Search

The depth–duration–frequency (DDF) model developed as part of the Flood Estimation Handbook (FEH) enables estimates of design rainfall to be made at any location in the UK. This study presents a shortcut method for estimating seasonal design rainfall for a summer and winter season, respectively, using a simple seasonal correction factor in combination with the FEH DDF estimate of design rainfall. The seasonal correction factors were derived for the summer and winter season for a range of durations. The observed correction factors were linked to standard average annual rainfall available from the FEH CD-ROM, thereby enabling the use of this method at any location in the UK. The results obtained in this study are shown to correspond well to the summary statistics on seasonal rainfall presented in the Flood Studies Report. [ABSTRACT FROM AUTHOR]

Seasonal predictions of river flow provide managers with the opportunity to plan the use of water resources in advance. Summer (June–August) river flows for the period 1961–2002 in 20 catchments in Great Britain were grouped into two clusters, largely representing the southeast and the northwest of Britain. The cluster average summer river flows were predicted by linear regression using the previous winter’s (December–February) river flow, North Atlantic sea surface temperatures (SSTs), North American land air temperature and global airflow indices as potential predictors. Separate regression equations were derived for the northwest and southeast regions. Both equations contain eight predictors. Predictors for the southeast are shifted somewhat southward compared to the predictors for the northwest. The most influential predictors are the temperature difference between the North Atlantic Current and the Norwegian Sea for the northwest and the SST in the central part of the subtropical gyre in the North Atlantic (Central Gyre) for the southeast. The regression equations fitted to all data explain 55% and 61% of the variance for the northwest and southeast regions, respectively. The corresponding cross-validation correlations between the predicted and observed series are 0.54 and 0.62. The models are fairly skilful at avoiding the false positive or negative prediction of extremes, but less skilled at predicting medium sized river flows. [ABSTRACT FROM AUTHOR]