Surface irrigation worldwide exhibits low efficiency due to excessively deep percolation and runoff. To optimize surface irrigation practice, two questions must be answered simultaneously: when to irrigate and how to irrigate (what stream size to use and for how long) to meet crop water requirements. Current surface irrigation optimization models are one-dimensional, meaning they can only simulate surface water flow, neglecting subsurface flow. They therefore indicate only how to irrigate, not when to irrigate. Furthermore, the required depth of irrigation is needed as input data for optimization, though this parameter is difficult to establish accurately. In addition, the effects of different irrigation practices on crop transpiration and yield, and variations in these across an irrigated field, are unknown. The present study adopts two one-dimensional simulation models – WinSRFR (for surface flow) and SWAP (for subsurface flow) – to determine when and how best to irrigate in a blocked-end furrow irrigation system for sugarcane cultivation in southwestern Iran. SWAP is used to determine irrigation schedule, required irrigation depth and crop transpiration along furrows, while WinSRFR is used to determine the soil infiltration function and evaluate and optimize irrigation practice. The results of the combined model approach indicate that some 2.7 times more water is applied than required in current practice, imposing high waterlogging stress on sugarcane crops. According to the SWAP simulations, the required irrigation depth is some 75 mm and the number of irrigation applications can be reduced from 26 to 19. The irrigation optimization in WinSRFR indicates that irrigation depth can be reduced from 162 mm to 81 mm, resulting in application efficiency increasing from 47% to 92%. Furthermore, as furrow slope increases from 0.02% to 0.12%, optimized stream size decreases from 3.0 to 1.6 l/s and optimized irrigation cutoff times increase from 3.2 to 6.5 hrs. Overall, optimization of irrigation practice using the combined model approach could result in 63% water conservation and 22% higher sugarcane yield.

由于过深的渗流和径流，全世界的地表灌溉效率低下。为了优化地表灌溉实践，必须同时回答两个问题：何时灌溉以及如何灌溉（使用何种河流大小和多长时间）以满足作物对水的需求。当前的地表灌溉优化模型是一维的，这意味着它们只能模拟地表水流量，而忽略地下流量。因此，它们仅指示如何灌溉，而不是何时灌溉。此外，需要灌溉所需的深度作为优化的输入数据，尽管该参数难以准确建立。此外，不同灌溉方式对作物蒸腾和产量的影响，以及灌溉田间这些变化的影响，都是未知的。本研究采用两个一维模拟模型——WinSRFR（用于地表流）和 SWAP（用于地下流）——来确定何时以及如何在伊朗西南部甘蔗种植的封闭端沟灌溉系统中进行最佳灌溉。SWAP 用于确定灌溉计划、所需灌溉深度和作物沿犁沟蒸腾，而 WinSRFR 用于确定土壤入渗功能并评估和优化灌溉实践。组合模型方法的结果表明，用水量比当前实践所需的水量多 2.7 倍，对甘蔗作物造成了严重的涝渍胁迫。根据 SWAP 模拟，所需的灌溉深度约为 75 毫米，灌溉应用的数量可以从 26 次减少到 19 次。WinSRFR 中的灌溉优化表明，灌溉深度可以从 162 毫米减少到 81 毫米，从而使应用效率从 47% 提高到 92%。此外，随着犁沟坡度从 0.02% 增加到 0.12%，优化的水流大小从 3.0 减少到 1.6 l/s，优化的灌溉截断时间从 3.2 增加到 6.5 小时。总体而言，使用组合模型方法优化灌溉实践可实现 63% 的节水和 22% 的甘蔗产量增加。