基于APSIM模型的高寒地区燕麦灌溉制度优化

Optimizing oat irrigation schedules in an alpine region using APSIM

提出青海海西地区燕麦生产优化灌溉制度,为当地燕麦优质高效生产提供技术参考。大田试验于2017-2018年在青海省乌兰县金泰牧场开展,2017年无灌溉处理,2018年设置4个灌溉处理,分别为仅开花期灌溉(I1),在分蘖期和拔节期灌溉(I2),在分蘖期、拔节期和开花期灌溉(I3),无灌溉(NI),灌溉处理每次的灌溉量均为50 mm。采用2018年I3处理的数据校准APSIM模型,用2018年其他灌溉处理和2017年的数据验证模型,然后模拟不同降水年型下不同灌溉情景的产量和水分利用,并提出青海海西地区燕麦草地优化灌溉制度。试验结果表明,2018年处理I3的产量和耗水量最高,且与其他处理差异显著,处理I2的水分利用效率最高。2017年无灌溉处理的干物质高于2018年的处理I2、I1和NI,并且水分利用效率比2018年的4个处理均高。模型校准过程中干物质产量和土壤水分的模拟值与实测值的均方根误差(RMSE)分别为0.94 t·hm^-2和4.96 mm,4个关键物候期(出苗期、开花期、灌浆期和收获期)的模拟值与实测值的RMSE分别为1、3、4和8 d。验证过程中2018年I1、I2和NI干物质产量和土壤储水量模拟值与实测值的RMSE分别为1.03 t·hm^-2和7.13 mm;2017年4个关键物候期的模拟值与实测值的RMSE为1、1、5和10 d。表明调参之后的"APSIM-燕麦"模拟水分和产量的可靠性较高。利用校准后的模型模拟了10种灌溉情景下燕麦的干物质产量、水分利用效率和灌溉水生产效率,结果表明,在2017(降水低于年平均降水量)和2018年(降水高于年平均降水量),情景8(在分蘖期和拔节期灌溉,每次的灌溉量为50 mm)均为最优灌溉方案。研究结果对于研究区的燕麦人工草地的节水灌溉管理具有一定的参考价值。

This research investigated optimization of irrigation schedules for oat production in an alpine region located in Haixi Prefecture of Qinghai Province,to provide technical guidance for oat production in the region.A field experiment was carried out at Jintai Farm in Wulan County,Qinghai Province,in 2017 and 2018.No irrigation was applied in 2017.Four irrigation treatments were set up in 2018:one irrigation at flowering stage(I 1),two irrigations at tillering stage and jointing stage(I 2),three irrigations at tillering stage,jointing stage and flowering stage(I 3),and no irrigation(N I).Water applied at each irrigation was 50 mm.The APSIM(Agricultural Production Systems Simulation)model was calibrated with the data in I 3 in 2018,and validated with the data from the other irrigation treatments in 2018 and from unirrigated plantings in 2017.The yield and water use under different irrigation scenarios in different years were then simulated,and the optimal irrigation schedule for oat cropping in the study region was modelled.It was found that the highest yield and water consumption(significantly different from other treatments)occurred in I 3 in 2018,while,the water use efficiency was highest in I 2.Dry matter yield in 2017 was higher,and water use efficiency greater than in the treatments I 2,I 1 and N I in 2018.The root mean squared error(RMSE)of the simulated and measured values of dry matter yield and soil moisture were 0.94 t·hm^-2 and 4.96 mm respectively in model calibration,and were 1,3,4 and 8 d in the four key phenological periods(sowing-emergence,sowing-flowering,sowing-filling and sowing-harvesting).The RMSE of the simulated and measured values of dry matter yield and soil moisture were 1.03 t·hm^-2 and 7.13 mm respectively in model validation,and were 1,1,5 and 10 d in the four key phenological periods in 2017.The results showed that APSIM-oat,after calibration,was reliable in simulating soil water content and yield.The calibrated model was used to simulate the dry matter yield,water use efficiency and irrigation water production efficiency of oats under 10 irrigation scenarios.The results showed that a scenario involving irrigation at tillering and jointing,with 50 mm water applied at each irrigation,was the optimal irrigation regime in 2017.(In 2017,annual rainfall was below average,while in 2018 annual rainfall was above average.The results have reference value for water-saving irrigation management of forage oat plantings in the study area.