The Iterative Empirical Rational Method for Flood Forecasting

Dams can store, hold, and recharge rainwater from the catchment to the river and help moderate floods. The dam operating schedule regulates when and how a dam should be filled and emptied. Most recent floods in India have occurred due to the sudden release of dam water. Flash floods have high intensity. A model for forecasting floods in advance is needed. The catchment area is the total area of land that collects water after rainfall and contributes runoff into the dam. The yield coefficient is the proportion of rainfall that runs off. The time required for water to reach from the catchment area to the dam, which is a time of concentration, is different for all catchment areas. The total inflow to the dam at a particular time is complex and challenging to predict accurately. In this paper, based on the proposed iteration method, the designed dam volume forecasting model uses data-driven yield coefficients derived from rainfall data and river gauge data of one dam in the Maharashtra state of India. Rainfall volumes, gate-discharged volume, overall yield coefficient, and time of concentration are calculated to forecast inflow and dam volumes. The model is applicable only when the reservoir is about $60 \%-80 \%$ full, when the soil is almost saturated, and the parameters like evaporation, seepage, and vegetation will not make any difference. The model gives less than 1 % error in nearly 90 % of the readings. It can be utilized for effective gate operations during flood control.