四川东北区域冬小麦水分生产函数模型初探
A Preliminary Study on the Water Production Function Model of Winter Wheat in Northeast Sichuan
DOI: 10.12677/HJAS.2022.1211164, PDF,    科研立项经费支持
作者: 孟祥薇:成都信息工程大学,四川 成都;正蓝旗气象局,内蒙古 锡林郭勒盟;张楚楠, 谢欣芮, 宋延杰, 杨 羽:成都信息工程大学,四川 成都
关键词: 蒸散量人工智能模型水分生产函数Evapotranspiration Artificial Intelligence Model Water Production Function
摘要: 作物水分生产函数对灌溉工程的规划设计和水资源的合理使用都具有重大作用。本文基于四川省东北区域的气象观测数据,明确了适合该区域的最优人工智能蒸散模型,并且进一步基于实际蒸散量和冬小麦实际产量,建立和评估了全生育期水分生产函数模型。结果表明:1) 当温度、湿度、风速、日照气象条件都具备时,逐步线性回归(SLR)、回归树(RT)、支持向量机(SVM)、广义神经模型(GRNN)都能够较好地模拟冬小麦产区的参考作物蒸散量,其中GRNN模型是最优方法;2) 在不同地区GRNN模型计算的蒸散量与冬小麦产量关系的拟合精度存在明显差异。绵阳地区冬小麦产量随着蒸散量的增加呈下降趋势,而在巴中地区冬小麦产量随着蒸散量的增加表现为先增加后减小。3) 对水分生产函数精度的验证表明在遂宁地区线性模型模拟的产量大部分偏小,而抛物线模型模拟的产量偏大;绵阳地区线性模型和抛物线模型整体而言模拟的产量都偏小;巴中地区抛物型模型对产量的模拟精度最好。
Abstract: Crop water production function plays an important role in the planning and design of irrigation projects and the rational use of water resources. Based on the meteorological observation data of northeast Sichuan Province and the actual yield of winter wheat, the optimal artificial intelligence evapotranspiration model suitable for the region is clarified. And further based on the actual evapotranspiration and the actual yield of winter wheat, water production function model in the whole growth period are established and evaluated. The results show that: 1) When temperature, humidity, wind speed, and sunshine meteorological conditions are available, stepwise linear regression (SLR), regression tree (RT), support vector machine (SVM), and generalized neural model (GRNN) can better simulate the reference crop evapotranspiration in winter wheat producing areas, of which GRNN model is the optimal method; 2) There were obvious differences in the simulation precision of the evapotranspiration calculated by the GRNN model and the yield relationship of winter wheat in different regions. The yield of winter wheat in Mianyang region showed a downward trend with the increase of evapotranspiration, while the yield of winter wheat in Bazhong region increased first and then decreased with the increase of evapotranspiration. 3) The verification of the precision of water production function shows that in the Suining region, the winter wheat’ yield simulated by the linear model is mostly small, while the winter wheat’ yield simulated by the parabolic model is larger. On the whole, the winter wheat’ yield simulated by linear and parabolic model are both smaller in the Mianyang region. The parabolic model in Bazhong area has the best simulation accuracy for the yield of winter wheat.
文章引用:孟祥薇, 张楚楠, 谢欣芮, 宋延杰, 杨羽. 四川东北区域冬小麦水分生产函数模型初探[J]. 农业科学, 2022, 12(11): 1188-1196. https://doi.org/10.12677/HJAS.2022.1211164

参考文献

[1] Alexander, L.V., Zhang, X., Peterson, T.C., et al. (2006) Global Observed Changes in Daily Climate Extremes of Tem-perature and Precipitation. Journal of Geophysical Research, 111, D05109.
[2] 王英, 曹明奎, 陶波, 等. 全球气候变化背景下中国降水量空间格局的变化特征[J]. 地理研究, 2006, 25(6): 1031-1040.
[3] 刘后胜. 基于支持向量机的水分生产函数研究[J]. 生物数学学报, 2019, 34(1): 105-110.
[4] 田丰. 冬小麦水分生产函数及水肥耦合关系试验研究[J]. 山西水利, 2008(2): 37-39.
[5] 杨路华, 夏辉, 侯振军, 等. 河北平原冬小麦三种水分生产函数的试验比较[J]. 河北农业大学学报, 2003, 26(Z1): 5-8.
[6] 彭致功, 刘钰, 许迪, 等. 基于RS数据和GIS方法的冬小麦水分生产函数估算[J]. 农业机械学报, 2014, 45(8): 167-171.
[7] 吴训, 许艳奇, 石建初, 等. 基于根系加权土壤水分有效性的冬小麦水分生产函数[J]. 农业工程学报, 2022, 38(8): 124-134.
[8] 曹秀清. 江淮丘陵区冬小麦水分生产函数模型初步分析[J]. 水利水电技术, 2011, 42(8): 72-74.
[9] 汪顺生, 李欢欢, 王康三, 等. 宽垄沟灌下冬小麦水分生产函数试验研究[J]. 排灌机械工程学报, 2017, 35(11): 987-992.
[10] Clempner, G. and Solomon, K. (1987) Accuracy and Geographic Transfer Ability of Crop Water Production Functions. In: James, L.G. and Enghsh, M.J., Eds., Irrigation Systems for the 21st Century, ASCE, New York, 285-292.
[11] 鲍玲玲, 杨永刚, 刘建军, 等. 基于5种人工智能模型计算重庆地区参考作物蒸散量[J]. 水土保持研究, 2021, 28(1): 85-92.
[12] 魏俊, 崔宁博, 陈雨霖, 等. 基于极限学习机模型的中国西北地区参考作物蒸散量预报[J]. 中国农村水利水电, 2018(8): 35-39.
[13] 邢立文, 崔宁博, 董娟. 基于LSTM深度学习模型的华北地区参考作物蒸散量预测研究[J]. 水利水电技术, 2019, 50(4): 64-72.
[14] 徐颖, 张皓杰, 崔宁博, 等. 基于不同ELM的西北旱区参考作物蒸散量模拟模型[J]. 中国农村水利水电, 2019(1): 6-12.
[15] 高文峰, 林文贤. 利用逐步线性回归方法拟合太阳辐射量的经验计算公式[J]. 新能源, 1996, 18(1): 27-30.
[16] 刘洋, 吕一河, 郑海峰, 等. 用回归树模型分析陕北黄土丘陵沟壑区气候因子对NDVI变异的影响[J]. 应用生态学报, 2010, 21(5): 1153-1158.
[17] Cortes, C. and Vapnik, V. (1995) Support-Vector Network. Machine Learning, 20, 273-297. [Google Scholar] [CrossRef
[18] Specht, D.F. (1993) The General Regression Neural Network Rediscov-ered. Neural Network, 6, 1033-1034. [Google Scholar] [CrossRef
[19] Allen, R.G., Pereira, L.S., Raes, D., et al. (1998) Crop Evap-otranspiration: Guidelines for Computing Crop Water Requirements-FAO Irrigation and Drainage Paper 56. FAO, Rome.
[20] Willmott, C.J., Ackleson, S.G., Davis, R.E., et al. (1985) Statistics for the Evaluation of Model Performance. Journal of Geophysical Research—Oceans, 90, 8995-9005. [Google Scholar] [CrossRef
[21] 张恒嘉. 几种大田作物水分−产量模型及其应用[J]. 中国生态农业学报, 2009, 17(5): 997-1001.
[22] 沈荣开, 张瑜芳, 黄冠华. 作物水分生产函数与农田非充分灌溉研究述评[J]. 水科学进展, 1995, 6(3): 248-254.