Climate Risk Management (Jan 2021)
Regulation of freshwater use to restore ecosystems resilience
Abstract
Concern about the impacts of water regulation upon the aquatic environment has led to increasingly stringent regulatory constraints on the quantity and timing of freshwater withdrawals. For the time being these regulatory constraints tend to be articulated in terms of limits upon withdrawals, partly because of limited knowledge of the condition and resilience of the aquatic ecosystems. A more sophisticated approach to regulation would be more directly related to indicators of ecological condition. Moreover, it would consider ecosystem response to climatic events not present in the historical record. In this paper we use a combination of empirical evidence of ecosystem condition with simulation to propose and test reductions to regulatory limits on river water withdrawals and downstream minimum flow requirements. The study uses multi-level linear regression to relate the Lotic-invertebrate Index for Flow Evaluation (LIFE) to antecedent flow statistics observed in the Lee catchment, England. The selected flow statistics included extreme low (Q90) and high (Q10) flows in the summer season (April-Sept), and the median flows observed in the winter season (Oct-Mar). The derived model is used to forecast the response of the macroinvertebrate index to future flow scenarios and demand forecasts, incorporating the uncertainties in ecosystem response. Simulation is used to evaluate the sensitivity of the indices to different regulatory limits. Results indicate that macroinvertebrate health will worsen under 21st Century climate conditions, and that the existing regulation policy must be modified to maintain historically observed LIFE scores into the future. The framework demonstrates how regulations could move from precautionary limits on withdrawals to an approach based on observations, forecasting and simulation, allowing planners to refine the trade-offs between river health and reliable water supply in the face of uncertainty.