In order to maintain and strengthen the road network throughout Japan against seismic slope failures, there is an urgent need to develop a risk assessment method that is applicable to the diverse geology of Japan, especially sedimentary rocks with a wide distribution. However, the existing method (Tanehira et al., 2022) is a model for volcanic rocks using the 2016 Kumamoto Earthquake as a case study, and sedimentary rocks are not the main target. Therefore, in this study, a new model for sedimentary rocks was developed using the 2004 Mid Niigata Prefecture Earthquake as a case study, and the conventional model and the proposed model were applied to volcanic and sedimentary rock areas, respectively, using the 2008 Iwate-Miyagi Nairiku earthquake as a validation case. As a result, the prediction accuracy of the number and locations of collapses varied greatly depending on the model applied, and the proposed model was confirmed to have superior accuracy in sedimentary rock areas compared to conventional models. The use of both models, depending on the geological conditions, has shown the possibility of conducting risk assessments of road networks against slope failures during earthquakes on a wider area.

  In airborne laser surveying, the generation of ground data from the original dataset typically involves an initial automated extraction followed by manual removal of unnecessary points by technicians. Consequently, post-survey processing time was required for such removal tasks when conducting terrain analysis. To address this issue, Kriging interpolation was applied to the automatically extracted data from the original dataset (referred to as near-ground data) to compute grid data. As a result, grid data comparable to those derived from ground data were obtained, effectively capturing unique micro-topographic features such as small cliffs and outcrops, which are of interest for on-site inspection by survey technicians. Furthermore, this method eliminates the need for visual inspection by technicians during ground data creation, significantly reducing processing time and labor requirements.