Evaluation of landslides susceptibility in Southeastern Tibet considering seismic sensitivity.

Southeastern Tibet features a complex geological environment and a high incidence of earthquakes. Earthquake-induced chain disasters pose a great threat to engineering construction and public safety in this area, and landslides are among the most frequent postearthquake disasters. To investigate the impact of earthquakes on landslides, this study constructed a comprehensive database for landslide susceptibility analysis based on various factors, including elevation, slope, slope direction, distances from roads and rivers, proximity to faults, land use patterns, rainfall patterns, and seismic parameters. By integrating the frequency ratio (FR) model with the analytic hierarchy process (AHP) model, this work delineated landslide susceptibility zones in southeastern Tibet. Subsequently, the susceptibility zoning layer was overlaid with the magnitude sensitivity layer and validated using ROC curve analysis to identify the earthquake magnitudes that exerted the greatest influence on landslides. Finally, by incorporating the distance from earthquake epicentres into our refined model framework, different monitoring levels for landslide susceptibility zoning were established. The AHP results show the relative importance of the landslide-influencing factors in southeastern Tibet can be ranked as follows: elevation, slope direction, distance from road, land use, distance from river, slope, and rainfall. The ROC values of the landslide models with seismic sensitivities of 1, 2 and 3 are 0.876, 0.883 and 0.877, respectively, indicating that earthquakes of magnitude 4 and above have a great influence on landslides in the study area. Through the overlay of the landslide susceptibility zoning map and the vector map of distance from seismic focal points, a correlation with the distance between landslide-prone areas and seismic focal points is identified. Within the extremely high susceptibility area and within 40 km of the focal point, there are 220 landslide points, accounting for approximately 34 % of the total landslides in the study area. Additionally, 133 landslide points are located in the extremely susceptible area and within 40-80 km of the focal point, representing approximately 20 % of the total landslides in the study area. The susceptible areas were assessed based on grades, resulting in the production of 4 maps depicting different levels of monitoring for landslide-prone areas. These maps are valuable tools for implementing landslide disaster prevention measures within the study area.

Investigation and monitoring of rotational landslides in El Mokkattam plateau Egypt, using integrated geological and geophysical techniques.

Landslides are a rare but hazardous geological phenomenon in Egypt, with the El Mokattam plateau situated in the eastern part of Cairo covering approximately 64 km2 and ranging in elevation from 50 to 205 m. This study aims to identify and monitor landslides in the area using various geophysical methods. Twelve Electrical resistivity tomography (ERT) profiles,twenty-two P-wave Seismic Refraction profiles, twenty-two Refraction microtremors profiles, three ground penetrating radar (GPR) profiles and borehole data were utilized to analyze the occurrence of landslides in the El Mokattam Plateau. Additionally, we employed a relatively new geophysical method, studying high-frequency microtremor sounds emitted from landslide collapses at 22 stations. Our analysis identified steep slopes, jointed or fractured rocks, and irrigation water as primary factors contributing to landslides, with irrigation water acting as a lubricant for clays and promoting ground sliding. Examination of high-frequency microtremor sounds revealed a potential correlation between vertical high-frequency spectra at 100 Hz and landslide collapses, which aids in the identification of landslide-prone zones. Therefore, we conclude that seismological studies, particularly spectral analysis of high-frequency and low-amplitude sounds (microtremors) emitted from soil, offer a promising approach for investigating landslides.

Impact of floods and landslides on rhizosphere bacterial communities: a high-throughput 16S rRNA gene sequencing study.

This study investigates the diversity and composition of soil bacterial communities in the rhizosphere of Attapadi and Nelliyampathy, prominent hill stations in Palakkad district, Kerala, India. The persistent flooding and landslides in 2018 and 2019 significantly impacted agricultural productivity in these regions. Utilizing high-throughput 16S rRNA gene sequencing (Illumina MiSeq), we conducted a comprehensive analysis of soil samples. Correlative assessments between soil parameters and microbial relative abundance at the phylum level revealed noteworthy positive associations. Notably, nitrogen (N) exhibited a positive relation with Crenarchaeota, Chloroflexi, Actinobacteriota, and Acidobacteriota; pH correlated with Firmicutes; organic carbon (OC) with WPS-2; and phosphorous with Proteobacteria. A total of 31,402 operational taxonomic units (OTUs) were identified, with the highest feature counts observed in undisturbed soils from Attapadi (AUD) and Nelliyampathy (NUD) (13,007 and 12,915, respectively). Disturbed soils in Nelliyampathy (ND) and Attapadi (AD) displayed a substantial decline in microbial diversity and composition, harbouring 1409 and 4071 OTUs, respectively. Alpha and beta diversity indices further underscored the more severe impairment of ND soils compared to AD soils. Interestingly, a majority of ND samples were landslide-affected (four out of five), while flood-affected soils accounted for four out of six AD samples. This indicates that landslides exert a more pronounced impact on microbial diversity and composition than floods. The observed decline in microbial count, composition, and diversity, even after 2 years of the disaster, raises concerns about potential threats to agricultural output. The findings emphasize the need for corrective measures, including the incorporation of microbial inoculum, to restore soil fertility in post-disaster landscapes.

Extensive pipeline location data resource: Integrating reported incidents, past environmental loadings, and potential geohazards for integrity evaluations in the U.S. Gulf of Mexico.

The U.S. Gulf of Mexico contains a complex network of existing, decommissioned, and abandoned oil and gas pipelines, which are susceptible to a number of stressors in the natural-engineered offshore system including corrosion, environmental hazards, and human error. The age of these structures, coupled with extreme weather events increasing in intensity and occurrence from climate change, have resulted in detrimental environmental and operational impacts such as hydrocarbon release events and pipeline damage. To support the evaluation of pipeline infrastructure integrity for reusability, remediation, and risk prevention, the U.S. Gulf of Mexico Pipeline and Reported Incident Datasets were developed and published. These datasets, in addition to supporting advanced analytics, were constructed to inform regulatory, industry, and research stakeholders. They encompass more than 490 attributes relating to structural information, incident reports, environmental loading statistics, seafloor factors, and potential geohazards, all of which have been spatially, and in some cases temporally matched to more than 89,000 oil and gas pipeline locations. Attributes were acquired or derived from publicly available, credible resources, and were processed using a combination of manual efforts and customized scripts, including big data processing using supercomputing resources. The resulting datasets comprise a spatial geodatabase, tabular files, and metadata. These datasets are publicly available through the Energy Data eXchange®, a curated online data and research library and laboratory developed by the U.S. Department of Energy's National Energy Technology Laboratory. This article describes the contents of the datasets, details the methods involved in processing and curation, and suggests application of the data to inform and mitigate risk associated with offshore pipeline infrastructure in the Gulf of Mexico.

Shifting from traditional landslide occurrence modeling to scenario estimation with a "glass-box" machine learning.

Extreme rainfall events represent one of the main triggers of landslides. As climate change continues to reshape global weather patterns, the frequency and intensity of such events are increasing, amplifying landslide occurrences and associated threats to communities. In this contribution, we analyze relationships between landslide occurrence and extreme rainfall events by using a "glass-box" machine learning model, namely Explainable Boosting Machine. What sets these models as a "glass-box" technique is their exact intelligibility, offering transparent explanations for their predictions. We leverage these capabilities to model the landslide occurrence induced by an extreme rainfall event in the form of spatial probability (i.e., susceptibility). In doing so, we use the heavy rainfall event in the Misa River Basin (Central Italy) on September 15, 2022. Notably, we introduce a rainfall anomaly among our set of predictors to express the intensity of the event compared to past rainfall patterns. Spatial variable selection and model evaluation through random and spatial routines are incorporated into our protocol. Our findings highlight the critical role of the rainfall anomaly as the most important variable in modeling landslide susceptibility. Furthermore, we leverage the dynamic nature of such a variable to estimate landslide occurrence under different rainfall scenarios.

Instrumental Monitoring of a Slow-Moving Landslide in Piedmont (Northwest Italy) for the Definition of Rainfall Thresholds.

The prediction and prevention of landslide hazard is a challenging topic involving the assessment and quantitative evaluation of several elements: geological and geomorphological setting, rainfalls, and ground motion. This paper presents the multi-approach investigation of the Nevissano landslide (Asti Province, Piedmont, NW Italy). It shows a continuous and slow movement, alongside few paroxysmal events, the last recorded in 2016. The geological and geomorphological models were defined through a field survey. An inventory of the landslide's movements and rainfall records in the period 2000-2016 was performed, respectively, through archive investigations and the application of "Moving Sum of Daily Rainfall" method, allowing for the definition of rain thresholds for the landslide activation (105 mm and 193 mm, respectively, in 3 and 30 days prior to the event). The displacements over the last 8 years (2016-2023) were monitored through an innovative in-continuum monitoring inclinometric system and Earth Observation (EO) data (i.e., relying on Interferometric Synthetic Aperture Radar, or InSAR data): it gave the opportunity to validate the rainfall thresholds previously defined. This study aims to provide information to public authorities for the appropriate management of the site. Moreover, the proposed workflow could be adopted as a guideline for investigating similar situations.

Analysis of deformation mechanism of rainfall-induced landslide in the Three Gorges Reservoir Area: Piansongshu landslide.

The Three Gorges Reservoir Area (TGRA) is characterized by unique geological features that increase its susceptibility to landslides. These slopes are especially prone to destabilization when influenced by external triggers like rainfall. This research focuses on the Piansongshu landslide within the TGRA, aiming at unraveling the complex internal deformation mechanisms of landslides triggered by rainfall and providing critical insights for their prevention and mitigation. The study begins with on-site geological surveys to meticulously examine the macroscopic signs and mechanisms of deformation. It then utilizes the GeoStudio numerical simulation software to assess the landslide's stability, focusing on the changes in internal seepage fields and stability under various rainfall scenarios. Results indicate that continuous rainfall leads to the formation of a temporary saturation zone on the slope, which gradually deepens. In regions with more pronounced deformation, the infiltration line at the leading edge of accumulation notably protrudes towards the surface. Notably, the stability coefficient of the secondary shear surface of the landslide fluctuates more significantly than that of the primary sliding surface. Higher rainfall intensity and longer duration are positively correlated with a more pronounced decrease in stability coefficients. The impact on stability also varies across different rainfall patterns. As rainfall infiltrates over time, the slope's safety factor gradually decreases. This reduction continues even post-rainfall, indicating a delayed restoration period before stability returns to a safe level. These results yield valuable data for forecasting and mitigating landslides.

Developing a regional scale construction and demolition waste landfill landslide risk rapid assessment approach.

In recent years, construction and demolition waste (CDW) landfills landslide accidents have occurred globally, with consequences varying due to surrounding environmental factors. Risk monitoring is crucial to mitigate these risks effectively. Existing studies mainly focus on improving risk assessment accuracy for individual landfills, lacking the ability to rapidly assess multiple landfills at a regional scale. This study proposes an innovative approach utilizing deep learning models to quickly locate suspected landfills and develop risk assessment models based on surrounding environmental factors. Shenzhen, China, with significant CDW disposal pressure, is chosen as the empirical research area. Empirical findings from this study include: (1) the identification of 52 suspected CDW landfills predominantly located at the administrative boundaries within Shenzhen, specifically in the Longgang, Guangming, and Bao'an districts; (2) landfills at the lower risk of landslides are typically found near the northern borders adjacent to cities like Huizhou and Dongguan; (3) landfills situated at the internal administrative junctions generally exhibit higher landslide risks; (4) about 70 % of these landfills are high-risk, mostly located in densely populated areas with substantial rainfall and complex topographies. This study advances landfill landslide risk assessments by integrating computer vision and environmental analysis, providing a robust method for governments to rapidly evaluate risks at CDW landfills regionally. The adaptable models can be customized for various urban and broadened to general landfills by adjusting specific indicators, enhancing environmental safety protocols and risk management strategies effectively.

Validation of the recycled backfill material for the landslide stabilization at a railway line.

In mountain areas landslides many times endanger safety of transport infrastructures, and these must be stabilized with retaining wall structures. In this paper the validation of a new composite as a backfill material for landslide stabilization with a large scale demo retaining wall is presented. The new composite was made from residues of paper industry, which uses for its production deinking process. New composite was validated with the laboratory tests, construction of small demo sites and at the end with a large demo retaining wall structure with a length of 50 m. It was concluded that the paper sludge ash and the paper sludge are in proportion 70:30, compacted on the optimal water content and maximum dry density, reached sufficient uniaxial compressive and shear strength. However, the composite's hydration processes required the definition of an optimal time between the composite mixing and installation. In 2019, the retaining wall structure from the new composite was successfully built. The large demo structure is an example of the knowledge transfer from the laboratory to the construction site, in which composite and installing technology could be verified.

A systematic review on rainfall thresholds for landslides occurrence.

The study of rainfall thresholds is vital in understanding the factors that trigger landslides, being one of the criteria applied to landslide early warning systems that aim to mitigate their consequences. These thresholds enable the prediction of landslide occurrences as a function of rainfall measurements. This work presents an overview of the parameters involved in defining rainfall thresholds based on scientific articles published between 2008 and 2021 that discuss the subject through statistical or physical methods. These articles provided data such as publication information, threshold types, details on the data used in the works, methodology, and application of the threshold in early warning systems. There was a significant increase in research papers on this theme during this period, possibly due to the strategies advocated by the Sendai Framework. However, some regions of the world severely affected by landslides are barely mentioned in these studies. The results indicate specific trends, such as those found in the methods used to define rainfall thresholds and the parameters relating to the database when a statistical approach was used. Certain deficiencies were found, such as those concerning geological-geotechnical conditions for categorizing thresholds, the time scales of rainfall data, rain gauge density, and the criteria to define the accumulated rainfall period to be considered.

Analysis of the spatial distribution of the landslides triggered by the 1923 Great Kanto Earthquake, Japan.

The Great Kanto Earthquake that occurred in the southern part of Kanto district, Japan, on September 1, 1923, was reported to have triggered numerous landslides (over 89,080 slope failures over an area of 86.32 km2). This study investigated the relationship between the landslide occurrence caused by this earthquake and geomorphology, geology, soil, seismic ground motion, and coseismic deformation. We found that a higher landslide density was mainly related to a larger absolute curvature and a higher slope angle, as well as to several geological units (Neogene plutonic rock, accretionary prism, and metamorphic rocks). Moreover, we performed decision tree analyses, which showed that slope angle, geology, and coseismic deformation were correlated to landslide density in that order. However, no clear correlation was found between landslide density and seismic ground motion. These results suggest that landslide density was greater in areas of large slope angle or fragile geology in the area with strong shaking enough to trigger landslides.

Comparison between mechanical and hydrological reinforcement effects of cultivated plants on shallow slope stability.

Root reinforcement, provided by plants in soil, can be exerted by a mechanical effect, increasing soil shear strength for the presence of roots, or by a hydrological effect, induced by plant transpiration. No comparisons have been still carried out between mechanical and hydrological reinforcements on shallow slope stability in typical agroecosystems. This paper aims to compare these effects induced by sowed fields and vineyards and to assess their effects towards the shallow slope staibility. Root mechanical reinforcement has been assessed through Root Bundle Model-Weibull. Root hydrological reinforcement has been evaluated using an empirical relationship with monitored or modelled pore water pressure. Each reinforcement has been inserted in a stability model to quantify their impacts on susceptibility towards shallow landslides. Considering the same environment, corresponding to a typical agroecosystem of northern Italian Apennines, land use has significant effects on saturation degree and pore water pressure, influencing hydrological reinforcement. Root hydrological reinforcement effect is higher in summer, although rainfall-induced shallow landslides rarely occur in this period due to dry soil conditions. Instead, in wet and cold periods, when shallow landslides can develop more frequently, the stabilizing contribution of mechanical reinforcement is on average higher than the hydrological reinforcement. In vineyards, the hydrological reinforcement effect could be observed also during autumn, winter and spring periods, giving a contribution to slope stability also in these conditions. This situation occurs when plants uptake enough water from soil to reduce significantly pore water pressure, guaranteeing values of hydrological reinforcement of 1-3 kPa at 1 m from ground, in agreement with measured mechanical root reinforcement (up to 1.6 kPa). These results suggest that both hydrological and mechanical effects of vegetation deserve high regard in susceptibility towards shallow landslides, helping in selection of the best land uses to reduce probability of occurrence of these failures over large territories.

Space-time data-driven modeling of precipitation-induced shallow landslides in South Tyrol, Italy.

Shallow landslides represent potentially damaging processes in mountain areas worldwide. These geomorphic processes are usually caused by an interplay of predisposing, preparatory, and triggering environmental factors. At regional scales, data-driven methods have been used to model shallow landslides by addressing the spatial and temporal components separately. So far, few studies have explored the integration of space and time for landslide prediction. This research leverages generalized additive mixed models to develop an integrated approach to model shallow landslides in space and time. We built upon data on precipitation-induced landslide records from 2000 to 2020 in South Tyrol, Italy (7400 km2). The slope unit-based model predicts landslide occurrence as a function of static and dynamic factors while seasonal effects are incorporated. The model also accounts for spatial and temporal biases inherent in the underlying landslide data. We validated the resulting predictions through a suite of cross-validation techniques, obtaining consistent performance scores above 0.85. The analyses revealed that the best-performing model combines static ground conditions and two precipitation time windows: a short-term cumulative precipitation of 2 days before the landslide event and a medium-term cumulative precipitation of 14 days. We demonstrated the model's predictive capabilities by predicting the dynamic landslide probabilities over historical data associated with a heavy precipitation event on August 4th and August 5th, 2016, and hypothetical non-spatially explicit precipitation (what-if) scenarios. The novel approach shows the potential to integrate static and dynamic landslide factors for large areas, accounting for the underlying data structure and data limitations.

Study of substrata of a slope susceptible to landslide in hilly environment using a geophysical method in the Nilgiris, India.

Landslides are one of the prevailing threats to life that cause huge loss to the environment. Around 3.7 million km2 of the area is exposed to landslides globally, and 820,000 km2 is at high risk for landslides in India. Rainfall and earthquakes are the two primary landslide-causing variables in India. The Nilgiris district which is in the south-western part of India is more prone to rainfall-induced landslides. This study intends to calculate the depth of the slip surface on a slope (Lovedale area, the Nilgiris) in the event of a future landslide using Multichannel Analysis of Surface Waves (MASW) and validate using bore log data. During November 2009 rainfall, a shallow landslide occurred at the toe of this slope. There is a greater chance that a landslide will occur again in the event of rainfall in the future. To comprehend how the sub-strata vary, and to forecast the depth of a prospective failure surface, the shear wave velocity (Vs) obtained from MASW proved beneficial. Slip surfaces, one at a shallow depth and another at a deeper depth, were found based on the shear wave velocity and bore log data. The importance of the MASW output in the engineering properties of soil was also studied. The compressional velocity (Vp) and shear wave velocity obtained from MASW were evaluated for their applicability in calculating the elastic moduli of soil. It was established that shear wave velocity was of greater significance than compressional velocity. The MASW results can be further used as a preliminary data for analysing the stability of the slope, reactivation of landslides, and landslide early warning system.

Laboratory Assessment of an In-Place Inclinometer Chain for Structural and Geotechnical Monitoring.

The necessity of early warning systems to ensure people's safety requires the usage of real-time monitoring instrumentation. To meet the required real-time monitoring performance, in-place inclinometer systems represent one of the most common solutions to obtain accurate measures over time. This paper presents the results of a laboratory tests campaign performed on the prototypes and preproduction samples of an in-place inclinometer chain for structural and geotechnical monitoring applications. First, each element sensor has been calibrated to reach a proper level of measure accuracy. Eventually, laboratory tests are carried out on both a single instrument (element) and on the complete measurement chain (system). The adopted centering device, obtained as a combination of a Cardan joint and four spring plungers avoids bending of elements by preventing fictitious displacement measurements and permits the creation of a kinematic chain that accommodates the displacements of a grooveless tube. A specially designed and constructed test set-up that permits assigning a movement to each node has been employed to test a specifically designed centering device and check the system stability over time. Different scenarios have been investigated to determine the accuracy and repeatability of the measures in replicating real cases. The results demonstrated the necessity of validating a measurement chain by analyzing its overall behavior and not limiting the study on the performances of a single element.

Structural and geomorphological study of Bradost, Chinara, Shireen and Sare Musa anticlines, Iraqi Kurdistan Region.

The Bradost and Chinara mountains are two well-known geomorphic features in the Iraqi Kurdistan Region (IKR), forming two anticlines, besides Shireen and Sare Musa anticlines, which are located north of the Bradost anticline, all four anticlines trend NW - SE. The four anticlines are dissected by the Greater Zab River that swings along its course within the anticlines due to tens of very old landslides and/or plunges. The four studied anticlines are dissected by different thrust faults, which extend for a few kilometers. The thrust faults trend NW - SE; however, locally they deflect from the main trend. The Lower Jurassic rocks are the oldest exposed rocks in the studied area, whereas the rocks of the Bekhme Formation form the carapace of the Bradost and Chinara anticlines. Different structural and geomorphological features were interpreted from satellite images and those which are accessible were checked in the field, all of them indicate the four anticlines exhibit lateral growth. We have measured different aspects to elucidate the type of folds. The four anticlines are Detachment folds, with shallow decollement, which ranges in depth between (100-250) m.

Ensemble of fuzzy-analytical hierarchy process in landslide susceptibility modeling from a humid tropical region of Western Ghats, Southern India.

The western flanks of the Western Ghats are one of the major landslide hotspots in India. Recent rainfall triggered landslide incidents in this humid tropical region necessitating the accurate and reliable landslide susceptibility mapping (LSM) of selected parts of Western Ghats for hazard mitigation. In this study, a GIS-coupled fuzzy Multi-Criteria Decision Making (MCDM) technique is used to evaluate the landslide-susceptible zones in a highland segment of the Southern Western Ghats. Fuzzy numbers specified the relative weights of nine landslide influencing factors that were established and delineated using the ArcGIS, and the pairwise comparison of these fuzzy numbers in the Analytical hierarchy process (AHP) system resulted in standardized causative factor weights. Thereafter, the normalized weights are assigned to corresponding thematic layers, and finally, a landslide susceptibility map is generated. The model is validated using the area under the curve values (AUC) and F1 scores. The result reveals that about 27% of the study area is classified as highly susceptible zones followed by 24% area in moderately susceptible zone, 33% in low susceptible, and 16% in a very low susceptible area. Also, the study shows that the plateau scarps in the Western Ghats are highly susceptible to the occurrence of landslides. Moreover, the predictive accuracy estimated by the AUC scores (79%) and F1 scores (85%) shows that the LSM map is trustworthy for future hazard mitigation and land use planning in the study area.

Modeling landslide activity and sediment connectivity after eruptions: Insights from the Blanco River (Chile).

Volcanic eruptions can disrupt entire river basins by affecting the hydro-geomorphic characteristics of channel networks and hillslopes. Reports suggest a pulsed and delayed increase in landslide activity following the eruptions, which, depending on the degree of linkage between hillslopes and channels, i.e. sediment connectivity, can represent a massive source of sediment input for the fluvial system. Therefore, predicting landslide occurrence and sediment connectivity is fundamental for management risk strategies, especially in such dynamic and complex environments. The aim of this work is to develop and offer a more reliable approach to map the areas susceptible to landslides and connected to the active channel in a catchment impacted by volcanic eruption. The analyses were carried out in the Blanco River catchment in southern Chile, affected by the Chaitén eruption (2008-09). A combined approach is presented, based on landslide susceptibility models, carried out multi-temporally (from 2010 to 2019), and a threshold-based sediment connectivity map. The results showed that the highest landslide occurrence was reported 4 years after the eruption, whereas the faster increase in the overall area affected was observed only after 7 years. Landslide susceptibility models showed high accuracy when applied in the same year, but were less accurate in predicting future occurrences. This result is ascribed to the dynamic conditions of the vegetation, regenerating quickly after the mass movements. Nevertheless, considering the potential sources of error, the combined landslide susceptibility-connectivity map for the year 2019 well-identified relevant areas for catchment management. The largest part of the catchment was found non-susceptible and disconnected, while areas classified as susceptible and connected represent only 3.1 %. The application of this novel approach allowed to unravel the geomorphic trajectory of the study area and, more importantly, can represent a benchmark for future applications in other catchments affected by large disturbances.

Application of Bagging, Boosting and Stacking Ensemble and EasyEnsemble Methods for Landslide Susceptibility Mapping in the Three Gorges Reservoir Area of China.

Since the impoundment of the Three Gorges Reservoir area in 2003, the potential risks of geological disasters in the reservoir area have increased significantly, among which the hidden dangers of landslides are particularly prominent. To reduce casualties and damage, efficient and precise landslide susceptibility evaluation methods are important. Multiple ensemble models have been used to evaluate the susceptibility of the upper part of Badong County to landslides. In this study, EasyEnsemble technology was used to solve the imbalance between landslide and nonlandslide sample data. The extracted evaluation factors were input into three bagging, boosting, and stacking ensemble models for training, and landslide susceptibility mapping (LSM) was drawn. According to the importance analysis, the important factors affecting the occurrence of landslides are altitude, terrain surface texture (TST), distance to residences, distance to rivers and land use. The influences of different grid sizes on the susceptibility results were compared, and a larger grid was found to lead to the overfitting of the prediction results. Therefore, a 30 m grid was selected as the evaluation unit. The accuracy, area under the curve (AUC), recall rate, test set precision, and kappa coefficient of a multi-grained cascade forest (gcForest) model with the stacking method were 0.958, 0.991, 0.965, 0.946, and 0.91, respectively, which a significantly better than the values produced by the other models.

ArcGIS-based protocol to calculate the area fraction of landslide for multiple catchments.

The area fraction of specific kinds of information in a catchment provides parameters to be utilized in catchment-scale analyses. For example, the area fraction of soil movement caused by landslides is an indicator for the estimation of the magnitude of landslides. However, catchment-scale analyses often require applying the same processing to higher numbers of study catchments, making it a time-consuming process. Here an ArcGIS-based method has been presented to reduce cumbersome procedures for the calculation of the area fraction of several target surface data. The method applies automated and iterative processing to multiple catchments, whose location and scale are defined by users. This method may prove to be useful for calculating the area fraction of parameters other than landslide area (e.g., specific land use or lithology) within a framework of catchment-scale analysis.•An Arcgis-based method to calculate the area fraction of landslide area in catchments.•Manual work is reduced by automated and iterative processing based on ModelBuilder.•It can be used to get the area fraction of several surface information in catchments.