Kamikamihorizawa Creek on the slopes of Mount Yakedake, Nagano Prefecture, was selected as an observation site for debris flows due to the high frequency of debris flow and instrumented with monitoring equipments in 1970: seven years after the last phreatic explosion of this volcano. The monitoring system was improved by adding speedometers, stage meters, seismometers and so on. Over the last 30 years, data were obtained from 88 debris-flow events that contained more than 200 episodes of debris-flow surges. The data indicated the following common features related to debris flows. 1) High-intensity rainstorms with a short duration of about 10 min were more likely to trigger debris flows than storms with large rainfall for longer durations. 2) The threshold of rainstorm intensities for debris flows increased due to the changes in hillslope hydrology. 3) The hydrograph of debris-flow surges showed mass and boulder focusing on the front due to mechanical processes. 4) Pulsation and superelevation were common features that were significant as hazard-inducing factors. 5) Monitoring of seismic signals from debris flows is useful for hazard warning as well as for estimating debris-flow hydrographs. 6) Evaluation of hazard-prone locations downstream in the fan is possible by analyzing the interaction of the flows and the fan morphology. 7) Moderation of debris-flow hydrographs is achievable using new net dam or flat-board breakers in the lower reaches of gullies.
This is a review of the national policy related to hazard maps for sediment-related disasters in Japan. Until the 1980s, we focused mainly on rainfall-triggered sediment-related disasters, including debris flow, deep-seated landslides, and steep-slope failures, and until 2001, Japan did not have any laws related to hazard maps for sediment-related disasters. The Sediment-related Disaster Prevention Act became effective in 2001 and required all prefectural governments to publish hazard maps. Hazard mapping for volcanic sediment-related disasters started in 1991, and hazard maps have now been published for all 29 volcanoes that would have significant social impacts in the event of eruption. Investigations aimed at assessing the susceptibility to massive collapse and shallow seismic landslides recently started based on new methods.
Measurement of sediment transport in mountain torrents is important to elucidate the nature of the process. Many obstacles have prevented long-term continuous measurement of natural river basins with a substantial scale, so only a few sporadic observations have been collected over short durations. Direct sampling, therefore, has recently been augmented by more indirect but stable methods (hereafter ‘indirect methods’) including a hydrophone pipe-microphone acoustic sediment-discharge measuring system (hereafter ‘hydrophone system’). Hydrophone systems count the times (hereafter ‘pulses’) at which bedload sediments strike a steel pipe-microphone acoustic sensor. A statistical analytical method can be used to calibrate and estimate bedload amounts sampled by a sediment pit with a linear combination of pulses and flow discharges. Sediment discharges exhibit striking variability in orders of magnitude during flood events. However, researchers have not explicitly taken into account the volatile nature of sediment transport phenomena, as most quantitative studies have applied unitary analytical forms. In this study, three hydrophone systems were installed in 100- and 200-km2-scale river basins together with sediment trap pits. During floods, most sediment (approximately 80%) occurred within the first one-quarter of sediment sequences reordered in a descending manner. Therefore, sediment-hydraulic quantities were dissected into two sets: the upper quarter and the lower three-quarters. Frequency distributions of sediment-hydraulic quantities for each set were compared closely to differentiate the two sets based on a set of cutoffs using only direct methods. Bimodal analytical fitting forms were drawn for each set. The results revealed that at the time scale of flood events, the quantitative accuracy of sediment discharge estimates can be improved by dissecting sediment discharge phenomena into periods of concentrated and of minor occurrences, incorporating sediment discharge variation during the duration of floods.