作物杂志,2019, 第5期: 159–165 doi: 10.16035/j.issn.1001-7283.2019.05.026

• 生理生化·植物营养·栽培耕作 • 上一篇    下一篇

WOFOST模型对山东省夏玉米发育期与产量模拟的适用性评价

董智强1,王萌萌2,李鸿怡1,薛晓萍1,潘志华3,侯英雨4,陈辰1,李楠1,李曼华1   

  1. 1 山东省气候中心,250031,山东济南
    2 德州市气象局,253078,山东德州
    3 中国农业大学资源与环境学院,100193,北京
    4 国家气象中心,100081,北京
  • 收稿日期:2019-03-08 修回日期:2019-06-06 出版日期:2019-10-15 发布日期:2019-11-07
  • 通讯作者: 李鸿怡
  • 作者简介:董智强,工程师,主要从事气象因子对农业生产影响的定量评估研究
  • 基金资助:
    山东省自然科学基金博士基金项目(ZR2018BD024);山东省气象局气象科学技术研究项目重点课题(2017sdqxz02);十三五山东重大气象工程项目(鲁发改农经〔2017〕97号);国家重点研发计划(2017YFD0301004)

Applicability Assessment of WOFOST Model of Growth and Yield of Summer Maize in Shandong Province

Dong Zhiqiang1,Wang Mengmeng2,Li Hongyi1,Xue Xiaoping1,Pan Zhihua3,Hou Yingyu4,Chen Chen1,Li Nan1,Li Manhua1   

  1. 1 Shandong Provincial Climate Center, Jinan 250031, Shandong, China
    2 Dezhou Meteorological Bureau, Dezhou 253078, Shandong, China
    3 College of Resources and Environmental Science, China Agricultural University, Beijing 100193, China
    4 National Meteorological Center, Beijing 100081, China
  • Received:2019-03-08 Revised:2019-06-06 Online:2019-10-15 Published:2019-11-07
  • Contact: Hongyi Li

摘要:

夏玉米作为山东省最主要的粮食作物,其生长发育与产量变化,对保障地区乃至全国的粮食安全具有举足轻重的作用。不同生育期内气象要素的不同组合对夏玉米发育进程与产量形成等将产生重要影响。WOFOST作物模型机理性较强、定量水平高,且更为高效,能够为客观、定量、动态地评估气象要素对夏玉米生产的影响提供技术支撑。为提高作物模型模拟的准确性,将山东省分为鲁西北、鲁中、鲁西南、鲁东南与半岛5个调参区域,并结合山东省10个夏玉米观测站2012-2014年玉米主栽品种的生长发育数据,开展模型的调参验证与适用性评价。研究结果表明,WOFOST模型对山东省各观测站点所有年份出苗期的模拟误差均不超过4d,决定系数(R 2)在0.43~0.99,归一化均方根误差(nRMSE)为0.3%~1.9%;针对开花期和成熟期,各观测站点绝大多数年份模拟误差均不超过5d,大多数观测站点R 2分别在0.77~0.99与0.51~0.99,各观测站点nRMSE分别在0.4%~2.3%与0.7%~3.2%;绝大多数观测站点产量模拟R 2在0.68~0.99,相对误差为0.8%~16.7%,绝大多数观测站点相对误差小于10%;nRMSE在1.2%~19.5%,均小于30%,大部分观测站点nRMSE小于10%。各评价指标均在可接受的范围内,WOFOST模型能够对山东省夏玉米发育期与产量进行较准确的模拟。

关键词: WOFOST模型, 适用性, 夏玉米, 发育期, 产量

Abstract:

Summer maize is the most important grain crop in Shandong province. The changes of its growth period and yield play an important role in ensuring regional and even national food security. Different combinations of meteorological factors during different growth periods will affect the development process and yield of summer maize significantly. WOFOST crop model has stronger mechanism, higher quantitative level and great efficiency. It can provide technical support for the objective, quantitative and dynamic assessment of the impacts of meteorological factors on summer maize production. To improve the accuracy of WOFOST model, Shandong province was segregated into five regions, including the northwest region, middle region, southwest region, southeast region and peninsula region. Based on the growth and development data of staple varieties from 2012 to 2014 of 10 summer maize observation stations in Shandong province, we completed the parameters adjustment and verification and applicability assessment of WOFOST model. The results showed that the simulation error of the period of emergence of all observation stations is no more than 4d. Its determination coefficient (R 2) is 0.43-0.99. And its normalized root mean square error (nRMSE) is 0.3%-1.9%. The simulation error of the period of flowering and the period of mature of all observation stations in most years are no more than 5d. Their R 2 of most stations are 0.77-0.99 and 0.51-0.99, respectively. And their nRMSE are 0.4%-2.3% and 0.7%-3.2%, respectively. For the simulation of yield, the R 2 of most stations is 0.68-0.99. Its relative error is 0.8%-16.7% with most stations less than 10%. Its nRMSE is 1.2%-19.5%, and all the stations are less than 30% with most stations less than 10%. All assessment indicators are within the acceptable limits. The WOFOST model can accurately simulate the growth period and yield of summer maize in Shandong province.

Key words: WOFOST model, Applicability, Summer maize, Growth period, Yield

表1

控制作物生理生长参数"

参数
Parameter
生物学意义
Biological significance
最小值
Minimum value
最大值
Maximum value
TSUMEM 播种到出苗期积温 0 170
TSUM1 出苗期到开花期积温 150 1 800
TUSM2 开花期到成熟期积温 400 1 550
TBASEM 出苗期最低温度阈值 -10 8
DLO 最优日长 6 18
DLC 临界日长 6 18
TEFFMX 出苗期最大有效温度 18 32

表2

控制单要素与综合要素影响潜在产量参数"

参数
Parameter
生物学意义
Biological significance
最小值
Minimum value
最大值
Maximum value
AMAXTB 最大CO2同化速率 1 70
SLATB 比叶面积 0.0007 0.0042
SPAN 35℃时叶片寿命 17 50
RGRLAI 叶面积最大相对增长量 0.007 0.500
LAIEM 出苗期叶面积指数 0.0007 0.3000
TDWI 初始总干物重 0.5 300.0
FLTB 叶分配系数 0 1
FSTB 茎分配系数 0 1
FRTB 根分配系数 0 1
FOTB 穗分配系数 0 1
TMPFTB 平均温度最大CO2
同化速率折减系数
0 1
RDRRTB 根相对死亡速率 0 0.02

表3

控制水分限制影响产量参数"

参数
Parameter
生物学意义
Biological significance
最小值
Minimum value
最大值
Maximum value
CFET 作物参考蒸散量校正因子 0.8 1.2
RDMCR 作物成熟最大根深度 50 400
PERDL 水分限制叶片最大相对死亡率 0 0.1
DEPNR 土壤水分耗竭的作物类群数 1 5

表4

各观测站点出苗期模拟验证"

站点
Station
年份
Year
绝对误差(d)
Absolute error
绝对误差平均值(d)
Average of
absolute error
R2 nRMSE
(%)
高密 2012 3 2.0 0.52 1.2
Gaomi 2013 1
2014 -2
菏泽 2012 0 1.7 0.43 1.4
Heze 2013 1
2014 -4
济阳 2012 3 2.7 0.99 1.7
Jiyang 2013 4
2014 1
胶州 2012 1 1.3 0.99 0.8
Jiaozhou 2013 1
2014 2
莒县 2012 1 0.3 0.96 0.3
Juxian 2013 0
2014 0
莱阳 2012 2 2.0 0.99 1.1
Laiyang 2013 2
2014 2
聊城 2012 0 1.3 0.94 1.1
Liaocheng 2013 -1
2014 -3
泰安 2012 -2 2.7 0.99 1.5
Tai′an 2013 -3
2014 -3
潍坊 2012 4 3.0 0.97 1.9
Weifang 2013 2
2014 3
淄博 2012 0 0.7 0.96 0.5
Zibo 2013 1
2014 1

表5

各观测站点开花期模拟验证"

站点
Station
年份
Year
绝对误差(d)
Absolute error
绝对误差平均值(d)
Average absolute error
R2 nRMSE
(%)
高密 2012 5 2.3 0.77 1.4
Gaomi 2013 -1
2014 -1
菏泽 2012 2 1.0 0.98 0.6
Heze 2013 0
2014 -1
济阳 2012 7 4.3 0.13 2.3
Jiyang 2013 5
2014 -1
胶州 2012 2 1.3 0.81 0.7
Jiaozhou 2013 -2
2014 0
莒县 2012 5 4.7 0.99 2.2
Juxian 2013 6
2014 3
莱阳 2012 4 1.3 0.10 1.0
Laiyang 2013 0
2014 0
聊城 2012 1 0.5 - 0.4
Liaocheng 2014 0
2012 -6 2.3 0.92 1.5
泰安 2013 0
Tai′an 2014 -1
2012 3 2.0 0.98 1.0
潍坊 2013 2
Weifang 2014 1
2012 3 3.3 0.17 1.5
淄博 2013 -4
Zibo 2014 3

表6

各观测站点成熟期模拟验证"

站点
Station
年份
Year
绝对误差(d)
Absolute error
绝对误差平均值(d)
Average absolute
error
R2 nRMSE
(%)
高密 2012 5 2.7 0.27 1.2
Gaomi 2013 -3
2014 0
菏泽 2012 1 1.7 0.96 0.7
Heze 2013 -2
2014 -2
济阳 2012 8 4.0 0.98 1.8
Jiyang 2013 -1
2014 -3
胶州 2012 4 5.3 0.57 2.0
Jiaozhou 2013 -7
2014 -5
莒县 2012 7 7.3 0.99 3.2
Juxian 2013 13
2014 -2
莱阳 2012 8 5.0 0.01 2.0
Laiyang 2013 -2
2014 -5
聊城 2012 -2 2.0 - 0.9
Liaocheng 2014 -2
2012 12 6.3 0.88 2.8
泰安 2013 2
Tai′an 2014 -5
2012 9 5.3 0.86 2.3
潍坊 2013 4
Weifang 2014 3
2012 4 5.3 0.51 2.3
淄博 2013 -3
Zibo 2014 9

表7

各观测站点产量模拟验证"

站点Station R2 相对误差Relative error (%) nRMSE (%)
高密Gaomi 0.02 16.7 19.5
菏泽Heze 0.99 11.3 11.4
济阳Jiyang 0.84 4.5 5.3
胶州Jiaozhou 0.71 5.8 7.6
莒县Juxian 0.83 9.1 16.6
莱阳Laiyang 0.97 7.2 10.1
聊城Liaocheng - 4.0 5.5
泰安Tai′an 0.69 0.8 1.2
潍坊Weifang 0.68 2.7 3.6
淄博Zibo 0.97 1.3 1.5
[1] 周广胜 . 气候变化对中国农业生产影响研究展望. 气象与环境科学, 2015,38(1):80-94.
doi: 10.3969/j.issn.1673-7148.2015.01.012
[2] Dong Z Q, Pan Z H, An P L , et al. A quantitative method for risk assessment of agriculture due to climate change. Theoretical and Applied Climatology, 2018,131(1/2):653-659.
[3] 李克让, 陈育峰 . 中国全球气候变化影响研究方法的进展. 地理研究, 1999,18(2):214-219.
[4] 赵俊芳, 郭建平, 马玉平 , 等. 气候变化背景下我国农业热量资源的变化趋势及适应对策. 应用生态学报, 2010,21(11):2922-2930.
[5] Diepen C A, Wolf J, Keulen H , et al. WOFOST:A simulation model of crop production. Soil Use and Management, 1989,5(1):16-24.
[6] 侯英雨, 何亮, 靳宁 , 等. 中国作物生长模拟监测系统构建及应用. 农业工程学报, 2018,34(21):165-175.
[7] 黄健熙, 贾世灵, 马鸿元 , 等. 基于WOFOST模型的中国主产区冬小麦生长过程动态模拟. 农业工程学报, 2017,33(10):222-228.
[8] Supit I Hooijper A A van Diepen C A , et al. System description of the WOFOST 6.0 crop simulation model implemented in CGMS, Luxembourg Office for Official Publications of the European Commission, 1994.
[9] 陈振林, 张建平, 王春乙 , 等. 应用WOFOST模型模拟低温与干旱对玉米产量的综合影响. 中国农业气象, 2007,28(4):440-442.
[10] 李秀芬, 马树庆, 赵慧颖 , 等. 基于WOFOST模型的内蒙古河套灌区玉米低温冷害评价. 中国农业气象, 2016,37(3):352-360.
[11] 栾庆祖, 叶彩华, 莫志鸿 , 等. 基于WOFOST模型的玉米干旱损失评估:以北京为例. 中国农业气象, 2014,35(3):311-316.
doi: 10.3969/j.issn.1000-6362.2014.03.012
[12] Penning De Vries F W T, Jansen D M, Ten Berge H F M , et al. Simulation of ecophysiological processes of growth in several annual crops. Wageningen: Centre for Agricultural Publishing and Documentation, 1989.
[13] Boogaard H L, Diepen C A, van Roetter R P , et al. User's guide for the WOFOST 7.1 crop growth simulation model and WOFOST Control Center 1.5. Wageningen:Wageningen University and Research Centre, 1998: 1-40.
[14] 马玉平, 王石立, 张黎 . 针对华北小麦越冬的WOFOST模型改进. 中国农业气象, 2005,26(3):145-149.
[15] 马玉平, 王石立, 张黎 , 等. 基于遥感信息的华北冬小麦区域生长模型及模拟研究. 气象学报, 2005,63(2):204-215.
doi: 10.11676/qxxb2005.020
[16] 张雪芬, 余卫东, 王春乙 , 等. WOFOST模型在冬小麦晚霜冻害评估中的应用. 自然灾害学报, 2006(S1):137-141.
[17] 张雪芬, 余卫东, 王春乙 . 基于作物模型灾损识别的黄淮区域冬小麦晚霜冻风险评估. 高原气象, 2012,31(1):277-284.
[18] 何亮, 侯英雨, 赵刚 , 等. 基于全局敏感性分析和贝叶斯方法的WOFOST作物模型参数优化. 农业工程学报, 2016,32(2):169-179.
[19] 邬定荣, 欧阳竹, 赵小敏 , 等. 作物生长模型WOFOST在华北平原的适用性研究. 植物生态学报, 2003,27(5):594-602.
[20] 谢文霞, 王光火, 张奇春 . WOFOST模型的发展及应用. 土壤通报, 2006,37(1):154-158.
[21] 孙琳丽, 侯琼, 马玉平 , 等. WOFOST模型在内蒙古河套灌区模拟玉米生长全程的适应性. 生态学杂志, 2016(3):800-807.
[22] 高永刚, 王育光, 殷世平 , 等. 世界粮食研究模型在黑龙江省作物产量预报中的应用. 中国农业气象, 2006,27(1):27-30.
[23] Bussay A, Velde M V D, Fumagalli D , et al. Improving operational maize yield forecasting in Hungary. Agricultural Systems, 2015,141:94-106.
[24] Ma G N, Huang J X, Wu W B , et al. Assimilation of MODIS-LAI into the WOFOST model for forecasting regional winter wheat yield. Mathematical and Computer Modelling, 2013,58(3/4):634-643.
[25] Saltelli A, Tarantola S, Campolongo F , et a1. Sensitivity analysis in practice:a guide to assessing scientific models. Hoboken:John Wiley and Sons, 2004: 20-78.
[26] Song Y L, Chen D L, Liu Y J , et al. The Influence of Climate Change on Winter Wheat during 2012-2100 under A2 and A1B Scenarios in China. Advances in Climate Change Research, 2012,3(3):138-146.