Developing generic reservoir operating rules for inclusion in the national-scale hydrological modelling of Great Britain

To meet growing water demand and to satisfy an increasing population, reservoirs are continually being integrated into river systems across the world. The presence of a reservoir can dictate the downstream flow regime, such that in many locations, understanding reservoir operations can be crucial to understanding the hydrological functioning of an impacted catchment. Consequently, over the last two decades, correctly representing reservoirs, and their operations, in hydrological modelling frameworks has become a key area of research for simulating water availability. Although substantial progress has been made in modelling reservoir operations (which control how water volumes are distributed across space and time), there is still no consensus on the best way to define, calibrate and evaluate operating rules within hydrological models. In most locations, data describing reservoir operating rules are not available, and timeseries of reservoir inflow, outflow and storage are often unpublished. Consequently, modelers must simplify and generalize sets of release rules from very little information, particularly where they are to be applied across large scales (e.g. across hundreds of reservoirs). Generic reservoir operating rules have typically been tested and developed using the Global Reservoir and Dam (GranD) database and thus are biased towards large irrigation reservoirs (which make up the majority of the dataset). Whilst operating rules have also been tested across many hydropower and multipurpose reservoirs, a gap remains for the definition of generic reservoir operating rules designed for smaller water supply reservoirs that can be applied nationally in countries such as Great Britain (GB).

In this study, we integrate a new generic reservoir simulation component into a national-scale hydrological model of Great Britain and compare simulation results from two modelling scenarios (with and without the new reservoir component). The first scenario, where reservoirs are omitted, is used as a benchmark representative of current modelling practices in GB (where none of the national-scale hydrological models include reservoirs), whilst the second uses a set of generic operating rules focused on simulating small, water resource reservoirs. In both scenarios, we use Multiscale Parameter Regionalisation (MPR) for model calibration. To assess the suitability of our operating rules for simulating future conditions and evaluating water availability during hydrological extremes, we test the consistency of model performance across the onset, duration and recovery from droughts. This study will demonstrate the importance of including reservoir representation in hydrological models of Great Britain, and will introduce a set of operating rules suitable for smaller reservoirs with a focus on water supply.