[1]张薇,韦群,吴天傲,等.基于GBDT算法的参考作物蒸散量模型在江苏省的预测[J].江苏农业学报,2020,(05):1169-1180.[doi:doi:10.3969/j.issn.1000-4440.2020.05.014]
　ZHANG Wei,WEI Qun,WU Tian-ao,et al.Prediction models of reference crop evapotranspiration based on gradient boosting decision tree(GBDT) algorithm in Jiangsu province[J].,2020,(05):1169-1180.[doi:doi:10.3969/j.issn.1000-4440.2020.05.014]
点击复制

基于GBDT算法的参考作物蒸散量模型在江苏省的预测()
分享到：

江苏农业学报[ISSN:1006-6977/CN:61-1281/TN]

卷:
期数:
2020年05期
页码:
1169-1180
栏目:
耕作栽培·资源环境
出版日期:
2020-10-31

文章信息/Info

Title:
Prediction models of reference crop evapotranspiration based on gradient boosting decision tree(GBDT) algorithm in Jiangsu province
作者:
张薇1; 韦群2; 吴天傲1; 林洁3; 邵光成1; 丁鸣鸣4
（1.河海大学农业科学与工程学院，江苏南京210098；2.南京市浦口区水务局，江苏南京211800；3.南京市江宁区水务局，江苏南京211100；4.南京市水务局，江苏南京210098）
Author(s):
ZHANG Wei1; WEI Qun2; WU Tian-ao1; LIN Jie3; SHAO Guang-cheng1; DING Ming-ming4
（1.College of Agricultural Science and Engineering，Hohai University， Nanjing 210098, China；2.Water Resources Bureau of Pukou Area in Nanjing City， Nanjing 211800, China；3.Water Resources Bureau of Jiangning Area in Nanjing City， Nanjing 211100, China；4.Nanjing Water Resources Bureau，Nanjing 210098, China）
关键词:
参考作物蒸散量; 梯度提升决策树（GBDT）算法; 随机森林（RF）算法; 可移植性分析; 敏感性分析
Keywords:
reference crop evapotranspiration; gradient boosting decision tree(GBDT) algorithm; random forest(RF) algorithm; portability analysis; sensitivity analysis
分类号:
S16
DOI:
doi:10.3969/j.issn.1000-4440.2020.05.014
文献标志码:
A
摘要:
选取江苏省6个气象站点1997-2016年的逐日气象资料，建立了3种基于树型算法的参考作物蒸散量（ET0）预测模型，包括梯度提升决策树（Gradient boosting decision tree，GBDT）、随机森林（Random forest，RF）和回归树（Regression tree）模型，以FAO-56 Penman-Monteith公式计算所得的ET0值作为标准值，对GBDT、RF、Regresssion tree模型和3种经验模型（EI-Sebail、Irmak、Hargreaves-Samani模型）的预测结果进行比较分析。结果表明：在相同气象因子输入组合下，GBDT、RF模型能取得较高的模拟精度，且明显高于Regression tree模型和经验模型，其中，气象参数组合为最高气温、最低气温和辐射的GBDT模型具有最高的模拟精度[全局评价指标（GPI）排名第1]；通过敏感性分析发现，辐射是对江苏省逐日ET0影响最为显著的气象因子，其直接通径系数为0.512，对决定系数（R2）的贡献度为0.740，显著高于其他气象因子；通过可移植性分析发现，气象因子组合为最高气温、最低气温和辐射的GBDT、RF模型在江苏省内6个站点相互交叉验证下仍具有较高的预测精度。因此，可以将GBDT、RF模型应用于江苏省气象资料缺乏时的ET0预测，为农业灌溉提供可靠依据。
Abstract:
Daily meteorological data from 1997 to 2016 in six meteorological stations of Jiangsu province were selected to establish three prediction models of reference crop evapotranspiration (ET0) based on tree algorithm, including gradient boosting decision tree (GBDT) model, random forest (RF) model and regression tree model. Taking ET0 value calculated by formula of FAO-56 Penman-Monteith as standard value, the prediction results of GBDT model, RF model, regression tree model and three empirical models (EI-Sebail model, Irmak model and Hargreaves-Samani model) were compared. The results showed that GBDT model and RF model could get high simulation accuracies under the combination of the same meteorological factor inputs, and the accuracies of GBDT model and RF model were significantly higher than regression tree model and empirical model. Among them, GBDT model with the meteorological parameters of maximum temperature, minimum temperature and radiation had the highest simulation accuracy (global performance indicator ranked No.1). Through sensitivity analysis, it was found that radiation was the most significant meteorological factor affecting the daily ET0 of Jiangsu province, its direct path coefficient was 0.512 and its contribution to the determination coefficient (R2) was 0.740, which were significantly higher than other meteorological factors. Through portability analysis, it was found that GBDT model and RF model with the meteorological parameters of maximum temperature, minimum temperature and radiation still had high prediction accuracies under cross-validation of six stations in Jiangsu province. Therefore, GBDT model and RF model can be applied for ET0 prediction in Jiangsu province when the meteorological data are absent and provide reliable evidence for agricultural irrigation.

参考文献/References:

[1]徐俊增,彭世彰,丁加丽,等. 基于蒸渗仪实测数据的日参考作物蒸发腾发量计算方法评价[J]. 水利学报, 2010, 41(12):1497-1505.
[2]冯禹,崔宁博,龚道枝,等. 基于极限学习机的参考作物蒸散量预测模型[J].农业工程学报,2015,31(S1):153-160.
[3]李晨,崔宁博,冯禹,等. 四川省不同区域参考作物蒸散量计算方法的适用性评价[J].农业工程学报,2016,32(4):127-134，316.
[4]ALLEN R G, PEREIRA L S, RAES D, et al. Crop evapotranspiration：Guidelines for computing crop water requirements[M]. Rome:FAO Irrigation and Drainage Paper 56,1998:1-15.
[5]FAN J L, YUE W J, WU L F, et al. Evaluation of SVM, ELM and four tree-based ensemble models for predicting daily reference evapotranspiration using limited meteorological data in different climates of China[J]. Agricultural and Forest Meteorology, 2018,263: 225-241.
[6]WU L F, FAN J L. Comparison of neuron-based, kernel-based, tree-based and curve-based machine learning models for predicting daily reference evapotranspiration [J]. PLoS One,2019,14(5): e0217520.
[7]KISI O. Applicability of Mamdani and Sugeno fuzzy genetic approaches for modeling reference evapotranspiration[J]. Journal of Hydrology, 2013, 504: 160-170.
[8]SHIH S F, SNYDER G H. Leaf area index and evapotranspiration of taro[J]. Agronomy Journal, 1985, 77(4):554-556.
[9]彭世彰,徐俊增. 参考作物蒸发蒸腾量计算方法的应用比较[J]. 灌溉排水学报, 2004,23(6):5-9.
[10]IRMAK S, IRMAK A, ALLEN R G, et al. Solar and net radiation-based equations to estimate reference evapotranspiration in humid climates[J]. Journal of Irrigation and Drainage Engineering, 2003, 129(5): 336-347.
[11]PRIESTLEY C H B, TAYLO R J. On the assessment of surface heat flux and evaporation using large-scale parameters[J]. Mon Weather Rev, 1972,100:81-92.
[12]HARGREAVES G H, SAMANI Z A. Reference crop evapotranspiration from temperature[J]. Appl Eng Agric, 1985,1:96-99.
[13]KISI O. Pan evaporation modeling using least square support vector machine, multivariate adaptive regression splines and M5 model tree[J]. J Hydrol,2015,528: 312-320.
[14]WANG L, KISI O, Zounemat-Kermani M, et al. Pan evaporation modeling using six different heuristic computing methods in different climates of China[J]. J Hydrol, 2017,544: 407-427.
[15]JOVIC S, NEDELJKOVIC B, GOLUBOVIC Z, et al. Evolutionary algorithm for reference evapotranspiration analysis[J]. Comput Electron Agric,2018,150: 1-4.
[16]LANDERAS G, ORTIZ-BARREDO A, LO′PEZ J J. Comparison of artificial neural network models and empirical and semi-empirical equations for daily reference evapotranspiration estimation in the Basque Country (Northern Spain) [J]. Agric Water Manag,2008,95: 553-565.
[17]FENG Y, CUI N B, ZHAO L, et al. Comparison of ELM, GANN, WNN and empirical models for estimating reference evapotranspiration in humid region of Southwest China[J]. Journal of Hydrology, 2016, 536: 376-383.
[18]MOUSAVI R, SABZIPARVAR A A, MAROFI S, et al. Calibration of the Angstrm-Prescott solar radiation model for accurate estimation of reference evapotranspiration in the absence of observed solar radiation[J]. Theoretical and Applied Climatology, 2015, 119(1/2):43-54.
[19]LADLANI I, HOUICHI L, DJEMILI L, et al. Estimation of daily reference evapotranspiration (ET0) in the North of Algeria using adaptive neuro-fuzzy inference system (ANFIS) and multiple linear regression (MLR) models: A comparative study[J]. Arabian Journal for Science and Engineering, 2014, 39(8):5959-5969.
[20]KUMAR M, RAGHUWANSHI N S, SINGH R, et al. Estimating evapotranspiration using artificial neural networks[J]. Journal of Irrigation and Drainage Engineering, 2002, 128(4): 224-233.
[21]张皓杰,崔宁博,徐颖,等. 基于ELM的西北旱区参考作物蒸散量预报模型[J]. 排灌机械工程学报, 2018, 36 (8):140-145.
[22]TRAORE S, WANG Y M, KERH T. Artificial neural network for modeling reference evapotranspiration complex process in Sudano-Sahelian zone[J]. Agricultural Water Management, 2010, 97(5): 707-714.
[23]TABARI H, KISI O, EZANI A, et al. SVM, ANFIS, regression and climate based models for reference evapotranspiration modeling using limited climatic data in a semi-arid highland environment[J]. Journal of Hydrology, 2012, 777: 78-89.
[24]ABDULLAH S S, MALEK M A, ABDULLAH N S, et al. Extreme learning machines: A new approach for prediction of reference evapotranspiration[J]. Journal of Hydrology, 2015, 527:184-195.
[25]HASSAN M A, KHALIL A, KASEB S, et al. Potential of four different machine-learning algorithms in modeling daily global solar radiation[J]. Renewable Energy, 2017, 111:52-62.
[26]FAN J, WANG X, WU L, et al. Comparison of support vector machine and extreme gradient boosting for predicting daily global solar radiation using temperature and precipitation in humid subtropical climates: A case study in China[J]. Energy Conversion & Management, 2018, 164:102-111.
[27]于玲,吴铁军. 集成学习:Boosting算法综述[J]. 模式识别与人工智能, 2004, 17(1):52-59.
[28]HASTIE T, TIBSHIRANI R, FRIEDMAN J. Ensemble Learning[M]//HASTIE T, TIBSHIRANI R, FRIEDMAN J. The Elements of Statistical Learning. Springer Series in Statistics. New York, NY: Springer, 2009: 605-624.
[29]BAUER E, KOHAVI R. An Empirical comparison of voting classification algorithms: Bagging, Boosting, and Variants[J]. Machine Learning, 1999, 36(1/2):105-139.
[30]DIETTERICH T G. An experimental comparison of three methods for constructing ensembles of decision trees: Bagging, Boosting, and Randomization[J]. Machine Learning, 2000, 40(2):139-157.
[31]MANIKUMARI N, MURUGAPPAN A, VINODHINI G. Time series forecasting of daily reference evapotranspiration by neural network ensemble learning for irrigation system[J]. IOP Conference Series: Earth and Environmental Science, 2017, 80:012069.
[32]FENG Y, CUI N, GONG D, et al. Evaluation of random forests and generalized regression neural networks for daily reference evapotranspiration modelling[J]. Agric Water Manage,2017, 193:163-173.
[33]王升,付智勇,陈洪松,等.基于随机森林算法的参考作物蒸发蒸腾量模拟计算[J].农业机械学报,2017,48(3):302-309.
[34]韩启迪,张小桐,申维.基于梯度提升决策树(GBDT)算法的岩性识别技术[J].矿物岩石地球化学通报,2018,37(6):1173-1180.
[35]郑凯文,杨超.基于迭代决策树(GBDT)短期负荷预测研究[J].贵州电力技术,2017,20(2):82-84,90.
[36]蔡文学,罗永豪,张冠湘,等.基于GBDT与Logistic回归融合的个人信贷风险评估模型及实证分析[J].管理现代化,2017,37(2):1-4.
[37]GORDON R B A D. Classification and regression trees[J]. Biometrics, 1984, 40(3):874.
[38]EVERITT B S. Classification and regression trees[M]//GOLDBERG J, FISCHER M. Encyclopedia of Statistics in Behavioral Science. Hoboken, NJ, USA:John Wiley& Sons, Ltd., 2005.
[39]BREIMAN L. Random Forests[J]. Machine Learning, 2001, 45(1):5-32.
[40]FRIEDMAN J H. Stochastic gradient boosting[J].Computational Statistics and Data Analysis,2002,38(4):367-378.
[41]汪彪,曾新民,刘正奇,等. 中国西北地区参考作物蒸散量的估算与变化特征[J]. 干旱气象, 2016, 34(2):243-251.
[42]冯禹,崔宁博,魏新平,等. 川中丘陵区参考作物蒸散量时空变化特征与成因分析[J].农业工程学报, 2014,30(14):78-86,339.

相似文献/References:

[1]尹起,周建平,许燕,等.基于粒子群优化（PSO）超限学习机预测新疆参考作物蒸散量[J].江苏农业学报,2021,(03):622.[doi:doi:10.3969/j.issn.1000-4440.2021.03.010]
　YIN Qi,ZHOU Jian-ping,XU Yan,et al.Prediction of reference crop evapotranspiration in Xinjiang based on particle swarm optimization(PSO) optimized extreme learning machine[J].,2021,(05):622.[doi:doi:10.3969/j.issn.1000-4440.2021.03.010]
[2]马钊,任传栋,刘静,等.基于不同LSTM模型和Hargreaves模型估算鲁中地区参考作物蒸散量[J].江苏农业学报,2022,38(06):1559.[doi:doi:10.3969/j.issn.1000-4440.2022.06.014]
　MA Zhao,REN Chuan-dong,LIU Jing,et al.Estimation of reference crop evapotranspiration in central Shandong by different LSTM models and Hargreaves models[J].,2022,38(05):1559.[doi:doi:10.3969/j.issn.1000-4440.2022.06.014]

备注/Memo

备注/Memo:
收稿日期：2020-02-10基金项目：国家自然科学基金项目（51879072）；江苏省研究生科研与实践创新计划项目（SJKY19_0523）；中央高校基本科研业务费专项资金项目（2019B68014）；江苏省水利科技项目（2015087）作者简介：张薇（1992-），女，福建龙岩人，硕士研究生，主要从事农业水资源高效利用研究，(E-mail)hallowinnie@163.com通讯作者：邵光成, (E-mail)sgctgzy@163.com
更新日期/Last Update: 2020-11-16