This study shows that cloud seeding should be planned by considering the dam reservoir characteristics as well as the dam basin characteristics. First, the collection efficiency of increased rainfall by cloud seeding is compared for three dam basins (Boryeong Dam, Yongdam Dam, and Namgang Dam basins) located in the western part of the Korean Peninsula. Second, the additional runoff volumes in those three basins from cloud seeding are compared with each other. Finally, the change in water supply capacity is evaluated by considering the dam reservoir operation and planned water supply. In this study, cloud seeding is simulated using the WRF−ARW model, and, additionally, four different rainfall data generated by considering the scenarios of a rainfall increase of 5, 10, 15, and 20% are used for more practical evaluation. The results in this study show that the situation in Boryeong Dam basin is better than in the other two dam basins. More active cloud seeding is necessary in the Yongdam Dam and Namgang Dam basins. However, it has also been found that cloud seeding alone cannot solve the water supply problems in those two dam basins. The above findings also indicate that cloud seeding should be carefully planned. It can vary dam-by-dam. Cloud seeding might be effective every season in one dam, but only in Spring in another dam basin, while in other dams, summer or fall season might be the best option. The target increase of rainfall is also an important issue. Just a mild increase could be better in one dam, but it can be important to secure much more rainfall in other dams. Even though the three dams considered in this study are located in practically the same climatic zone, the conditions required for cloud seeding differ completely.

This study investigates the mechanisms underlying floodplain degradation in dam reservoirs from a socio-hydrological perspective. Specifically, it aims to understand how human behavior, flood dynamics, and policy interventions interact to influence long-term changes in floodplain conditions. To achieve this, the study developed a dynamic modeling framework using differential equations that incorporate the memory of inundation, impact perception, economic incentives, and policy penalties. The model is applied to the Andong Dam in Korea, and simulations are performed under both historical period and future climate conditions (RCP 4.5 and 8.5). The impact of the strengthened penalties on the degradation control is also analyzed. The simulation results reproduce the historical period degradation trend and reveal that low inundation frequency leads to rapid floodplain conversion due to dominant economic incentives. In contrast, more frequent inundation under future climate scenarios helps maintain public memory and reduce degradation rates. However, floodplain degradation was fully suppressed only when the penalty level became comparable to the expected agricultural gain.