River basin planning in Alberta has relied on the use of computer modeling since the early 1980s. Typical modeling studies rely on a single time step operational framework, where water allocation decisions are made for individual model time steps, without taking into account seasonal forecasts or the corresponding demand hedging rules that are often implemented by farming communities. This kind of modeling often leads to premature depletion of reservoir storage during dry years, producing model results that represent worse decisions than those that irrigators would make by using the rule-of-thumb. This paper critically reviews the current modeling practice, and provides insight into possible improvements in modeling through the use of multiple time step optimization in combination with optimal demand hedging, which is found as part of the model solution. A case study focuses on potential storage expansions in the Western Irrigation District of Southern Alberta. Improvements with the multiple time step optimization approach also shed new light on important water management decisions made in the past and the value of a revised definition of irrigation failure criteria. Finally, the selected modeling approach reveals significant potential for capital cost savings related to future infrastructure development, and suggests that investing in digital infrastructure – better forecasting and reservoir management tools – may be more productive than investment in additional physical infrastructure.

自1980年代初以来，艾伯塔省的流域规划一直依靠计算机模型的使用。典型的建模研究依赖于单个时间步长操作框架，在该框架中，将为单个模型时间步长做出配水决策，而不考虑季节性预报或农业社区通常执行的相应需求对冲规则。这种模型通常会导致干旱年份水库存储过早耗尽，所产生的模型结果比灌溉者使用经验法则所做出的决策更差。本文严格审查了当前的建模实践，并通过结合使用多个时间步长优化和最佳需求套期，深入了解了建模的可能改进，这是模型解决方案的一部分。案例研究着眼于艾伯塔省南部西部灌溉区的潜在存储扩展。多时间步长优化方法的改进也为过去做出的重要水管理决策和修订后的灌溉失灵标准的价值提供了新的启示。最后，选择的建模方法显示了与未来基础设施开发相关的资本成本节省的巨大潜力，并表明对数字基础设施的投资（更好的预测和储层管理工具）可能比对其他物理基础设施的投资更具生产力。多时间步长优化方法的改进也为过去做出的重要水管理决策和修订的灌溉失灵标准的价值提供了新的启示。最后，选择的建模方法显示了与未来基础设施开发相关的资本成本节省的巨大潜力，并表明对数字基础设施的投资（更好的预测和储层管理工具）可能比对其他物理基础设施的投资更具生产力。多时间步长优化方法的改进也为过去做出的重要水管理决策和修订后的灌溉失灵标准的价值提供了新的启示。最后，选择的建模方法显示了与未来基础设施开发相关的资本成本节省的巨大潜力，并表明对数字基础设施的投资（更好的预测和储层管理工具）可能比对其他物理基础设施的投资更具生产力。