Integrated Reservoir Operations using coupled Hydro-Met Multi-Model system for flood forecasting and mitigations for Pune, India

Meteorology and Hydrological extreme events, such as heavy rainfall and associated Flooding is one of the increasing disasters in India for last two decades. Due to heavy reservoir discharge, Impact of rapid Urbanization, unauthorized encroachments across riverbanks extreme flood events are likely to be more common and severe in the future, potentially impacting millions of people.

Pune one the fastest growing megacities in India facing frequent riverine flooding and associated disaster causing huge property losses in millions and causalities. The city is located at the leeward side of Sahyadri mountain range, with 7 reservoirs on the upstream side of the catchment, which control the flows in the rivers impacting the downstream Urban catchment. The reservoirs spillway discharges causes riverine flooding along with contribution from free catchment runoff, which usually occur concurrently. Estimation of reservoir inflows and subsequent spillway discharges is needed for integrated reservoir operations to execute effective flood control measures. To understand these severe flood disasters associated with reservoir operation ensemble multi model simulations were carried for Pune catchment for flood mitigation.

In current study, coupled meteorology model WRF with integrated high resolution (10m) hydrology model HEC-HMS and Hydraulic Model HEC-RAS was developed. High resolution CartoSAT, Digital Elevation Model (DEM) and generated 1m DTM was used to develop both hydraulic and hydrology models. The geometric data for dam structures and gates/spillways have been incorporated in developed models. Gates were operated based on reservoir rule curves for spillway discharge and riverine flood simulations. Spatially distributed high-resolution WRF (1.5 Km) forecasted (72 Hrs.) gridded rainfall data with temporal resolution of 15 mins has been used for forecasting the flood condition in the city. 3D buildings have been incorporated in the terrain to recognize water depth and flooding in the city, which can be visualized through 2-dimensional Rasmapper and 3-dimensional viewer.  The performance of the models has been validated on the basis of statistical error functions (NSE, RSR, PBIAS and R²). Pune flood disaster events for the year 2019 and 2022 were simulated by developed flood forecasting system with reservoir operations. The model output (water level, spread and discharge) were validated using observed flood data from Pune Municipal corporation and dam discharges from water Resource department.

The developed multi-model flood forecasting framework will help the reservoir authorities to perform reservoir operations effectively in future to minimize the downstream flood conditions. Also the disaster management authorities will plan flood mitigation plans with sufficient lead time.