DTM Generation and Avalanche Hazard Mapping using Large Format Digital Photogrammetric Data and Geomatics Technique

Journal of Remote Sensing & GIS Volume 4, Issue 2, ISSN: 2230 -7990 __________________________________________________________________________________________ DTM Generation and Avalanche Hazard Mapping using Large Format Digital Photogrammetric Data and Geomatics Technique Snehmani1, Mritunjay Kumar Singh1, 2*, R. D. Gupta2, A. Ganju1 1 Snow Avalanche Study Establishment-Research and Development Center (SASE-RDC) Him Parisar, Chandigarh (UT), India 2 Motilal Nehru National Institute of Technology (MNNIT), Allahabad (UP), India Abstract The main objective of the study is Digital Terrain Model (DTM) generation from aerial photogrammetric data and identify and map the potential avalanche prone zones in Manali region. Avalanche is a dynamic hazardous phenomenon in the snow-bound mountainous terrain. Mapping of avalanche prone terrain is crucial to minimize the avalanche hazard. Nowadays, airborne data capturing technology, such as large-format Photogrammetry, has opened new vistas for the mapping of complex and inaccessible mountainous areas. In the present study, large format digital Photogrammetry data of 20 cm ground sample distance (GSD) have been used to generate high-resolution and accurate Digital Elevation Model and ortho-images. Digital terrain model along with its derivative terrain products and land cover map generated from land cover classification of derived ortho-photo is analyzed to locate the probable avalanche zone. The terrain characteristics, snow-pack condition and prevailing meteorological conditions are the groups of variables that influence the occurrence of avalanche. Amongst these, the terrain characteristics is the most influencing factor, and easier to map due to its stable nature along the time. Therefore advanced geo-informatics techniques have been used by mixing terrain property, Digital Elevation Model (DEM) and satellite imagery to determine the different geographical factors that affect the avalanche triggering. Also the derived information was combined in Analytic Hierarchy Process to extract a map of the avalanche prone zones in the study area standard mapping techniques as coarseresolution data are not very appropriate for such studies. Keywords: photogrammetry, DTM, snow, avalanche *Author for Correspondence: E-mail: jay_rsgis@yahoo.co.in High mountain areas are strongly affected by different types of hazards due to their abrupt landscape and mass movement related processes [1]. The sensitivity of these areas particularly glacier regions with respect to “climatic and anthropogenic” conditions is aggravated. The mapping of avalanche hazard information such as location, extent and spatial pattern is essential for avalanche mitigation measure planning and many further issues in avalanche research. curvature, roughness or vegetation cover, b) Meteorological parameters such as wind, temperature, amount of fresh snowfall or humidity, c) Snowpack parameters such as existence of weak layer, bonding between layers, free water content or grain size and grain forms [2]. The event of avalanche vulnerability also depends on a variety of factors such as earth vibration, extreme precipitation and man-made turbulence, i.e., noise and heavy movement of a skier or snowboarder. The avalanche hazard depends on different parameters. We can roughly classify these parameters in three groups a) Terrain parameters such as slope, exposition, In addition to this, there is triggering of avalanche, initiated by further loading caused by humans or naturally by fresh snow or abrupt warming [3]. INTRODUCTION JoRSG (2013) 4-13 © STM Journals 2013. All Rights Reserved Page 4 DTM Generation and Avalanche Hazard Mapping Snehmani et al. __________________________________________________________________________________________ For the numerical simulation of snow avalanches with up-to-date tools such as ELBA+ [4], accurate information on the location of the release zones and high quality DEM data are crucial [5]. It is also the base for large-scale snow avalanche hazard mapping. Today avalanche experts base their release zone identification on long-term experience, field visits and experiences from local people [6], as well as on databases of historic avalanche events. Due to sparse settlement of most cold mountain regions in the world, the coverage with high spatial resolution geo-datasets is usually insufficient for detailed natural hazard analysis. However, up-to-date remote sensing technology is used now more and more to close such data gaps. In particular, airborne digital photogrammetry is able to generate high spatial resolution (approx. 1 m), high quality DEM derived from true color and false color ortho-imagery [7]. In many regions, digital Photogrammetry is the only feasible source for such geo-datasets as base for alpine natural hazard investigations. The avalanche regions are generally remote areas where ground based collection of data and monitoring is difficult. Remote sensing could be innovative tool, which can monitor larger areas by satellite images and DEMs. Currently, 30 m ASTER and 90 m SRTM DEM is available for most of the world. This readily available spatial data lack the resolution to identify small avalanche starting zones, smooth slope and aspect calculations for larger and identifiable slide paths. Therefore, large format digital Photogrammetry along with geomatics techniques have great potential for monitoring and assessing avalanche hazard. One of the primary products of airborne digital data is very accurate and precise bare ground DEMs. Ice velocities studies [8–10] have amply shown that airborne digital data can be used to monitor snow glacial movement, snow accumulation and predict the onset of avalanche. The data set can further be employed to estimate the risk from a particular avalanche. The analysis of airborne digital data have played an important role for avalanche and other high mountains hazard research in recent years in different mountain ranges of earth such as Alps [7]. However, avalanche hazard mapping based on airborne DEM has not been attempted in Indian Himalayas. Therefore, we introduce avalanche hazard mapping based on airborne DEM for Indian Himalayas. In the present study, aerial digital Photogrammetry data of 20 cm GSD have been used to generate high resolution and accurate Digital Elevation Model and related derivative along with ortho-photos of the study area. All the generated data have been further processed and used in AHP model as input for demarcating avalanche hazard zones. Study Area The study area (Figure 1) falls in the state of Himachal Pradesh covering Manali town and its nearby areas, i.e., Palchan, Solang, Dhundi, Teling etc.; a sub-basin of Beas river lying in the Pir-Panjal range in the NWHimalayas. This is high mountainous area with mean altitude of approximately 4430 m asl (above sea level); covering approximately 100 sq km area. Previous studies suggest that area receives moderate to heavy snowfall in winter and Manali-Leh highway is threatened by a large number of avalanche tracts. The identified avalanche tracts start from MSP1 (Manali South Portal) close to Manali going to MSP13, right at the entrance of Rohtang tunnel. The tracks MNP1 (Manali North Portal) to MNP5 are located across the Rohtang pass in the Lahaul-spiti of Himachal Pradesh. The study area falls in SoI toposheet no. 52H/3 at 1:50,000 scale. The climate in Manali is predominantly cold during winters and moderately cool during summers. The temperature ranges from 4°C (39 °F) to 30°C (86 °F) over the year. The average temperature during summer is between 14°C (57 °F) and 20°C (68 °F), and between −7°C (19.4 °F) and 10°C (50 °F) in winter. Data Sources Once the area for survey is decided, permission is obtained from authorized organizations in India. Rockwell Commander C-690 Jetprop twin-engined aircraft have been used for the aerial survey campaign from Chandigarh in the 1st week of Dec’09 and got the flying window for only 3 (2nd, 6th, and 7th Dec '09) days. UltraCamXp (UCXp) camera JoRSG (2013) 4-13 © STM Journals 2013. All Rights Reserved Page 5 Journal of Remote Sensing & GIS Volume 4, Issue 2, ISSN: 2230 -7990 __________________________________________________________________________________________ was used for data capturing photogrammetric survey. for the Fig. 1: Extent of Test Site Manali and Near-by Area. Accurate flight planning for airborne survey is essential for a total Quality Assurance experience. Since expenses related to a flight mission are very high, proper mission planning at this stage is very important. The acquisition took place in clear conditions with good visibility, but a reasonable level of air turbulence. Every capture image has individual details files containing all the information such as time of acquisition, lat/long, date, direction, etc. The data were collected flying between 6000 m and 7200 m asl in order to product imagery with a nominal GSD of 0.25 m. Consecutive photos in each flight line is having an average forward overlap of 80%–90% to ensure full stereoscopic coverage. Side lap between adjacent parallel flight lines is a minimum of 60%. Flight data are post-processed to provide exact positional information for the airborne sensors and camera systems. The orientation data provided by the Inertial-measuring unit (IMU) when combined with Global positioning system (GPS) data effectively eliminates the need for aerial triangulation in airborne photography and enables scanners to be used as mapping tools. Advanced control technologies such as GPS and IMU provide a significant contribution to mapping accuracy, efficiency of operations and cost savings. Total 23 Ground control points (GCPs) have been collected covering the study area. Time consumed on each point was around one hour so that we can more accurately measure the point. Positional dilution of precession (PDOP) for each point was less than 3. Solution types for all the points were fixed and RMS error was less than 0.05. Wherever possible, we have marked these GCPs with paint for identification on aerial photographs. Two GPS base-station receivers were operated for the duration of the data acquisition. One receiver located at the site SCHOOL ROOF and the other at the site SASE VALLEY. This data were collected for use during the processing of the camera position and orientation system (POS) data. The POS was determined using the collected GPS/IMU datasets and Applanix POSPac software. This work was all undertaken in WGS84 coordinate system. Single baseline GPS processing was used to compute the coordinate. Methodology Data Processing Methodology for Digital Photogrammetric Data A two-step process has been followed to make airborne digital images ready for photogrammetric mapping. This work has been done using the Vexcel software, Office Processing Centre (OPC). For each exposure made by the UCXp camera it creates 13 sub images. Nine of these images are panchromatic and the other four are the red, green, blue and infrared channels. These collections of images are referred to as the lvl0 images. In the first step (Level 1) the 13 images are stitched together and corrected for lens distort and vignetting. In the second step (Level 2 & 3) the radiometric adjustments are made for sun hot spot effects and exposure effects and the pan-sharpening of the red, green, blue and infrared images. Figure 2, represents Work flow of the data processing. Level 3 data are the photogrammetric mapping ready aerial photography. These images are in TIFF file format, 16bit per channel and contain four-color bands R/G/B/NIR. Figure 2, describes the workflow JoRSG (2013) 4-13 © STM Journals 2013. All Rights Reserved Page 6 DTM Generation and Avalanche Hazard Mapping Snehmani et al. __________________________________________________________________________________________ adopted for data processing. The bore-sight calibration photos were processed using Applanix POSMMS CalQC software. The results confirmed that the system was operating within specification and established that the misalignment angles were –1.23, 4.29,–0.13 arc minutes. Fig. 2: Workflow of the Data Processing. Perform Aerial Triangulation Aerial triangulation is a complex process that includes all the information related to aerial photo capturing i.e. flying height, focal length, scale, yaw, pitch and roll with each image in the block based on pre-determined specifications, performing interior orientation, collecting and measuring all tie, check, and control points visible on the photographs, and applying a least square block adjustment. This process calculates exterior orientation parameters for photographs and threedimensional coordinates for all considered object points. The accuracy of any image matching method is reliant on the availability of “good” estimated values for exterior orientation parameters. Whereas some of these parameters are known approximately (photo base, average terrain elevation, and average flying height). Aerial Triangulation was performed to extend horizontal and vertical control from relatively few ground surveyed control points (14) to each unknown ground point included in the solution. The supplemental control points are called pass points, and they are used to control subsequent photogrammetric mapping. In order to easily deal with the image files of larger sizes, a full set of Gaussian overviews were generated for these images and overviews were used as a replacement to the actual images. Relative and Absolute Orientation In Relative Orientation a set of tie and pass or control points were measured. Tie or pass points that are used to bridge the images were collected in such a way that they are welldefined and well-distributed throughout the model. Control and check points were measured as pass points, these were carried through to the absolute orientation process, where they were used as control points. a suitable tie-point pattern is created using the Project Management utility to generate a well distributed image point configuration. We have taken 5 point pattern. Strips were tied together in a 4 or 6 image mode, viewing 2 or 3 images from each strip. During the Absolute Orientation process, control and check points were measured. A minimum of three control points were measured to compute an Absolute Orientation. The control points measured as pass points, were carried through to the Absolute Orientation process, and were used as control points. Once all the points were measured, they were reviewed, modified till the sigma RMS values were adjusted to the tolerance limits. Bundle Adjustment In this process, a mathematical technique had been performed, to determine the positions and orientations of each image, as they existed at the time of image capturing. The process was concluded, after ensuring that the RMS values are within tolerances. Bundle block JoRSG (2013) 4-13 © STM Journals 2013. All Rights Reserved Page 7 Journal of Remote Sensing & GIS Volume 4, Issue 2, ISSN: 2230 -7990 __________________________________________________________________________________________ adjustment was performed on the project using all Photogrammetry measurement and 14 control points with the standard deviation of 1 cm. DTM Generation The Planimetric and Height data were captured in blocks pre-determined by the Compilers within the model area to enable other compiler to continue the same model and in the recommended sequence of Hydrology, Road network, Structures and utilities. Break lines will be collected with sufficient frequency and with a density of points to adequately represent the terrain features where there is abrupt change in elevation and where break line are necessary to represent the true geographical character of the terrain. Ridgelines, valleys, cliffs, ravines and any steep change in the elevation must be collected as a break line. Necessary care will be taken to ensure that, break lines must not cross over one another. The break line data will ensure that all the model area is covered with accurate height information. Mass points are collected as a regular grid of points that are manually measured. The grid intervals shall be selected in such a way that the terrain is clearly represented without any loss of undulations. The grid distance less terrain areas shall be increased to facilitate faster compilation. The contours are then generated at 1 m intervals mentioned for the different mapping areas and are edited to meet the accuracy standards. The total file area is divided into grids and the checking of data is carried out grid wise. A checklist of the features that are captured in the project is prepared and the file is checked to ensure that all the features are captured without missing and wrong interpretation. The DTM data is checked thoroughly by generating contours and the areas with improper contours are manually edited. Ortho-Photo Production and Mosaicking Ortho Process mainly involves the rectification of Aerial photos using digital terrain data that represents real ground as closely as possible. An orthorectified image (or ortho-photo) is one where each pixel represents a true ground location and all eometric, terrain, and sensor distortions have been removed to within a specified accuracy. Orthorectification is the process of transforming the central perspective of a photograph to an orthogonal view of the ground. This process removes the sensor tilt and terrain relief effects. Scale is constant throughout the ortho-photo, regardless of elevation, thus providing accurate measurements of distance and direction. AT files with respect to the coordinate system and Projection parameters & then the DEM data (Figure 3) is checked for its completeness with respect to layers and also for elevations without Zero values and utilized for the orthorectification process. The elevation data were then applied and suitable ortho properties were set with the parameters to proceed with the ortho-rectification. After the generation of the rectified orthos, they were all loaded and checked for proper overlap. Then the seams were captured along the overlap areas of adjacent images. Care was taken to ensure that the collection was carried out along the edge lines, roads without crossing any structures. Once the seam line collection and editing was completed, we checked whether all seam line consists of assign polygons. In the mosaicking process, the images were stitched together after seam line editing and were made into a single complete Tile, by assigning proper tonal balancing settings. The Mosaic Tile was taken and removal of hair, dust was carried out. Figure 4, shows the orthorectified, color balanced, mosaicked data of the study area. Fig. 3: Extracted DTM of the Study Area. JoRSG (2013) 4-13 © STM Journals 2013. All Rights Reserved Page 8 DTM Generation and Avalanche Hazard Mapping Snehmani et al. __________________________________________________________________________________________ Generation of Thematic Maps The thematic maps such as slope (Figure 5a) and aspect (Figure 5b) have been calculated using Jones [13], 4-Cell Method from derived DEM (Figure 3), generated from digital photogrammetric data. Curvature (Figure 7) has been calculated from the Zevenbergen and Thorne [14], approach, which uses a higherorder 9-parameter polynomial going through all 9 neighboring points. Land cover (Figure 7) maps are generated based on the developed ortho-photos. Fig. 4: Ortho-photo of the Study Area. Analytic Hierarchy Process (AHP) In the present study, AHP developed by Saaty [11], had been used to calculate the weighting factors with the help of preference matrix. In the preference matrix all, the known significant criteria had been compared against each other with reproducible preference factors. AHP is the aggregation of criterion approach. This method facilitates to make a judgment in logical manner. At each level, a preference matrix is built through comparison, to access the decision maker preferences between the criteria of considered level. All the available, parameters had been used in this method to check consistency ratio. Consistency ratio is the measure of how much variation is allowed and it must be less than 10%. Satty and Vargas [12], suggested a scale for comparison consisting of values ranging from 1 to 9 which describe the intensity of importance (preference/dominance). Table 1, gives the scale of comparison: Fig. 5a: Slope Map of the Study Area. Table 1: Scale of comparison in AHP. Intensity of Description importance 1 Equal importance 3 Moderate importance 5 Strong importance 7 Very strong importance 9 Extreme importance 2,4,6,8 Intermediate values Fig. 5b: Aspect Map of the Study Area. JoRSG (2013) 4-13 © STM Journals 2013. All Rights Reserved Page 9 Journal of Remote Sensing & GIS Volume 4, Issue 2, ISSN: 2230 -7990 __________________________________________________________________________________________ Fig. 6: Land Cover Binary Map. considered parameters) or objectives in order of importance with reference to snow avalanche. Thus, numeric estimation of the risk, i.e. the adopted values by each parameter involves. Figure 5a and Figure 5b, shows slope and aspect map derived from the DEM (Figure 3), generated from digital photogrammetric data. The primary parameter is slope. A range of slope between 25 degrees to 30 degrees was considered the most potential for avalanche formation. It is widely accepted that avalanches release from slopes between 300 and 500 [2, 3]. DEM has been classified in to 9 elevation zones based on previous avalanche zone knowledge about the area. Influence of land cover on Snow avalanche has been taken into consideration. Ranking Method was used to rank the criteria or objectives in order of importance with reference to snow avalanche and AHP is adopted for the avalanche hazard zonation mapping. Figure 8, represent avalanche hazard area mapped. Weighted summation method has been applied based on a combination of standardized criterion scores and weights. The evaluation scores is calculated for each option by multiplying each criterion score by the corresponding criterion weights and adding the products given in Table 9. Table 2: Weights given to Slope Categories. Class <12 12–25 25–45 >45 Fig. 7: Curvature Map. Weight Assignment Procedures Weights and ratings can be determined based on the subjective expert's opinions as well as based on the objective analysis. In present study, factors and their categories were assigned numerical values based on the field experience or knowledge of the experts and study. Weight 1 4 9 3 Table 3: Weights given to Land Cover. Land Cover Snow Barren Grass Trees Weight 9 6 4 1 Table 4: Weights given to Aspect Categories. In this case, Ranking Method was used to rank the criteria (Tables 2–6 are showing weightages give to different classes and Table 7, represents criteria weights for all the JoRSG (2013) 4-13 © STM Journals 2013. All Rights Reserved Aspect Flat N NE E SE S SW W NW Weight 1 9 9 3 7 3 1 1 4 Page 10 DTM Generation and Avalanche Hazard Mapping Snehmani et al. __________________________________________________________________________________________ Table 5: Weights given to Profile Curvature. Class Weight <0.2 1 0.2 – 1 4 >1 9 Table 6: Weights given to Elevation Zones. Aspect 4458 – 4744 4744–5030 3887–4172 4172–4458 3601–3887 3345–3601 0–2743 2743–3029 3029–3315 Weight 1 2 3 4 5 6 7 8 9 Table 7: Criteria weights for all the considered Parameters. Criteria LULC Curvature Slope Aspect Elevation Weight 0.1145 0.1202 0.481 0.1159 0.1684 Table 8: Risk Values for all the considered Parameters. Parameters Slope Altitude Curvature Ground Cover Aspect Risk Factor 0.275 0.225 0.2 0.2 0.1 The sum of the products calculated for each option becomes the decisive factor for the evaluation of score. The higher the evaluations score, the more suitable avalanche prone area of the given option in the ranking. Table 9: Suitability Index for Avalanche Release Areas. Suitability Index MCE Value Area Range (in Hectare) No Risk/ Very Low 1.0-2.4 500 Low 2.4-3.1 1400 Moderate 3.1-3.8 4100 High 3.8-4.2 3000 Severe 4.2-4.7 700 Very Severe 4.7-5.0 300 The suitability map is grouped into six classes. The area (in Hectare) of each class is given in Table 9. The areas “Very Low/No Avalanche” for avalanche trigger sites (value 1-2.4) at lower elevation in valley, except some area at higher elevation near to ridges/spurs, where the slopes area less than 120 (i.e. occur at the flattest areas). The “Low Avalanche Prone Area” avalanche initiation (values 2.4-3.1) is restricted to the valleys between the mountains at lower elevations and on concave surfaces. The “Moderate Avalanche Prone Areas” (values 3.1-3.8) tend to occur on all aspect of slope between 450-550 and concave surface. The “Highly Avalanche Prone Areas” (values 3.8-4.2) tend to occur on slopes of all aspects and on convex surfaces. The “Very Highly Avalanche Prone Areas” (values 4.2-4.7) occur at mountain peaks, along ridges and on non-forested slopes and bare slopes between critical angles of 250-450, at all aspects and on convex surfaces. RESULTS AND DISCUSSION The identification of potential avalanche hazard zones is difficult. In many alpine areas around the world such as the Indian Himalayas nearly no information on past avalanche events are recorded and most of the other required information is not, or only very limited available. Furthermore, the dimension of terrain affected by avalanches is vast and most regions are partly or completely inaccessible. But high spatial remote sensing sensors are able to map large areas even in otherwise inaccessible terrain. Such datasets get more and more available, and have proven their ability to generate high quality DEM and aerial ortho-imagery [7]. For areas such as the Indian Himalayas, these datasets are the only available base for the identification of avalanche hazard zones. By only using terrain information and neglecting weather and snowpack information, this approach is clearly limited. However, these terrestrial parameters are important input to identify avalanche hazard zones, which have to be further investigated, especially if large, poorly accessible areas have to be assessed. The results obtained for the Manali area demonstrate the value this study. Furthermore, JoRSG (2013) 4-13 © STM Journals 2013. All Rights Reserved Page 11 Journal of Remote Sensing & GIS Volume 4, Issue 2, ISSN: 2230 -7990 __________________________________________________________________________________________ this study enables the planning of avalanche mitigation measures. The polygon area of snow avalanche hazard zone demarcated in this study vary from already registered avalanche hazard zones in terms of shape, size and spatial extent. The number of avalanche hazard zones demarcated in this study have increased in comparison to already registered avalanche hazard zones. In this study, high-resolution airborne digital data have been used for the first time, so the outcome obtained is more accurate and reliable as compare to previous studies of concerned study area. AHP was used over the other methods such as multi-criteria and statistical methods because of its robustness. Here the sub-criterion can also be taken in for calculation. For making decision over many criterion and subcriterions of AHP becomes handy to assign the desired weightings to all. Thus advantage of this method, we can compare all the factors and sub-factors with each other. In AHP, the consistency ratio is numerical index to check the consistency of the comparison matrix. The AHP results generated were analyzed with reference to the registered avalanche sites. The registered avalanche sites were prepared from direct field survey and research work done at SASE. The results show good correlation with the registered avalanche sites. This study is very useful for avalanche hazard zonation, safety assessment and planning for infrastructure in difficult mountainous regions. Two new potential avalanche sites have also been identified which are located near Manali north portal. A mathematical process of AHP has been used to find out the avalanche sites considering terrain factors. The other two factors such as meteorological parameters and the snow pack properties were not considered. As Himalayas show complex variation in the meteorological conditions in mere aerial distance of 100 km and thus the snow pack properties also change over that particular range. The registered avalanche sites were marked on basis of occurrence of avalanche in the study area. Hence the difference in the proposed and registered sites occurs. In AHP the results completely depend on the preference matrix generated by comparing the factors with each other. And expertise in assigning the proper weightage in AHP would come only with experience. To generate more accurate results time-to-time ground survey is very essential. In addition, the values obtained from the satellite data are the values at that particular instant when the satellite passes over the study area. CONCLUSION The avalanche hazard mapping by using high resolution DEM, Land cover map along with AHP and numerical simulation model can be used in avalanche hazard management by taking the appropriate actions for precaution, warning systems for prevention of disastrous consequences, monitoring of avalanche hazards and assessment of damage. Fig. 8: Avalanche Prone Map of Manali Area. This study has scope for future researches to be carried out in this field. As the avalanche database will grow, future goals become more ambitious. For maintaining the current database, more spatial information may be added and application can be written to access and analyze the data. This study deals with the terrain factors causing avalanche and other factors for future scope are: Snow pack JoRSG (2013) 4-13 © STM Journals 2013. All Rights Reserved Page 12 DTM Generation and Avalanche Hazard Mapping Snehmani et al. __________________________________________________________________________________________ properties and meteorological conditions can be incorporated. Taking more ground observation values from observatories avalanche simulation and run-out study can be done to know the impact for snow cover type, which can be of avalanche in different climate conditions. ACKNOWLEDGEMENT The authors are thankful to Dr. Rakesh Bhambri, (Scientist- Wadia Institute of Himalayan Geology) for his valuable contribution in compiling the manuscript. REFERENCES 1. Huggel C. et al. Evaluation of QuickBird and IKONOS imagery for assessment of high-mountain hazards, EARSeL eProceedings 2005. 2. Schweizer J., et al. Snow avalanche formation. Rev. Geophys. 2003; 41(4): 1016p. 3. McClung D. M., Schaerer P. The Avalanche Handbook, The Mountaineers Books Seattle 2006; 342p. 4. Keiler M., et al. Avalanche risk assessment - a multi-temporal approach, results from Galtur, Austria, Natural Hazards and Earth Systems Sciences 2006; 6: 637–651p. 5. Buhler Y., et al. Sensitivity of snow avalanche simulations to digital elevation model quality and resolution, Annals of Glaciology 2011; 52 (58): 72–80p. 6. Rudolf-Miklau F., Sauermoser S., Handbuch Technischer Lawinenschutz, Berlin, Ernst & Sohn 2011. 7. Bühler Y., et al. High resolution DEM generation in high-alpine terrain using airborne remote sensing techniques. Transactions in GIS (in press) 2012. 8. Kääb A., Photogrammetric reconstruction of glacier mass-balance using a kinematic ice-flow model. A 20-year time series on Grubengletscher, Swiss Alps, Annals of Glaciology 2001; 31: 45–52p. 9. Strozzi T., et al. Detecting and quantifying mountain permafrost creep from in situ inventory, space-borne radar interferometry and airborne digital photogrammetry. International Journal of Remote Sensing 2004; 25 (15): 2919– 2931p. 10. Baltsavias E. P., et al. Digital surface modelling by airborne laser scanning and digital photogrammetry for glacier monitoring. Photogrammetric Record 2001; 17 (98): 243–273p. 11. Saaty T.L., A scaling method for priorities in hierarchical structures, J. Math. Social 1977: 15p. 12. Saaty T.L., and Vargas L.G. Prediction, projection and forecasting; Kluwer academic publishers: Dordrecht 1991. 13. Jones KH. A comparison of algorithms used to compute hill slope as a property of the DEM. Computers and Geosciences 1998; 24 (4): 315–323p. 14. Zevenbergen L.W., Thorne C.R. Quantitative analysis of land surface topography. Earth, Surface Processes and Landforms 1987; 12: 47–56p. JoRSG (2013) 4-13 © STM Journals 2013. All Rights Reserved Page 13