; Zare, M. Landslide susceptibility mapping by comparing weight of evidence, fuzzy logic, and frequency ratio methods. Geomat. Nat. Hazards Risk 2015, 7, 1731–1752. [Google Scholar] [CrossRef]
  55. Lin, L.; Lin, Q.; Wang, Y. Landslide susceptibility mapping on a global scale using the method of logistic regression. Nat. Hazards Earth Syst. Sci. 2017, 17, 1411–1424. [Google Scholar] [CrossRef]
  56. Zhao, Z.; He, Y.; Yao, S.; Yang, W.; Wang, W.; Zhang, L.; Sun, Q. A comparative study of different neural network models for landslide susceptibility mapping. Adv. Space Res. 2022, 70, 383–401. [Google Scholar] [CrossRef]
  57. Nadim, F.; Kjekstad, O.; Peduzzi, P.; Herold, C.; Jaedicke, C. Global landslide and avalanche hotspots. Landslides 2006, 3, 159–173. [Google Scholar] [CrossRef]
  58. SGC. Documento Metodológico de la Zonificación de Susceptibilidad y Amenaza por Movimientos en Masa Escala 1:100.000; Servicio Geológico Colombiano (SGC): Bogotá, Colombia, 2013. [Google Scholar] [CrossRef]
  59. CEDEX; Centro de Estudios y Experimentación de Obras Públicas. Ministerio de Transportes y Movilidad Sostenible. Available online: https://ceh.cedex.es/web/Imp_CClimatico_Pmax.htm (accessed on 12 April 2024).
  60. IGN. Instituto Geográfico Nacional. Mapas de Sismicidad y Peligrosidad. Available online: https://www.ign.es/web/ign/portal (accessed on 12 April 2024).
  61. Pachauri, A.; Pant, M. Landslide hazard mapping based on geological attributes. Eng. Geol. 1992, 32, 81–100. [Google Scholar] [CrossRef]
  62. Dai, F.C.; Lee, C.F. Landslide characteristics and slope instability modeling using GIS, Lantau Island, Hong Kong. Geomorphology 2002, 42, 213–228. [Google Scholar] [CrossRef]
  63. Santacana, N. Análisis de la susceptibilidad del terreno a la formación de deslizamientos superficiales y grandes deslizamientos mediante el uso de sistemas de información geográfica. Aplicación a la cuenca alta del río Llobregat. Ph.D. Thesis, UPC, Departament d’Enginyeria del Terreny, Cartogràfica i Geofísica, Barcelona, Spain, 2001. [Google Scholar]
  64. van Zuidam, R.A. Aerial Photo-Interpretation in Terrain Analysis and Geomorphologic Mapping; Smits Publishers: The Hague, The Netherlands, 1986. [Google Scholar]
  65. Felicisimo, A.M. Modelos Digitales del Terreno. Oviedo: Pentalfa. 1994. Available online: http://www.etsimo.uniovi.es/~feli (accessed on 11 November 2016).
  66. ESRI. GIS Dictionary. Available online: https://support.esri.com/en-us/gis-dictionary/geometric-interval-classification (accessed on 15 May 2024).
  67. IPCC. Climate Change 2013: The Physical Science Basis Contribution of Working Group I to the Fifth Assessment Report of the Intergovernmental Panel on Climate Change; Stocker, T.D.-K., Ed.; Cambridge University Press: New York, NY, USA; Cambridge, UK, 2013; pp. 91, 106, 184. [Google Scholar]
  68. Bathrellos, G.D.; Gaki-Papanastassiou, K.; Skilodimou, H.D.; Papanastassiou, D.; Chousianitis, K.G. Potential suitability for urban planning and industry development using natural hazard maps and geological–geomorphological parameters. Environ. Earth Sci. 2012, 66, 537–548. [Google Scholar] [CrossRef]
  69. Glade, T.; Crozier, M.; Smith, P. Applying Probability Determination to Refine Landslide-triggering Rainfall Thresholds Using an Empirical “Antecedent Daily Rainfall Model”. Pure Appl. Geophys. 2000, 157, 1059–1079. [Google Scholar] [CrossRef]
  70. Li, J.; Convertino, M. Inferring ecosystem networks as information flows. Sci. Rep. 2021, 11, 7094. [Google Scholar] [CrossRef] [PubMed]
  71. de Moel, H.; Jongman, B.; Kreibich, H.; Merz, B.; Penning-Rowsell, E.; Ward, P.J. Flood risk assessments at different spatial scales. Mitig. Adapt. Strat. Glob. Chang. 2015, 20, 865–890. [Google Scholar] [CrossRef] [PubMed]
  72. CRED; Centre for Research on the Epidemiology of Disasters. Disasters in Numbers 2022. Available online: https://www.cred.be/publications (accessed on 19 April 2024).
  73. Miteco. AdapteCCa.es- Visor de Escenarios de Cambio Climático. Ministerio para la Transición Ecológica y el Reto Demográfico—Miteco. Available online: https://escenarios.adaptecca.es/ (accessed on 23 June 2024).
  74. Aemet. Valores climatológicos normales. Agencia Estatal de Meteorología—Aemet. Available online: https://qrcd.org/5gqV (accessed on 23 June 2024).
  75. Dávid, L. Quarrying: An anthropogenic geomorphological approach. Acta Montan. Slovaca 2008, 13, 66–74. [Google Scholar]
  76. Szabó, J.; Dávid, L.; Lóczy, D. Anthropogenic Geomorphology: A Guide to Man-Made Landforms; Springer Science & Business Media: Dordrecht, The Netherlands, 2010. [Google Scholar] [CrossRef]
Remotesensing 16 02478 g001
Figure 1. Geographical location of the study area in the region of Pontevedra, Spain.
Figure 1. Geographical location of the study area in the region of Pontevedra, Spain.
Remotesensing 16 02478 g001
Remotesensing 16 02478 g002
Figure 2. (A) Map of geological units for the study region at a scale of 1:50,000, modified from sheets 151, 152, 184, and 185 of the Geological Institute of Spain [34,35,36,37]. (B) Map of geomorphological units for the study region, modified from Martínez-Graña [38].
Figure 2. (A) Map of geological units for the study region at a scale of 1:50,000, modified from sheets 151, 152, 184, and 185 of the Geological Institute of Spain [34,35,36,37]. (B) Map of geomorphological units for the study region, modified from Martínez-Graña [38].
Remotesensing 16 02478 g002
Remotesensing 16 02478 g003
Figure 3. Flow diagram for the calculation of mass movement susceptibility and hazard. DEM: digital elevation model, ROC: receiver operating characteristic.
Figure 3. Flow diagram for the calculation of mass movement susceptibility and hazard. DEM: digital elevation model, ROC: receiver operating characteristic.
Remotesensing 16 02478 g003
Remotesensing 16 02478 g004
Figure 4. Inventory of mass movements obtained from the photo interpretation process on satellite images contained in the Google Earth® (Pontevedra, Spain) platform. The red line corresponds to the study area.
Figure 4. Inventory of mass movements obtained from the photo interpretation process on satellite images contained in the Google Earth® (Pontevedra, Spain) platform. The red line corresponds to the study area.
Remotesensing 16 02478 g004
Remotesensing 16 02478 g005
Figure 5. (A) Distribution of the weights of evidence calculated for the lithological variable. (B) Distribution of the weights of evidence calculated for the distance to faults variable.
Figure 5. (A) Distribution of the weights of evidence calculated for the lithological variable. (B) Distribution of the weights of evidence calculated for the distance to faults variable.
Remotesensing 16 02478 g005
Remotesensing 16 02478 g006
Figure 6. Distribution of the weights of evidence calculated for the morphogenetic variable.
Figure 6. Distribution of the weights of evidence calculated for the morphogenetic variable.
Remotesensing 16 02478 g006
Remotesensing 16 02478 g007
Figure 7. (A) Distribution of the weights of evidence calculated for the slope variable. (B) Distribution of the weights of evidence calculated for the orientation variable. (C) Distribution of the weights of evidence calculated for the relative relief variable. (D) Distribution of the weights of evidence calculated for the rugosities variable.
Figure 7. (A) Distribution of the weights of evidence calculated for the slope variable. (B) Distribution of the weights of evidence calculated for the orientation variable. (C) Distribution of the weights of evidence calculated for the relative relief variable. (D) Distribution of the weights of evidence calculated for the rugosities variable.
Remotesensing 16 02478 g007
Remotesensing 16 02478 g008
Figure 8. (A) Distribution of the weights of evidence calculated for the current land cover status variable. (B) Distribution of the weights of evidence calculated for the land use variable.
Figure 8. (A) Distribution of the weights of evidence calculated for the current land cover status variable. (B) Distribution of the weights of evidence calculated for the land use variable.
Remotesensing 16 02478 g008
Remotesensing 16 02478 g009
Figure 9. Mass movement susceptibility map reclassified into 5 categories and obtained by means of the bivariate method.
Figure 9. Mass movement susceptibility map reclassified into 5 categories and obtained by means of the bivariate method.
Remotesensing 16 02478 g009
Remotesensing 16 02478 g010
Figure 10. ROC curve distribution, its AUC evaluation, and confusion matrix results.
Figure 10. ROC curve distribution, its AUC evaluation, and confusion matrix results.
Remotesensing 16 02478 g010
Remotesensing 16 02478 g011
Figure 11. (A) Maximum annual daily precipitation map, RCP 4.5 model, and a Tr of 100 years for Spain. Modified map from CEDEX [59]. (B) Seismic hazard map, PGA values in gravity acceleration for Spain. Modified map from IGN [60].
Figure 11. (A) Maximum annual daily precipitation map, RCP 4.5 model, and a Tr of 100 years for Spain. Modified map from CEDEX [59]. (B) Seismic hazard map, PGA values in gravity acceleration for Spain. Modified map from IGN [60].
Remotesensing 16 02478 g011
Remotesensing 16 02478 g012
Figure 12. Mass movement hazard map based on the susceptibility model and the triggers of rainfall and earthquake; this map is reclassified into 4 categories.
Figure 12. Mass movement hazard map based on the susceptibility model and the triggers of rainfall and earthquake; this map is reclassified into 4 categories.
Remotesensing 16 02478 g012
Table 1. Area distribution of mass movement susceptibility categories and their relationship to the mass movement inventory.
Table 1. Area distribution of mass movement susceptibility categories and their relationship to the mass movement inventory.
CategoryArea [m2]%Movements Area [m2]%
Very low20,548,89021.600
Low28,314,90429.800
Medium13,891,37514.6170.03
High14,413,41715.29661.76
Very high17,917,57318.853,94398.21
Total95,086,15910054,926100
Table 2. Ranking of rainfall and acceleration for the combination matrix with susceptibility values.
Table 2. Ranking of rainfall and acceleration for the combination matrix with susceptibility values.
Rainfall ClassificationAcceleration ClassificationSusceptibility
Range [mm]ClassificationRange [PGA]ClassificationCategoryClassification
−6.8–610.02–0.0641Very low1
6.1–1220.064–0.1082Low2
13–1830.108–0.1523Medium3
19–2440.152–0.1964High4
25–3050.196–0.245Very high5
Table 3. Distribution of the areas in hazardous condition due to mass movements for each of the municipalities present in the study area.
Table 3. Distribution of the areas in hazardous condition due to mass movements for each of the municipalities present in the study area.
MunicipalityArea [m2]Low%Medium%High%Very High%
A Illa de Arousa5,085,203299,5435.891,993,79439.211,572,72630.931,219,14023.97
Cambados15,375,6174,559,53029.657,815,24450.832,028,21013.19972,6336.33
Meaño6,057,9361,589,02526.233,139,28451.82823,27613.59506,3518.36
O Grove19,624,754338,7081.737,351,49437.463,868,74219.718,065,81041.10
Ribadumia644,234430,56166.83186,67928.9826,9944.1900.00
Sanxenxo34,537,5846,470,56018.7316,543,19847.905,224,00315.136,299,82318.24
Vilagarcía de Arousa751,403479,68263.84239,12831.8232,5934.3400.00
Vilanova de Arousa11,930,4376,283,56952.674,228,14335.44724,7396.07693,9865.82
Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods, instructions or products referred to in the content.

Share and Cite

MDPI and ACS Style

Valencia Ortiz, J.A.; Nieto, C.E.; Martínez-Graña, A.M. Evaluation of Mass Movement Hazard in the Shoreline of the Intertidal Complex of El Grove (Pontevedra, Galicia). Remote Sens. 2024, 16, 2478. https://doi.org/10.3390/rs16132478

AMA Style

Valencia Ortiz JA, Nieto CE, Martínez-Graña AM. Evaluation of Mass Movement Hazard in the Shoreline of the Intertidal Complex of El Grove (Pontevedra, Galicia). Remote Sensing. 2024; 16(13):2478. https://doi.org/10.3390/rs16132478

Chicago/Turabian Style

Valencia Ortiz, Joaquín Andrés, Carlos Enrique Nieto, and Antonio Miguel Martínez-Graña. 2024. "Evaluation of Mass Movement Hazard in the Shoreline of the Intertidal Complex of El Grove (Pontevedra, Galicia)" Remote Sensing 16, no. 13: 2478. https://doi.org/10.3390/rs16132478

APA Style

Valencia Ortiz, J. A., Nieto, C. E., & Martínez-Graña, A. M. (2024). Evaluation of Mass Movement Hazard in the Shoreline of the Intertidal Complex of El Grove (Pontevedra, Galicia). Remote Sensing, 16(13), 2478. https://doi.org/10.3390/rs16132478

Note that from the first issue of 2016, this journal uses article numbers instead of page numbers. See further details here.

Article Metrics

Back to TopTop