Reservoir flood estimation: the way ahead

Abstract

Design floods for reservoirs in the UK and Ireland are currently estimated using a method that in many respects has not changed since the Flood Studies Report of 1975. Although estimates of design rainfall in the UK have been updated, other aspects of the design method and the estimation of probable maximum precipitation (PMP) are dated. Methods for river flood estimation have moved on since the 1970s and there are new and longer-term sources of hydro-meteorological data. Research studies have shown instances of both PMP and probable maximum flood (PMF) estimates using existing methods being exceeded. In this paper we give an overview of aspects of the design flood estimation procedure that are in need of an update and others that tend not to be correctly applied, such as allowance for snowmelt. We identify discrepancies between the different methods used to calculate percentage runoff and time to peak for the 10 000-year flood and the PMF. We discuss the pros and cons of adopting the newer revitalised flood hydrograph rainfall−runoff model for reservoir safety work in the UK. We offer suggestions for development of an up-to-date method for reservoir flood estimation that builds on existing methods, in an international context.