Abstract
Abstract. Classical statistical methods for flood frequency estimation assume stationarity in the gauged data. However, recent focus on climate change and, within UK hydrology, severe floods in 2009 and 2015 has raised the profile of statistical analyses that include trends. This paper considers how parameter estimates for the generalised logistic distribution vary through time in the UK. The UK Benchmark Network (UKBN2) is used to allow focus on climate change separate from the effects of land-use change. We focus on the sensitivity of parameter estimates to adding data, through fixed-width moving window and fixed-start extending window approaches, and on whether parameter trends are more prominent in specific geographical regions. Under stationary assumptions, the addition of new data tends to further the convergence of parameters to some final value. However, addition of a single data point can vastly change non-stationary parameter estimates. Little spatial correlation is seen in the magnitude of trends in peak flow data, potentially due to the spatial clustering of catchments in the UKBN2. In many places, the ratio between the 50-year and 100-year flood is decreasing, whereas the ratio between the 2-year and 30-year flood is increasing, presenting as a flattening of the flood frequency curve.
Highlights
Over the last decade, the United Kingdom has seen several extreme flood events, as a result of significant winter storm events in 2009 and 2015–2016 (Barker et al, 2016; Defra, 2016)
The generalised logistic distribution (GLO) is fitted, using maximum likelihood estimators, to the annual maximum (AMAX) series of peak river flow based on 15 min readings for stations in UKBN2
A distribution based on an AMAX record in which just one event is much larger than QMED will have strongly negative κ, while a record with several sized events much larger than QMED will result in a strongly positive κ, which could suggest a possible maximum flow rate at the station
Summary
The United Kingdom has seen several extreme flood events, as a result of significant winter storm events in 2009 and 2015–2016 (Barker et al, 2016; Defra, 2016). The 2015–2016 storms took place over the Lake District in north-west England, and during the event record observations of 24 h and 2 d rainfalls were seen (Marsh et al, 2016; Spencer et al, 2018) This has added weight to various questions about whether this frequency of extreme events is indicative of some change in the nature of the flooding due to changes in rainfall patterns as a result of climate change or due to land-use changes and river channel alterations (IPCC, 2014). Within statistical flood frequency estimation, one common assumption is that the time series of annual maxima or threshold exceedances (peaks over threshold) is stationary: the underlying modelling distribution is constant in time This may not be wholly appropriate in all cases. Taking this non-stationarity into account may be crucial in flood risk management (Reynard, et al, 2017) due to the potential for underestimates of reliability of defence structures or overspending due to the failure to account for a reduction in flood estimates. Spencer et al (2018) use up-to-date National River Flow Archive (NRFA) data to look into whether the record-breaking events are reason for practitioners to adopt non-stationary assumptions, highlighting historical data and local data as ways to supplement the systematic data, being used as evidence for trends and to improve associated uncertainties
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
