Bayliss, A.C., Reed, D.W. 2001. The use of historical data in flood frequency estimation. Report to MAFF, 87pp.
For full report: The use of historical data in flood frequency estimation
Flood studies regularly require the estimation of the peak discharge for a specified return period that is substantially longer than the available gauged record. Historical data can be used to augment a flood frequency analysis by providing information on floods that predate the period of systematic gauging.
This report gives guidance on locating and evaluating historical flood information. It reviews methods for incorporating historical data into flood frequency estimates and includes a case study that serves to illustrate the procedures described.
The benefits of extending the relatively short gauged records that exist for most rivers in the UK, using historical flood data, are briefly discussed in the introduction (Chapter 1). The review of historical information can lead to a better understanding of the factors that lead to extreme flood events. The analysis of information on historic floods can also provide better insight into the seasonality of flood occurrence and the effects of land-use change on catchment flood regime. The chapter concludes by introducing the concept of reviewing flood frequency estimates through the incorporation of historical data.
The prospect of looking for information on flood events that occurred decades, or even centuries, before systematic river flow gauging became established, may appear daunting. Chapter 2 gives guidance on those sources most likely to provide useful data.
The format and reliability of historical flood data are likely to be extremely varied. Chapter 3 encourages a systematic and rigorous evaluation of all historical information gathered. The evaluation, according to completeness and authenticity, is recorded in a simple tick-box procedure.
The review of methods for incorporating historical data into flood frequency estimates (Chapter 4) is relatively wide ranging. It considers the use of palæoflood data as well as historical data, and refers to some of the formal and informal methods that have been suggested for incorporating historical data into flood frequency estimates. Key issues surrounding the use of graphical methods are summarised and comments on the options that appear most useful are made.
A case study - the River Avon at Evesham - is used to illustrate the guidance given (Chapter 5). Historical flood data are collated from a number of sources. A rigorous evaluation of the information, in terms of completeness and authenticity, is undertaken using a tick-box procedure. Flood frequency curves are constructed using the site and pooled analysis procedures described in the Flood Estimation Handbook, and are subsequently reviewed using historical flood information.
Chapter 6 provides a brief discussion of the research and the guidance given. The Report concludes by presenting the principal historical datasets, used in the case study, in Appendices A to H.