Landslide susceptibility maps are helpful tools to identify areas potentially prone to future lan... more Landslide susceptibility maps are helpful tools to identify areas potentially prone to future landslide occurrence. As more and more national and provincial authorities demand for these maps to be computed and implemented in spatial planning strategies, several aspects of the quality of the landslide susceptibility model and the resulting classified map are of high interest. In this study of landslides in Lower Austria, we focus on the model form uncertainty to assess the quality of a flexible statistical modelling technique, the generalized additive model (GAM). The study area (15 850 km 2 ) is divided into 16 modelling domains based on lithology classes. A model representing the entire study area is constructed by combining these models. The performances of the models are assessed using repeated k-fold cross-validation with spatial and random subsampling. This reflects the variability of performance estimates arising from sampling variation. Measures of spatial transferability and thematic consistency are applied to empirically assess model quality. We also analyse and visualize the implications of spatially varying prediction uncertainties regarding the susceptibility map classes by taking into account the confidence intervals of model predictions. The 95 % confidence limits fall within the same susceptibility class in 85 % of the study area. Overall, this study contributes to advancing open communication and assessment of model quality related to statistical landslide susceptibility models.
Safe operations of forest practices in mountainous regions require effective development planning... more Safe operations of forest practices in mountainous regions require effective development planning to mitigate hazards posed by landslides. British Columbia, Canada, has for the past 2 decades implemented landslide risk management policies aimed at reducing the impacts of the forestry industry on landslides. Consequently, it is required that timber harvesting sites be evaluated for their potential or existing impacts on terrain stability. Statistical landslide susceptibility modelling can enhance this evaluation by geographically highlighting potential hazardous areas. In addition, these statistical models can also improve our understanding of regional landslide controlling factors. The purpose of this research was to explore the regional effects of forest harvesting activities, topography, precipitation and geology on landslides initiated during an extreme rainfall event in November 2006 on Vancouver Island, British Columbia. These effects were analyzed with a nonparametric statistical method, the generalized additive model (GAM). Although topography was the strongest predictor of landslide initiation, low density forest interpreted as regrowth areas and proximity to forest service roads were jointly associated with a 6- to 9-fold increase in the odds of landslide initiation, while accounting for other environmental confounders. This result highlights the importance of continuing proper landslide risk management to control the effects of forest practices on landslide initiation.
Statistical and now machine learning prediction methods have been gaining popularity in the field... more Statistical and now machine learning prediction methods have been gaining popularity in the field of landslide susceptibility modeling. Particularly, these data driven approaches show promise when tackling the challenge of mapping landslide prone areas for large regions, which may not have sufficient geotechnical data to conduct physically-based methods. Currently, there is no best method for empirical susceptibility modeling. Therefore, this study presents a comparison of traditional statistical and novel machine learning models applied for regional scale landslide susceptibility modeling. These methods were evaluated by spatial k-fold cross-validation estimation of the predictive performance, assessment of variable importance for gaining insights into model behavior and by the appearance of the prediction (i.e. susceptibility) map. The modeling techniques applied were logistic regression (GLM), generalized additive models (GAM), weights of evidence (WOE), the support vector machine (SVM), random forest classification (RF), and bootstrap aggregated classification trees (bundling) with penalized discriminant analysis (BPLDA). These modeling methods were tested for three areas in the province of Lower Austria, Austria. The areas are characterized by different geological and morphological settings.
Random forest and bundling classification techniques had the overall best predictive performances. However, the performances of all modeling techniques were for the majority not significantly different from each other; depending on the areas of interest, the overall median estimated area under the receiver operating characteristic curve (AUROC) differences ranged from 2.9 to 8.9 percentage points. The overall median estimated true positive rate (TPR) measured at a 10% false positive rate (FPR) differences ranged from 11 to 15pp. The relative importance of each predictor was generally different between the modeling methods. However, slope angle, surface roughness and plan curvature were consistently highly ranked variables. The prediction methods that create splits in the predictors (RF, BPLDA and WOE) resulted in heterogeneous prediction maps full of spatial artifacts. In contrast, the GAM, GLM and SVM produced smooth prediction surfaces. Overall, it is suggested that the framework of this model evaluation approach can be applied to assist in selection of a suitable landslide susceptibility modeling technique.
The potential of surface roughness to quantify geomorphological landforms and processes has been ... more The potential of surface roughness to quantify geomorphological landforms and processes has been enhanced with the availability of high-resolution digital terrain models (DTM). Recent studies that attempt to identify landslide features with surface roughness have suggested that this measure of topographic heterogeneity may also be applied to estimate the relative age of landslides. This is a provisional study that explores the potential of this relationship by assessing the ability of surface roughness to act as a proxy for relative landslide age. The surface roughness for a set of twelve dated landslides in the Swabian Alb that occurred between 1789 and 1985 was calculated from a 1 m2 spatial resolution LiDAR DTM with three algorithms: root-mean-square-height (RMSH), standard deviation of slope (SDS), and direction cosine eigenvalue ratios (DCE). Scale-dependence was analysed by calculating surface roughness for a range of moving window sizes (3 × 3, 5 × 5, 9 × 9 and 15 × 15), and surface roughness for each landslide was summarized by the median and upper quartile. Only weak correlations (best Spearman's rho 0.58) were present between landslide age and surface roughness. This correlation becomes weaker with increasing moving window size. Given weak observed associations and discussed challenges pertaining to the complexities of landslide morphology change over time, we currently find that surface roughness alone may not be justifiable to act as a proxy for landslide age for our study region. Furthermore, we recommend future studies should focus on addressing possible natural and anthropogenic factors such as land use change that may alter surface roughness. These studies may focus on one of the three roughness measures used here as they are strongly correlated. This article is protected by copyright. All rights reserved.
Landslide susceptibility maps are helpful tools to identify areas potentially prone to future lan... more Landslide susceptibility maps are helpful tools to identify areas potentially prone to future landslide occurrence. As more and more national and provincial authorities demand for these maps to be computed and implemented in spatial planning strategies, several aspects of the quality of the landslide susceptibility model and the resulting classified map are of high interest. In this study of landslides in Lower Austria, we focus on the model form uncertainty to assess the quality of a flexible statistical modelling technique, the generalized additive model (GAM). The study area (15 850 km2) is divided into 16 modelling domains based on lithology classes. A model representing the entire study area is constructed by combining these models. The performances of the models are assessed using repeated k-fold cross-validation with spatial and random subsampling. This reflects the variability of performance estimates arising from sampling variation. Measures of spatial transferability and thematic consistency are applied to empirically assess model quality. We also analyse and visualize the implications of spatially varying prediction uncertainties regarding the susceptibility map classes by taking into account the confidence intervals of model predictions. The 95% confidence limits fall within the same susceptibility class in 85% of the study area. Overall, this study contributes to advancing open communication and assessment of model quality related to statistical landslide susceptibility models.
Results of stable isotope measurements (δ2H, δ18O) of daily grab samples, taken from the Danube R... more Results of stable isotope measurements (δ2H, δ18O) of daily grab samples, taken from the Danube River at Tulln (river km 1963) during 2012, show seasonal and short-term variations depending on the climatic/hydrological conditions and changes in the catchment area (temperature changes, heavy rains and snow melt processes). Isotope ratios in river water clearly reflect the isotopic composition of precipitation water in the catchment area since evaporation influences play a minor role. Average δ2H and δ18O values in 2012 are−78‰ and−11.0‰, respectively, deuterium excess averages 10‰. The entire variation amounts to 1.8‰ in δ18O and 15‰ in δ2H. Quick changes of the isotopic composition within a few days emphasise the necessity of daily sampling for the investigation of hydrological events, while monthly grab sampling seems sufficient for the investigation of long-term hydro-climatic trends. 3H results show peaks (half-width 1–2 days, up to about 150 TU) exceeding the regional environmental level of about 9 TU, probably due to releases from nuclear power plants.
Empirically-based models of landslide distribution and susceptibility are currently the most comm... more Empirically-based models of landslide distribution and susceptibility are currently the most commonly used approach for mapping probabilities of landslide initiation and analyzing their association with natural and anthropogenic environmental factors. In general, these models statistically estimate susceptibility based on the predisposition of an area to experience a landslide given a range of environmental factors, which may include land use, topography, hydrology and other spatial attributes. Novel statistical approaches include the generalized additive model (GAM), a non-parametric regression technique, which is used in this study to explore the relationship of landslide initiation to topography, rainfall and forest land cover and logging roads on Vancouver Island, British Columbia. The analysis is centered on an inventory of 639 landslides of winter 2006/07. Data sources representing potentially relevant environmental conditions of landslide initiation are based on: terrain analysis derived from a 20-m CDED digital elevation model; forest land cover classified from Landsat TM scenes for the summer before the 2006 rainy season; geostatistically interpolated antecedent rainfall patterns representing different temporal scales of rainfall (a major storm, winter and annual rainfall); and the main lithological units of surface geology. In order to assess the incremental effect of these data sources to predict landslide susceptibility, predictive performances of models based on GAMs are compared using spatial cross-validation estimates of the area under the ROC curve (AUROC), and variable selection frequencies are used to determine the prevalence of non-parametric associations to landslides. In addition to topographic variables, forest land cover (e.g., deforestation), and logging roads showed a strong association with landslide initiation, followed by rainfall patterns and the very general lithological classification as less important controls of landscape-scale landslide activity in this area. Annual rainfall patterns are found not to contribute significantly to model prediction improvement and may lead to model overfitting. Comparisons to generalized linear models (i.e., logistic regression) indicate that GAMs are significantly better for modeling landslide susceptibility. Overall, based on the model predictions, the most susceptible 4% of the study area had 29 times higher density of landslide initiation points than the least susceptible 73% of the study area (0.156 versus 0.005 landslides/km2).
Physically based models are commonly used as an integral step in landslide hazard assessment. Geo... more Physically based models are commonly used as an integral step in landslide hazard assessment. Geomorphic principles can be applied to a broad area, resulting in first order assessment of landslide susceptibility. New techniques are now available that may result in the increased accuracy of such models. We investigate the possibility to enhance landslide susceptibility modeling by integrating two physically-based landslide models, the Factor of Safety (FS) and the Shallow Stability model (SHALSTAB), with traditional empirical-statistical methods that utilize terrain attribute information derived from a digital elevation model and land use characteristics related to forest harvesting. The model performance is measured by the area under the receiver operating characteristic curve (AUROC) and sensitivity at 90% and 80% specificity both estimated by bootstrap resampling. Our study examines 278 landslide initiation points in the Klanawa Watershed located on Vancouver Island, British Columbia, Canada. We use a generalized additive model (GAM)and a logistic regression model (GLM) combining physical landslide models, terrain attributes and land use data, and GAMs and GLMs using only subsets of these variables.In this study, all empirical and combined physical-empirical models outperformthe physically-based models, with GAMs often performing significantly better than GLMs. The strongest predictive performance is achieved by the GAMsusing terrain attributes in combination with land use data. Variables representing physically-based models do not significantly improve the empirical models, but they may allow for a better physical interpretation of empirical models. Also, based on bootstrap variable-selection frequencies,land use data, FS, slope and plan/profile curvature are relatively the most important predictor variables.
With so many techniques now available for landslide susceptibility modelling, it can be challengi... more With so many techniques now available for landslide susceptibility modelling, it can be challenging to decide on which technique to apply. Generally speaking, the criteria for model selection should be tied closely to end users' purpose, which could be spatial prediction, spatial analysis or both. In our research, we focus on comparing the spatial predictive abilities of landslide susceptibility models. We illustrate how spatial cross-validation, a statistical approach for assessing spatial prediction performance, can be applied with the area under the receiver operating characteristic curve (AUROC) as a prediction measure for model comparison. Several machine learning and statistical techniques are evaluated for prediction in Lower Austria: support vector machine, random forest, bundling with penalized linear discriminant analysis, logistic regression, weights of evidence, and the generalized additive model. In addition to predictive performance, the importance of predictor variables in each model was estimated using spatial cross-validation by calculating the change in AUROC performance when variables are randomly permuted. The susceptibility modelling techniques were tested in three areas of interest in Lower Austria, which have unique geologic conditions associated with landslide occurrence.
Overall, we found for the majority of comparisons that there were little practical or even statistically significant differences in AUROCs. That is the models' prediction performances were very similar. Therefore, in addition to prediction, the ability to interpret models for spatial analysis and the qualitative qualities of the prediction surface (map) are considered and discussed. The measure of variable importance provided some insight into the model behaviour for prediction, in particular for “black-box” models. However, there were no clear patterns in all areas of interest to why certain variables were given more importance over others.
Landslides pose threats not only for specific localities, they are also influencing
larger areas... more Landslides pose threats not only for specific localities, they are also influencing
larger areas and consequently require spatial analysis methods for assessing the
susceptibility to landslides. The availability of landslide susceptibility maps and
their consideration in spatial planning practises is a major step towards avoidance
of damage due to landslides.
In this contribution we present a study design developed with the objective to
determine the best suited landslide susceptibility maps for debris and earth slides
and for rock falls for spatial planning in Lower Austria. Therefore, two different
methods were tested for slide and rock fall susceptibility modelling respectively.
Landslide susceptibility maps are helpful tools to identify areas potentially prone to future lan... more Landslide susceptibility maps are helpful tools to identify areas potentially prone to future landslide occurrence. As more and more national and provincial authorities demand for these maps to be computed and implemented in spatial planning strategies, several aspects of the quality of the landslide susceptibility model and the resulting classified map are of high interest. In this study of landslides in Lower Austria, we focus on the model form uncertainty to assess the quality of a flexible statistical modelling technique, the generalized additive model (GAM). The study area (15 850 km 2 ) is divided into 16 modelling domains based on lithology classes. A model representing the entire study area is constructed by combining these models. The performances of the models are assessed using repeated k-fold cross-validation with spatial and random subsampling. This reflects the variability of performance estimates arising from sampling variation. Measures of spatial transferability and thematic consistency are applied to empirically assess model quality. We also analyse and visualize the implications of spatially varying prediction uncertainties regarding the susceptibility map classes by taking into account the confidence intervals of model predictions. The 95 % confidence limits fall within the same susceptibility class in 85 % of the study area. Overall, this study contributes to advancing open communication and assessment of model quality related to statistical landslide susceptibility models.
Safe operations of forest practices in mountainous regions require effective development planning... more Safe operations of forest practices in mountainous regions require effective development planning to mitigate hazards posed by landslides. British Columbia, Canada, has for the past 2 decades implemented landslide risk management policies aimed at reducing the impacts of the forestry industry on landslides. Consequently, it is required that timber harvesting sites be evaluated for their potential or existing impacts on terrain stability. Statistical landslide susceptibility modelling can enhance this evaluation by geographically highlighting potential hazardous areas. In addition, these statistical models can also improve our understanding of regional landslide controlling factors. The purpose of this research was to explore the regional effects of forest harvesting activities, topography, precipitation and geology on landslides initiated during an extreme rainfall event in November 2006 on Vancouver Island, British Columbia. These effects were analyzed with a nonparametric statistical method, the generalized additive model (GAM). Although topography was the strongest predictor of landslide initiation, low density forest interpreted as regrowth areas and proximity to forest service roads were jointly associated with a 6- to 9-fold increase in the odds of landslide initiation, while accounting for other environmental confounders. This result highlights the importance of continuing proper landslide risk management to control the effects of forest practices on landslide initiation.
Statistical and now machine learning prediction methods have been gaining popularity in the field... more Statistical and now machine learning prediction methods have been gaining popularity in the field of landslide susceptibility modeling. Particularly, these data driven approaches show promise when tackling the challenge of mapping landslide prone areas for large regions, which may not have sufficient geotechnical data to conduct physically-based methods. Currently, there is no best method for empirical susceptibility modeling. Therefore, this study presents a comparison of traditional statistical and novel machine learning models applied for regional scale landslide susceptibility modeling. These methods were evaluated by spatial k-fold cross-validation estimation of the predictive performance, assessment of variable importance for gaining insights into model behavior and by the appearance of the prediction (i.e. susceptibility) map. The modeling techniques applied were logistic regression (GLM), generalized additive models (GAM), weights of evidence (WOE), the support vector machine (SVM), random forest classification (RF), and bootstrap aggregated classification trees (bundling) with penalized discriminant analysis (BPLDA). These modeling methods were tested for three areas in the province of Lower Austria, Austria. The areas are characterized by different geological and morphological settings.
Random forest and bundling classification techniques had the overall best predictive performances. However, the performances of all modeling techniques were for the majority not significantly different from each other; depending on the areas of interest, the overall median estimated area under the receiver operating characteristic curve (AUROC) differences ranged from 2.9 to 8.9 percentage points. The overall median estimated true positive rate (TPR) measured at a 10% false positive rate (FPR) differences ranged from 11 to 15pp. The relative importance of each predictor was generally different between the modeling methods. However, slope angle, surface roughness and plan curvature were consistently highly ranked variables. The prediction methods that create splits in the predictors (RF, BPLDA and WOE) resulted in heterogeneous prediction maps full of spatial artifacts. In contrast, the GAM, GLM and SVM produced smooth prediction surfaces. Overall, it is suggested that the framework of this model evaluation approach can be applied to assist in selection of a suitable landslide susceptibility modeling technique.
The potential of surface roughness to quantify geomorphological landforms and processes has been ... more The potential of surface roughness to quantify geomorphological landforms and processes has been enhanced with the availability of high-resolution digital terrain models (DTM). Recent studies that attempt to identify landslide features with surface roughness have suggested that this measure of topographic heterogeneity may also be applied to estimate the relative age of landslides. This is a provisional study that explores the potential of this relationship by assessing the ability of surface roughness to act as a proxy for relative landslide age. The surface roughness for a set of twelve dated landslides in the Swabian Alb that occurred between 1789 and 1985 was calculated from a 1 m2 spatial resolution LiDAR DTM with three algorithms: root-mean-square-height (RMSH), standard deviation of slope (SDS), and direction cosine eigenvalue ratios (DCE). Scale-dependence was analysed by calculating surface roughness for a range of moving window sizes (3 × 3, 5 × 5, 9 × 9 and 15 × 15), and surface roughness for each landslide was summarized by the median and upper quartile. Only weak correlations (best Spearman's rho 0.58) were present between landslide age and surface roughness. This correlation becomes weaker with increasing moving window size. Given weak observed associations and discussed challenges pertaining to the complexities of landslide morphology change over time, we currently find that surface roughness alone may not be justifiable to act as a proxy for landslide age for our study region. Furthermore, we recommend future studies should focus on addressing possible natural and anthropogenic factors such as land use change that may alter surface roughness. These studies may focus on one of the three roughness measures used here as they are strongly correlated. This article is protected by copyright. All rights reserved.
Landslide susceptibility maps are helpful tools to identify areas potentially prone to future lan... more Landslide susceptibility maps are helpful tools to identify areas potentially prone to future landslide occurrence. As more and more national and provincial authorities demand for these maps to be computed and implemented in spatial planning strategies, several aspects of the quality of the landslide susceptibility model and the resulting classified map are of high interest. In this study of landslides in Lower Austria, we focus on the model form uncertainty to assess the quality of a flexible statistical modelling technique, the generalized additive model (GAM). The study area (15 850 km2) is divided into 16 modelling domains based on lithology classes. A model representing the entire study area is constructed by combining these models. The performances of the models are assessed using repeated k-fold cross-validation with spatial and random subsampling. This reflects the variability of performance estimates arising from sampling variation. Measures of spatial transferability and thematic consistency are applied to empirically assess model quality. We also analyse and visualize the implications of spatially varying prediction uncertainties regarding the susceptibility map classes by taking into account the confidence intervals of model predictions. The 95% confidence limits fall within the same susceptibility class in 85% of the study area. Overall, this study contributes to advancing open communication and assessment of model quality related to statistical landslide susceptibility models.
Results of stable isotope measurements (δ2H, δ18O) of daily grab samples, taken from the Danube R... more Results of stable isotope measurements (δ2H, δ18O) of daily grab samples, taken from the Danube River at Tulln (river km 1963) during 2012, show seasonal and short-term variations depending on the climatic/hydrological conditions and changes in the catchment area (temperature changes, heavy rains and snow melt processes). Isotope ratios in river water clearly reflect the isotopic composition of precipitation water in the catchment area since evaporation influences play a minor role. Average δ2H and δ18O values in 2012 are−78‰ and−11.0‰, respectively, deuterium excess averages 10‰. The entire variation amounts to 1.8‰ in δ18O and 15‰ in δ2H. Quick changes of the isotopic composition within a few days emphasise the necessity of daily sampling for the investigation of hydrological events, while monthly grab sampling seems sufficient for the investigation of long-term hydro-climatic trends. 3H results show peaks (half-width 1–2 days, up to about 150 TU) exceeding the regional environmental level of about 9 TU, probably due to releases from nuclear power plants.
Empirically-based models of landslide distribution and susceptibility are currently the most comm... more Empirically-based models of landslide distribution and susceptibility are currently the most commonly used approach for mapping probabilities of landslide initiation and analyzing their association with natural and anthropogenic environmental factors. In general, these models statistically estimate susceptibility based on the predisposition of an area to experience a landslide given a range of environmental factors, which may include land use, topography, hydrology and other spatial attributes. Novel statistical approaches include the generalized additive model (GAM), a non-parametric regression technique, which is used in this study to explore the relationship of landslide initiation to topography, rainfall and forest land cover and logging roads on Vancouver Island, British Columbia. The analysis is centered on an inventory of 639 landslides of winter 2006/07. Data sources representing potentially relevant environmental conditions of landslide initiation are based on: terrain analysis derived from a 20-m CDED digital elevation model; forest land cover classified from Landsat TM scenes for the summer before the 2006 rainy season; geostatistically interpolated antecedent rainfall patterns representing different temporal scales of rainfall (a major storm, winter and annual rainfall); and the main lithological units of surface geology. In order to assess the incremental effect of these data sources to predict landslide susceptibility, predictive performances of models based on GAMs are compared using spatial cross-validation estimates of the area under the ROC curve (AUROC), and variable selection frequencies are used to determine the prevalence of non-parametric associations to landslides. In addition to topographic variables, forest land cover (e.g., deforestation), and logging roads showed a strong association with landslide initiation, followed by rainfall patterns and the very general lithological classification as less important controls of landscape-scale landslide activity in this area. Annual rainfall patterns are found not to contribute significantly to model prediction improvement and may lead to model overfitting. Comparisons to generalized linear models (i.e., logistic regression) indicate that GAMs are significantly better for modeling landslide susceptibility. Overall, based on the model predictions, the most susceptible 4% of the study area had 29 times higher density of landslide initiation points than the least susceptible 73% of the study area (0.156 versus 0.005 landslides/km2).
Physically based models are commonly used as an integral step in landslide hazard assessment. Geo... more Physically based models are commonly used as an integral step in landslide hazard assessment. Geomorphic principles can be applied to a broad area, resulting in first order assessment of landslide susceptibility. New techniques are now available that may result in the increased accuracy of such models. We investigate the possibility to enhance landslide susceptibility modeling by integrating two physically-based landslide models, the Factor of Safety (FS) and the Shallow Stability model (SHALSTAB), with traditional empirical-statistical methods that utilize terrain attribute information derived from a digital elevation model and land use characteristics related to forest harvesting. The model performance is measured by the area under the receiver operating characteristic curve (AUROC) and sensitivity at 90% and 80% specificity both estimated by bootstrap resampling. Our study examines 278 landslide initiation points in the Klanawa Watershed located on Vancouver Island, British Columbia, Canada. We use a generalized additive model (GAM)and a logistic regression model (GLM) combining physical landslide models, terrain attributes and land use data, and GAMs and GLMs using only subsets of these variables.In this study, all empirical and combined physical-empirical models outperformthe physically-based models, with GAMs often performing significantly better than GLMs. The strongest predictive performance is achieved by the GAMsusing terrain attributes in combination with land use data. Variables representing physically-based models do not significantly improve the empirical models, but they may allow for a better physical interpretation of empirical models. Also, based on bootstrap variable-selection frequencies,land use data, FS, slope and plan/profile curvature are relatively the most important predictor variables.
With so many techniques now available for landslide susceptibility modelling, it can be challengi... more With so many techniques now available for landslide susceptibility modelling, it can be challenging to decide on which technique to apply. Generally speaking, the criteria for model selection should be tied closely to end users' purpose, which could be spatial prediction, spatial analysis or both. In our research, we focus on comparing the spatial predictive abilities of landslide susceptibility models. We illustrate how spatial cross-validation, a statistical approach for assessing spatial prediction performance, can be applied with the area under the receiver operating characteristic curve (AUROC) as a prediction measure for model comparison. Several machine learning and statistical techniques are evaluated for prediction in Lower Austria: support vector machine, random forest, bundling with penalized linear discriminant analysis, logistic regression, weights of evidence, and the generalized additive model. In addition to predictive performance, the importance of predictor variables in each model was estimated using spatial cross-validation by calculating the change in AUROC performance when variables are randomly permuted. The susceptibility modelling techniques were tested in three areas of interest in Lower Austria, which have unique geologic conditions associated with landslide occurrence.
Overall, we found for the majority of comparisons that there were little practical or even statistically significant differences in AUROCs. That is the models' prediction performances were very similar. Therefore, in addition to prediction, the ability to interpret models for spatial analysis and the qualitative qualities of the prediction surface (map) are considered and discussed. The measure of variable importance provided some insight into the model behaviour for prediction, in particular for “black-box” models. However, there were no clear patterns in all areas of interest to why certain variables were given more importance over others.
Landslides pose threats not only for specific localities, they are also influencing
larger areas... more Landslides pose threats not only for specific localities, they are also influencing
larger areas and consequently require spatial analysis methods for assessing the
susceptibility to landslides. The availability of landslide susceptibility maps and
their consideration in spatial planning practises is a major step towards avoidance
of damage due to landslides.
In this contribution we present a study design developed with the objective to
determine the best suited landslide susceptibility maps for debris and earth slides
and for rock falls for spatial planning in Lower Austria. Therefore, two different
methods were tested for slide and rock fall susceptibility modelling respectively.
Uploads
Papers by Jason Goetz
Random forest and bundling classification techniques had the overall best predictive performances. However, the performances of all modeling techniques were for the majority not significantly different from each other; depending on the areas of interest, the overall median estimated area under the receiver operating characteristic curve (AUROC) differences ranged from 2.9 to 8.9 percentage points. The overall median estimated true positive rate (TPR) measured at a 10% false positive rate (FPR) differences ranged from 11 to 15pp. The relative importance of each predictor was generally different between the modeling methods. However, slope angle, surface roughness and plan curvature were consistently highly ranked variables. The prediction methods that create splits in the predictors (RF, BPLDA and WOE) resulted in heterogeneous prediction maps full of spatial artifacts. In contrast, the GAM, GLM and SVM produced smooth prediction surfaces. Overall, it is suggested that the framework of this model evaluation approach can be applied to assist in selection of a suitable landslide susceptibility modeling technique.
Conference Presentations by Jason Goetz
Overall, we found for the majority of comparisons that there were little practical or even statistically significant differences in AUROCs. That is the models' prediction performances were very similar. Therefore, in addition to prediction, the ability to interpret models for spatial analysis and the qualitative qualities of the prediction surface (map) are considered and discussed. The measure of variable importance provided some insight into the model behaviour for prediction, in particular for “black-box” models. However, there were no clear patterns in all areas of interest to why certain variables were given more importance over others.
larger areas and consequently require spatial analysis methods for assessing the
susceptibility to landslides. The availability of landslide susceptibility maps and
their consideration in spatial planning practises is a major step towards avoidance
of damage due to landslides.
In this contribution we present a study design developed with the objective to
determine the best suited landslide susceptibility maps for debris and earth slides
and for rock falls for spatial planning in Lower Austria. Therefore, two different
methods were tested for slide and rock fall susceptibility modelling respectively.
Random forest and bundling classification techniques had the overall best predictive performances. However, the performances of all modeling techniques were for the majority not significantly different from each other; depending on the areas of interest, the overall median estimated area under the receiver operating characteristic curve (AUROC) differences ranged from 2.9 to 8.9 percentage points. The overall median estimated true positive rate (TPR) measured at a 10% false positive rate (FPR) differences ranged from 11 to 15pp. The relative importance of each predictor was generally different between the modeling methods. However, slope angle, surface roughness and plan curvature were consistently highly ranked variables. The prediction methods that create splits in the predictors (RF, BPLDA and WOE) resulted in heterogeneous prediction maps full of spatial artifacts. In contrast, the GAM, GLM and SVM produced smooth prediction surfaces. Overall, it is suggested that the framework of this model evaluation approach can be applied to assist in selection of a suitable landslide susceptibility modeling technique.
Overall, we found for the majority of comparisons that there were little practical or even statistically significant differences in AUROCs. That is the models' prediction performances were very similar. Therefore, in addition to prediction, the ability to interpret models for spatial analysis and the qualitative qualities of the prediction surface (map) are considered and discussed. The measure of variable importance provided some insight into the model behaviour for prediction, in particular for “black-box” models. However, there were no clear patterns in all areas of interest to why certain variables were given more importance over others.
larger areas and consequently require spatial analysis methods for assessing the
susceptibility to landslides. The availability of landslide susceptibility maps and
their consideration in spatial planning practises is a major step towards avoidance
of damage due to landslides.
In this contribution we present a study design developed with the objective to
determine the best suited landslide susceptibility maps for debris and earth slides
and for rock falls for spatial planning in Lower Austria. Therefore, two different
methods were tested for slide and rock fall susceptibility modelling respectively.