Home Machine learning methods for landslide mapping studies: A comparative study of SVM and RF algorithms in the Oued Aoulai watershed (Morocco)
Article Open Access

Machine learning methods for landslide mapping studies: A comparative study of SVM and RF algorithms in the Oued Aoulai watershed (Morocco)

  • Latifa Ladel ORCID logo , Mohamed Mastere ORCID logo , Shuraik Kader ORCID logo , Velibor Spalević ORCID logo and Branislav Dudic ORCID logo EMAIL logo
Published/Copyright: March 3, 2025
Download Article (PDF)
Become an author with De Gruyter Brill
Submit Manuscript Author Information
Open Geosciences
From the journal Open Geosciences Volume 17 Issue 1

Abstract

Effective management of watershed risks and landslides necessitates comprehensive landslide susceptibility mapping. Support vector machine (SVM) and random forest (RF) machine learning models were used to map the landslide susceptibility in Morocco’s Taounate Province. Detailed landslide inventory maps were generated based on aerial pictures, field research, and geotechnical survey reports. Factor correlation analysis carefully eliminated redundant factors from the original 14 landslide triggering factors. As a result, 30% of the sites were randomly chosen for testing, whereas 70% of the landslide locations were randomly picked for model training. The RF model achieved an area under the curve (AUC) of 94.7%, categorizing 30.07% of the region as low susceptibility, while the SVM model reached an AUC of 80.65%, indicating high sensitivity in 53.5% of the locations. These results provide crucial information for local authorities, supporting sound catchment planning and development strategies.

Keywords: management; landslide susceptibility; machine learning models; factor correlation analysis; area under the curve; Morocco

1 Introduction

Landslides are one of the world’s natural geohazards. They change the terrain and disrupt people’s lives and communities wherever they occur [1,2,3]. The severity of climate change, unexpected construction projects, and recent global economic expansion have all intensified the social repercussions of landslides [4,5]. Statistical analysis by the Disaster Epidemiology Research Center shows that more than 17% of the natural disaster deaths are caused by landslides and other mass movements [6,7]. In northern Morocco, landslides are among the most prevalent natural disasters following floods and droughts, which are regarded as the most serious danger to the country’s socioeconomic growth [8,9].

While traditional methods of landslide susceptibility mapping (LSM) have provided valuable insights, they often rely heavily on expert knowledge and can be limited by data availability. Recent studies employing machine learning (ML) techniques, such as random forest (RF) and support vector machine (SVM), have shown promise in improving the accuracy and handling complex datasets. However, there remains a lack of comparative analyses that assess the performance of these models in specific regions with diverse geological characteristics. This study aims to address this gap by evaluating the effectiveness of SVM and RF models in mapping landslide susceptibility in the Oued Aoulai watershed, highlighting their respective advantages and limitations.

The early warning system effectively mitigates the risk of catastrophic landslides, encompassing both spatial and temporal failures [10,11,12]. As a result, numerous landslide prediction approaches have gained prominence in recent years, focusing on LSM methods, among others [13,14]. (i) Deterministic methods, which include “expert knowledge-based methods” and geomorphologic mapping. These methods rely on the expertise of geologists, geomorphologists, and other experts to assess site vulnerability based on field observations and existing knowledge [15]. (ii) Statistical tools include logistic regression (LR), weight of evidence, and frequency ratio (FR) [16,17]. (iii) Artificial intelligence and ML include techniques such as decision tree (DT), SVM, and artificial neural network [18,19,20].

The models were validated using a separate dataset that included both historical landslide occurrences and environmental factors relevant to the study area. The RF model achieved an area under the curve (AUC) value of 94.7%, while the SVM model reached an AUC of 80.65%. This validation dataset was composed of 30% of the total landslide inventory, ensuring a robust evaluation of model performance.

ML methods offer a more effective approach to LSM compared to traditional expert opinion-based and analytical methodologies due to their ability to process large datasets and identify complex patterns [21,22,23]. Comprehending the mechanism of landslides is necessary in order to appropriately implement LSM. After that, one must look at the connection between causative variables and landslide occurrence [24].

Despite the growing use of ML models in LSM, there remains a lack of comparative studies that systematically evaluate the performance of different algorithms in specific geographic contexts, such as the Oued Aoulai watershed in Morocco. This study aims to fill this gap by not only comparing the efficacy of SVM and RF models but also by assessing their contributions to effective risk-reduction strategies based on local geological and environmental conditions.

This research aims to evaluate the use of prediction rate and FR models in determining the landslide risk of the Oued Aoulai watershed in Taounate Province, Morocco. The research objective is to assess the benefits and drawbacks of these models as well as their ability to provide efficient risk-reduction techniques. This study employs SVM and RF models due to their complementary strengths in handling LSM. SVM is selected for its effectiveness in high-dimensional spaces, where it can accurately classify and predict landslide occurrences by identifying optimal decision boundaries. Conversely, RF is chosen for its robustness and ability to aggregate multiple DTs, which enhances the predictive accuracy and reduces the likelihood of overfitting. Together, these models provide a comprehensive approach to analyzing landslide susceptibility in the Oued Aoulai watershed. There have been many landslides in the study area in the past. To the best of our knowledge, state-of-the-art technologies have been integrated into this research together with a thorough evaluation of the available scientific literature. The next step is to make systematic geographical links between certain characteristics and landslide episodes. Performance curves and matrices were used to evaluate the accuracy of RF and SVM models.

2 Methods

In this study, factors influencing landslide susceptibility were selected based on expert judgment and a thorough review of relevant literature, resulting in an initial list of 14 potential triggering factors. When they are selected one by one using the slope unit as the base unit, the RF and SVM models yield different ratios of the susceptibility map. A factor correlation analysis was conducted to identify multicollinearity among these factors, leading to the elimination of two redundant variables. Additionally, a sensitivity analysis was performed to assess the contribution of each factor to landslide occurrences, enabling us to prioritize the most significant variables for modeling. Figure 1 depicts the flowchart of this study, describing the essential steps involved in developing the landslide status factor, inventory, and comparison, in addition to applying different ratios and models. These methods ensured a robust selection of factors for the study area, enhancing the reliability of the susceptibility maps.

Figure 1 Flowchart of the research methodology.
Figure 1

Flowchart of the research methodology.

2.1 Study area

The Oued Aoulai catchment located in the Taounate Province is a sub-catchment of the Oued Ouergha catchment. It divides the Taounate Province into northern and southern parts and represents the case study for this research. In the northern part of Taounate, between latitudes 34°48′39.60″N and 34°33″16.77″N and longitudes 4°56′20.63E̅ and 4°58′11.20E̅, the Oued Aoulai catchment is located in a region (Figure 2) with an area of 370 km2, a length of 27 km, and a width of 19 km and controlled by the Ghafsay and Ratba weather stations. It is home to a diverse population around 91,900, whose livelihoods depend heavily on agriculture and natural resources. The region has experienced numerous landslides, which pose significant risks to local communities by damaging the infrastructure, disrupting the transportation networks, and affecting the agricultural productivity [25,26]. Previous research studies in the region indicate that landslides have not only led to economic losses but also threatened the safety and well-being of residents [27,28]. Given this context, effective LSM becomes crucial for informing risk management strategies and enhancing community resilience.

Figure 2 Geographic setting of the research area.
Figure 2

Geographic setting of the research area.

Part of the western Mediterranean’s Alpine chain is the Rif chain, which is located in northern Morocco and is distinguished by its intricate geology. Its present structure was determined by the movement of internal zones toward the WSW and by the African and European NS convergence during the Neogene [29,30]. It consists of three major structural areas, such as inner areas, flysch zones, and outer areas, from which ancient furrows and folds are derived from north to south.

The research region is a part of the Mesorifan zone in the middle Rif area, which is made up of Neogene basins (Rhafsai, Tafrant, Taounate Bouhadi, and Dhar Souk), all of which have a synclinal structure framed by NE–SW, EW, NW accidents–SE [31]. The whole outline is an arc with a concavity-directed NE. The Aoulai watershed is part of the Tafrant–Rhafsai synclinal basin formed from Triassic to Quaternary rocks.

The facies are composed of Jurassic limestone, marl limestone, marl, thick Cretaceous marl, and marl limestone; the Tertiary cover began in the Eocene (white marl containing chert rock), then the Lower Miocene (alternating marl and sandy limestone) and the Upper Miocene, consisting of blue marl and marl sandstone. Quaternary is prominent and widespread in terraces or dispersed parts of the basin.

2.2 Application of various ratios and models

2.2.1 RF model

Several independent DTs that are all capable of producing results make up the RF model [32]. Based on DTs, the RF model is a categorization approach [33]. Data scientists frequently utilize the supervised learning method RF. Every DT starts with a randomly chosen sample, uses it for training, and then replaces it with the original data [34]. The RF model consists of multiple DTs that are trained independently using a method known as bootstrap aggregating or bagging. In this process, each tree is constructed using a random subset of the training data, generated by sampling with replacement. This means that some observations may be used multiple times in a single tree, while others may not be included at all. By averaging the predictions from all the DTs, the RF model reduces the variance associated with individual trees, thus minimizing the risk of overfitting to the training data. This ensemble approach enhances the model’s generalization capability on unseen data. It is a reliable model that combines the outputs of several DTs into a single result and may be applied to regression and classification applications. Since the RF selection of each split node depends on two data objects – out-of-bag (OOB) and neighbor – RF is known for its high accuracy in predicting outliers [35]. When these DTs are combined, the result is the RF. In terms of outcome, this is how the RF performs better than the DT model (equation (1)).

(1) H ( X ) = a v k max ∑ i = 1 k I ( h i ( X ) = Y ) ,

where a v k shows the average, and the indicator function is denoted by I.

The RF model has many advantages, such as (i) determining correlations between variables without generating any central hypotheses [36], (ii) being a better ML program for assessing hierarchical situations in big data [36], and (iii) ranking the independent variables according to importance [37]. This approach also has the advantage of meticulously eliminating the problem of overfitting the data by (i) building multiple trees, (ii) bootstrapping observations, and (iii) dividing the nodes into optimal divisions within a randomized subset [38].

2.2.2 SVM model

The SVM family of automatic learning algorithms addresses issues with anomaly detection, regression, and classification. Vapnik was the one who first suggested the SVM for regression [39]. In a high-dimensional or infinite space, the SVM generates a single hyperplane or a group of hyperplanes. It is applicable to both regression and classification.

Their foundation lies in the quest for the ideal marginal hypersurface that, to the greatest extent feasible, distinguishes or categorizes the data while remaining as remote from every observation as feasible. The idea is to identify a discriminative function or classifier with the largest potential generalization capacity that is known as predictive quality [20].

Frequently, SVMs depend on “kernels” to get around this. By projecting the data into a feature space – a vector space of higher dimension – or by bringing them back to the two-dimensional space, these mathematical functions enable the separation of the data [40]. The nonlinear decision boundary was explained by a kernel function K (equation (2)) [39].

(2) K ( x i → , x j → ) = Ø ( x i → ) T Ø ( x j → ) .

The function ∅ was used to map the training vector ( x i → ) into a high-dimensional space. To deal with nonlinear problems, three kernel functions were regularly introduced. This study used three types of kernel functions listed below: the linear kernel function (equation (3)), radial basis kernel function (equation (4)), and polynomial kernel function (equation (5)). Using geographic information system (GIS) methods, the environmental parameters were determined.

(3) K ( x i → , x j → ) = x i → T × x j → ,

(4) K ( x → i , x → j ) = exp { − / δ ² ‖ x i → − x j → ‖ ² } ,

where δ ² is the radial basis function’s bandwidth.

(5) K ( x i → , x j → ) = ( γ X i T + r ) d ,

when γ > 0, the polynomial kernel has degree d.

2.3 Landslide inventory, risk variables, and data sources

For this investigation, the following were the main sources of data: (1) published survey data and landslide reports, (2) Landsat 8 remote sensing images, (3) a 15 m resolution digital elevation model (DEM), and (4) topographic and geology maps at a scale of 1:50,000.

2.3.1 Landslide inventory

Using a geology inventory to find and classify landslides is a helpful method for LSM. Precise mapping of landslide susceptibility was established by gathering pertinent reports and photos, as well as by carrying out a survey at a 1:50,000 scale. There have been numerous recorded landslides in the area, including rock blades, debris flows, and debris avalanches. In the present study, the landslide dataset comprises a total of 200 documented landslide events within the Oued Aoulai watershed, with 140 landslides used for model training and 60 for validation. The dataset includes various types of landslides, such as debris flows, rockfalls, and translational slides, which are prevalent in the region. Key characteristics of the landslides include their geographical distribution, with occurrences primarily concentrated in steep slopes and areas with specific geological formations. This diverse dataset allows for a comprehensive evaluation of the models’ performance and enhances the reliability of the susceptibility mapping. This study used the fieldwork, data from satellite image analysis, and historical records to create a map of active landslides. In this way, a list of almost 1,000 landslides in the research region was generated.

2.3.2 Landslide causative factors

The factors considered for LSM in the Oued Aoulai watershed were briefly outlined by using in-depth literature review, brainstorming, and discussion with experts. The data sources for these factors include satellite imagery for land cover classification, topographic maps for slope analysis, and geological surveys for understanding subsurface conditions. The factors were processed using GIS software, which facilitated spatial analysis and integration of various datasets, alongside statistical analysis tools for correlation and sensitivity assessments. This streamlined approach ensures that the factors are relevant and effectively contribute to the susceptibility mapping process.

The features of the research area guided the subjective selection of landslide-causing elements [41]. The topographic position index (TPI), topographic wetness index (TWI), lithology, distance from faults, slope gradient, slope aspect, soil texture, curvature, precipitation, land use and land cover (LULC), distance from highways, and normalized difference vegetation index (NDVI) were among the 14 landslide-causing components that were investigated in this study. The selection of these factors was based on expert judgment, a review of the literature, and the availability of data at the appropriate scale by referring to successful research outcomes from the published literature such as Sachdeva et al. and Bashir et al. [42,43]. Slope inclination – landslides occur more frequently on hillside slopes that are steeper [44]. Since the slope gradient affects the soil moisture content and is required for subsurface flow [45], this has a direct bearing on how frequently landslides occur.

NDVI is recognized as an important variable in landslide susceptibility modeling, even though it significantly enhances the cohesiveness and shear strength of lithologic mass and immobilizes vast amounts of water [44,46]. Using the January 2023 images, equation (6) was applied to determine the present research region’s NDVI value.

(6) NDVI = ( NIR − R ) / ( NIR + R ) ,

where the electromagnetic spectrum’s red band is denoted by R and the near-infrared band is represented by NIR.

Elevation is a topographical feature that affects how unstable slopes are. Almost all landslide susceptibility evaluations employ it often [47]. In terms of LULC, landslides are mostly caused in the study area by anthropogenic activities such as infrastructure development and urbanization. Roads and structures must be constructed because of these activities, which may alter the stability of the slope and the original geological conditions [48,49]. Earlier models included data on land use and cover [50,51,52] to assess vulnerability to natural disasters such as avalanches, flooding, wildfires, and landslides.

Soil texture describes each soil profile’s materials and physical properties, with soil texture data taken from laboratory analyses. Tectonic deformations that might result in landslides are defined as weak spots in the rock, characterized by a drop in resistivity and can be brittle (faults, shears, etc.) or ductile folds [53,54,55]. Thus, the elimination of faults may serve as a warning indicator for landslides. The runoff from rivers contributes significantly to undercutting phenomena that increase the pore water pressure in the vicinity and trigger landslides [56]. As such, it has a major impact on the susceptibility to landslides [57].

One essential terrain variable used in many different kinds of geomorphometric research is curvature [58]. There are two types of curvatures: plane curvature and profile curvature. It is well established that the profile curvature controls these materials’ downslope acceleration and deceleration to affect material deposition, while surface runoff’s convergence and dispersion are directly impacted by plane curvature [59].

The TPI is the difference between a single cell’s height and the average height of its neighbors [60,61]. The TWI is one of the characteristics that demonstrates the hydrological process linked to the build-up of water flow based on the regulation of the slope factor in a region [62]. Slope affects the hydrological cycle in observable ways in locations with high relief.

In terms of Aspect, the slope gradient can affect landslides both directly and indirectly in several processes, including the orientation of discontinuities, wind and precipitation patterns, soil moisture concentration, root development evapotranspiration, hydrological processes, vegetation, and solar radiation [63].

One of the significant anthropogenic activities that alters the morphology of natural slopes is the road building. This changes how close the slopes are to roads and forecasts the intensity of the landslides [64]. Precipitation data can be used to calculate the amount of water that accumulates and causes landslides, as well as the movement of materials and soil [65]. Lithology is important to LMS and influences various types of landslides. Numerous mechanical and physical properties, including the kind, degree of weathering resistance, density, durability, and permeability, are present in these units [66].

3 Results and discussion

3.1 Factors contributing to landslides

Building upon established techniques in the field, such as those demonstrated by [61] in their LR model case study, our approach integrates multiple ML models, including SVMs and RF. These methodologies are supported by recent findings in the literature [62], which highlight the efficacy of ML techniques in accurately predicting landslide susceptibility based on environmental factors.

Slope inclination: The slope map (Figure 3a) for this inquiry was created using the DEM and ranged from 0 to 55. The map was constructed at a resolution of 12 m.

Figure 3 Maps of factors influencing landslide occurrence: DT faults, (a) slope, (b) NDVI, (c) elevation, (d) LULC, (e) soil type, and (f) distance from faults.
Figure 3

Maps of factors influencing landslide occurrence: DT faults, (a) slope, (b) NDVI, (c) elevation, (d) LULC, (e) soil type, and (f) distance from faults.

NDVI: The current research region’s NDVI values range from −0.13 to 0.67, where a positive number implies vegetated regions and a negative one denotes the barren land (Figure 3b).

Elevation: The elevation map in this study was made in accordance with the DEM’s classification. It was situated above the sea level at a height of 198–1,583 m (Figure 3c).

LULC: The land use map used in this study was derived from the “Sentinel-2 10-Meter LULC” dataset. Six distinct land use types were identified in the research region: trees, water, crops, cultivated land, bare soil, and grassland (Figure 3d).

Soil type: The research region had an average soil composition of 18% sand, 32% silt, and 50% clay. Figure 3e illustrates the distribution of soil texture in the research region.

Distance from faults: Using GIS software, Euclidean Distance tool, a map was created, and the distance to faults was plotted. The 1:50,000 scale geology maps of the research area provided the faults employed in this investigation (Figure 3f). There could be anywhere from 0 to 400 m between faults.

Distance from rivers: The hydrographic network for this study was created using a DEM with a resolution of 12 m. The Euclidean distance approach in GIS software was used to derive the river’s distance, which spans from 0 to 500 m.

Curvature: The plane and profile curvatures were derived from the DEM using GIS software. In the study area, the profile curvature exhibits a convex, flat shape, with values ranging from −4.43 to 4.66. Convex curvatures indicate a decelerating flow, while concave curvatures indicate an accelerating flow. The area’s plane curvature often depicts the flow’s convergence and divergence along a surface since it is perpendicular to the direction of the sharpest gradient. The three plane curvatures of the research region are concave, convex, and flat, and they range from −4.39 to 4.64 (Figure 4a).

Figure 4 Maps of the landslide conditioning factor: (a) curvature, (b) TPI, (c) TWI, (d) Distance from river, (e) aspect, and (f) DT roads.
Figure 4

Maps of the landslide conditioning factor: (a) curvature, (b) TPI, (c) TWI, (d) Distance from river, (e) aspect, and (f) DT roads.

TPI: Topographic landforms like valleys, slopes, and ridges can be identified by applying particular thresholds to the TPI values. Because hills are the source of most landslide ruts, the TPI may be utilized to produce a landslide susceptibility map. The TPI for the study area spans from −25.95 to 32.01 (Figure 4b).

TWI: The terrain wetness index highlights the significant influence of topography and soil moisture content on landslide probability. The research area’s TWI (Figure 4c) varies from 2.41 to 21.6.

Aspect: The slope gradient in the research area was utilized to generate nine categories using the DEM: flat (1), north, northeast, east, southeast, south, southeast, west, and northeast (Figure 4e).

Distance from roads: The elimination of highways could be a sign that landslides are about to happen. In the current experiment, the range of highway distances (0–500 m) was generated using the Euclidean distance tool within the GIS software (Figure 4f).

Rainfall: The rainfall condition map is illustrated in Figure 5a, and Figure 5b shows the geographical distribution of average precipitation data over 10 years.

Figure 5 Landslide conditioning factor maps: (a) rainfall and (b) lithology.
Figure 5

Landslide conditioning factor maps: (a) rainfall and (b) lithology.

Lithology: The Moroccan Ministry of Mines and Geology provided two geological maps (scale, 1/50,000) from which the lithology used in this study was derived; Rhafsay-kelaa des sles and Tafrannt de l’Ouerra-Moulay Bouchta (Figure 5b).

3.2 Multicollinearity analysis

Determining the multicollinearity of land cover factors (LCFs) is essential in any multivariable landslide simulation. The extent to which independent variables are interdependent can significantly impact a model’s overall accuracy. Excessive multicollinearity, for example, may reduce the predictive power of the model, emphasizing the need for rigorous assessment to ensure the stability and reliability of simulation results [67,68].

The 14 components’ multicollinearity was tested using variance inflation factors (VIFs) and tolerances. Multicollinearity is defined as a tolerance of less than 0.2 or a VIF of 5.0 or more [63,64]. These thresholds are commonly used in statistical analysis to identify potential multicollinearity issues among predictor variables. A tolerance value below 0.2 indicates that the variable shares a significant amount of variance with other predictors, which can lead to unreliable coefficient estimates and inflated standard errors. Similarly, a VIF exceeding 5.0 suggests that the variable is highly correlated with one or more other predictors, warranting further investigation to ensure the stability and reliability of the model. Investigation of the conditioning variables indicates that every landslide conditioning variable utilized in the LSM for this case study is trustworthy and safe to apply (Table 1).

Table 1

Multicollinearity diagnosis results

Variable VIF Tolerance
Curvature 2.147012 0.46576
DT rivers 2.095214 0.47728
DT roads 1.816736 0.55044
Slope 1.47478 0.67807
DT faults 1.461652 0.68416
Aspect 1.461243 0.68435
Rainfall 1.440201 0.69435
TWI 1.31355 0.7613
TPI 1.280253 0.7811
Soil type 1.240115 0.80638
Lithology 1.213983 0.82373
DEM 1.181745 0.84621
NDVI 1.161018 0.86131
LULC 1.146891 0.87192

The result demonstrates that the lowest TOL value is somewhat greater. Elevation reaches the maximum VIF of 2.147012 and the lowest TOL of 0.46576, indicating that the maximum VIF is likewise far below the cutoff level.

3.3 Pearson correlation coefficient

The Pearson correlation coefficient, abbreviated “r,” is a popular statistical metric for determining the direction and strength of a linear relationship between two variables. Equation (7) defines how to calculate this coefficient and gives a numerical representation of the degree of linear relationship between the two parameters:

(7) ρ x , y = Cov ( x , y ) σ x σ y = [ ( x − μ x ) ( y − μ y ) ] σ x σ y ,

where

  1. Cov(x,y) represents the covariance of x and y between two elements; σ x and σ y represent the standard deviations of x and y, respectively, and μ x and μ y represent the means of x and y, respectively.

Equation (3) depicts the mathematical expectation. A Pearson correlation coefficient absolute value greater than 0.5 often indicates a significant link between the two parameters [68,69]. The correlation coefficient matrix of this study reveals no strong correlations among the 14 conditioning factors, indicating that no conditioning factors need to be eliminated (Figure 6).

Figure 6 Matrix with Pearson correlation coefficients for various elements.
Figure 6

Matrix with Pearson correlation coefficients for various elements.

3.4 Mapping the susceptibility of landslides

Figure 7 compares two landslide susceptibility maps created with RF and SVM models. From extremely low to extremely high susceptibility classifications, five were identified (Figure 8). Specific vulnerable zones were identified using the outcomes of each landslide susceptibility model. The proportional percentages of the area that each vulnerability class was assigned in the models were carefully selected to give thorough information and enhance the comparison analysis’s granularity.

Figure 7 Generated landslide susceptibility maps using the SVM and RF.
Figure 7

Generated landslide susceptibility maps using the SVM and RF.

Figure 8 Area % for landslide susceptibility classes using the applied MLTs.
Figure 8

Area % for landslide susceptibility classes using the applied MLTs.

3.5 Model performance and validation

The current study used a variety of statistical and graphical performance measures to assess the prediction performance of the SVM and RF models. For classification problems, the confusion matrix was frequently utilized. False positive (FP), false negative (TN), true positive (TP), and false negative (FN) values were used to identify counts of genuine and predicted values. TP stands for the number of properly categorized genuine landslide pixels, TN for accurately classified non-landslide pixels, FP for accurately classified non-landslide pixels, and FN for accurately classified actual landslide pixels. The following calculations were performed: accuracy, precision, recall, confusion matrix, mean absolute error (MAE), and root mean square error (RMSE).

AUC values from receiver operating characteristic (ROC) curves are often used to calculate the classification performance. The AUC is particularly valuable in the context of imbalanced data, as it provides a single metric that captures the model’s ability to differentiate between classes, regardless of their distribution. This is crucial in landslide susceptibility modeling, where the number of susceptible versus non-susceptible locations may not be evenly distributed. A higher AUC indicates better model performance across all classification thresholds [70,71], making it a preferred metric for evaluating the efficacy of our models. The greater the AUC score, the better the model forecasts landslides and non-landslides. The AUC prediction curve was generated using the landslide inventory training dataset, while the AUC validation curve and performance matrices were created using the validation dataset.

Figure 9 displays the training accuracies of the SE-FR and SE-SVM models, which were 94.7 and 80.65 AUC, respectively. According to the confusion matrices, the SE-FR and SE-SVM models have 0.946 and 0.906 accuracy, 0.904 and 0.863 precision, and 0.982 and 0.902 recall, respectively. Table 2 displays the MAE and RMSE values for the SE-FR and SE-SVM models, which were 0.053 and 0.193, 0.230 and 0.439, respectively. According to the studies of Youssef et al., Aguirre-Gutiérrez et al., and Sestras et al. [8,72,73], evaluating a model’s performance using a single metric, such as AUC, is not necessarily meaningful, because high AUC does not always imply a high degree of accuracy in spatial predictions.

Figure 9 Implemented model’s stacked ROC curves.
Figure 9

Implemented model’s stacked ROC curves.

Table 2

Models’ RMSE and AUC

Performance indicators Precision Accuracy Recall MAE RMSE AUC
Model RF 0.904 0.946 0.982 0.053 0.230 94.7
SVM 0.863 0.906 0.902 0.193 0.439 80.65

Both models exhibit good prediction, recall, and accuracy scores, and their respective MAE and RMSE errors fall within acceptable bounds, according to the prediction matrix results. These statistical and visual performance measures show that both models have respectable error values and a strong potential for prediction. As a result, it was discovered that RF performed better than the SVM model when utilizing the AUC technique. However, when the RMSE technique was used, the SVM fared better than the RF model.

4 Conclusions

LSM is a crucial geomatic technique for effective risk assessment and watershed management because different regions of Morocco are always at risk due to factors such as lithology, climate, geology, land use/cover, vegetation, anthropogenic interventions, and other related factors. The current study employed two ML models to evaluate the specified region in terms of landslide scenarios: SVM and RF. This was done since creating viable and realistic susceptibility maps is essential for preventive actions and treatments. According to the thorough investigation and analytical approach, the RF model classified 30.07% of the research region as low susceptible to landslides, 8.88% as moderately susceptible, and 61.06% as extremely susceptible. On the other hand, the SVM model determined that 35.41% of the region was susceptible to landslides at a low level, 11.08% at a moderate level, and 53.5% at a high level. The performance of the models was tested using the AUC and RMSE measurements, and the results were favorable. The RF model had an AUC of 94.7%, while the SVM model had an AUC of 80.65%. In the end, the RF and SVM models provide strong and trustworthy findings for mapping the susceptibility of landslides. Since the resulting hazard mapping scenarios give crucial information for informed watershed planning and development strategies, they are of great value to local authorities.

,

  1. Funding information: No funding was received for this research.

  2. Author contributions: Latifa Ladel and Mohamed Mastere – conceived and designed the studies; Latifa Ladel, Mohamed Mastere, Shuraik Kader, Velibor Spalević, and Branislav Dudic – performed the analysis; Latifa Ladel, Mohamed Mastere, Shuraik Kader, Velibor Spalević, and Branislav Dudic – analyzed and interpreted the data; Latifa Ladel, Mohamed Mastere, Shuraik Kader, Velibor Spalević, and Branislav Dudic – contributed materials, analysis tools, or data; Latifa Ladel, Mohamed Mastere, Shuraik Kader, Velibor Spalević, and Branislav Dudic – preparation of draft; Mohamed Mastere, Shuraik Kader, and Velibor Spalević – internal reviewers; Mohamed Mastere – project administration. All authors have read and agreed to the submitted version of the manuscript.

  3. Conflict of interest: The authors have no relevant financial or non-financial interests to disclose.

  4. Ethical approval: Not Applicable.

  5. Consent to participate: Not Applicable.

  6. Consent to publish: Not Applicable.

  7. Data availability statement: The data and materials will be available on request.

References

[1] Benzougagh B, Meshram SG, Baamar B, Dridri A, Boudad L, Sadkaoui D, et al. Relationship between landslide and morpho-structural analysis: A case study in Northeast of Morocco. Appl Water Sci. 2020;10(7):1–10. 10.1007/s13201-020-01258-4.Search in Google Scholar

[2] Fan X, Scaringi G, Korup O, West AJ, van Westen CJ, Tanyas H, et al. Earthquake‐induced chains of geologic hazards: Patterns, mechanisms, and impacts. Rev Geophysics. 2019;57(2):421–503. 10.1029/2018RG000626.Search in Google Scholar

[3] Zejak D, Spalević V, Popović V, Markoski M, Dudić B, Ouallali A, et al. Analysis of the presence of heavy metals in the soils of the hilly-mountainous areas of Balkan Peninsula with the assessment of its potential for the fruit growing: Case study of the Ljubovidja river basin, Polimlje, Montenegro. In Proceedings, 26 International Eco-Conference and 12 Safe Food, Novi Sad, 21-23 September 2022. Novi Sad: Ecological Movement of Novi Sad; 2022.Search in Google Scholar

[4] Guzzetti F, Mondini AC, Cardinali M, Fiorucci F, Santangelo M, Chang KT. Landslide inventory maps: New tools for an old problem. Earth-Sci Rev. 2012;112(1):42–66. 10.1016/j.earscirev.2012.02.001.Search in Google Scholar

[5] Li Y, Liu X, Han Z, Dou J. Spatial proximity-based geographically weighted regression model for landslide susceptibility assessment: a case study of Qingchuan area, China. Appl Sci. 2020;10(3):1107. 10.3390/app10031107.Search in Google Scholar

[6] Bartholdson S, von Schreeb J. Natural disasters and injuries: What does a surgeon need to know? Curr Trauma Rep. 2018;4:103–8. 10.1007/s40719-018-0125-3.Search in Google Scholar PubMed PubMed Central

[7] Sadkaoui D, Brahim B, Kader S, Agharroud K, Mihraje AI, Aluni K, et al. Evaluation of tectonic activity using morphometric indices: Study of the case of Taïliloute ridge (middle-Atlas region, Morocco). J Afr Earth Sci. 2024;213:105219. 10.1016/j.jafrearsci.2024.105219.Search in Google Scholar

[8] Youssef B, Bouskri I, Brahim B, Kader S, Brahim I, Abdelkrim B, et al. The contribution of the frequency ratio model and the prediction rate for the analysis of landslide risk in the Tizi N’tichka area on the national road (RN9) linking Marrakech and Ouarzazate. Catena. 2023;232:107464. 10.1016/j.catena.2023.107464.Search in Google Scholar

[9] Ouallali A, Kader S, Bammou Y, Aqnouy M, Courba S, Beroho M, et al. Assessment of the erosion and outflow intensity in the rif region under different land use and land cover scenarios. Land. 2024;13:141. 10.3390/land13020141.Search in Google Scholar

[10] Casagli N, Intrieri E, Carlà T, Di Traglia F, Frodella W, Gigli G, et al. Monitoring and early warning systems: Applications and perspectives. Understanding and reducing landslide disaster risk: Volume 3 Monitoring and Early Warning 5th. United Kingdom: Elsevier; 2021. p. 1–21. 10.1007/978-3-030-60311-3_1.Search in Google Scholar

[11] Bammou Y, Benzougagh B, Abdessalam O, Brahim I, Kader S, Spalevic V, et al. Machine learning models for gully erosion susceptibility assessment in the Tensift catchment, Haouz Plain, Morocco for sustainable development. J Afr Earth Sci. 2024;213:105229. 10.1016/j.jafrearsci.2024.105229.Search in Google Scholar

[12] Bammou Y, Benzougagh B, Igmoullan B, Ouallali A, Kader S, Spalevic V, et al. Optimizing flood susceptibility assessment in semi-arid regions using ensemble algorithms: A case study of Moroccan High Atlas. Nat Hazards. 2024;120:7787–816. 10.1007/s11069-024-06550-z.Search in Google Scholar

[13] Corominas J, van Westen C, Frattini P, Cascini L, Malet JP, Fotopoulou S, et al. Recommendations for the quantitative analysis of landslide risk. Bull Eng Geol Environ. 2014;73:209–63. 10.1007/s10064-013-0538-8.Search in Google Scholar

[14] Nwazelibe VE, Unigwe CO, Egbueri JC. Testing the performances of different fuzzy overlay methods in GIS-based landslide susceptibility mapping of Udi Province, SE Nigeria. Catena. 2023;220:106654. 10.1016/j.catena.2022.106654.Search in Google Scholar

[15] Aghdam IN, Varzandeh MHM, Pradhan B. Landslide susceptibility mapping using an ensemble statistical index (Wi) and adaptive neuro-fuzzy inference system (ANFIS) model at Alborz Mountains (Iran). Environ Earth Sci. 2016;75:1–20. 10.1007/s12665-015-5233-6.Search in Google Scholar

[16] Wang L-J, Sawada K, Moriguchi S. Landslide susceptibility analysis with logistic regression model based on FCM sampling strategy. Computers Geosci. 2013;57:81–92. 10.1016/j.cageo.2013.04.006.Search in Google Scholar

[17] Bashir O, Bangroo SA, Shafai SS, Senesi N, Naikoo NB, Kader S, et al. Unlocking the potential of soil potassium: Geostatistical approaches for understanding spatial variations in Northwestern Himalayas. Ecol Inform. 2024;81:102592. 10.1016/j.ecoinf.2024.102592.Search in Google Scholar

[18] San BT. An evaluation of SVM using polygon-based random sampling in landslide susceptibility mapping: The Candir catchment area (western Antalya, Turkey). Int J Appl Earth Geoinf. 2014;26:399–412. 10.1016/j.jag.2013.09.010.Search in Google Scholar

[19] Chang Z, Du Z, Zhang F, Huang F, Chen J, Li W, et al. Landslide susceptibility prediction based on remote sensing images and GIS: Comparisons of supervised and unsupervised machine learning models. Remote Sens. 2020;12(3):502. 10.3390/rs12030502.Search in Google Scholar

[20] Benzougagh B, Meshram SG, Fellah BE, Mastere M, El Basri M, Ouchen I, et al. Mapping of land degradation using spectral angle mapper approach (SAM): the case of Inaouene watershed (Northeast Morocco). Modeling Earth Syst Environ. 2023;10:1–11. 10.1007/s40808-023-01711-8.Search in Google Scholar

[21] Pourghasemi HR, Rahmati O. Prediction of the landslide susceptibility: Which algorithm, which precision? Catena. 2018;162:177–92. 10.1016/j.catena.2017.11.022.Search in Google Scholar

[22] Hu X, Wu S, Zhang G, Zheng W, Liu C, He C, et al. Landslide displacement prediction using kinematics-based random forests method: A case study in Jinping Reservoir Area, China. Eng Geol. 2021;283:105975. 10.1016/j.enggeo.2020.105975.Search in Google Scholar

[23] Bashir O, Bangroo SA, Shafai SS, Senesi N, Kader S, Alamri S. Geostatistical modeling approach for studying total soil nitrogen and phosphorus under various land uses of North-Western Himalayas. Ecol Inform. 2024;80:102520. 10.1016/j.ecoinf.2024.102520.Search in Google Scholar

[24] Guo C, Montgomery DR, Zhang Y, Wang K, Yang Z. Quantitative assessment of landslide susceptibility along the Xianshuihe fault zone, Tibetan Plateau, China. Geomorphology. 2015;248:93–110. 10.1016/j.geomorph.2015.07.012.Search in Google Scholar

[25] Loudyi D, Hasnaoui MD, Fekri A. Flood risk management practices in Morocco: Facts and challenges. Wadi Flash Floods. 2022;35:35–94.10.1007/978-981-16-2904-4_2Search in Google Scholar

[26] Sahrane R, El Kharim Y, Bounab A. Investigating the effects of landscape characteristics on landslide susceptibility and frequency-area distributions: The case of Taounate province, Northern Morocco. Geocarto Int. 2022;37(27):17686–712. 10.1080/10106049.2022.2134462.Search in Google Scholar

[27] Sahrane R, Bounab A, El Kharim Y. Investigating the effects of landslides inventory completeness on susceptibility mapping and frequency-area distributions: Case of Taounate province, Northern Morocco. Catena. 2023;220:106737. 10.1016/j.catena.2022.106737.Search in Google Scholar

[28] Sahrane R, Bounab A, Obda I, Obda O, El Hamdouni R, EL Kharim Y. Assessing the reliability of landslides susceptibility models with limited data: Impact of geomorphological diversity and technique selection on model performance in Taounate Province, Northern Morocco. Earth Syst Environ. 2024;9:1–25. 10.1007/s41748-024-00455-4.Search in Google Scholar

[29] Leblanc D, Olivier P. Role of strike-slip faults in the Betic-Rifian orogeny. Tectonophysics. 1984;101(3):345–55. 10.1016/0040-1951(84)90120-3.Search in Google Scholar

[30] Bahrouni N, Masson F, Meghraoui M, Saleh M, Maamri R, Dhaha F, et al. Active tectonics and GPS data analysis of the Maghrebian thrust belt and Africa-Eurasia plate convergence in Tunisia. Tectonophysics. 2020;785:228440. 10.1016/j.tecto.2020.228440.Search in Google Scholar

[31] Ait Brahim L, Chotin P. Genèse et déformation des bassins néogènes du Rif central (Maroc) au cours du rapprochement Europe-Afrique. Geodinamica Acta. 1989;3(4):295–304. 10.1080/09853111.1989.11105194.Search in Google Scholar

[32] Liu R, Li L, Pirasteh S, Lai Z, Yang X, Shahabi H. The performance quality of LR, SVM, and RF for earthquake-induced landslides susceptibility mapping incorporating remote sensing imagery. Arab J Geosci. 2021;14:1–15. 10.1007/s12517-021-06573-x.Search in Google Scholar

[33] Youssef AM, Pourghasemi HR, Pourtaghi ZS, Al-Katheeri MM. Landslide susceptibility mapping using random forest, boosted regression tree, classification and regression tree, and general linear models and comparison of their performance at Wadi Tayyah Basin, Asir Region, Saudi Arabia. Landslides. 2016;13:839–56. 10.1007/s10346-015-0614-1.Search in Google Scholar

[34] Xu K, Zhao Z, Chen W, Ma J, Liu F, Zhang Y, et al. Comparative study on landslide susceptibility mapping based on different ratios of training samples and testing samples by using RF and FR-RF models. Nat Hazards Res. 2023;20:273–97. 10.1016/j.nhres.2023.07.004.Search in Google Scholar

[35] Breiman L. Random forests. Mach Learn. 2001;45:5–32. 10.1023/A:1010933404324.Search in Google Scholar

[36] Darminto M, Chu H-J. Mapping landslide release area using random forest model. In IOP Conference Series: Earth and Environmental Science. United Kingdom: Elsevier; 2019. 10.1088/1755-1315/389/1/012038.Search in Google Scholar

[37] Youssef AM, Pradhan B, Tarabees E. Integrated evaluation of urban development suitability based on remote sensing and GIS techniques: contribution from the analytic hierarchy process. Arab J Geosci. 2011;4:463–73. 10.1007/s12517-009-0118-1.Search in Google Scholar

[38] Merghadi A, Yunus AP, Dou J, Whiteley J, ThaiPham B, Bui DT, et al. Machine learning methods for landslide susceptibility studies: A comparative overview of algorithm performance. Earth-Sci Rev. 2020;207:103225. 10.1016/j.earscirev.2020.103225.Search in Google Scholar

[39] Cortes C, Vapnik V. Support-vector networks. Mach Learn. 1995;20:273–97. 10.1007/BF00994018.Search in Google Scholar

[40] Giovanis DG, Shields MD. Data-driven surrogates for high dimensional models using Gaussian process regression on the Grassmann manifold. Computer Methods Appl Mech Eng. 2020;370:113269. 10.1016/j.cma.2020.113269.Search in Google Scholar

[41] Zhou C, Yin K, Cao Y, Ahmed B, Li Y, Catani F, et al. Landslide susceptibility modeling applying machine learning methods: A case study from Longju in the Three Gorges Reservoir area, China. Computers Geosci. 2018;112:23–37. 10.1016/j.cageo.2017.11.019.Search in Google Scholar

[42] Sachdeva S, Bhatia T, Verma AK. A novel voting ensemble model for spatial prediction of landslides using GIS. Int J Remote Sens. 2020;41(3):929–52. 10.1080/01431161.2019.1654141.Search in Google Scholar

[43] Bashir O, Bangroo SA, Shafai SS, Shah TI, Kader S, Jaufer L, et al. Mathematical vs. machine learning models for particle size distribution in fragile soils of North-Western Himalayas. J Soils Sediment. 2024;13(3):385. 10.1007/s11368-024-03820-y.Search in Google Scholar

[44] Youssef AM, Pourghasemi HR. Landslide susceptibility mapping using machine learning algorithms and comparison of their performance at Abha Basin, Asir Region, Saudi Arabia. Geosci Front. 2021;12(2):639–55. 10.1016/j.gsf.2020.05.010.Search in Google Scholar

[45] Magliulo P, Di Lisio A, Russo F, Zelano A. Geomorphology and landslide susceptibility assessment using GIS and bivariate statistics: A case study in southern Italy. Nat Hazards. 2008;47(3):411–35. 10.1007/s11069-008-9230-x.Search in Google Scholar

[46] Riaz MT, Basharat M, Brunetti MT. Assessing the effectiveness of alternative landslide partitioning in machine learning methods for landslide prediction in the complex Himalayan terrain. Prog Phys Geogr: Earth Environ. 2023;47(3):315–47. 10.1177/03091333221113660.Search in Google Scholar

[47] Feizizadeh B, Blaschke T, Nazmfar H. GIS-based ordered weighted averaging and Dempster–Shafer methods for landslide susceptibility mapping in the Urmia Lake Basin, Iran. Int J Digital Earth. 2014;7(8):688–708. 10.1080/17538947.2012.749950.Search in Google Scholar

[48] Xiao T, Yin K, Yao T, Liu S. Spatial prediction of landslide susceptibility using GIS-based statistical and machine learning models in Wanzhou County, Three Gorges Reservoir, China. Acta Geochim. 2019;38:654–69. 10.1007/s11631-019-00341-1.Search in Google Scholar

[49] Sestras P, Bilașco Ș, Roșca S, Dudic B, Hysa A, Spalević V. Geodetic and UAV monitoring in the sustainable management of shallow landslides and erosion of a susceptible urban environment. Remote Sens. 2021;13(3):385. 10.3390/rs13030385.Search in Google Scholar

[50] Choubin B, Rahmati O, Soleimani F, Alilou H, Moradi E, Alamdari N. Regional groundwater potential analysis using classification and regression trees. In: Pourghasemi HR, Gokceoglu C, editors. Spatial modeling in GIS and R for earth and environmental sciences. United Kingdom: Elsevier; 2019. p. 485–98. 10.1016/B978-0-12-815226-3.00022-3.Search in Google Scholar

[51] Hysa A, Spalevic V, Dudic B, Roșca S, Kuriqi A, Bilașco Ș, et al. Utilizing the available open-source remotely sensed data in assessing the wildfire ignition and spread capacities of vegetated surfaces in Romania. Remote Sens. 2021;13(14):2737. 10.3390/rs13142737.Search in Google Scholar

[52] Debnath J, Sahariah D, Nath N, Saikia A, Lahon D, Islam MN, et al. Modelling on assessment of flood risk susceptibility at the Jia Bharali River basin in Eastern Himalayas by integrating multicollinearity tests and geospatial techniques. Modeling Earth Syst Environ. 2023;139–140:1–27. 10.1007/s40808-023-01912-1.Search in Google Scholar

[53] Lee S, Chwae U, Min K. Landslide susceptibility mapping by correlation between topography and geological structure: the Janghung area, Korea. Geomorphology. 2002;46(3):149–62. 10.1016/S0169-555X(02)00057-0.Search in Google Scholar

[54] Kanungo DP, Arora MK, Sarkar S, Gupta RP. A comparative study of conventional, ANN black box, fuzzy and combined neural and fuzzy weighting procedures for landslide susceptibility zonation in Darjeeling Himalayas. Eng Geol. 2006;85(3):347–66. 10.1016/j.enggeo.2006.03.004.Search in Google Scholar

[55] Bucci F, Santangelo M, Cardinali M, Fiorucci F, Guzzetti F. Landslide distribution and size in response to Quaternary fault activity: the Peloritani Range, NE Sicily, Italy. Earth Surf Process Landf. 2016;41(5):711–20. 10.1002/esp.3898.Search in Google Scholar

[56] Hadji R, Boumazbeur A, Limani Y, Baghem M, Chouabi AM, Demdoum A. Geologic, topographic and climatic controls in landslide hazard assessment using GIS modeling: A case study of Souk Ahras region, NE Algeria. Quat Int. 2013;302:224–37. 10.1016/j.quaint.2012.11.027.Search in Google Scholar

[57] Pradhan B, Mansor S, Pirasteh S, Buchroithner MF. Landslide hazard and risk analyses at a landslide prone catchment area using statistical based geospatial model. Int J Remote Sens. 2011;32(14):4075–87. 10.1080/01431161.2010.484433.Search in Google Scholar

[58] Romstad B, Etzelmüller B. Mean-curvature watersheds: A simple method for segmentation of a digital elevation model into terrain units. Geomorphology. 2012;139-140:293–302. 10.1016/j.geomorph.2011.10.031.Search in Google Scholar

[59] Prasicek G, Otto JC, Montgomery DR, Schrott L. Multi-scale curvature for automated identification of glaciated mountain landscapes. Geomorphology. 2014;209:53–65. 10.1016/j.geomorph.2013.11.026.Search in Google Scholar PubMed PubMed Central

[60] Mascandola C, Luzi L, Felicetta C, Pacor F. A GIS procedure for the topographic classification of Italy, according to the seismic code provisions. Soil Dyn Earthq Eng. 2021;148:106848. 10.1016/j.soildyn.2021.106848.Search in Google Scholar

[61] Dahal MS, Chhetri A, Ghalley H, Pasang S, Patra M. Chapter 45 – Landslide susceptibility mapping along the Thimphu-Phuentsholing highway using machine learning. Pourghasemi HR, editor. Computers in earth and environmental sciences. Switzerland: MDPI; 2022. p. 601–17. 10.1016/B978-0-323-89861-4.00038-5.Search in Google Scholar

[62] Wu Z, Baartman JEM, Pedro Nunes J, López-Vicente M. Intra-annual sediment dynamic assessment in the Wei River Basin, China, using the AIC functional-structural connectivity index. Ecol Indic. 2023;146:109775. 10.1016/j.ecolind.2022.109775.Search in Google Scholar

[63] Quan H-C, Lee B-G. GIS-based landslide susceptibility mapping using analytic hierarchy process and artificial neural network in Jeju (Korea). KSCE J Civ Eng. 2012;16(7):1258–66. 10.1007/s12205-012-1242-0.Search in Google Scholar

[64] Kalantar B, Ueda N, Saeidi V, Janizadeh S, Shabani F, Ahmadi K, et al. Deep neural network utilizing remote sensing datasets for flood hazard susceptibility mapping in Brisbane, Australia. Remote Sens. 2021;13(13):2638.10.3390/rs13132638Search in Google Scholar

[65] Huo A, Yang L, Peng J, Cheng Y, Jiang C. Spatial characteristics of the rainfall induced landslides in the Chinese Loess Plateau. Hum Ecol Risk Assessment: An Int J. 2020;26(9):2462–77. 10.1080/10807039.2020.1728517.Search in Google Scholar

[66] Sun D, Ding Y, Zhang J, Wen H, Wang Y, Xu J, et al. Essential insights into decision mechanism of landslide susceptibility mapping based on different machine learning models. Geocarto Int. 2022;51(3):1–29. 10.1080/10106049.2022.2146763.Search in Google Scholar

[67] Arabameri A, Saha S, Roy J, Chen W, Blaschke T, Tien Bui D. Landslide susceptibility evaluation and management using different machine learning methods in the Gallicash River Watershed, Iran. Remote Sens. 2020;12(3):475. 10.3390/rs12030475.Search in Google Scholar

[68] Wang Y, Feng L, Li S, Ren F, Du Q. A hybrid model considering spatial heterogeneity for landslide susceptibility mapping in Zhejiang Province, China. Catena. 2020;188:104425. 10.1016/j.catena.2019.104425.Search in Google Scholar

[69] Bammou Y, Benzougagh B, Igmoullan B, Kader S, Ouallali A, Spalevic V, et al. Spatial mapping for multi-hazard land management in sparsely vegetated watersheds using machine learning algorithms. Environ Earth Sci. 2024;83(15):447. 10.1007/s12665-024-11741-9.Search in Google Scholar

[70] Zhiyong F, Changdong L, Wenmin Y. Landslide susceptibility assessment through TrAdaBoost transfer learning models using two landslide inventories. Catena. 2023;222:106799. 10.1016/j.catena.2022.106799.Search in Google Scholar

[71] Sbihi A, Mastere M, Benzougagh B, Spalevic V, Sestras P, Radovic M, et al. Assessing landslide susceptibility in northern Morocco: A geostatistical mapping approach in Al Hoceima-Ajdir. J Afr Earth Sci. 2024;218:105361. 10.1016/j.jafrearsci.2024.105361.Search in Google Scholar

[72] Aguirre-Gutiérrez J, Carvalheiro LG, Polce C, van Loon EE, Raes N, Reemer M, et al. Fit-for-purpose: species distribution model performance depends on evaluation criteria–Dutch hoverflies as a case study. PLoS One. 2013;8(5):e63708. 10.1371/journal.pone.0063708.Search in Google Scholar PubMed PubMed Central

[73] Sestras P, Mircea S, Roșca S, Bilașco Ș, Sălăgean T, Dragomir LO, et al. GIS based soil erosion assessment using the USLE model for efficient land management: A case study in an area with diverse pedo-geomorphological and bioclimatic characteristics. Not Botanicae Horti Agrobotanici Cluj-Napoca. 2023;51(3):13263. 10.15835/nbha51313263.Search in Google Scholar
Received: 2024-05-03
Revised: 2024-10-05
Accepted: 2024-11-16
Published Online: 2025-03-03

© 2025 the author(s), published by De Gruyter

This work is licensed under the Creative Commons Attribution 4.0 International License.

Articles in the same Issue

  1. Research Articles
  2. Seismic response and damage model analysis of rocky slopes with weak interlayers
  3. Multi-scenario simulation and eco-environmental effect analysis of “Production–Living–Ecological space” based on PLUS model: A case study of Anyang City
  4. Remote sensing estimation of chlorophyll content in rape leaves in Weibei dryland region of China
  5. GIS-based frequency ratio and Shannon entropy modeling for landslide susceptibility mapping: A case study in Kundah Taluk, Nilgiris District, India
  6. Natural gas origin and accumulation of the Changxing–Feixianguan Formation in the Puguang area, China
  7. Spatial variations of shear-wave velocity anomaly derived from Love wave ambient noise seismic tomography along Lembang Fault (West Java, Indonesia)
  8. Evaluation of cumulative rainfall and rainfall event–duration threshold based on triggering and non-triggering rainfalls: Northern Thailand case
  9. Pixel and region-oriented classification of Sentinel-2 imagery to assess LULC dynamics and their climate impact in Nowshera, Pakistan
  10. The use of radar-optical remote sensing data and geographic information system–analytical hierarchy process–multicriteria decision analysis techniques for revealing groundwater recharge prospective zones in arid-semi arid lands
  11. Effect of pore throats on the reservoir quality of tight sandstone: A case study of the Yanchang Formation in the Zhidan area, Ordos Basin
  12. Hydroelectric simulation of the phreatic water response of mining cracked soil based on microbial solidification
  13. Spatial-temporal evolution of habitat quality in tropical monsoon climate region based on “pattern–process–quality” – a case study of Cambodia
  14. Early Permian to Middle Triassic Formation petroleum potentials of Sydney Basin, Australia: A geochemical analysis
  15. Micro-mechanism analysis of Zhongchuan loess liquefaction disaster induced by Jishishan M6.2 earthquake in 2023
  16. Prediction method of S-wave velocities in tight sandstone reservoirs – a case study of CO2 geological storage area in Ordos Basin
  17. Ecological restoration in valley area of semiarid region damaged by shallow buried coal seam mining
  18. Hydrocarbon-generating characteristics of Xujiahe coal-bearing source rocks in the continuous sedimentary environment of the Southwest Sichuan
  19. Hazard analysis of future surface displacements on active faults based on the recurrence interval of strong earthquakes
  20. Structural characterization of the Zalm district, West Saudi Arabia, using aeromagnetic data: An approach for gold mineral exploration
  21. Research on the variation in the Shields curve of silt initiation
  22. Reuse of agricultural drainage water and wastewater for crop irrigation in southeastern Algeria
  23. Assessing the effectiveness of utilizing low-cost inertial measurement unit sensors for producing as-built plans
  24. Analysis of the formation process of a natural fertilizer in the loess area
  25. Machine learning methods for landslide mapping studies: A comparative study of SVM and RF algorithms in the Oued Aoulai watershed (Morocco)
  26. Chemical dissolution and the source of salt efflorescence in weathering of sandstone cultural relics
  27. Molecular simulation of methane adsorption capacity in transitional shale – a case study of Longtan Formation shale in Southern Sichuan Basin, SW China
  28. Evolution characteristics of extreme maximum temperature events in Central China and adaptation strategies under different future warming scenarios
  29. Estimating Bowen ratio in local environment based on satellite imagery
  30. 3D fusion modeling of multi-scale geological structures based on subdivision-NURBS surfaces and stratigraphic sequence formalization
  31. Comparative analysis of machine learning algorithms in Google Earth Engine for urban land use dynamics in rapidly urbanizing South Asian cities
  32. Study on the mechanism of plant root influence on soil properties in expansive soil areas
  33. Simulation of seismic hazard parameters and earthquakes source mechanisms along the Red Sea rift, western Saudi Arabia
  34. Tectonics vs sedimentation in foredeep basins: A tale from the Oligo-Miocene Monte Falterona Formation (Northern Apennines, Italy)
  35. Investigation of landslide areas in Tokat-Almus road between Bakımlı-Almus by the PS-InSAR method (Türkiye)
  36. Predicting coastal variations in non-storm conditions with machine learning
  37. Cross-dimensional adaptivity research on a 3D earth observation data cube model
  38. Geochronology and geochemistry of late Paleozoic volcanic rocks in eastern Inner Mongolia and their geological significance
  39. Spatial and temporal evolution of land use and habitat quality in arid regions – a case of Northwest China
  40. Ground-penetrating radar imaging of subsurface karst features controlling water leakage across Wadi Namar dam, south Riyadh, Saudi Arabia
  41. Rayleigh wave dispersion inversion via modified sine cosine algorithm: Application to Hangzhou, China passive surface wave data
  42. Fractal insights into permeability control by pore structure in tight sandstone reservoirs, Heshui area, Ordos Basin
  43. Debris flow hazard characteristic and mitigation in Yusitong Gully, Hengduan Mountainous Region
  44. Research on community characteristics of vegetation restoration in hilly power engineering based on multi temporal remote sensing technology
  45. Identification of radial drainage networks based on topographic and geometric features
  46. Trace elements and melt inclusion in zircon within the Qunji porphyry Cu deposit: Application to the metallogenic potential of the reduced magma-hydrothermal system
  47. Pore, fracture characteristics and diagenetic evolution of medium-maturity marine shales from the Silurian Longmaxi Formation, NE Sichuan Basin, China
  48. Study of the earthquakes source parameters, site response, and path attenuation using P and S-waves spectral inversion, Aswan region, south Egypt
  49. Source of contamination and assessment of potential health risks of potentially toxic metal(loid)s in agricultural soil from Al Lith, Saudi Arabia
  50. Regional spatiotemporal evolution and influencing factors of rural construction areas in the Nanxi River Basin via GIS
  51. An efficient network for object detection in scale-imbalanced remote sensing images
  52. Effect of microscopic pore–throat structure heterogeneity on waterflooding seepage characteristics of tight sandstone reservoirs
  53. Environmental health risk assessment of Zn, Cd, Pb, Fe, and Co in coastal sediments of the southeastern Gulf of Aqaba
  54. A modified Hoek–Brown model considering softening effects and its applications
  55. Evaluation of engineering properties of soil for sustainable urban development
  56. The spatio-temporal characteristics and influencing factors of sustainable development in China’s provincial areas
  57. Application of a mixed additive and multiplicative random error model to generate DTM products from LiDAR data
  58. Gold vein mineralogy and oxygen isotopes of Wadi Abu Khusheiba, Jordan
  59. Prediction of surface deformation time series in closed mines based on LSTM and optimization algorithms
  60. 2D–3D Geological features collaborative identification of surrounding rock structural planes in hydraulic adit based on OC-AINet
  61. Spatiotemporal patterns and drivers of Chl-a in Chinese lakes between 1986 and 2023
  62. Land use classification through fusion of remote sensing images and multi-source data
  63. Nexus between renewable energy, technological innovation, and carbon dioxide emissions in Saudi Arabia
  64. Analysis of the spillover effects of green organic transformation on sustainable development in ethnic regions’ agriculture and animal husbandry
  65. Factors impacting spatial distribution of black and odorous water bodies in Hebei
  66. Large-scale shaking table tests on the liquefaction and deformation responses of an ultra-deep overburden
  67. Impacts of climate change and sea-level rise on the coastal geological environment of Quang Nam province, Vietnam
  68. Reservoir characterization and exploration potential of shale reservoir near denudation area: A case study of Ordovician–Silurian marine shale, China
  69. Seismic prediction of Permian volcanic rock reservoirs in Southwest Sichuan Basin
  70. Application of CBERS-04 IRS data to land surface temperature inversion: A case study based on Minqin arid area
  71. Geological characteristics and prospecting direction of Sanjiaoding gold mine in Saishiteng area
  72. Research on the deformation prediction model of surrounding rock based on SSA-VMD-GRU
  73. Geochronology, geochemical characteristics, and tectonic significance of the granites, Menghewula, Southern Great Xing’an range
  74. Hazard classification of active faults in Yunnan base on probabilistic seismic hazard assessment
  75. Characteristics analysis of hydrate reservoirs with different geological structures developed by vertical well depressurization
  76. Estimating the travel distance of channelized rock avalanches using genetic programming method
  77. Landscape preferences of hikers in Three Parallel Rivers Region and its adjacent regions by content analysis of user-generated photography
  78. New age constraints of the LGM onset in the Bohemian Forest – Central Europe
  79. Characteristics of geological evolution based on the multifractal singularity theory: A case study of Heyu granite and Mesozoic tectonics
  80. Soil water content and longitudinal microbiota distribution in disturbed areas of tower foundations of power transmission and transformation projects
  81. Oil accumulation process of the Kongdian reservoir in the deep subsag zone of the Cangdong Sag, Bohai Bay Basin, China
  82. Investigation of velocity profile in rock–ice avalanche by particle image velocimetry measurement
  83. Optimizing 3D seismic survey geometries using ray tracing and illumination modeling: A case study from Penobscot field
  84. Sedimentology of the Phra That and Pha Daeng Formations: A preliminary evaluation of geological CO2 storage potential in the Lampang Basin, Thailand
  85. Improved classification algorithm for hyperspectral remote sensing images based on the hybrid spectral network model
  86. Map analysis of soil erodibility rates and gully erosion sites in Anambra State, South Eastern Nigeria
  87. Identification and driving mechanism of land use conflict in China’s South-North transition zone: A case study of Huaihe River Basin
  88. Evaluation of the impact of land-use change on earthquake risk distribution in different periods: An empirical analysis from Sichuan Province
  89. A test site case study on the long-term behavior of geotextile tubes
  90. An experimental investigation into carbon dioxide flooding and rock dissolution in low-permeability reservoirs of the South China Sea
  91. Detection and semi-quantitative analysis of naphthenic acids in coal and gangue from mining areas in China
  92. Comparative effects of olivine and sand on KOH-treated clayey soil
  93. YOLO-MC: An algorithm for early forest fire recognition based on drone image
  94. Earthquake building damage classification based on full suite of Sentinel-1 features
  95. Potential landslide detection and influencing factors analysis in the upper Yellow River based on SBAS-InSAR technology
  96. Assessing green area changes in Najran City, Saudi Arabia (2013–2022) using hybrid deep learning techniques
  97. An advanced approach integrating methods to estimate hydraulic conductivity of different soil types supported by a machine learning model
  98. Hybrid methods for land use and land cover classification using remote sensing and combined spectral feature extraction: A case study of Najran City, KSA
  99. Streamlining digital elevation model construction from historical aerial photographs: The impact of reference elevation data on spatial accuracy
  100. Analysis of urban expansion patterns in the Yangtze River Delta based on the fusion impervious surfaces dataset
  101. A metaverse-based visual analysis approach for 3D reservoir models
  102. Late Quaternary record of 100 ka depositional cycles on the Larache shelf (NW Morocco)
  103. Integrated well-seismic analysis of sedimentary facies distribution: A case study from the Mesoproterozoic, Ordos Basin, China
  104. Study on the spatial equilibrium of cultural and tourism resources in Macao, China
  105. Urban road surface condition detecting and integrating based on the mobile sensing framework with multi-modal sensors
  106. Application of improved sine cosine algorithm with chaotic mapping and novel updating methods for joint inversion of resistivity and surface wave data
  107. The synergistic use of AHP and GIS to assess factors driving forest fire potential in a peat swamp forest in Thailand
  108. Dynamic response analysis and comprehensive evaluation of cement-improved aeolian sand roadbed
  109. Rock control on evolution of Khorat Cuesta, Khorat UNESCO Geopark, Northeastern Thailand
  110. Gradient response mechanism of carbon storage: Spatiotemporal analysis of economic-ecological dimensions based on hybrid machine learning
  111. Comparison of several seismic active earth pressure calculation methods for retaining structures
  112. Mantle dynamics and petrogenesis of Gomer basalts in the Northwestern Ethiopia: A geochemical perspective
  113. Study on ground deformation monitoring in Xiong’an New Area from 2021 to 2023 based on DS-InSAR
  114. Paleoenvironmental characteristics of continental shale and its significance to organic matter enrichment: Taking the fifth member of Xujiahe Formation in Tianfu area of Sichuan Basin as an example
  115. Equipping the integral approach with generalized least squares to reconstruct relict channel profile and its usage in the Shanxi Rift, northern China
  116. InSAR-driven landslide hazard assessment along highways in hilly regions: A case-based validation approach
  117. Attribution analysis of multi-temporal scale surface streamflow changes in the Ganjiang River based on a multi-temporal Budyko framework
  118. Review Articles
  119. Humic substances influence on the distribution of dissolved iron in seawater: A review of electrochemical methods and other techniques
  120. Applications of physics-informed neural networks in geosciences: From basic seismology to comprehensive environmental studies
  121. Ore-controlling structures of granite-related uranium deposits in South China: A review
  122. Shallow geological structure features in Balikpapan Bay East Kalimantan Province – Indonesia
  123. A review on the tectonic affinity of microcontinents and evolution of the Proto-Tethys Ocean in Northeastern Tibet
  124. Special Issue: Natural Resources and Environmental Risks: Towards a Sustainable Future - Part II
  125. Depopulation in the Visok micro-region: Toward demographic and economic revitalization
  126. Special Issue: Geospatial and Environmental Dynamics - Part II
  127. Advancing urban sustainability: Applying GIS technologies to assess SDG indicators – a case study of Podgorica (Montenegro)
  128. Spatiotemporal and trend analysis of common cancers in men in Central Serbia (1999–2021)
  129. Minerals for the green agenda, implications, stalemates, and alternatives
  130. Spatiotemporal water quality analysis of Vrana Lake, Croatia
  131. Functional transformation of settlements in coal exploitation zones: A case study of the municipality of Stanari in Republic of Srpska (Bosnia and Herzegovina)
  132. Hypertension in AP Vojvodina (Northern Serbia): A spatio-temporal analysis of patients at the Institute for Cardiovascular Diseases of Vojvodina
  133. Regional patterns in cause-specific mortality in Montenegro, 1991–2019
  134. Spatio-temporal analysis of flood events using GIS and remote sensing-based approach in the Ukrina River Basin, Bosnia and Herzegovina
  135. Flash flood susceptibility mapping using LiDAR-Derived DEM and machine learning algorithms: Ljuboviđa case study, Serbia
  136. Geocultural heritage as a basis for geotourism development: Banjska Monastery, Zvečan (Serbia)
  137. Assessment of groundwater potential zones using GIS and AHP techniques – A case study of the zone of influence of Kolubara Mining Basin
  138. Impact of the agri-geographical transformation of rural settlements on the geospatial dynamics of soil erosion intensity in municipalities of Central Serbia
  139. Where faith meets geomorphology: The cultural and religious significance of geodiversity explored through geospatial technologies
  140. Applications of local climate zone classification in European cities: A review of in situ and mobile monitoring methods in urban climate studies
  141. Complex multivariate water quality impact assessment on Krivaja River
  142. Ionization hotspots near waterfalls in Eastern Serbia’s Stara Planina Mountain
  143. Shift in landscape use strategies during the transition from the Bronze age to Iron age in Northwest Serbia
  144. Assessing the geotourism potential of glacial lakes in Plav, Montenegro: A multi-criteria assessment by using the M-GAM model
  145. Flash flood potential index at national scale: Susceptibility assessment within catchments
https://doi.org/10.1515/geo-2022-0740
Keywords for this article
management; landslide susceptibility; machine learning models; factor correlation analysis; area under the curve; Morocco
Creative Commons
BY 4.0

Articles in the same Issue

  1. Research Articles
  2. Seismic response and damage model analysis of rocky slopes with weak interlayers
  3. Multi-scenario simulation and eco-environmental effect analysis of “Production–Living–Ecological space” based on PLUS model: A case study of Anyang City
  4. Remote sensing estimation of chlorophyll content in rape leaves in Weibei dryland region of China
  5. GIS-based frequency ratio and Shannon entropy modeling for landslide susceptibility mapping: A case study in Kundah Taluk, Nilgiris District, India
  6. Natural gas origin and accumulation of the Changxing–Feixianguan Formation in the Puguang area, China
  7. Spatial variations of shear-wave velocity anomaly derived from Love wave ambient noise seismic tomography along Lembang Fault (West Java, Indonesia)
  8. Evaluation of cumulative rainfall and rainfall event–duration threshold based on triggering and non-triggering rainfalls: Northern Thailand case
  9. Pixel and region-oriented classification of Sentinel-2 imagery to assess LULC dynamics and their climate impact in Nowshera, Pakistan
  10. The use of radar-optical remote sensing data and geographic information system–analytical hierarchy process–multicriteria decision analysis techniques for revealing groundwater recharge prospective zones in arid-semi arid lands
  11. Effect of pore throats on the reservoir quality of tight sandstone: A case study of the Yanchang Formation in the Zhidan area, Ordos Basin
  12. Hydroelectric simulation of the phreatic water response of mining cracked soil based on microbial solidification
  13. Spatial-temporal evolution of habitat quality in tropical monsoon climate region based on “pattern–process–quality” – a case study of Cambodia
  14. Early Permian to Middle Triassic Formation petroleum potentials of Sydney Basin, Australia: A geochemical analysis
  15. Micro-mechanism analysis of Zhongchuan loess liquefaction disaster induced by Jishishan M6.2 earthquake in 2023
  16. Prediction method of S-wave velocities in tight sandstone reservoirs – a case study of CO2 geological storage area in Ordos Basin
  17. Ecological restoration in valley area of semiarid region damaged by shallow buried coal seam mining
  18. Hydrocarbon-generating characteristics of Xujiahe coal-bearing source rocks in the continuous sedimentary environment of the Southwest Sichuan
  19. Hazard analysis of future surface displacements on active faults based on the recurrence interval of strong earthquakes
  20. Structural characterization of the Zalm district, West Saudi Arabia, using aeromagnetic data: An approach for gold mineral exploration
  21. Research on the variation in the Shields curve of silt initiation
  22. Reuse of agricultural drainage water and wastewater for crop irrigation in southeastern Algeria
  23. Assessing the effectiveness of utilizing low-cost inertial measurement unit sensors for producing as-built plans
  24. Analysis of the formation process of a natural fertilizer in the loess area
  25. Machine learning methods for landslide mapping studies: A comparative study of SVM and RF algorithms in the Oued Aoulai watershed (Morocco)
  26. Chemical dissolution and the source of salt efflorescence in weathering of sandstone cultural relics
  27. Molecular simulation of methane adsorption capacity in transitional shale – a case study of Longtan Formation shale in Southern Sichuan Basin, SW China
  28. Evolution characteristics of extreme maximum temperature events in Central China and adaptation strategies under different future warming scenarios
  29. Estimating Bowen ratio in local environment based on satellite imagery
  30. 3D fusion modeling of multi-scale geological structures based on subdivision-NURBS surfaces and stratigraphic sequence formalization
  31. Comparative analysis of machine learning algorithms in Google Earth Engine for urban land use dynamics in rapidly urbanizing South Asian cities
  32. Study on the mechanism of plant root influence on soil properties in expansive soil areas
  33. Simulation of seismic hazard parameters and earthquakes source mechanisms along the Red Sea rift, western Saudi Arabia
  34. Tectonics vs sedimentation in foredeep basins: A tale from the Oligo-Miocene Monte Falterona Formation (Northern Apennines, Italy)
  35. Investigation of landslide areas in Tokat-Almus road between Bakımlı-Almus by the PS-InSAR method (Türkiye)
  36. Predicting coastal variations in non-storm conditions with machine learning
  37. Cross-dimensional adaptivity research on a 3D earth observation data cube model
  38. Geochronology and geochemistry of late Paleozoic volcanic rocks in eastern Inner Mongolia and their geological significance
  39. Spatial and temporal evolution of land use and habitat quality in arid regions – a case of Northwest China
  40. Ground-penetrating radar imaging of subsurface karst features controlling water leakage across Wadi Namar dam, south Riyadh, Saudi Arabia
  41. Rayleigh wave dispersion inversion via modified sine cosine algorithm: Application to Hangzhou, China passive surface wave data
  42. Fractal insights into permeability control by pore structure in tight sandstone reservoirs, Heshui area, Ordos Basin
  43. Debris flow hazard characteristic and mitigation in Yusitong Gully, Hengduan Mountainous Region
  44. Research on community characteristics of vegetation restoration in hilly power engineering based on multi temporal remote sensing technology
  45. Identification of radial drainage networks based on topographic and geometric features
  46. Trace elements and melt inclusion in zircon within the Qunji porphyry Cu deposit: Application to the metallogenic potential of the reduced magma-hydrothermal system
  47. Pore, fracture characteristics and diagenetic evolution of medium-maturity marine shales from the Silurian Longmaxi Formation, NE Sichuan Basin, China
  48. Study of the earthquakes source parameters, site response, and path attenuation using P and S-waves spectral inversion, Aswan region, south Egypt
  49. Source of contamination and assessment of potential health risks of potentially toxic metal(loid)s in agricultural soil from Al Lith, Saudi Arabia
  50. Regional spatiotemporal evolution and influencing factors of rural construction areas in the Nanxi River Basin via GIS
  51. An efficient network for object detection in scale-imbalanced remote sensing images
  52. Effect of microscopic pore–throat structure heterogeneity on waterflooding seepage characteristics of tight sandstone reservoirs
  53. Environmental health risk assessment of Zn, Cd, Pb, Fe, and Co in coastal sediments of the southeastern Gulf of Aqaba
  54. A modified Hoek–Brown model considering softening effects and its applications
  55. Evaluation of engineering properties of soil for sustainable urban development
  56. The spatio-temporal characteristics and influencing factors of sustainable development in China’s provincial areas
  57. Application of a mixed additive and multiplicative random error model to generate DTM products from LiDAR data
  58. Gold vein mineralogy and oxygen isotopes of Wadi Abu Khusheiba, Jordan
  59. Prediction of surface deformation time series in closed mines based on LSTM and optimization algorithms
  60. 2D–3D Geological features collaborative identification of surrounding rock structural planes in hydraulic adit based on OC-AINet
  61. Spatiotemporal patterns and drivers of Chl-a in Chinese lakes between 1986 and 2023
  62. Land use classification through fusion of remote sensing images and multi-source data
  63. Nexus between renewable energy, technological innovation, and carbon dioxide emissions in Saudi Arabia
  64. Analysis of the spillover effects of green organic transformation on sustainable development in ethnic regions’ agriculture and animal husbandry
  65. Factors impacting spatial distribution of black and odorous water bodies in Hebei
  66. Large-scale shaking table tests on the liquefaction and deformation responses of an ultra-deep overburden
  67. Impacts of climate change and sea-level rise on the coastal geological environment of Quang Nam province, Vietnam
  68. Reservoir characterization and exploration potential of shale reservoir near denudation area: A case study of Ordovician–Silurian marine shale, China
  69. Seismic prediction of Permian volcanic rock reservoirs in Southwest Sichuan Basin
  70. Application of CBERS-04 IRS data to land surface temperature inversion: A case study based on Minqin arid area
  71. Geological characteristics and prospecting direction of Sanjiaoding gold mine in Saishiteng area
  72. Research on the deformation prediction model of surrounding rock based on SSA-VMD-GRU
  73. Geochronology, geochemical characteristics, and tectonic significance of the granites, Menghewula, Southern Great Xing’an range
  74. Hazard classification of active faults in Yunnan base on probabilistic seismic hazard assessment
  75. Characteristics analysis of hydrate reservoirs with different geological structures developed by vertical well depressurization
  76. Estimating the travel distance of channelized rock avalanches using genetic programming method
  77. Landscape preferences of hikers in Three Parallel Rivers Region and its adjacent regions by content analysis of user-generated photography
  78. New age constraints of the LGM onset in the Bohemian Forest – Central Europe
  79. Characteristics of geological evolution based on the multifractal singularity theory: A case study of Heyu granite and Mesozoic tectonics
  80. Soil water content and longitudinal microbiota distribution in disturbed areas of tower foundations of power transmission and transformation projects
  81. Oil accumulation process of the Kongdian reservoir in the deep subsag zone of the Cangdong Sag, Bohai Bay Basin, China
  82. Investigation of velocity profile in rock–ice avalanche by particle image velocimetry measurement
  83. Optimizing 3D seismic survey geometries using ray tracing and illumination modeling: A case study from Penobscot field
  84. Sedimentology of the Phra That and Pha Daeng Formations: A preliminary evaluation of geological CO2 storage potential in the Lampang Basin, Thailand
  85. Improved classification algorithm for hyperspectral remote sensing images based on the hybrid spectral network model
  86. Map analysis of soil erodibility rates and gully erosion sites in Anambra State, South Eastern Nigeria
  87. Identification and driving mechanism of land use conflict in China’s South-North transition zone: A case study of Huaihe River Basin
  88. Evaluation of the impact of land-use change on earthquake risk distribution in different periods: An empirical analysis from Sichuan Province
  89. A test site case study on the long-term behavior of geotextile tubes
  90. An experimental investigation into carbon dioxide flooding and rock dissolution in low-permeability reservoirs of the South China Sea
  91. Detection and semi-quantitative analysis of naphthenic acids in coal and gangue from mining areas in China
  92. Comparative effects of olivine and sand on KOH-treated clayey soil
  93. YOLO-MC: An algorithm for early forest fire recognition based on drone image
  94. Earthquake building damage classification based on full suite of Sentinel-1 features
  95. Potential landslide detection and influencing factors analysis in the upper Yellow River based on SBAS-InSAR technology
  96. Assessing green area changes in Najran City, Saudi Arabia (2013–2022) using hybrid deep learning techniques
  97. An advanced approach integrating methods to estimate hydraulic conductivity of different soil types supported by a machine learning model
  98. Hybrid methods for land use and land cover classification using remote sensing and combined spectral feature extraction: A case study of Najran City, KSA
  99. Streamlining digital elevation model construction from historical aerial photographs: The impact of reference elevation data on spatial accuracy
  100. Analysis of urban expansion patterns in the Yangtze River Delta based on the fusion impervious surfaces dataset
  101. A metaverse-based visual analysis approach for 3D reservoir models
  102. Late Quaternary record of 100 ka depositional cycles on the Larache shelf (NW Morocco)
  103. Integrated well-seismic analysis of sedimentary facies distribution: A case study from the Mesoproterozoic, Ordos Basin, China
  104. Study on the spatial equilibrium of cultural and tourism resources in Macao, China
  105. Urban road surface condition detecting and integrating based on the mobile sensing framework with multi-modal sensors
  106. Application of improved sine cosine algorithm with chaotic mapping and novel updating methods for joint inversion of resistivity and surface wave data
  107. The synergistic use of AHP and GIS to assess factors driving forest fire potential in a peat swamp forest in Thailand
  108. Dynamic response analysis and comprehensive evaluation of cement-improved aeolian sand roadbed
  109. Rock control on evolution of Khorat Cuesta, Khorat UNESCO Geopark, Northeastern Thailand
  110. Gradient response mechanism of carbon storage: Spatiotemporal analysis of economic-ecological dimensions based on hybrid machine learning
  111. Comparison of several seismic active earth pressure calculation methods for retaining structures
  112. Mantle dynamics and petrogenesis of Gomer basalts in the Northwestern Ethiopia: A geochemical perspective
  113. Study on ground deformation monitoring in Xiong’an New Area from 2021 to 2023 based on DS-InSAR
  114. Paleoenvironmental characteristics of continental shale and its significance to organic matter enrichment: Taking the fifth member of Xujiahe Formation in Tianfu area of Sichuan Basin as an example
  115. Equipping the integral approach with generalized least squares to reconstruct relict channel profile and its usage in the Shanxi Rift, northern China
  116. InSAR-driven landslide hazard assessment along highways in hilly regions: A case-based validation approach
  117. Attribution analysis of multi-temporal scale surface streamflow changes in the Ganjiang River based on a multi-temporal Budyko framework
  118. Review Articles
  119. Humic substances influence on the distribution of dissolved iron in seawater: A review of electrochemical methods and other techniques
  120. Applications of physics-informed neural networks in geosciences: From basic seismology to comprehensive environmental studies
  121. Ore-controlling structures of granite-related uranium deposits in South China: A review
  122. Shallow geological structure features in Balikpapan Bay East Kalimantan Province – Indonesia
  123. A review on the tectonic affinity of microcontinents and evolution of the Proto-Tethys Ocean in Northeastern Tibet
  124. Special Issue: Natural Resources and Environmental Risks: Towards a Sustainable Future - Part II
  125. Depopulation in the Visok micro-region: Toward demographic and economic revitalization
  126. Special Issue: Geospatial and Environmental Dynamics - Part II
  127. Advancing urban sustainability: Applying GIS technologies to assess SDG indicators – a case study of Podgorica (Montenegro)
  128. Spatiotemporal and trend analysis of common cancers in men in Central Serbia (1999–2021)
  129. Minerals for the green agenda, implications, stalemates, and alternatives
  130. Spatiotemporal water quality analysis of Vrana Lake, Croatia
  131. Functional transformation of settlements in coal exploitation zones: A case study of the municipality of Stanari in Republic of Srpska (Bosnia and Herzegovina)
  132. Hypertension in AP Vojvodina (Northern Serbia): A spatio-temporal analysis of patients at the Institute for Cardiovascular Diseases of Vojvodina
  133. Regional patterns in cause-specific mortality in Montenegro, 1991–2019
  134. Spatio-temporal analysis of flood events using GIS and remote sensing-based approach in the Ukrina River Basin, Bosnia and Herzegovina
  135. Flash flood susceptibility mapping using LiDAR-Derived DEM and machine learning algorithms: Ljuboviđa case study, Serbia
  136. Geocultural heritage as a basis for geotourism development: Banjska Monastery, Zvečan (Serbia)
  137. Assessment of groundwater potential zones using GIS and AHP techniques – A case study of the zone of influence of Kolubara Mining Basin
  138. Impact of the agri-geographical transformation of rural settlements on the geospatial dynamics of soil erosion intensity in municipalities of Central Serbia
  139. Where faith meets geomorphology: The cultural and religious significance of geodiversity explored through geospatial technologies
  140. Applications of local climate zone classification in European cities: A review of in situ and mobile monitoring methods in urban climate studies
  141. Complex multivariate water quality impact assessment on Krivaja River
  142. Ionization hotspots near waterfalls in Eastern Serbia’s Stara Planina Mountain
  143. Shift in landscape use strategies during the transition from the Bronze age to Iron age in Northwest Serbia
  144. Assessing the geotourism potential of glacial lakes in Plav, Montenegro: A multi-criteria assessment by using the M-GAM model
  145. Flash flood potential index at national scale: Susceptibility assessment within catchments
Sign up now to receive a 20% welcome discount
Institutional Access
How does access work?
Have an idea on how to improve our website?
Please write us.
© 2025 De Gruyter Brill
Downloaded on 27.10.2025 from https://www.degruyterbrill.com/document/doi/10.1515/geo-2022-0740/html?lang=en&srsltid=AfmBOor7NxFZa4ArdVOUtqYJMPHACh_8KUKEW3F18inKAdNAv9qdphvP
Scroll to top button