Increased rainfall-runoff drives flood hazard intensification of Central Himalayan Rivers
Date:
Ivo Pink, Sim Reaney, Isabella Bovolo, and Richard Hardy Durham University, Department of Geography, Durham, United Kingdom (ivopink2@gmail.com)
The development of flood adaptation and mitigation strategies to climatic changes requires frameworks to predict potential future design floods (e.g. the 1% Annual Exceedance Probability (AEP)) and their associated uncertainty. We present a modelling framework that predicts potential future flood hazards from probabilistic climate scenarios. The framework assesses the drivers of change and the predictive uncertainty and is implemented for the Central Himalayan Karnali River.
The modelling framework applies a continuous hydrological model within a Generalised Likelihood Uncertainty Estimation (GLUE) with climate projections of the latest generation of climate models (12 CMIP6 models). We then conduct a Flood Frequency Analysis (FFA) to estimate the probabilities of extreme flows and use a bootstrapping approach to estimate the uncertainty related to the internal variability.
We project an intensification of flood hazards with time and emissions. The 1% AEP flood magnitude is projected to increase by 23% (medium-emission scenario SSP245) and 26% (high-emission scenario SSP585) in the near-future (2020 – 2059), and by 40% (SSP245) and 79% (SSP585) in the far-future (2060 – 2099) compared to the baseline period (1975 – 2014). Consequently, the flood magnitude of the baseline 1% AEP event is projected to occur once every 11 years (SSP245) and 3 years (SSP585) in the far-future. This intensification is to >90% driven by rainfall-runoff increases and is, thus, attributed to changes in the precipitation characteristics. The baseflow and glacier melt contributions remain similar while snowmelt contributions decrease due to an earlier onset of the melting season.
We analyse the standard deviation (SD) to assess the uncertainty contribution of the modelling components. The hydrological model is a main source of uncertainty, but its contribution is independent of time, emissions and event frequency (SD: 21 – 24%). The uncertainty related to the climate projections increases with time and emissions from 14 – 18% in the baseline period to 22 – 28% (SSP245) and 29 – 32% (SSP585) in the far-future. The uncertainty introduced by the FFA is generally independent of time and emissions and increases with the event frequency from 8-9% for the 10% AEP event to 11 – 16% for the 1% AEP event. However, we detect that the uncertainty increases with the increasing flow difference between the rarest and more frequent events and is, thus, sensitive towards the projected precipitation extremes.
We conclude that flood hazards intensify with GHG emissions in the Central Himalayan River system because of changes in the monsoon precipitation characteristics. The projections of potential future hazards to guide flood adaptation and mitigation strategies need to consider the uncertainty in the climate projections, the internal variability, and the hydrological simulations.
How to cite: Pink, I., Reaney, S., Bovolo, I., and Hardy, R.: Increased rainfall-runoff drives flood hazard intensification of Central Himalayan Rivers, EGU General Assembly 2025, Vienna, Austria, 27 Apr–2 May 2025, EGU25-10581, https://doi.org/10.5194/egusphere-egu25-10581, 2025.