Flood management is an extremely essential consequence in the world; closing conditions of reservoir largely affect the water release decision to control flood downstream. This paper focuses on the mapping of entire reservoir operation scenario using 19 assorted independent variables and further does include with predictive analytics for five dependent variables. Multivariate regression analysis is used in coordination with cross-checking of data sets using various statistical measures. 2295 sample reading of reservoir operation is considered to formulate mathematical models, and its statistical interpretation is also presented. Overall, five mathematical models are presented; all models fitted well with coefficient of correlation in the range of 0.946–0.967. Most influential (Sensitive) variables are sought out from the mathematical models to formulate future strategy of flood management.