Avalanche Susceptibility Mapping Research Articles

Human-centered avalanche susceptibility mapping (H-CASM): shifting the cartographic emphasis of backcountry avalanche susceptibility maps

ABSTRACT With backcountry winter travel increasing in popularity, the importance of accessible avalanche safety information is crucial. Widely accessible avalanche backcountry maps use slope shading to show where avalanches are likely to start. This paper presents a new methodology called Human-Centered Avalanche Susceptibility Mapping (H-CASM), which shifts the cartographic emphasis of backcountry avalanche maps from the hazard of an avalanche starting in a specific location to risk to a human traveling in the terrain. Utilizing ArcGIS, RAMMS avalanche runout model, and Python, we developed H-CASM as an open-source code to incorporate possible avalanche runouts, terrain traps, and connected slopes while trying to reduce instances of false certainty in visualizations. To provide a more comprehensive representation of susceptible terrain, we integrated aspect, distance to ridges, and vector ruggedness in addition to slope for start zone susceptibility using high-resolution one-meter Digital Elevation Models (DEMs). Avoiding the use of discrete susceptibility categories, we applied continuous gradient shading reducing false certainty in the final visualization. This proof of concept sets a new precedent for the future of backcountry avalanche susceptibility maps.

Avalanche Susceptibility Mapping by Investigating Spatiotemporal Characteristics of Snow Cover Based on Remote Sensing Imagery along the Pemo Highway—A Critical Transportation Road in Tibet, China

The Pemo Highway is a critical transportation road to Medog County in the Tibet Plateau (TP). Since its completion in 2021, the Pemo Highway has been prone to frequent avalanches due to heavy rainfall and snowfall. Despite the lack of monitoring stations along the highway and limited research conducted in this area, remote sensing imagery provides valuable data for investigating avalanche hazards along the highway. In this paper, we first investigated the spatiotemporal characteristics of snow cover along the Pemo Highway over the past two years based on the GEE platform. Second, we integrated snow, topography, meteorology, and vegetation factors to assess avalanche susceptibility in January, February, and March 2023 along the highway using the AHP method. The results reveal that the exit of the Duoshungla Tunnel is particularly susceptible to avalanches during the winter months, specifically from January to March, with a significant risk observed in March. Approximately 3.7 km in the direction of the tunnel exit to Lager is prone to avalanche hazards during this period. The recent “1.17 avalanche” event along the Pemo Highway validates the accuracy of our analysis. The findings of this paper provide timely guidance for implementing effective avalanche prevention measures on the Pemo Highway.

Snow avalanche susceptibility mapping using novel tree-based machine learning algorithms (XGBoost, NGBoost, and LightGBM) with eXplainable Artificial Intelligence (XAI) approach

Snow avalanche susceptibility mapping using novel tree-based machine learning algorithms (XGBoost, NGBoost, and LightGBM) with eXplainable Artificial Intelligence (XAI) approach

Snow avalanche susceptibility mapping from tree-based machine learning approaches in ungauged or poorly-gauged regions

Mapping avalanche susceptibility is essential for disaster management, especially in ungauged or poorly-gauged regions. Most existing machine-learning methods have a shortfall in multi-source, heterogeneous spatial data requirements and overfitting. Faced with these deficiencies, this study has developed four novel tree-based machine-learning models (Catboost, LightGBM, RF, and XGBoost) to predict avalanche susceptibility, along with the data support of remote sensing images (e.g., Superview-1, Sentinel-2, and Calibrated Enhanced-Resolution Passive Microwave Daily Brightness Temperature), Meteorological data (e.g., WRF simulation), topography, and limited observation in Tianshan Mountains, China. There are 21 causative factors in the selection, training, and validation processes. Then, the SHAP value is introduced to quantify the variable-based share of contribution to the possibility of avalanches on both global and local scales. The results demonstrate the following: (1) All four models are potent in assessing avalanche susceptibility due to similar patterns, with Catboost being the best with its eight indices: Accuracy (0.9249), TPR (0.9920), FAR (0.0080), TNR (0.8904), PPR (0.8229), NPR (0.0046), CSI (0.8175), and HSS (0.8404) superior to others; (2) The Catboost provides a reliable susceptibility map illustrating moderate and high susceptible areas up to 54.12% of the total, mainly concentrated in the west and southeast; (3) Topographic factors (distance to rivers, aspect, and relative slope position) and meteorology (precipitation) are the most effective for susceptibility modeling. The results above indicate the significant potential of tree-based machine-learning with multi-source and heterogeneous spatial data in obtaining high-quality susceptibility maps without too much field observation. It is of great importance for avalanche prevention in regions with complex terrain conditions and sparse data.

Towards Linking the Sustainable Development Goals and a Novel-Proposed Snow Avalanche Susceptibility Mapping

Towards Linking the Sustainable Development Goals and a Novel-Proposed Snow Avalanche Susceptibility Mapping

Integrating Traditional Knowledge with GIS for Snow Avalanche Susceptibility Mapping in Kargil-Ladakh Region of Trans-Himalayan India

Integrating Traditional Knowledge with GIS for Snow Avalanche Susceptibility Mapping in Kargil-Ladakh Region of Trans-Himalayan India

SNOW AVALANCHE SUSCEPTIBILITY MAPPING FOR DAVOS, SWITZERLAND

Abstract. Snow avalanches are among destructive hazards occurring in mountainous regions and spatial distribution (susceptibility) of their occurrences needs to be considered for spatial planning and disaster risk mitigation efforts. The susceptibility assessment is the first step in avalanche disaster management and can be carried out using high resolution geospatial data and machine learning (ML) algorithms. In this study, we have assessed the snow avalanche susceptibility in Davos, Switzerland using an inventory delineated on satellite imagery in a previous study. The conditioning factors used for the avalanche susceptibility assessment include elevation, slope, plan curvature, profile curvature, aspect, topographic position index, topographic ruggedness index, topographic wetness index, land use and land cover, lithology, distance to road, and distance to the river. Two ML algorithms, the logistic regression (LR) and the random forest (RF), were comparatively assessed using validation data split from the training data (30/70). The prediction performances of both models were assessed based on the area under the receiver operating characteristic curve (ROC-AUC) value. Although the AUC value obtained from the LR method was relatively low (0.74), the value obtained from the RF (0.96) demonstrated high performance and usability of this approach. The results indicate that the RF method can successfully produce an avalanche susceptibility map for the region, although potential improvements may be possible by investigating various input features and ML algorithms as well as by classifying the starting and runout zones of the avalanche data separately. Furthermore, the accuracy is expected to increase by using a larger training dataset.

Application of machine learning methods for snow avalanche susceptibility mapping in the Parlung Tsangpo catchment, southeastern Qinghai-Tibet Plateau

Determining the snow avalanche-prone areas is fundamental for risk mitigation in the snowy mountains of the Qinghai-Tibet Plateau, especially in the context of current climate warming. We use the Parlung Tsangpo catchment in southeastern Tibet as an example, collect 381 snow avalanches through field surveys. On this basis, two sample sets, three combinations of conditioning factor, and four machine learning algorithms involving Support vector machine (SVM), K-nearest neighbors (KNN), Classification and Regression Tree (CART), and Multilayer perceptron (MLP) are compared to produce snow avalanche susceptibility maps. The Kappa coefficient and ROC curve are used for the accuracy test. The results show that four basic models all have good prediction capability, and the most robust one is SVM, while the comparative sample set and two comparative combinations of factors underperform. The high snow avalanche-prone areas are mainly distributed at the Ranwu-Yupu section along the National Road 318 in the upper narrow valley section of Parlung Tsangpo River, the ridges of the mountains on both sides of the middle and lower reaches, and the narrow valley sections of the tributaries. The finding not only verifies the applicability of these four machine learning algorithms in the snow avalanche susceptibility mapping, but also points out the key areas for the local snow avalanche disaster reduction and prevention. This study can provide a good reference for other snowy mountainous regions lacking well-documents of snow avalanches.

Evaluating novel hybrid models based on GIS for snow avalanche susceptibility mapping: A comparative study

Snow avalanches cause economic losses in many parts of the world, especially in mountainous areas. Due to its changing climates and the complex topography of mountainous regions in Iran, avalanches are significant problems. Therefore, modeling snow avalanches to predict them is essential for sustainable and safe development, appropriate infrastructure, planning for residential development, and loss prevention. This study assesses the performances of multi-criteria decision-making models and machine-learning hybrid models for snow avalanche susceptibility mapping in northwestern Iran's Sirvan watershed based on the weight of analytic network process, weighted linear combination, and frequency ratio. Using multicollinearity analysis and Pearson correlation, eleven snow avalanche conditioning factors were deemed appropriate for use. The areas under the curve (AUC), root mean square error (RMSE), seed cell area index, and snow avalanche relative index were used to validate snow avalanche susceptibility maps. The results showed that all hybrid models performed very well for the determination of the susceptibility of locations to avalanching. Based on the results of the validation analyses, the analytic network process-frequency ratio hybrid model was best among the models (AUC = 0.941, RMSE = 0.3). The resulting maps provide strong evidence to guide efforts to prevent the exposure to snow avalanches in the Sirvan watershed, and these hybrid models can be tested in other regions affected by snow avalanches.

Avalanche susceptibility mapping with the use of frequency ratio, fuzzy and classical analytical hierarchy process for Uzungol area, Turkey

As expressed in many studies in literature, the dynamic nature of snow avalanches is affected by the meteorological, environmental, topographical and human-related factors of the region. To better appreciate and forecast the snow avalanche, one of the hydro-geological disasters, becomes possible thanks to the approaches that take these parameters into consideration. They are commonly seen in mountainous areas, cause loss of life and property when they reach settlements, energy transmission lines, highways, ski resorts and other structures. Field studies, GIS, remote sensing, and various numerical avalanche modeling programs are used in determining avalanche-prone zones. There are various methods for producing avalanche susceptibility maps. In this study, susceptibility maps were created using the Frequency Ratio (Fr), Analytical Hierarchy Process (AHP), and Fuzzy-AHP (FAHP) methods for avalanche potential, in the tourism region of Uzungol, Trabzon located in northeastern Turkey. There has been no literature so far in the application of Fuzzy-AHP methods on avalanche susceptibility analysis, which has been used in other fields. This work will be the first case study on this subject. The results were compared according to existing land and inventory data; the success rates of the methods were found to be 89.12% for Fr, 71.15% for AHP and 73.24% for FAHP, and the effectiveness of the Fuzzy-AHP method was further tested. Comparing the results of the different methods, it has been found out that the Fr method is the most successful, while the AHP and FAHP methods provide approximate results.

GIS-based spatial modeling of snow avalanches using four novel ensemble models

Snow avalanches can destroy lives and infrastructure and are very important phenomena in some regions of the world. This study maps snow avalanche susceptibility in Sirvan Watershed, Iran, using a new approach. Two statistical models – belief function (Bel) and probability density (PD) – are combined with two learning models – multi-layer perceptron (MLP) and logistic regression (LR) – to predict avalanche susceptibility using remote sensing data in a geographic information system (GIS). A snow avalanche inventory map was generated from Google Earth imagery, regional documentation, and field surveys. Of 101 avalanche locations, 71 (70%) were used to train the models and 30 (30%) were used to validate the resulting models. Fourteen snow avalanche conditioning factors were used as independent variables in the predictive modeling process. First, the weight of Bel and PD techniques were applied to each class of factors. Then, they were combined with two MLP and LR learning models for snow avalanche susceptibility mapping (SASM). The results were validated using positive predictive values, negative predictive values, sensitivity, specificity, accuracy, root-mean-square error, and area-under-the-curve (AUC) values. Thus, the AUCs for the PD-LR, Bel-LR, Bel-MLP, and PD-MLP hybrid models are 0.941, 0.936, 0.931 and 0.924, respectively. Based on the validation results, the PD-LR hybrid model achieved the best accuracy among the models. This hybrid modeling approach can provide accurate and reliable evaluations of snow avalanche-prone areas for management and decision making.

Snow avalanche susceptibility of the Circo de Gredos (Iberian Central System, Spain)

ABSTRACT We present a detailed snow avalanche susceptibility map at scale 1:20,000 of the Circo de Gredos in the Sierra de Gredos (Iberian Central System, Spain). This cirque-shaped landscape is one of the most popular spots for winter sports in the region. However, no snow avalanche activity assessment has been conducted to date. We have, therefore, produced a snow avalanche susceptibility map based on aerial and satellite imagery, newspaper reviews and field work, including avalanche features recognition and interviews with frequent backcountry users. We extracted the spatial distribution of necessary and enhancer factors for triggering slab, wet and loose snow avalanches from a digital elevation model. Finally, calculations to evaluate each snow avalanche type susceptibility were performed using a Geographical Information System. By combining our map collection, we concluded that most of the area in the Circo de Gredos is highly susceptible to snow avalanches, especially slab and wet snow types.

Towards an Ensemble Machine Learning Model of Random Subspace Based Functional Tree Classifier for Snow Avalanche Susceptibility Mapping

Snow avalanche as a natural disaster severely affects socio-economic and geomorphic processes through damaging ecosystems, vegetation, landscape, infrastructures, transportation networks, and human life. Modeling the snow avalanche has been seen as an essential approach for understanding the mountainous landscape dynamics to assess hazard susceptibility leading to effective mitigation and resilience. Therefore, the main aim of this study is to introduce and implement an ensemble machine learning model of random subspace (RS) based on a classifier, functional tree (FT), named RSFT model for snow avalanche susceptibility mapping at Karaj Watershed, Iran. According to the best knowledge of literature, the proposed model, RSFT, has not earlier been introduced and implemented for snow avalanche modeling and mapping over the world. Four benchmark models, including logistic regression (LR), logistic model tree (LMT), alternating decision tree (ADT), and functional trees (FT) models were used to check the goodness-of-fit and prediction accuracy of the proposed model. To achieve this objective, the most important factors among many climatic, topographic, lithologic, and hydrologic factors, which affect the snow accumulation and snow avalanche occurrence, were determined by the information gain ratio (IGR) technique. The goodness-of-fit and prediction accuracy of the models were evaluated by some statistical-based indexes including, sensitivity, specificity, accuracy, kappa, and area under the ROC curve, Friedman and Wilcoxon sign rank tests. Results indicated that the ensemble proposed model (RSFT), had the highest performance (Sensitivity = 94.1%, Specificity = 92.4%, Accuracy = 93.3%, and Kappa = 0.782) rather than the other soft-computing benchmark models. The snow avalanche susceptibility maps indicated that the high and very high susceptibility avalanche areas are located in the north and northeast parts of the study area, which have a higher elevation with more precipitation and lower temperatures.

Snow avalanche susceptibility in the eastern hillside of the Aramo Range (Asturian Central Massif, Cantabrian Mountains, NW Spain)

ABSTRACTA detailed snow avalanche susceptibility map of the eastern hillside in the Aramo Range (Cantabrian Mountains) is presented at a scale of 1:25,000. The Aramo Range is one of the major middle-altitude mountains of the Asturian Central Massif. Although it has of moderate height (maximum altitude of 1791 m a.s.l.), its eastern slope presents unusual snow avalanche activity. Specifically, a hundred of snow avalanche tracks have been mapped based on meticulous fieldwork and supported by interviews with local people, searches in newspaper archives, photointerpretation, and calculations based on the digital terrain model and geographic information system. As a result, a susceptibility map has been elaborated, which shows the suitability of combining fieldwork and geographic information technology. The composition consists of two maps that detail how the susceptibility mapping is obtained. The section analysed is limited to the eastern slope of the Aramo Range, whose total surface area is 1555.62 ha.

Analogy between grid-based modeling of landslide and avalanche using GIS with surface flow analysis

Mountainous areas with steep slopes are vulnerable to landslide and often to avalanche according to the climate condition. In the rainy season, heavy rainfall causes landslides, and avalanche can be a serious threat to mountainous areas in winter. However, avalanche has not been emphasized relatively as compared with landslides in some nations like Korea. This paper estimates the landslide and avalanche hazard of the mountainous area with distinct seasons such as Provo Canyon in Utah and Seorak Mountain in Korea. To predict susceptibility of landslides and avalanches, several geomorphological factors were considered. These predictive factors were derived from digital elevation map, and the grid-based modeling was applied for landslide and avalanche susceptibility mapping within a geographical information system (GIS). To simulate debris flow and avalanche paths from the high potential areas, GIS-based surface flow analysis was used. As a result, this study provides information about ares prone to natural hazards, and it can be useful ancillary data for people attempting to avoid potentially hazardous areas.

Geospatial Modelling and Mapping of Snow Avalanche Susceptibility

The main objective of this study was to mapping the potential avalanche susceptible zones in Siachen region of the Western Indian Himalaya. A geospatial analytical hierarchy process (AHP) model based on expert judgements has been applied in this study. Five most important terrain based avalanche occurrence parameters including slope, aspect, curvature, elevation and ground cover are employed in the present model. The ASTER GDEM and Landsat 8 OLI imagery are used to generate the avalanche occurrence parameters. A pairwise comparison matrix is computed to estimate the weight values for input terrain parameters. These weight values are then assigned to each respective avalanche occurrence parameter and employed in a geospatial AHP model to generate an avalanche susceptibility map. Finally, an avalanche inventory has been utilized to validate the results. The avalanche susceptibility map has been compared with the avalanche inventory map by calculating the area under the receiver operating characteristics curve (ROC-AUC) technique.

Snow Avalanche Susceptibility Mapping using Remote Sensing and GIS in Nubra-Shyok Basin, Himalaya, India

Objectives: To produce an avalanche susceptibility map using static terrain factor accountable for avalanche occurrence in Nubra-Shyok basin, Himalaya, India. Methods/Statistical Analysis: Expert judgements based Analytical hierarchy process model was applied in the present study. The weight values for each avalanche contributory factor (aspect, elevation, slope, curvature and ground cover), was assigned by comparing possible pairs in a matrix. Receiver operator characteristics- Area under the Curve technique was used to validate the results. Findings: The calculated weight values of the slope factor is 0.47, which is found most significant, while the weight value of elevation factor is 0.25, were also found to be significant factor. The weight value of Ground cover was 0.05 and creates a least importance. The curvature and aspect weight values were 0.12 and 0.11 respectively, and produce intermediate degree of importance. The avalanche susceptibility map of the Nubra-Shyok basin was divided into five zones: - (I) Very High (II) Moderate-to-High (III) Moderate (IV) Low (V) Very Low, and covers 27.5% , 28.0% 19.0%, 20.0%, and 4.8%, of the total area respectively. ROC-AUC method was applied to verify the study results. The value of Area under the curve (0.9112) shows that results of the present study are acceptable and 91.12% avalanche pixels are properly categorized. Application/Improvements: This study will help in safe movement across the Nubra-Shyok basin, India.

GIS-based MCDA–AHP modelling for avalanche susceptibility mapping of Nubra valley region, Indian Himalaya

Avalanches are behind the majority of fatalities and heavy damage to property in snow-covered mountainous terrain like Himalaya. Recognizing avalanche susceptible areas and publication of avalanche susceptibility maps assist decision-makers and planners to execute suitable measures to reduce the avalanche risk. The present study is an attempt to prepare an avalanche susceptibility map of the Nubra valley region using multi-criteria decision analysis–analytical hierarchy process model in GIS environment. The most prominent avalanche occurrence factors used in this model are slope, aspect, curvature, elevation, terrain roughness and ground cover. ASTER GDEM V2 and Landsat 8 satellite imagery were used to generate considered factors. For validation of the results, prediction rate/accuracy is calculated using the avalanche inventory map of documented avalanche locations. To calculate the prediction accuracy, area under the ROC curve (ROC-AUC) method has been used. The prediction accuracy of the validation results using ROC-AUC shows 91%.

Fuzzy–frequency ratio model for avalanche susceptibility mapping

ABSTRACTAvalanche activities in the Indian Himalaya cause the majority of fatalities and responsible for heavy damage to the property. Avalanche susceptibility maps assist decision-makers and planners to execute suitable measures to reduce the avalanche risk. In the present study, a probabilistic data-driven geospatial fuzzy–frequency ratio (fuzzy–FR) model is proposed and developed for avalanche susceptibility mapping, especially for the large undocumented region. The fuzzy–FR model for avalanche susceptibility mapping is initially developed and applied for Lahaul-Spiti region. The fuzzy–FR model utilized the six avalanche occurrence factors (i.e. slope, aspect, curvature, elevation, terrain roughness and vegetation cover) and one referent avalanche inventory map to generate the avalanche susceptibility map. Amongst 292 documented avalanche locations from the avalanche inventory map, 233 (80%) were used for training the model and remaining 59 (20%) were used for validation of the map. The avalanche susceptibility map is validated by calculating the area under the receiver operating characteristic curve (ROC-AUC) technique. For validation of the results using ROC-AUC technique, the success rate and prediction rate were calculated. The values of success rate and prediction rate were 94.07% and 91.76%, respectively. The validation of results using ROC-AUC indicated the fuzzy–FR model is appropriate for avalanche susceptibility mapping.

Creation of the snow avalanche susceptibility map of the Krkonoše Mountains using GIS

This article deals with the development of the snow avalanche susceptibility map in the Czech part of the Krkonoše Mountains using the free Geographic Information System (GIS) GRASS. The area susceptibility map consists of two components: the morphological risk map, which is derived from the digital terrain model (DTM) and describes the slope steepness, aspect and curvature of the slope, and the protecting vegetation influence map, which is based on supervised image classification (spectrozonal aerial photos) and takes into consideration the importance of vegetation cover. The final map also includes starting zones calculated on the basis of significant changes in slope steepness and approximate shapes of avalanche paths based on these zones. In the map development, the layer of measured paths of avalanche cadastre in the Czech part of the Krkonoše Mountains was used, partly to gain the morphological characteristics of starting zones and partly to check the quality of the map.