Flow, frequency, and uncertainty estimation for an extreme historical flood event in the Highlands of Scotland, UK

This paper explores the uncertainties associated with estimating the flow and frequency of a very large historical flood event, the August 1829 flood, on the River Findhorn in the Scottish Highlands, United Kingdom. For different, but plausible, roughness values, the Manning's equation is used to estimate the flow from a historic flood level at an ungauged site (Findhorn Bridge) (Werritty and McEwen, 2003 ). Catchment area is used to scale the flow estimates downstream to a gauging station site (Shenachie), where they are considered against flood frequency curves obtained using the continuous simulation methodology of Cameron et al. ( 1999 , 2000a ). Operating within the uncertainty framework of generalised likelihood uncertainty estimation (GLUE) (Beven and Binley, 1992 ), this methodology utilises a stochastic rainfall model to drive the rainfall‐runoff model TOPMODEL for a series of continuous ten thousand year simulations with an hourly timestep. A range of possible return periods is obtained for the 1829 flood and the uncertainties associated with the estimation of both the flow and frequency of that event are discussed.