Spatial Prediction of Landslide Hazard Using Fuzzy k-means and Dempster-Shafer Theory

Landslide databases and input parameters used for modeling landslide hazard often contain imprecisions and uncertainties inherent in the decision‐making process. Dealing with imprecision and uncertainty requires techniques that go beyond classical logic. In this paper, methods of fuzzy k‐means classification were used to assign digital terrain attributes to continuous landform classes whereas the Dempster‐Shafer theory of evidence was used to represent and manage imprecise information and to deal with uncertainties. The paper introduces the integration of the fuzzy k‐means classification method and the Dempster‐Shafer theory of evidence to model landslide hazard in roaded and roadless areas illustrated through a case study in the Clearwater National Forest in central Idaho, USA. Sample probabilistic maps of landslide hazard potential and uncertainties are presented. The probabilistic maps are intended to help decision‐making in effective forest management and planning.     

IRS-LISS-III and PAN data analysis for landslide susceptibility mapping using heuristic approach in active tectonic region of Himalaya

Structurally disturbed zones of Himalaya are among the worst landslide affected regions in the world. Although landslides are induced/triggered either by torrential rain during monsoon or by seismic activity in the region, the inherent terrain conditions characterize the prevailing basic conditions susceptible to landslides. Using remotely sensed data and Geographic Information System (GIS), geological and terrain factors can be integrated for preparation of factor maps and demarcation of areas susceptible to landslides. Moderate to high resolution data products available from Indian Remote Sensing satellites have been utilized for deriving geological and terrain factor maps, which were integrated using knowledge driven heuristic approach in Integrated Land and Water Information System (ILWIS) GIS. The resultant map shows division of the area into landslide susceptibility classes ranked in terms of hazard potential in one of the structurally disturbed zones in western Himalaya around Rishikesh.     

Spatial Prediction of Landslide Hazard Using Logistic Regression and ROC Analysis

An empirical modeling of road related and non‐road related landslide hazard for a large geographical area using logistic regression in tandem with signal detection theory is presented. This modeling was developed using geographic information system (GIS) and remote sensing data, and was implemented on the Clearwater National Forest in central Idaho. The approach is based on explicit and quantitative environmental correlations between observed landslide occurrences, climate, parent material, and environmental attributes while the receiver operating characteristic (ROC) curves are used as a measure of performance of a predictive rule. The modeling results suggest that development of two independent models for road related and non‐road related landslide hazard was necessary because spatial prediction and predictor variables were different for these models. The probabilistic models of landslide potential may be used as a decision support tool in forest planning involving the maintenance, obliteration or development of new forest roads in steep mountainous terrain.     

Fuzzy mapping of tropical land cover along an environmental gradient from remotely sensed data with an artificial neural network

Remote sensing is the only feasible means of mapping and monitoring land cover at regional to global scales. Unfortunately the maps are generally derived through the use of a conventional 'hard' classification algorithm and depict classes separated by sharp boundaries. Such approaches and representations are often inappropriate particularly when the land cover being represented may be considered to be fuzzy. The definition of boundaries between classes can therefore be difficult from remotely sensed data, particularly for continuous land cover classes which are separated by a fuzzy boundary which may also vary spatially in time. In this paper a neural network was used to derive fuzzy classifications of land cover along a transect crossing the transition from moist semi-deciduous forest to savanna in West Africa in February and December 1990. The fuzzy classifications revealed both sharp and gradual boundaries between classes located along the transect. In particular, the fuzzy classifications enabled the definition of important boundary properties, such as width and temporal displacement.     

Mapping the porosity of international border to pedestrian traffic: a comparative data classification approach to a study of the border region in Austria,Italy, and Slovenia

National borders play an important role in everyday life. Interest in border studies has increased with recent changes in geographical locations of the border or the fluctuation of the permeability of the border between some countries, such as in the European Union. Whether the nations are trying to increase traffic flow of the border or to implement stricter border control, having appropriate information of the border is crucial for effective policymaking.

The objective of this research was to identify areas of high porosity, or high permeability, for pedestrians along the southern national border region in Carinthia, Austria using terrain, land use, and road data along with geocomputational methods. Two unsupervised classification methods, the fuzzy K-means clustering and the Self-Organizing Map, were applied to segment the border into homogeneous zones according to topographic and infrastructural attributes. The fuzzy K-means clustering method was chosen for its ability to allow for a continuous approach to classification. With this method, an object can belong, with different degrees of membership, to multiple classes, which is a more realistic reflection of the natural world than discrete clustering, where each object can only belong to one class. However, the fuzzy K-means clustering method does have disadvantages, i.e. the user must determine the number of classes and the input parameters are required to be in continuous format. The second classification method, the Self-Organizing Map, is a type of artificial neural network and was chosen for its ability to automatically determine the number of classes and handle categorical data. The Self-Organizing Map is unique because it can transform high dimensional data into low dimensional display while preserving the topology and spatial distribution of the input parameters. The results of the two classification methods suggest that the fuzzy K-means classification is more effective than the Self-Organizing Map for this situation. However, more research is needed to determine the fit of these algorithms for particular spatial data classification tasks.

The results obtained from this research provide an insight into the permeability of the border region of Carinthia, Slovenia, and Italy to pedestrian traffic and can be potentially useful for decision making processes for tourism development and road transportation management in that region. Furthermore, the approach presented in this article can be applied to other national borders to identify zones permeable to pedestrian traffic.     

Exploring uncertainties in terrain feature extraction across multi-scale,multi-feature,and multi-method approaches for variable terrain

Terrain analysis uses different workflows to extract features from terrain models for the purpose of understanding topographic patterns and processes. However, the results of different workflows often conflict, leading to uncertainties about feature locations. Instead of relying upon a single workflow, we suggest that a fusion of information from multiple workflows better informs terrain analysis. From terrain data with different degrees of variability, we extracted terrain features related to the set of topographic surface network feature classes {peaks, pits, saddles, ridges, courses} using workflows from free, open-source, and commercial software. A multi-scale analysis produced terrain features with fuzzy membership values for various feature classes and revealed that terrain locations can exhibit characteristics of all classes. Multi-feature maps were created by determining at each location the dominant and second-ranked features, and an uncertainty value. Our multi-method approach incorporated all of the workflows’ multi-scale results and again produced multi-feature maps that increased the confidence of some features and reduced the signal of dissimilar results. We also found that high variability terrain produced crisper features in both spatial extent and membership strength. Our overall conclusion is that multi-scale, multi-feature, and multi-method analyses clarify terrain feature uncertainty.     

基于地形画像的相似区域推荐

在开展滑坡危险性分析、山区公路规划以及工业园区选址等工作时,研究区域的地形都是首要考虑的因素,为研究区域推荐几个与之地形相似的区域作为参考对象能够使得研究工作快速展开。本文提取了研究区的宏观地形特征与直方图统计特征,对研究区域的地形进行精细刻画;而后基于随机森林搭建区域分类模型,根据分类结果进行粗略推荐;最后基于粗略推荐结果,计算研究区域与其他区域之间的相似度,进行精细化推荐,为研究区域推荐5个与之地形最相似的参考区域。推荐结果显示,参考区域与研究区域的地形高度相似,在开展滑坡、泥石流灾害研究以及道路规划等工作时,可以参考推荐区域的研究结果。     

Developing detailed age-specific thematic maps for coffee (Coffea arabica L.) in heterogeneous agricultural landscapes using random forests applied on Landsat 8 multispectral sensor

Coffee is a commodity of international trade significance, and its value chain can benefit from age-specific thematic maps. This study aimed to assess the potential of Landsat 8 OLI to develop these maps. Using field-collected samples with the random forest classifier, splitting coffee into three age classes (Scheme A) was compared with running the classification with one compound coffee class (Scheme B). Higher overall classification accuracy was obtained in Scheme B (90.3% for OLI and 86.8% for ETM+) than in Scheme A (86.2% for OLI and 81.0% for ETM+). The NIR band of OLI was the most important band in intra-class discrimination of coffee. Landsat 8 OLI mapped area closely matched farm records (R^2?=?0.88) compared to that of Landsat 7 ETM+ (R^2?=?0.78). It was concluded that Landsat 8 OLI data can be used to produce age-specific thematic maps in coffee production areas although disaggregating coffee classes reduces overall accuracy.     

Spatial and temporal change in landslide hazard by future climate change scenarios using probabilistic-based frequency ratio model

The aims of this study were to apply, verify and compare a frequency ratio model for landslide hazards, considering future climate change and using a geographic information system in Inje, Korea. Data for the future climate change scenario (A1B), topography, soil, forest, land cover and geology were collected, processed and compiled in a spatial database. The probability of landslides in the study area in target years in the future was then calculated assuming that landslides are triggered by a daily rainfall threshold. Landslide hazard maps were developed for the two study areas, and the frequency ratio for one area was applied to the other area as a cross-check of methodological validity. Verification results for the target years in the future were 82.32–84.69%. The study results, showing landslide hazards in future years, can be used to help develop landslide management plans.     

Approaches for delineating landslide hazard areas using different training sites in an advanced artificial neural network model

The current paper presents landslide hazard analysis around the Cameron area, Malaysia, using advanced artificial neural networks with the help of Geographic Information System (GIS) and remote sensing techniques. Landslide locations were determined in the study area by interpretation of aerial photographs and from field investigations. Topographical and geological data as well as satellite images were collected, processed, and constructed into a spatial database using GIS and image processing. Ten factors were selected for landslide hazard including: 1) factors related to topography as slope, aspect, and curvature; 2) factors related to geology as lithology and distance from lineament; 3) factors related to drainage as distance from drainage; and 4) factors extracted from TM satellite images as land cover and the vegetation index value. An advanced artificial neural network model has been used to analyze these factors in order to establish the landslide hazard map. The back-propagation training method has been used for the selection of the five different random training sites in order to calculate the factor’s weight and then the landslide hazard indices were computed for each of the five hazard maps. Finally, the landslide hazard maps (five cases) were prepared using GIS tools. Results of the landslides hazard maps have been verified using landslide test locations that were not used during the training phase of the neural network. Our findings of verification results show an accuracy of 69%, 75%, 70%, 83% and 86% for training sites 1, 2, 3, 4 and 5 respectively. GIS data was used to efficiently analyze the large volume of data, and the artificial neural network proved to be an effective tool for landslide hazard analysis. The verification results showed sufficient agreement between the presumptive hazard map and the existing data on landslide areas.     

Approaches for delineating landslide hazard areas using different training sites in an advanced artificial neural network model

The current paper presents landslide hazard analysis around the Cameron area, Malaysia, using advanced artificial neural networks with the help of Geographic Information System (GIS) and remote sensing techniques. Landslide locations were determined in the study area by interpretation of aerial photographs and from field investigations. Topographical and geological data as well as satellite images were collected, processed, and constructed into a spatial database using GIS and image processing. Ten factors were selected for landslide hazard including: 1) factors related to topography as slope, aspect, and curvature; 2) factors related to geology as lithology and distance from lineament; 3) factors related to drainage as distance from drainage; and 4) factors extracted from TM satellite images as land cover and the vegetation index value. An advanced artificial neural network model has been used to analyze these factors in order to establish the landslide hazard map. The back-propagation training method has been used for the selection of the five different random training sites in order to calculate the factor’s weight and then the landslide hazard indices were computed for each of the five hazard maps. Finally, the landslide hazard maps (five cases) were prepared using GIS tools. Results of the landslides hazard maps have been verified using landslide test locations that were not used during the training phase of the neural network. Our findings of verification results show an accuracy of 69%, 75%, 70%, 83% and 86% for training sites 1, 2, 3, 4 and 5 respectively. GIS data was used to efficiently analyze the large volume of data, and the artificial neural network proved to be an effective tool for landslide hazard analysis. The verification results showed sufficient agreement between the presumptive hazard map and the existing data on landslide areas.     

Mapping East African tropical forests and woodlands — A comparison of classifiers

In mapping the forest–woodland–savannah mosaic of Budongo Forest Reserve, Uganda, four classification methods were compared, i.e. Maximum Likelihood classifier (MLC), Spectral Angle Mapper (SAM), Maximum Likelihood combined with an Expert System (MaxExpert) and Spectral Angle Mapper combined with an Expert System (SAMExpert). The combination of conventional classifiers with an Expert System proved to be an effective approach for forest mapping. This was also the first time that the SAMExpert had been used in the mapping of tropical forests. SAMExpert not only maps with high accuracy, but is also fast and easy to use, making it attractive for use in less developed countries. Another advantage is that it can be executed on a standard PC set up for image processing.Combining the conventional classifiers (MLC and SAM) with the Expert System significantly improved the classification accuracy. The highest overall accuracy (94.6%) was obtained with SAMExpert. The MaxExpert approach yielded a map with an accuracy of 85.2%, which was also significantly higher than that obtained using the conventional MLC approach.The SAMExpert classifier accurately mapped individual classes. Of the four classes of woodland mapped, the Open Woodland (with Terminalia) and Wooded Grassland classes were more accurately mapped using SAMExpert. The Open Woodland had been previously identified by ecologists, but had never been mapped.     

Macro Landslide Hazard Zonation Mapping - Case Study from Bodi – Bodimettu Ghats Section,Theni District,Tamil Nadu - India

Landslide is a common natural hazard that usually occurs in mountainous areas. Rapid urban development and high traffic intensity movements have been hampered to a great extent by phenomenon of landslides. In Ghat section, vertical cuttings and steep slopes are induced slope failures. An assessment of landslide hazards is therefore a prerequisite for sustainable development of the hilly region. In the present study, Macro Landslide Hazard Zonation was carried out in the Bodi – Bodimettu ghats section, Western Ghats, Theni district. The slope spreads over an area of about 10.09 sq km encompassing Puliuttu Ar. sub-watershed. The study was made with help of different types of data including Survey of India topographic map, geology map, important inherent factors like lithology, structure, slope morphometry, relative relief, land use/land cover and hydrogeological conditions using Bureau of Indian Standard (BSI 14496 (Part 2):1998) and related thematic maps. Based on the thematic layers, landslide hazard evaluation factor (LHEF) and total estimated hazard (TEHD) were calculated and the macro hazard zonation map was prepared. Landslide Hazard Zonation (LHZ) of the terrain shows that out of 17 facets, facets 1 to 5 and 8 falls under Moderate Hazard zone category and facets 6, 7 and 9 to 17 under the High Hazard zone category. The field study with further analysis for hazard concluded that about 68% of the total area falls in the high hazard zone.     

Effects of Landslide Sampling Strategies on the Prediction Skill of Landslide Susceptibility Modelings

In this study, landslide susceptibility assessments were achieved using logistic regression, in a 523 km² area around the Eastern Mediterranean region of Southern Turkey. In reliable landslide susceptibility modeling, among others, an appropriate landslide sampling technique is always essential. In susceptibility assessments, two different random selection methods, ranging 78–83% for the train and 17–22% validation set in landslide affected areas, were applied. For the first, the landslides were selected based on their identity numbers considering the whole polygon while in the second, random grid cells of equal size of the former one was selected in any part of the landslides. Three random selections for the landslide free grid cells of equal proportion were also applied for each of the landslide affected data set. Among the landslide preparatory factors; geology, landform classification, land use, elevation, slope, plan curvature, profile curvature, slope length factor, solar radiation, stream power index, slope second derivate, topographic wetness index, heat load index, mean slope, slope position, roughness, dissection, surface relief ratio, linear aspect, slope/aspect ratio have been considered. The results showed that the susceptibility maps produced using the random selections considering the entire landslide polygons have higher performances by means of success and prediction rates.     

Photo-interpretation and site assessment survey — A case study

A study was undertaken to find out the possibility of using the available aerial photos on scale 1 : 50 000 for selection of areas suitable for industrial plantation. Micro relief or terrain & vegetation density classes were interpreted on air photos and a PI map was prepared. The PI map was then overlaid on land facet & soil facet maps and the result shows that PI map is a useful tool for planning of raw material supply to the forest based industry. The growth data is to be obtained from field sample plots laid out in each stratum.     

Lightning flash mensuration using video from the space shuttle Columbia (STS‐32)

Abstract

Environmental data are often utilized to guide interpretation of spectral information based on context, however, these are also important in deriving vegetation maps themselves, especially where ecological information can be mapped spatially. A vegetation classification procedure is presented which combines a classification of spectral data from Landsat‐5 Thematic Mapper (TM) and environmental data based on topography and fire history. These data were combined utilizing fuzzy logic where assignment of each pixel to a single vegetation category was derived comparing the partial membership of each vegetation category within spectral and environmental classes. Partial membership was assigned from canopy cover for forest types measured from field sampling. Initial classification of spectral and ecological data produced map accuracies of less than 50% due to overlap between spectrally similar vegetation and limited spatial precision for predicting local vegetation types solely from the ecological information. Combination of environmental data through fuzzy logic increased overall mapping accuracy (70%) in coniferous forest communities of northwestern Montana, USA.     

Seismic hazard maps of Yangon and its surrounding areas

The purpose of this study is to develop probabilistic seismic hazard maps for Yangon and its surrounding areas including ‘Peak Ground Acceleration’ values for 2% and 10 % probability of the exceedance in 50 years at rock sites. The present study area is situated between the latitudes of N 13°37′ and N 20°2′ and the longitudes of E 93°35′ and E 99°5′. The study areas are focused on nine source zones centered around Yangon with the radius of about 200 km. The probabilistic seismic hazard maps are created by ArcGIS-9.3 software.     

Capturing ephemeral forest dynamics with hybrid time-series and composite mapping in the Republic of Panama

Measuring and progressing toward international goals of curbing deforestation and improving livelihoods of people who depend on forests requires nuanced understanding of forests and the processes surrounding deforestation and degradation. Despite rapid improvements in Earth Observation technology, monitoring of tropical forests remains hindered by persistent cloud cover, heterogeneous landscapes, long wet seasons, and small and ephemeral clearings masked by rapid growth. A hybrid method is presented that combines elements of both time-series and compositing approaches to best overcome these obstacles to map forest cover and change in the Republic of Panama based on Landsat imagery. The resulting Panama Vegetation-Cover Time-Series (PVCTS) maps depict forest cover in Panama from 1990 to 2016 at 30 m resolution. Acknowledging the fuzzy boundary between forest and non-forest classes, these maps employ a hierarchical classification scheme that reflects the natural process of regeneration and can accommodate different definitions of forest and deforestation. Classification accuracy is 97–98 % between forest/non-forest categories and 76–81 % for deforestation events. The maps show a slight greening of Panama from 1990 to 2016 caused by expansion of young secondary growth. The annual rate of deforestation in mature forest has remained around -0.6 %/yr, although young forests have matured at a similar rate such that there is no net loss of forest. While estimates of total forest cover are similar to official national estimates depending on forest definition, there is little agreement in location of deforestation events.     

Classification of remote sensing images from urban areas using a fuzzy possibilistic model

The classification of very high-resolution remotely sensed images from urban areas is addressed. Previous studies have shown the interest of exploiting the local geometrical information of each pixel to improve the classification. This is performed using the derivative morphological profile (DMP) obtained with a granulometric approach, using opening and closing operators. For each pixel, this DMP constitutes the feature vector on which the classification is based. In this letter, we present an interpretation of the DMP in terms of a fuzzy measurement of the characteristic size and contrast of each structure. This fuzzy measure can be compared to predefined possibility distributions to derive a membership degree for a set of given classes. The decision is taken by selecting the class with the highest membership degree. This model is illustrated and validated in a classification problem using IKONOS images.     

Automated mapping of Hammond's landforms

We automated a method for mapping Hammond's landforms over large landscapes using digital elevation data. We compared our results against Hammond's published landform maps, derived using manual interpretation procedures. We found general agreement in landform patterns mapped by the manual and the automated approaches, and very close agreement in characterization of local topographic relief. The two approaches produced different interpretations of intermediate landforms, which relied upon quantification of the proportion of landscape having gently sloping terrain. This type of computation is more efficiently and consistently applied by computer than human. Today's ready access to digital data and computerized geospatial technology provides a good foundation for mapping terrain features, but the mapping criteria guiding manual techniques in the past may not be appropriate for automated approaches. We suggest that future efforts center on the advantages offered by digital advancements in refining an approach to better characterize complex landforms.