EGU25-15446, updated on 15 Mar 2025
https://doi.org/10.5194/egusphere-egu25-15446
EGU General Assembly 2025
© Author(s) 2025. This work is distributed under
the Creative Commons Attribution 4.0 License.
Improving Hydrological Process Representation in the Ganges River Basin Using a Data-Assimilation Approach
Supriya Tiwari1, Ehsan Forootan1, Bhaskar R. Nikam2, and Maike Schumacher1
Supriya Tiwari et al.
  • 1Geodesy Group, Department of Sustainability and Planning, Aalborg University, Aalborg, Denmark
  • 2Earth Observation Applications & Disaster Management Support Programme Office, Indian Space Research Organisation, Bengaluru, India

The Ganges River Basin with an area of 1,087,300 km2 is the most populous in the world. In recent years, the increasing severity of hydrological extremes, driven by climate change and human activities, has made water resources increasingly unpredictable, alarming water risks in the region. By understanding how, where, and when these changes affect water resources, we can better prepare and respond to the needs of ecosystems and communities in a rapidly changing climate.

Hydrological models have achieved varying degrees of success in simulating water cycle responses. In particular, they often struggle to accurately capture the non-linearity and complexity of processes in highly heterogeneous basins, such as the Ganges. This challenge is further exacerbated by factors such as changing weather patterns, variability in temperature throughout the basin, and other effects induced by climate change. The limited availability of representative and compatible input data, combined with uncertainties in meteorological forcing data, empirical parameters, initial conditions, and structural errors resulting from simplifications, leads to an incomplete understanding of the underlying physical processes within the basin.

In this study, we propose a Data Assimilation (DA) framework to improve hydrological simulations of the Variable Infiltration Capacity (VIC) land surface model within the Ganges River Basin. The DA is formulated to use the Ensemble Kalman Filter (EnKF) as its merger and satellite-based daily Surface Soil Moisture (SSM) data as observations. Uncertainties in meteorological inputs, such as precipitation and temperature, and model parameters are utilized to generate ensemble spreads, leading to a representative estimation of model uncertainty. Numerical evaluations are performed to examine the influence of this daily SSM DA on sub-monthly, monthly, seasonal, and multi-year variations of the key model outputs, including evapotranspiration, surface runoff, and base-flow. The findings aim to support the development of a satellite-fed hydrological system for the Ganges that further strengthens water management and reduces disaster risks.

Keywords: Variable Infiltration Capacity (VIC), Surface Soil Moisture (SSM), Data Assimilation (DA), Ensemble Kalman Filter (EnKF)

How to cite: Tiwari, S., Forootan, E., Nikam, B. R., and Schumacher, M.: Improving Hydrological Process Representation in the Ganges River Basin Using a Data-Assimilation Approach, EGU General Assembly 2025, Vienna, Austria, 27 Apr–2 May 2025, EGU25-15446, https://doi.org/10.5194/egusphere-egu25-15446, 2025.