Evaluation of Landslide Susceptibility of Şavşat District of Artvin Province (Turkey) Using Machine Learning Techniques

The aim of this study is to produce landslide susceptibility maps of Şavşat district of Artvin Province using machine learning (ML) models and to compare the predictive performances of the models used. Tree-based ensemble learning models, including random forest (RF), gradient boosting machines (GBM), and extreme gradient boosting (XGBoost), were used in the study. A landslide inventory map consisting of 85 landslide polygons was used in the study. The inventory map comprises 32,777 landslide pixels at 30 m resolution. Randomly selected 70% of the landslide pixels were used for training the models and the remaining 30% were used for the validation of the models. In susceptibility analysis, altitude, aspect, curvature, distance to drainage network, distance to faults, distance to roads, land cover, lithology, slope, slope length, and topographic wetness index parameters were used. The validation of the models was conducted using success and prediction rate curves. The validation results showed that the success rates for the GBM, RF, and XGBoost models were 91.6%, 98.4%, and 98.6%, respectively, whereas the prediction rate were 91.4%, 97.9%, and 98.1%, respectively. Therefore, it was concluded that landslide susceptibility map produced with XGBoost model can help decision makers in reducing landslide-associated damages in the study area.

Landslide Inventory along a National Highway Corridor in the Hissar-Allay Mountains, Central Tajikistan

An increasing amount of landslides leading to significant human and economic consequences is a primary concern for the government of Tajikistan and local authorities. Based on the Committee on Emergency Situations data, from 1996 to 2018, there were 3460 emergencies and more than 1000 fatalities because of earthquake-triggered and rainfall-induced landslides in the region. In addition, landslides caused severe damage to houses and infrastructure facilities due to the population’s lack of landslide hazard knowledge. Therefore, current research focuses on developing a regional-scale landslide inventory map in the Hissar–Allay region, central Tajikistan, where the population density is much higher than at other mountainous territories. In recent decades, the enhancements in geographic information systems, the open access to high-resolution remote sensing data, and an extensive field survey allowed us to identify 922 landslides possible along the highway corridor in the Hissar–Allay region. Based on Varnes’s system, these landslides are classified into four categories: debris flows, rockfalls, shallow landslides, and complex (deep-seated) landslides, considering landslides morphology, geology, deformation of slopes, degree and aspect of slopes, and weathered and disintegrated zones on slopes in the study area. The results show that 8.24% of the total study area is affected by landslides. Along the highway corridor in the Hissar–Allay region there are 96 bodies of deep-seated landslides and 216 rockfall catchments, 273 debris flow catchments, and 313 shallow landslides. Thus, shallow landslides are the most frequent type of movement. In addition, landslide frequency-area distribution analysis shows that shallow landslides are frequent with an area of 1.88E+04 m2; most frequent debris flow channels have a place of 5.58E+05 m2; rockfalls, for its part, are rife with an area of 1.50E+05 m2, and frequent complex landslides have an area of 4.70E+06 m2. Furthermore, it was found out that slopes consist of Silurian formation comprise shales, pebbles, sands, loams, and limestones, metamorphic clays are exposed to landslides more than other geological formations because of the layered structure and their broad spatial distribution in the study area. As the first applied research to compile a landslide inventory map in the Hissar–Allay region on the regional scale, our study provides a sound basis for future explorations of landslide susceptibility, hazard, and risk assessment for this region.

UPDATING LANDSLIDE INVENTORY MAPS USING HIGH RESOLUTION DIGITAL ORTHOPHOTOS AND DIGITAL SURFACE AND ELEVATION MODELING: THE CASE STUDY OF BRUSQUE CITY, SANTA CATARINA STATE, BRAZIL

This study reports the updating of the landslide inventory map of Brusque city, State of Santa Catarina, Southern Brazil. Twenty-six digital orthophotos acquired in 2010 with a ground resolution of 0.4 meters were analyzed together with shaded relief images obtained by Digital Surface and Digital Elevation modelling with spatial resolution of 1 meter. These remote sensing products were treated, analyzed and visualized in a Geographic Information System – GIS environment. The landslide inventory included a total of 500 landslides, corresponding to a mean density of 1.76 landslides per km2. The total area of landslide occurrences is 0.81 km2, which corresponds to 0.29% of the study area. 0.22 km2 of the total area landslides occur inside the urban perimeter and 0.59 km2 outside Brusque. The geological context and the occurrence of landslides were analyzed together: 277 landslides affect altered metamorphic rocks, 179 landslides granite, and 44 landslides unconsolidated sediments. The updated landslide inventory map showed that 80% of mapped landslides occur in areas of high and moderate susceptibility.

Factors Contributing to Landslide Susceptibility of the Kope Formation, Cincinnati, Ohio

ABSTRACT The objective of this study was to evaluate the factors that contribute to the high frequency of landslides in the Kope Formation and the overlying colluvial soil present in the Cincinnati area, southwestern Ohio. The Kope Formation consists of approximately 80 percent shale inter-bedded with 20 percent limestone. The colluvium that forms from the weathering of the shale bedrock consists of a low-plasticity clay. Based on field observations, LiDAR data, and information gathered from city and county agencies, we created a landslide inventory map for the Cincinnati area, identifying 842 landslides. From the inventory map, we selected 10 landslides that included seven rotational and three translational slides for detailed investigations. Representative samples were collected from the landslide sites for determining natural water content, Atterberg limits, grain size distribution, shear strength parameters, and slake durability index. For the translational landslides, strength parameters were determined along the contact between the bedrock and the overlying colluvium. The results of the study indicate that multiple factors contribute to landslide susceptibility of the Kope Formation and the overlying colluvium, including low shear strength of the colluvial soil, development of porewater pressure within the slope, human activity such as loading the top or cutting the toe of a slope, low to very low durability of the bedrock that allows rapid disintegration of the bedrock and accumulation of colluvial soil, undercutting of the slope toe by stream water, and steepness of the slopes.

A comparison of different Artificial Intelligence and Machine Learning methods for Gully Erosion Susceptibility Mapping in the Upper Narmada Basin

Gully erosion (GE) is one of the most important mechanisms of soil loss worldwide. In this study, various machine learning techniques such as Classification and Regression Trees (CART), Random Forest (RF) and Artificial Neural Networks (ANN), have been used to ascertain gully erosion susceptibility (GES) in the Upper Narmada Basin (UNB). The mapping and analysis was achieved using R programming and ArcGIS 10.8 software. Initially, a gully inventory map (GIM) of 1501 gully locations was prepared from Sentinel-2 and Google Earth images and extensive field surveys. Out of the 1501 gullies in the study area, 1051 gully locations (about 70%) were used for training and 450 gully locations (about 30%) were used for validating the models. For GES modeling, 12 gully conditioning factors (GCFs) were used and the relationships between these GCFs and gully erosion were evaluated. The GES maps were prepared using the CART, RF and ANN models and divided into three susceptibility-based classes: low, moderately and highly susceptible GE classes. A large part of the study area was found to be highly susceptible to GE. Subsequent validation tests proved the high efficacy of these models in ascertaining the GES. The RF model was found to perform best compared to the others in this respect with an AUC-ROC value of 0.78. This model can therefore be used not only in the UNB but in other such areas to evaluate the GES zones and thereby aid in framing suitable measures to mitigate soil loss through gully erosion.

Waterlogging Risk Assessment: An Undervalued Disaster Risk in Coastal Urban Community of Chattogram, Bangladesh

In recent years, rainfall-induced waterlogging has become a common hazard in the highly urbanized coastal city of Chattogram, Bangladesh, resulting in a high magnitude of property damage and economic loss. Therefore, the primary objective of this research was to prepare a waterlogging inventory map and understand the spatial variations of the risk by means of hazard intensity, exposure, and vulnerability of waterlogging. In this research, the inventory map and factors influencing waterlogging hazards were determined from a participatory survey, and other spatial data, including land elevation, population, and structural data, were collected from secondary sources. The analytical hierarchy process was applied to measure the hazard intensity, and the exposure and vulnerability were estimated by overlaying the spatial data onto the hazard intensity map. A total of 58 locations were identified as waterlogging affected, which covered ~8.42% of the city area. We showed that ~3.03% of the city area was greatly vulnerable to waterlogging in terms of their social, infrastructure, critical facilities, economic, and environmental vulnerabilities. The obtained waterlogging risk index map suggested that ~2.71% of the study area was at very high risk, followed by moderate (~0.15%), low (~3.89%), and very low (~1.67%). The risk analysis presented in this study was a simple method that can be applied to assess the relative risk of waterlogging in different regions, and the results were applicable to the prevention and mitigation of waterlogging for Chattogram City.

Geospatial Modeling of Landslide Vulnerability and Simulating Spatial Correlation with Associated Factors in Bandarban District

The present study aims at identifying and predicting landslide vulnerable areas in Bandarban District of Chittagong Hill Tracts (CHT) using weighted overlaying of the multiple geospatial layers to determine landslide hazard areas. The historical landslide inventory map was prepared using Google Earth image and through PRA technique. Then ten landslide triggering factors including landuse, rainfall, slope, elevation, cut-fill, soil types, geology, distance to rivers, roads and stream orders, population density, income, education of the inhabitants were chosen as effective factors on a landslide in the study area. Subsequently, the landslide vulnerability map was constructed using the weighted overlay model in Geographic Information System (GIS). Bandarban District has 348 landslides vulnerable locations. Among them, 6 are extremely vulnerable and 342 are highly vulnerable to landslides. Model results show that the Upazila Ruma and Thanchi are extremely vulnerable to landslides. About 91 percent of the landslides will occur within 10 degrees of slope, about 65 percent will occur within 50 meters elevation. The model shows that there is a strong relationship between landslides and physical, economic and social variables. The Dhaka University Journal of Earth and Environmental Sciences, Vol. 8(2), 2019, P 51-56

Landslide susceptibility zonation in the Tartagal River basin, Sierras Subandinas, Salta, Argentina

On February 2009 intense rainfall triggered landslides in the Tartagal River basin that evolved into a debris flow that caused severe flooding in the town of Tartagal, Salta, Argentina. Based on these events, this paper presents a first attempt to map the landslides susceptibility in the Tartagal River basin. First, we elaborated an inventory map by using a 10 m pixel SPOT image acquired just after the disaster. Second, we evaluated a set of conditioning factors, which included lithology, slope and curvature; we derived the topographical variables from a 12.5 m pixel digital elevation model (DEM) based on a stereo-pair of satellite images ALOS-PRISM. Finally, we used these conditioning factors and the 2009 landslides inventory map as input for a heuristic model to elaborate the susceptibility map. The results indicated that landslides affected an area of 8 km2 and that at least 2.2x106 m3 of material were removed. The susceptibility map identified zones of low, moderate, high and very high susceptibility that occupied 18, 22, 25 and 17 km2, respectively. Accuracy assessment using data covering landslides occurred in 2006 showed that 95% of them fell within the high and very high susceptibility areas. The results presented herein provide vital baseline information for future studies and may contribute for the development of landslide hazard mitigation strategies.

Waterlogging Risk Assessment: An Undervalued Disaster Risk in Coastal Urban Community of Chattogram, Bangladesh

In recent years, rainfall-induced waterlogging has become a common hazard in the highly urbanized coastal city of Chattogram, Bangladesh resulting in high magnitude of property damage and economic loss. Therefore, the primary objective of this research is to prepare a waterlogging inventory map and understand the spatial variation of the risk by means of hazard intensity, exposure, and vulnerability of waterlogging. In this research, the inventory map and factors influencing waterlogging hazard were determined from a participatory survey and other spatial data including land elevation, population, and structural data were collected from secondary sources. Analytical Hierarchy Process was applied to measure the hazard intensity and the exposure and vulnerability were estimated by overlaying the spatial data onto the hazard intensity map. A total of 58 locations in 22 wards have been identified as waterlogging affected, which covers ~8.42% of the city area. Obtained waterlogging vulnerability index map suggests that ward no. 5, 6, 16, 17, and 33 are greatly vulnerable to waterlogging in terms of their social, infrastructure, critical facilities, economic and environmental vulnerability. We show that ~2.71% of the study area is at very high risk, while the risk score is considerably higher for ward no. 5, 8, 17, 19, and 33.

Rapid mapping of landslides in the Western Ghats (India) triggered by 2018 extreme monsoon rainfall using a deep learning approach

Rainfall-induced landslide inventories can be compiled using remote sensing and topographical data, gathered using either traditional or semi-automatic supervised methods. In this study, we used the PlanetScope imagery and deep learning convolution neural networks (CNNs) to map the 2018 rainfall-induced landslides in the Kodagu district of Karnataka state in the Western Ghats of India. We used a fourfold cross-validation (CV) to select the training and testing data to remove any random results of the model. Topographic slope data was used as auxiliary information to increase the performance of the model. The resulting landslide inventory map, created using the slope data with the spectral information, reduces the false positives, which helps to distinguish the landslide areas from other similar features such as barren lands and riverbeds. However, while including the slope data did not increase the true positives, the overall accuracy was higher compared to using only spectral information to train the model. The mean accuracies of correctly classified landslide values were 65.5% when using only optical data, which increased to 78% with the use of slope data. The methodology presented in this research can be applied in other landslide-prone regions, and the results can be used to support hazard mitigation in landslide-prone regions.