Hydrological time series are often asymmetric in time, insomuch as ris
es are more rapid than recessions, as well as having highly skewed mar
ginal distributions. A two-stage transformation is proposed for deseas
onalized series. Rises are stretched, and recessions are squashed unti
l the series is symmetric over time. An autoregressive moving average
(ARMA) model is then fitted to the natural logarithms of this new seri
es. The residuals from the ARMA model are represented by a double mixt
ure of Weibull and exponential distributions. The method is demonstrat
ed with 24 years of daily flows from the River Cherwell in the south o
f England, and a 40-year record from the upper reaches of the Thames.
Seasonal estimates of flood risk are given, and these can be condition
ed on catchment wetness at the time of prediction.