Epistemic Uncertainty in Hydrological Modeling

Epistemic uncertainty in hydrological modelling refers to the lack of knowledge about the true processes governing the movement and storage of water in catchments. Unlike aleatory uncertainty, which arises from inherent natural variability, epistemic uncertainty is associated with incomplete or imperfect understanding of system structure, parameter values and observational data quality. Sources include uncertain rainfall inputs, simplified representations of evapotranspiration or soil-water interactions, and the choice of conceptual model structure. These knowledge gaps can lead to equifinality, where multiple models or parameter sets perform similarly against calibration data yet diverge in predictive behaviour. Addressing epistemic uncertainty is critical for flood forecasting, water-resource management and climate-change impact assessments, as it shapes the confidence of decision‐makers. Over the past decade, frameworks such as Generalised Likelihood Uncertainty Estimation, information-theoretic calibration and hypothesis-testing with explicit limits of acceptability have been developed to make assumptions transparent, to audit model choices and to reduce disinformation in model conditioning. Practical applications span from national‐scale benchmarking of model ensembles to event-based mass-balance tests, all aimed at quantifying and, where possible, reducing the impact of incomplete knowledge on hydrological predictions.

Research from Nature Portfolio

No recent Nature Portfolio content available.

Epistemic Uncertainty in Hydrological Modeling publication trend

The graph below shows the total number of articles in epistemic uncertainty in hydrological modeling across all publications each year (not limited to Nature Index journals).

Technical terms

Epistemic uncertainty: Uncertainty arising from incomplete knowledge of hydrological processes, model structures or parameters that can, in principle, be reduced through additional information or research.

Aleatory uncertainty: Inherent natural variability or randomness in hydrological systems that cannot be reduced by further study.

Equifinality: The phenomenon whereby multiple model structures or parameter sets yield similarly acceptable performance, complicating the identification of a single “best” model.

Limits of acceptability: Predefined bounds on model outputs used to identify behavioural simulations in the presence of epistemic uncertainties when testing models as hypotheses.

Ensemble modelling: An approach that jointly considers multiple models or parameter realisations to characterise predictive uncertainty and improve robustness of hydrological forecasts.

References

1. A short history of philosophies of hydrological model evaluation and hypothesis testing. Wiley Interdisciplinary Reviews Water (2024).
2. Facets of uncertainty: epistemic uncertainty, non-stationarity, likelihood, hypothesis testing, and communication. Hydrological Sciences Journal (2016).
3. Benchmarking the predictive capability of hydrological models for river flow and flood peak predictions across over 1000 catchments in Great Britain. Hydrology and Earth System Sciences (2019).
4. Disinformative data in large-scale hydrological modelling. Hydrology and Earth System Sciences (2013).
5. Concepts of Information Content and Likelihood in Parameter Calibration for Hydrological Simulation Models. Journal of Hydrologic Engineering (2014).
6. Towards a methodology for testing models as hypotheses in the inexact sciences. Proceedings of the Royal Society A (2019).