1. What precipitation dataset used?
The observed gridded precipitation dataset used is from Pai et al., 2014. It was developed using station-based rainfall observations from more than 6900 gauge stations in India. The dataset covers the 1901-2020 period and has been widely used for hydrological studies. It captures the key features of the summer monsoon variability and orographic rainfall over the western Ghats and foothills of the Himalayas.
read more
2. How was the H08-CaMa-Flood combined model calibrated and evaluated for India's eighteen major river basins?
The H08-CaMa-Flood combined model was calibrated for India's eighteen major river basins by adjusting four parameters for each river basin, which include singlelayer soil depth, gamma, bulk transfer coefficient, and tau. The model performance was evaluated using the coefficient of determination (R 2 ) and Nash-Sutcliffe Efficiency (NSE) for daily streamflow and reservoir live storage. Additionally, the simulated and satellite-based observed flood occurrences were compared. The model was forced with observed meteorological forcing from India Meteorological Department (IMD) at 0.25deg spatial resolution to conduct simulations from 1901 to 2020. The generated flood depths at 6 arc-minutes (0.1deg) were further downscaled to 1 arc-minute (~200m) resolution using the downscaling module available within the CaMa-Flood. The C-ratio was used to estimate the potential dam effect along a river, and the sub-basins prone to flooding due to dam operations were identified. The exposed rail and road infrastructure affected by floods were also estimated.
read more
3. What framework is used for flood risk assessment?
The common framework adopted by the United Nations in the Global Assessment Reports of the United Nations Office for Disaster Risk Reduction, 2011, 2013, is used for flood risk assessment. This framework considers hazard, exposure, and vulnerability. Similar frameworks have been used in previous studies, such as those by C.M. R. Mateo et al. (2014), Tanoue (2020), and Winsemius et al. (2013). The flood risk assessment helps identify hotspots and prioritize climate adaptation, as stated by de Moel et al. (2015). Vulnerability, one of the three components, is a degree of damage to a particular object at flood risk, ranging from 0 to 1. The vulnerability index for each district is obtained from capacity, and the exposure is defined as assets and population in a flood-exposed area resulting in flood damage. The Global Human Settlement Layers (GHSL) population dataset is used for exposure estimation, and the hazard is estimated as the exceedance probability of a flooded area exceeding half of the historical maximum flooded area in the last 50 years.
read more
4. How did the H08 and CaMa-Flood models perform in simulating flood extent?
The H08 and CaMa-Flood models exhibited satisfactory performance in simulating flood extent against satellite-based observations. However, the model overestimates the flood extent in the Ganga basin, potentially due to cloud contamination and dense vegetation cover affecting satellite-based flood estimates. Conversely, the model underestimates the flood occurrence in the upstream region of the Brahmaputra River, possibly due to limitations in model parameterization and limited observed flow in transboundary river basins. Despite these discrepancies, the simulated flood extent shows good agreement with reported flood data from EM-DAT and DFO databases, as well as with reported floods in cities within the Brahmaputra and Ganga River basins. These findings suggest that the models can be used for sub-basin level risk assessment, despite the limitations in streamflow and flood extent observations.
read more