Burundi is susceptible to future water-related disasters, but examining the influence of climate change on regional hydroclimatic features is challenging due to a lack of local data and adaptation planning. This study investigated the influence of climate change on hydroclimate-focused changes in the climatology of heavy precipitation (and streamflow) means and extremes based on the multi-model ensemble mean of earth system models in the sixth phase of the Coupled Model Intercomparison Project (CMIP). For runoff analysis, hydrologic responses to future climate conditions were simulated using the Soil and Water Assessment Tool (SWAT) model over the Ruvubu River basin, Burundi. Temperature increases by 5.6 °C, with strong robustness, under future climate conditions. The mean annual precipitation (and runoff) undergoes large seasonal variations, with weak robustness. Precipitation (and streamflow) changes between the wet and dry seasons differ in signal and magnitude. However, alterations in both the amount and frequency of precipitation reveal the intensification of the water cycle due to anthropogenic climate change. Thus, the highest variability in the maximum daily streamflow is shown in months of long wet seasons, especially in the far future (2085). Without considering the regional climate characteristics and shared socioeconomic pathway (SSP) scenarios, this behavior is expected to be enhanced in 2085 (compared with 2045) and increase the severity of extreme precipitation and flood risk. Climate change will cause alterations in the magnitude and seasonal distributions of extreme precipitation (and streamflow). These findings could be important for flood planning and mitigation measures to cope with climate change in Burundi.

The transboundary river basin is a great challenge for water management and disaster reduction due to its specific characteristics. In this study, upstream impacts from natural and artificial sources on the downstream discharge on the Imjin river basin, the well-known transboundary region in the Korean peninsula, were evaluated using a hydrological model integrating a dam operation module at an hourly timescale. The module uses a concept of the AutoROM method as the operational rule to update the dam storage and decide water release. Dam storages were translated into water levels using a water level–storage curve. To quantify the impact of hydraulic structures on the Northern Imjin river basin, change in discharge was analyzed in four flood events (2009, 2010, 2011, and 2012). Dam failure scenarios were developed under conditions of the 2010 flood event, in which the releases of 100%, 80%, 50%, and 20% of water storage of Hwanggang dam were simulated. The results indicate that the amount of water released from upstream dams is the main cause of floods in the downstream region. To reduce the risk of floods in the downstream river basin, an optimal dam operation module and information on upstream dams play an important role and contribute to the effective use of water resources.