Dynamically induced streamflow variability in UK river catchments

Understanding historical and future changes to seasonal and extreme flow regimes is crucial for both water resources planning and flood risk management. Historically, inter-annual to multi-decadal variability in seasonal flow has been influenced by variability in atmospheric circulation over the UK, and by long-term changes in the mean state of atmospheric circulation. Having trust in future hydrological projections therefore requires (1) a thorough understanding of the representation of this atmospheric circulation induced variability in climate models, and (2) confidence that climate models are capable of reproducing periods of atmospheric circulation patterns associated with wet or dry conditions.

In this work, we apply a novel dynamical adjustment methodology based on synoptic-scale weather patterns to a century long reconstruction of seasonal standardised streamflow index (SSI) for catchments in the UK. This methodology isolates the influence of atmospheric circulation variability on SSI, exhibiting clear seasonality and spatial patterns. In some catchments, this ‘dynamical’ SSI component explains a high proportion of variability in total seasonal SSI.

Using the same synoptic-scale weather patterns, we find that UKCP18 climate models underestimate seasonal variability in the dynamical component of SSI. We demonstrate that the differences in distribution between observations and model simulations must be due to differences in weather pattern frequency and/or clustering, rather than rainfall biases. Our findings raise questions about the suitability of climate models in projecting streamflow trends and understanding future seasonal extremes.