Generation of synthetic seasonal hydrographs for a large river basin

This paper describes a methodology for the generation of synthetic seasonal stage hydrographs with a number of flood waves for a large braided river basin based on statistical analysis of the historical stage records. The synthetic seasonal hydrographs in a river is required for different purposes such as assessing the hydraulic performances of various river training structures, morphological predictions, environmental impact analysis. The typical stage hydrograph of such a river has two components: flood waves and seasonal (monsoonal) response. Using historical stage records, flood waves in a seasonal stage record were identified and their characteristics were approximated using Maxwell distribution. The extracted characteristics of flood waves such as time of occurrence and successive flood lifts were analysed with various probability distribution function to find out their best distribution. The frequency analysis of the annual maximum flood lift was carried out. Beside this, seasonal responses were also approximated using Maxwell distribution. A relationship between the seasonal lift and total monsoonal rainfall was established. For a given total seasonal rainfall and return period of annual maximum flood wave lift, synthetic seasonal hydrograph is generated by superimposing both seasonal response and flood waves. The generated hydrographs are evaluated by comparing the cumulative frequency function of river stage and relative frequency of daily stage changes (rise/fall) for three seasons with different flood wave return periods, at two river gauging stations for the river Brahmaputra, India.