Project: T112700 #0225
 
Published January, 2014
Comparison of Flood Hazard Estimation Methods for Dam Safety - Phase 1 Task 2

This project provides a structured review of the available and evolving methods for estimating flood flows, where they are used and, importantly, the assumptions and limitations they contain. The project outputs are provided in two task reports. The Task 1 report provides a review of the regulatory frameworks in place around the world and their impact on the estimation of flood flows. The Task 2 report focuses on the estimation approaches themselves and is covered here.

The Task 2 report provides a reference for those seeking to understand the various approaches for the estimation of flood flows and their associated strengths and weaknesses. The report includes:

1. An overview of flood flow estimation – introduces the differing analysis paradigms (deterministic and probabilistic), simulation frameworks (deterministic and stochastic) and model development philosophies (statistical inference, deductive and inductive approaches).

2. A typology of hydrological models – classifies hydrological models based on the way they characterize the spatial and temporal variability and the physical processes of interest.

3. A review of approaches to: 

  • Extreme rainfall estimation, including:
    • Assumptions of limiting physical plausibility (e.g. Probable Maximum Precipitation, PMP)
    • Statistical analysis of multiple observations (e.g. design rainfall)
    • Hybrid and coupled statistical and process based (e.g. weather simulation models)
  • Extreme flood flow estimation, including:
    • Direct observation of single events (e.g. historical event reconstruction, paleo floods and geomorphological evidence as well as hydrograph scaling techniques).
    • Statistical analysis of multiple observations (e.g. frequency and extreme value analysis; site, regional and multi-fractal method as well as sampling and probabilistic approaches).
    • Assumptions of limiting physical plausibility (e.g. Probable Maximum Flood, PMF).
    • Hybrid and coupled statistical and process based (e.g. single event and continuous simulation).
    • Data-Based Mechanistic modeling approaches 
  • Whole system simulation - briefly explores potential for towards whole system models that represent rainfall-run-off, channel and structural performance, inundation and consequences.
  • Dealing with non-stationarity and future change – explores approaches for detecting trends and reflecting climate change.
  • Handling uncertainties – identifies the sources of uncertainties and explores methods of uncertainty and sensitivity analysis.
  • Gaining confidence in estimated flows – explores ways of gaining confidence in the absence of objective observations or quantified uncertainty analysis.

4. The applicability of different approaches.

5. Conclusions, recommendations and references

Keywords: Floods, flows, extreme estimates, risk, regulation, uncertainty, dam safety

Project: T112700 #0225
 
Published May, 2014
Comparison of Flood Hazard Estimation Methods for Dam Safety - Phase 1 Task 1

This project provides a structured review of the available and evolving methods for estimating flood flows, where they are used and, importantly, the assumptions and limitations they contain. The project outputs are provided in two task reports.

The Task 1 report provides a review of the regulatory frameworks in place around the world and their impact on the estimation of flood flows, and is covered here. A second, Task 2 report focuses on the estimation approaches themselves and is reported separately.The Task 1 review has highlighted that no explicit relationship exists between formal regulations and the methods for estimating the flood hazard. This finding is consistent with the conclusions of the ICOLD Committee on Dam Safety’s draft report on Dam Safety Regulation. The review also confirms that standards-based approaches continue to dominate formal regulations. In those jurisdictions where the supporting guidance promotes the use of risk approaches to help decide whether or not the legislation or regulations have been met, this does have a number of implications for flood hazard estimation. The most relevant of these include the need to:

(i) Account for aleatory uncertainty (arising from natural variability in climate and hydrologic inputs) and estimate a full distribution of flood flows (rather than a single design flow event characteristic of a traditional standards-based approach).

(ii) Explicitly recognize the epistemic uncertainty (arising from errors in data, models and model structure).

(iii) Recognize the stochastic nature of the system response and accept that some uncertainties are irreducible and cannot be resolved through ever more detailed data and modeling.

Keywords:Regulation, Dam Safety, Flood Hazard, Risk, Dam Classification

NOTE: TASK 1 Report was not published or released outside the sponsor group, as per the following disclaimer:

DISCLAIMER

The project sponsors are in agreement with the statements given in the Abstract, Executive Summary, Conclusions and Recommendations, however some of the statements and opinions expressed in the main body of the Report are not in alignment with the above sections of the Report and are not supported by the sponsors.

CEATI and any other person involved in the production and distribution of this report do not warrant the accuracy, reliability, currency or completeness of those views or statements and do not accept any legal liability whatsoever arising from any reliance on the views, statements and subject matter of the report.