中国农业气象

• 论文 • 上一篇    下一篇

 玉米叶面积指数估算通用模型

 栾青,郭建平,马雅丽,张丽敏,王婧瑄   

  1.  1.山西省气候中心,太原 030006;2.中国气象科学研究院,北京 100081;3.南京信息工程大学气象灾害预警预报与评估协同创新中心,南京 210044;4.葫芦岛市气象局,葫芦岛 125000
  • 出版日期:2020-08-20 发布日期:2020-08-19
  • 作者简介:栾青,E-mail:luanqing2003@163.com
  • 基金资助:
     国家自然科学基金(31571559);中国气象科学研究院科技发展基金(2019KJ006)

 A General Model for Estimating Leaf Area Index of Maize

 LUAN Qing, GUO Jian-ping, MA Ya-li, ZHANG Li-min, WANG Jing-xuan   

  1.  1. Shanxi Climate Center, Taiyuan 030006, China; 2.Chinese Academy of Meteorological Sciences, Beijing 100081; 3.Collaborative Innovation Center on Forecast and Evaluation of Meteorological Disasters, Nanjing University of Information Science and Technology, Nanjing 210044; 4.Huludao Meteorological Bureau, Huludao 125000
  • Online:2020-08-20 Published:2020-08-19
  • Supported by:
     

摘要:  基于2018年黑龙江哈尔滨、吉林榆树、辽宁锦州、新疆乌兰乌苏、甘肃西峰、河北固城6个农业气象试验站不同属性品种玉米的分期播种试验资料,以当地常年大田实际播种期为界,提前10d播种为第1播期,正常播种为第2播期,比正常晚10d播种为第3播期,晚20d为第4播期,以第1播期、第3播期和第4播期实测值计算的有效积温相对值为自变量,采用修正的Logistic方程,构建了通用的玉米叶面积指数估算模型,进一步利用有效积温相对值对模型在三叶期和七叶期的残差进行订正,并用2018年6个农业气象试验站及2019年吉林榆树、甘肃西峰和山东泰安3个农业气象试验站,8个不同品种玉米的分期播种试验资料对模型进行检验。结果显示:以多属性品种玉米有效积温相对值为自变量的RLAI拟合曲线完全符合修正的Logistic方程变化规律,模型拟合优度(R2)达到0.93,通过了0.01水平的显著性检验,具有较高的精度。玉米全生育期不同品种模拟RLAI与实测计算RLAI的相关性较高,通过了0.01水平的显著性检验,相关系数均超过0.9,平均相对误差介于13.8%~27.6%。不同生育期模拟RLAI与实测计算RLAI的平均相对误差介于9.4%~30.7%,七叶期最高,乳熟期最低。说明以不同属性玉米品种、土壤性质、管理措施、种植密度下的试验资料为基础构建的LAI估算模型,较以往基于单站、单品种、单播期或单站多品种LAI估算模型更具普适性,适用于大多数属性品种玉米的LAI模拟。

关键词:  , 玉米, 叶面积指数, Logistic曲线拟合, 估算模型

Abstract:  In order to build a more general model for estimating leaf area index of maize, in this paper, we used the staged seeding test data (based on the local actual field sowing date, the first sowing date was 10 days earlier than normal, the second sowing date was normal, the third sowing date was 10 days later, and the fourth sowing date was 20 days later) with different varieties of maize in six agro-meteorological experiment stations in 2018, including Harbin of Heilongjiang province, Yushu of Jilin province, Jinzhou of Liaoning province, Wulanwusu of Xinjiang province, Xifeng of Gansu province and Gucheng of Hebei province. Taking the relative integrated temperature of the first, third and fourth sowing period as independent variables and the relative value of leaf area index (RLAI) as the dependent variable, the modified Logistic equation was used to construct the estimation model of maize leaf area index. The fitting curve of the model showed that the simulated RLAI in the three-leaf stage and the seven-leaf stage of maize were higher than the measured RLAI, and a significant linear correlation between the residual of the model in these two stages and the relative values of the effective integrated temperature. Therefore, the relative values of the effective integrated temperature were used to fix the residuals of the model in these two stages. The model was tested using the data of 8 different varieties in six agro-meteorological experiment stations in 2018 and 3 agro-meteorological experiment stations (Yushu of Jilin province, Xifeng of Gansu province and Tai’an of Shandong province) in 2019. The results showed that the RLAI fitting curve with the relative integrated temperature of the multi-attribute varieties of maize as independent variables was completely in line with the modified Logistic equation. The model fitting coefficient of determination (R2) reached 0.93, and passed the significance test of 0.01 level with high accuracy. The test results showed that the simulated RLAI of different varieties of maize had a high correlation with the measured RLAI. The correlation coefficient exceeded 0.9 and passed the significance test of 0.01 level. The average relative error of different varieties ranged from 13.8% to 27.6%. The average relative error between simulated RLAI and measured RLAI at different growth stages was between 9.4% and 30.7%, with the highest in the seven-leaf stage and the lowest in the milk-ripe stage. In general, the estimation model constructed based on relative values, eliminated the differences in maize variety attributes, soil properties, management measures, planting density, etc. It has a wider applicability than the previous LAI estimation model based on single station, single variety, single sowing period, or multiple varieties in single station, is suitable for most varieties LAI simulation of maize.

Key words:  Maize, Leaf area index, Logistic curve fitting, Estimation model

中图分类号: