Only short periods of gauged record, generally less than 30 years, are available to assess the flood discharge that might be expected with a return period of 100 years. There is a requirement therefore for extrapolation of a flood frequency curve well beyond the limits of the record duration. However there is no a priori basis for judging which is the best flood frequency distribution or fitting method to use. In most instances but especially when there are outliers in the observed data set, extrapolation using different distributions can lead to widely different estimates. There is no means of judging from the data which of these estimated discharges (and the associated levels on the flood plain) is most likely to be correct and the usual solution is take a mean discharge from a range of different distributions and fitting methods.
Some confirmation to the slope and shape of the flood frequency distribution is provided by making comparisons of flood frequency at several stations within and on adjacent catchments, thus effectively adding to the number of station years of record. However, there is a degree of dependence between the floods experienced on adjacent catchments and, even when considered regionally, the flood record sample may not give a good indication of the distribution of the flood population and in particular for long return periods.
Examination of historical information provides an alternative means of assessing high return period flood discharges and levels. It has the disadvantage that the information is rarely as precise as for the gauged record especially in the assessment of discharge from a reported level. Nevertheless, there are instances where the control at a point on a river has remained with little change over a period of centuries e.g. at a bridge crossing or a mill weir. Some natural sections have also changed little. At these points it may be possible to convert levels to discharge and to combine with the gauged record, as for example was done for the River Wear (Archer, 1987) and for the River Tyne (Archer, 1993).
The main outcome of this review is a description of the extent of flooding during the major floods that have occurred over the period from the mid seventeenth century. This is presented as a flood chronicle with a summary description of each event. Continue reading