This study used questionnaire surveys to analyze the relation between community-level flood response and relief in Thailand and Myanmar, including their respective socioeconomic backgrounds.

 Survey results show common points in the two countries: in Thailand and Myanmar, the poor groups’ disaster preparation rate is lower, water inundation is deeper and remains longer, and the poor groups’ houses are more vulnerable than those of non-poor groups. Great differences exist between the two countries in terms of the evacuation rate, the disaster preparedness rate, and the support reception rate.

 Myanmar has an underdeveloped social infrastructure, vulnerable housing, no household savings, and goods distributed in evacuation centers, all of which cause a high evacuation rate of the poor group. Thailand has a higher support reception rate. Moreover, the kinds of support are more diverse than in Myanmar. Thailand’s poor and nonpoor groups’ receptions to support are equivalent, but Myanmar’s non-poor group’s reception rate is low because their main target for emergency relief is the poor group.

 Results show that the Thai social infrastructure is developed adequately, as a middle-income country, to support residents’ daily life during floods. Its early warning and emergency support by government can function at some level. Nevertheless, disparities persist between poor and non-poor in terms of housing, disaster preparedness, and rehabilitation. Consequently, improvement of the economic status of poor people might strengthen disaster resilience in communities effectively. In Myanmar, as a less-developed country with widespread poverty and underdeveloped social infrastructure, knowledge of disaster risk reduction and local government capacity produce a low level of disaster prevention. Not only poverty reduction policy, but also multidimensional approaches are necessary to improve the situation.

 The Tohoku Earthquake and Tsunami in 2011 caused severe damage not only to coastal structures, but also to riverine structures because of long-distance tsunami propagation into coastal rivers that empty into the Pacific Ocean. Although numerous investigations have been conducted of tsunami waves, few reports of the relevant literature have described studies of tsunami propagation into river channels. To elucidate tsunami propagation into a river quantitatively, a numerical simulation based on shallow water equations was solved numerically using the MacCormack scheme, with subsequent comparison to laboratory experiment data. Differences between calculated and experimentally obtained results were evaluated in terms of the root mean square error. Results demonstrate that the present numerical simulation shows good agreement with experimental data in a wave flume. Furthermore, results show that geographical characteristics in the river channel, such as sandbar and estuarine sand spit, strongly affect tsunami propagation processes in a river, with lowered water levels occuring along with the late arrival of the tsunami peak.

 Rain-gauge stations have been operated by many agencies, but temporal and spatial characteristics of the rain-gauge network have not been studied sufficiently. We studied the observation system and some characteristics of the rain-gauge network throughout Japan using available observed data of the Japan Meteorological Agency (JMA), the Water and Disaster Management Bureau under the Ministry of Land, Infrastructure, Transport and Tourism (MLIT), and local governments (L-Gov).
 Analysis results obtained using available observed data show different installed characteristics of rain-gauge networks among L-Gov observation systems. The JMA rain-gauge network decreased to 58 rain-gauge stations during 2008-2014. Most abandoned rain-gauge stations shared common features: rain-gauge stations were installed at high altitude, with high observation and maintenance costs. Regarding historical stored rainfall data of rain-gauge networks, approximately 90 % of JMA and MLIT rain-gauge stations have less than a 40-year historical store of data, with no missing rainfall data (hourly, daily, and annual).