Uncertainty and sensitivity analysis in 3D coupled hydrodynamics and water quality models: a literature review

Código

I-EBHE0158

Autores

Luana Siebra Andrade, Talita Fernanda das Graças Silva, Veber Afonso Figueiredo Costa

Tema

WG 2.6: Water systems analysis for integrated planning & management

Resumo

Lakes and reservoirs show significant spatial and temporal variability in hydrodynamics and water quality processes, especially when situated inside or near urban areas. Technical and funding limitations make it difficult to collect high-resolution data necessary for monitoring these processes through long time spans. As a result, hydrodynamic models have become a useful tool, widely utilized to investigate and predict changes in hydrodynamic and water quality processes in aquatic ecosystems. Hydrodynamic models with different dimensionalities have been developed and coupled with biochemical models for describing different processes and environmental issues in lentic systems. 3D hydrodynamic coupled with water quality models are strongly parametrized and require extensive data to accurately represent physical, chemical and biological processes. The accuracy and reliability of these models depend on the incorporation of uncertainties (i.e. inputs, parametric and structure) and the understanding of the sensitivity of parameters. In 3D hydrodynamic models, the sensitivity analysis can guide calibration process by identifying the parameters that influence the most the model outputs. In turn, uncertainty analysis quantifies the variability in outputs resulting from uncertainties in the inputs, parameters and model structure. Although both analyses are crucial, misinterpretation in terminology and objectives can lead to an improper application of techniques. This work aims to investigate the state of the art in sensitivity and uncertainty analysis of 3D hydrodynamic models of lakes and reservoirs, focusing on the limitations and opportunities of commonly applied methods. Sensitivity and uncertainty analyses are fundamental for improving 3D models and enhancing their predictive capability, thereby providing reliable results for water resources management.