Protecting environmental flows to achieve long-term water security

In this work, we propose a new approach to diagnose if a water allocation scheme is compatible with long-term water security at the catchment scale, and suggest steps to achieve such compatibility. We argue that when the remaining flow of a river after upstream withdrawals is not sufficient to safeguarding ecological river functions, the basin is at extreme risk of water scarcity, which indicates that the water management is failing. To test this, we analysed the water scarcity risks and the safeguarded environmental flows (e-flows) in 277 basins across a wide range of hydro-climatic conditions in Chile (17-55 degrees S). For each basin, water scarcity risks were assessed based on water stress indices (WSIs, computed as the ratio of withdrawals to water availability), considering two water-use scenarios: (i) WSImax, where total withdrawals correspond to the maximum consumptive water allowed by the law, i.e., where only the e-flows protected by law remain in the river, and (ii) WSIalloc, where total withdrawals correspond to the actual allocated consumptive water uses within the basins. Further, we evaluated the adequacy of the water management system to protect ecological river functions by contrasting the e-flows protected in Chile with those safeguarded in six other countries.The water allocation system in Chile incorporated the protection of minimum e-flows in 2005 and established that these do not exceed 20% of the mean annual streamflow, except in some exceptional cases. This upper limit is consistently lower than the e-flows safeguarded in other countries, where 20%-80% of the mean annual streamflow are protected. This turns out in WSImax values between 80% and 100% in all basins, well above the threshold associated with over-committed basins under extreme risk of water scarcity (70% typically). When moving from the legally allowed to the actually allocated water use scenario, we found contrasting results: about 70% of the basins show low water scarcity risk (WSIalloc <40%), while an 18% have WSIalloc above 100%, indicating the allocation is going beyond current law limits and even beyond physical limits.Our results reveal that the link between e-flows, water allocation and water security has not been adequately incorporated in the current law. E-flows stipulated by law are insufficient to fulfil environmental requirements, while placing the basins under extreme risk of water scarcity if the total allowed withdrawals were exerted. To move towards a system that can effectively achieve long-term water security, we recommend: (i) To define tolerable water scarcity risks for basins, considering environmental requirements. (ii) To translate those risks into measurable basin indices to measure water security, such as the WSI. (iii) To set maximum water use limits (or minimum e-flows) within the basins that are compatible to the water security goals. If, under current and projected water availability conditions, the existing withdrawals exceed these limits, water managers should be able to adapt total consumption to the required limits.