Flood inference simulation using surrogate modelling for the Yellow River multiple reservoir system

The Yellow River, in China, is one of the largest hydro systems in the world. Flooding is a major problem for the river, and therefore over the last 50 years a large number of interventions have been made in its reaches and tributaries, in order to control the flooding events in the lowland area, downstream of the Huayuankou hydrological station. The development of new technologies and approaches to decision support has raised possibilities for creating new ways of managing the river and reducing loss of life, in the case of flooding, for the people living within the embankment area of the river. Given the importance of the river for the development of economic activity in China, it is essential to increase the understanding of the general flooding processes triggered by several reservoir operation scenarios, and then, after applying them to a flooding model of a specific area, to test the findings. The main goal of the research presented here is to investigate and develop the statistical inference between the operation of reservoirs on the Yellow River and a set of variables related to the downstream flooding, such as the total flooding volume and the peak discharge. The research shows that it is possible to use such inference models as decision support tools, by reducing the number of explanatory variables to be included in the simulations carried out to determine the appropriate reservoir operation.