A fast and efficient algorithm to map prerequisites of landslides in sensitive clays based on detailed soil and topographical information

We present an algorithm developed for GIS-applications in order to produce maps of landside susceptibility in postglacial and glacial sediments in Sweden. The algorithm operates on detailed topographic and Quaternary deposit data. We compare our algorithm to two similar computational schemes based on a global visibility operator and a shadow-casting algorithm. We find that our algorithm produces more reliable results in the vicinity of stable material than the global visibility algorithm. We also conclude that our algorithm is more computationally efficient than the other two methods, which is important when we may want to assess the effects of uncertainty in the data by evaluating many different models. Our method also provides the possibility to take other data into account. We show how different soil types with different geotechnical properties may be modelled. Our algorithm may also take depth information, i.e. the thicknesses of the deposits into account. We thus propose that our method may be used to provide more refined maps than the overview maps in areas where more detailed geotechnical/geological data have been acquired. The efficiency of our algorithm suggests that it may replace any global visibility operators used in other applications or processing schemes of gridded map data.