In 2022, 14 people died due to wind and flood damage. At locations where casualties were caused by floods, rivers, and landslides, any or all of the seven rainfall indicators exceeded 100%. Based on the past 4 heavy rain events from 2018 to 2021, we created an evaluation function for the number of casualties per population using 7 indicators of heavy rain. In doing so, the use of disaster risk population, which uses disaster risk information such as estimated flood inundation areas, has improved the accuracy of the estimates. As a result of the estimation using this evaluation function, the estimated results for three cases were roughly consistent with the actual values. In the cases of heavy rain from August 3 and Typhoon No. 15, the estimated results were slightly overestimated compared to the actual values.

View full abstract

Sediments accumulate at the bottom of water distribution tanks, which are a component of water supply systems. In this study, vibration experiments were conducted using a small rigid water tank to confirm whether the sediments would be lifted by the water flow in the tanks when sloshing occurs during an earthquake. Two types of experiments were conducted under the same shaking conditions: one in which only tracers were placed in the tank to visualize the water flow during sloshing, and the other in which sediments were placed in the tank to measure sediment uplift and turbidity. When sloshing was excited by a sinusoidal wave with a first-order period of the oscillation, it was confirmed that sediment began to rise at the bottom of the tank near the side walls when the water level exceeded a certain level, and turbidity also increased. The turbidity increase due to sediment rising reached a maximum with a time lag after the end of the shaking.

View full abstract

In areas affected by the Great East Japan Earthquake, disaster storyteller learning is actively being carried out to convey experiences related to disasters and disaster response. On the other hand, the continuous influence and effects of those who heard it have not been quantitatively grasped. In this paper, we investigated the changes and effects on the listeners in the learning of disaster storytellers, in which they listened to stories of disaster experiences. We conducted a verification using questionnaire survey data for 1247 students of elementary, junior high and high schools who listened to an online disaster storyteller learning program. As a result, 66.3% of the listeners carried out a ripple effect activity in their homes by telling their families what they had heard from the storytellers, and that the change in awareness that emphasized the power to live in a disaster was maintained for more than one month. In addition, it was clarified that behavioral changes after listening to the storyteller are related to immersion in the storyteller's story and acquisition of new disaster knowledge for listeners.

View full abstract

Typhoon Hagibis in 2019 caused extensive damage throughout eastern Japan, especially the Chikuma River breaking its banks in Nagano Prefecture, which had a major socioeconomic impact. In addition, many small and medium-scale rivers were also inundated. However, hydrological data are scarce for many of these medium and small-scale rivers, making it difficult to accurately assess external forces. The purpose of this study is to estimate the actual conditions of flood damage caused by 2019 Typhoon Hagibis in the Namezu River, a right tributary of the Chikuma River, for which hydrological data were extremely scarce. Using limited hydrological data, the spatiotemporal changes in flood damage by comparing the results of numerical simulations by the RRI model coupled with iRIC Nays2DFlood model and damage surveys was clarified. In particular, we estimated that the maximum flow rate of the Namezu River was 478 m³/s by reproducing the actual inundation results in 2019 flood event.

View full abstract

The purpose of this paper is to clarify the issues involved in the resumption of childcare services and the evacuation and handover of childcare facilities at nursery schools, kindergartens and certified children centers damaged by the heavy rain in August 2019, the heavy rain in July 2020, and the heavy rain in August 2021. Regarding the reopening of childcare facilities, it was confirmed that there were problems with facilities and equipment, such as securing a safe and sanitary environment and space for childcare facilities during repair work, providing school lunches, and providing childcare at alternative facilities. Regarding evacuation and handover, issues related to (1) lack of information gathering means, (2) preparing for a prolonged evacuation, (3) ensuring the safety of parents and staff were found, and (4) communication between childcare facilities and parents.

View full abstract

In this study, we conducted a questionnaire survey on volcanoes among Nagano Prefecture's certified mountain guides to understand the current situation and identify issues related to hazard maps, eruption warning levels, responses to eruption phenomena, and obtaining volcano information. As a result, about half of the guides who guide mountaineers to one or more of the four volcanoes (Mt. Asama, Mt. Yakedake, Mt. Norikura, and Mt. Ontake) do not know the current eruption warning level, and about 30 to 50% of them cannot imagine how to respond to eruptive phenomena. It is indicated that the guides do not have sufficient knowledge of volcanic information and eruption phenomena as guides of active volcanoes, which require appropriate evacuation guidance in an emergency.

View full abstract

This study aims to clarify the current status of residents' reading of volcanic hazard maps for Mt. Fuji. To this end, we conducted a questionnaire survey of residents in the area at the foot of Mt. Fuji. As a result, the following three findings were obtained. First, the difficulty in deciphering the maps varies depending on the conditions of volcanic phenomena shown on the maps. Second, residents cannot read the volcano hazard maps properly because they show multiple hazards overlapping. Third, the older generation may be unable to decipher hazard information on hazard maps compared to the younger generation.

View full abstract

In view of the current situation where many victims are caused by frequent heavy rain disasters, various perspectives must be considered to promote resident evacuation. In this study, we used machine learning methods and XAI (eXplainable AI) to clarify the factors and their interactions that affect resident evacuation behavior. The data used for the analysis were the survey results of resident evacuation behavior during four disasters: the July 2018 heavy rain, Typhoon No. 19 in 2019, the July 2020 heavy rain, and the August 2020 heavy rain. We clarified how much the combination of factors related to the behavior and environment during evacuation affect the prediction of evacuation by performing PD analysis which is one kind of the XAI to these data. For each disaster, it was found that the combination of past evacuation experience, damage experience at home during the disaster, and damage such as landslides and flooding at home during the disaster affected evacuation behavior. We clarify that combinations of factors that do not influence evacuation behavior by themselves, such as having a value close to 0 or a negative value for each factor alone had an effect on evacuation behavior by focusing on combinations of factors with strong interaction effects.

View full abstract

We need to establish more effective evacuation measures for combined inundation by simultaneous multiple inflow of sediment and flood in densely populated areas close to mountainous terrain and rivers. In this study, using the numerical model considering combined inundation by simultaneous multiple inflow, we investigate the safeties of evacuation passes to several refuge sites during the combined inundation. The safeties are based on a simple relationship about the required time to move from “evacuation starting points” to “refuge sites” and the difference time of inundation starting between these two points. As a result, since we estimated the distribution of potential for suffering flood damage during the combined inundation, we could identify the more safety evacuation passes and refuge sites. In addition, our investigations could be estimated the magnitude of the influence by each flooding factor on the combined inundation.

View full abstract