This paper proposes a new method for assessing the sediment disaster risk using a deep-learning-based analysis of Digital Elevation Models (DEMs). The following three features are involved in this method : 1) conducting the risk assessment to each grid cell of DEMs, 2) calculating multiple terrain quantities related to sediment disasters, 3) incorporating values of the terrain quantities from surrounding cells for each cell. Three terrain quantities, gradient, Laplacian, and specific height difference, were used as explanatory variables of the sediment disaster risk. The author generated a data set of the explanatory variables by incorporating adjacent 7 by 7 grid cells into each cell. Based on deep learning training with this data set, a hierarchical neural network consisting of the five-layer structure with three intermediate layers was selected as the optimal model. The study area was set in the upstream region of the Tsurugi River catchment, Hofu City, Yamaguchi Prefecture. The model appropriately detected about 85% of landslide sites induced by the 2005 Hofu heavy rain disaster. The model analysis also showed that the sediment disaster risk became higher in 0-order basin valley, which is conventionally considered landslide-prone areas in mountain regions.

The policy of “preventive maintenance of infrastructures” is to exercise preventive works in early stages of deterioration before infrastructure facilities cease to be functional. It aims at reduction and leveling of total cost of maintenance over life cycle of facilities. As of now, preventive maintenance is successfully assumed and put into practice by administrators of Sabo-related facilities. Priorities for improving its effectiveness are ; 1) enhanced precision of deterioration projection of each facility given that efficiency of preventive works depends on the shape of deterioration curve. 2) disposition of engineers qualified for on-site inspection and evaluation of deterioration. Provision of database of results of inspection, evaluation and preventive works executed by administrators is expected to contribute to both.

The purposes of this study are to develop a numerical model of shallow water equations with Bingham shear stress model, to evaluate the accuracy by reproducing dam break phenomena with different viscous liquids, and to clarify the differences between flow profiles computed by the numerical model with Bingham model and flow profiles computed by the numerical model with Manning model. By conducting flume experiments, we obtained the time changes of flow profiles with water and two type of viscous liquids. We reproduced the time changes of experimental flow profiles by computing with both the numerical model with Bingham model and the numerical model with Manning model. The results said that the numerical model with Bingham model could reproduce the experimental flow profiles including both rarefaction waves and shock waves. And the numerical model with Bingham model could reproduce more accurately than the numerical model with Manning model.

At the confluence, a tributary accompanied with large sediment production and supplying make to increase amount of water and sediment flowing into main river. In addition, rising riverbeds and water levels at the confluence may cause trouble such as sediment floods. In promoting Sabo works, to comprehensively understand the sediment dynamics at the confluence joining a tributary with large sediment production and supplying is crucial issue. In this study, flume experiment considering bed load transport from tributary with large sediment production from mountainous basin were performed. Our flume experiment results suggest that ; the water flowing down from upstream of the main river prominently drifts into opposite side of tributary and the riverbed around the side joining tributary shows the deposition trends, and the riverbed around opposite side joining tributary shows washed trend. Furthermore, the trends of riverbed is more prominent as the flow rate ratio of tributary hydrograph to the main river is larger, and the time gap of peak flow between main stream and tributary is larger. The sedimentary and eroded areas near the confluence are similar to the areas of flowing water from the main and tributary. Hereafter, in order to establish a model that easily predicts the flow conditions and riverbed fluctuations at the confluence, it is necessary to quantitatively relate the inflow rate and the flow area of main and tributary, and planner distribution of riverbed height.