基于时序影像及不同模型的玉米早期估产研究

doi:10.3964/j.issn.1000-0593(2023)08-2627-11

摘要
参考文献
相关文章 (15)

全文: PDF (14159 KB)
输出: BibTeX | EndNote (RIS)

摘要：针对目前粮食产量定量评估模型泛化能力不足、预测时间滞后以及早期估产时间窗口难以确定等问题，以Sentinel-2遥感数据和实测玉米产量作为数据源，开展县域尺度玉米估产及早期最优估产时间窗口确定研究。基于玉米生长期内的时序影像数据集，通过玉米产量实测数据与影像植被指数建立相关关系，并采用MLRM（多元线性回归模型），GPR（高斯过程回归模型），LSTM（长短期记忆人工神经网络模型），建立玉米时序估产模型。实验结果表明，基于LSTM在NDVI、GNDVI、以及GN（NDVI与GNDVI组合）这三种植被指数作为参数建立的时序估产模型中，无论在估产精度，模型可靠性、产量异常值捕捉、以及早期最优估产时间窗口确定等方面均优于基于GPR、MLRM建立的时序估产模型。同时基于LSTM时序估产模型，采用截止到抽雄期的NDVI时序影像数据作为参数，其结果的决定系数R²可达0.83、均方根误差RMSE为0.26 t·ha^-1、相对分析误差RPD为3.52；GNDVI时序影像数据作为参数，其结果的决定系数R²为0.79、均方根误差RMSE为0.30 t·ha^-1、相对分析误差RPD为2.87；以GN时序影像数据作为参数，其结果决定系数R²为0.83、均方根误差RMSE为0.27 t·ha^-1、相对分析误差RPD为3.05；以NDVI作为LSTM模型参数的估产效果最优，相较于玉米收获期可提前2个月就能预测当年的玉米产量，对于县域尺度玉米产量预报具有一定的现实意义，同时也为类似作物的估产研究提供相关参考。

关键词：产量预测；玉米生育期；植被指数；Sentinel-2；长短期记忆人工神经网络模型

Abstract：For the inadequate generalization ability of the quantitative evaluation model of crop yield, the lag of forecasting time and the difficulty of establishing the optimum lead yield estimation time, this paper takes Sentinel-2 remote sensing data and the measured maize yield as the data source to research the establishment of county-scale maize yield estimation and optimum lead yield estimation time. Based on the time-series image data of maize growth-satges, through building the correlation between maize yield measured data and vegetation index, the time-series maize yield estimation model was established by MLRM (multivariable linear regression model), GPR (Gaussian process regression model) and LSTM (Long short-term memory artificial neural network model). The experimental results show that LSTM is superior to GPR and MLRM in terms of the accuracy, and reliability of the yield prediction model, the capture of the abnormal yield value, and the optimum lead yield estimation time in the time series yield estimation model established with NDVI, GNDVI and GN ( NDVI and GNDVI combination ) as parameters. At the same time, based on the LSTM estimation model, the NDVI time-series image data up to tasseling stage were used as parameters and the yield prediction results showed that the R²(determination coefficient) was 0.83, RMSE(root mean square error) was 0.26 t·ha^-1, RPD(relative percent deviation) was 3.52; The GNDVI time-series image data up to tasseling stage were used as parameters, and the yield prediction results showed that the R² was 0.79, RMSE was 0.30 t·ha^-1, RPD was 2.87; The GN time-series image data up to tasseling stage were used as parameters, and the yield prediction results showed that the R² was 0.83, RMSE was 0.27 t·ha^-1, RPD was 3.05. Using the NDVI time-series image data as the LSTM model parameter has the optimal yield estimation, and the maize yield could be predicted 2 months in advance compared with the maize harvest stage. As a result, we developed a crop yield forecasting method in this study to predict crop yield for county-scale. It has practical significance for maize yield forecasting and provides a relevant reference for similar crop yield estimation research.

Key words：Yield forcasting; Maize growth-satges; Vegetation Index; Sentinel-2; LSTM

收稿日期: 2022-04-09 修订日期: 2022-07-20

中图分类号:

TP79

基金资助: 中国科学院战略性先导科技专项项目课题（XDA28010503），面向城市多元信息三维空间高效表达与分析关键技术研究项目（E133S30201）资助

通讯作者: 张树清 E-mail: zhangshuqing@neigae.ac.cn

作者简介: 刘照，1991年生, 中国科学院东北地理与农业生态研究所博士研究生 e-mail: liuzhao@iga.ac.cn

引用本文:

刘照，李华朋，陈慧，张树清. 基于时序影像及不同模型的玉米早期估产研究[J]. 光谱学与光谱分析, 2023, 43(08): 2627-2637.
LIU Zhao, LI Hua-peng, CHEN Hui, ZHANG Shu-qing. Maize Yield Forecasting and Associated Optimum Lead Time Research Based on Temporal Remote Sensing Data and Different Model. SPECTROSCOPY AND SPECTRAL ANALYSIS, 2023, 43(08): 2627-2637.

链接本文:

https://www.gpxygpfx.com/CN/10.3964/j.issn.1000-0593(2023)08-2627-11 或 https://www.gpxygpfx.com/CN/Y2023/V43/I08/2627

[1] HAN Dong-hui, ZHAO Jin-yuan, HU Qi, et al（韩冬荟, 赵金媛, 胡琦, 等）. Journal of China Agricultural University（中国农业大学学报）, 2021, 26(3): 188.
[2] Sergio M, Vicente-Serrano, Jose M, et al. International Journal of Remote Sensing, 2006, 27(3): 511.
[3] Liu L Y, Wang J H, Bao Y S, et al. International Journal of Remote Sensing, 2006, 27(4): 737.
[4] Pan H, Chen Z, Wit A D, et al. Sensors (Basel, Switzerland), 2019, 19(14): 3161.
[5] JIN An-hua, WANG Peng-xin, QI Xuan, et al（靳安华, 王鹏新, 齐璇, 等）. Transactions of the Chinese Society of Agricultural Engineering（农业工程学报）, 2012, 28(6): 162.
[6] REN Jian-qiang, CHEN Zhong-xin, TANG Hua-jun, et al（任建强, 陈仲新, 唐华俊, 等）. Transactions of the Chinese Society of Agricultural Engineering（农业工程学报）, 2011, 27(8): 257.
[7] Araya A, Hoogenboom G, Luedeling E, et al. Agricultural and Forest Meteorology, 2015, 214: 252.
[8] Anothai J, Soler C M T, Green A, et al. Agricultural and Forest Meteorology, 2013, 176: 64.
[9] Becker-Reshef I, Vermote E, Lindeman M, et al. Remote Sensing of Environment, 2010, 114(6): 1312.
[10] Sun J, Lai Z, Di L, et al. IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing, 2020, 13: 5048.
[11] Huang J, Tian L, Liang S, et al. Agricultural and Forest Meteorology, 2015, 204: 106.
[12] Xie Y, Wang P, Bai X, et al. Agricultural and Forest Meteorology, 2017, 246: 194.
[13] WANG Peng-xin, QI Xuan, LI Li, et al（王鹏新, 齐璇, 李俐, 等）. Journal of Agricultural Machinery（农业机械学报）, 2019, 50(7): 237.
[14] YU Hai-yang, CHEN Sheng-bo, YANG Bei-ping, et al（于海洋, 陈圣波, 杨北萍, 等）. Global Geology（世界地质）, 2020, 39(1): 208.
[15] Zhang J, Zhu Y, Zhang X, et al. Journal of Hydrology, 2018, 561: 918.
[16] Jiang H, Hu H, Zhong R, et al. Global Change Biology, 2020, 26(3): 1754.
[17] Greff K, Srivastava R K, Koutník J, et al. IEEE Transactions on Neural Networks and Learning Systems, 2016, 28(10): 2222.
[18] Lin T, Zhong R, Wang Y, et al. Environmental Research Letters, 2020, 15(3): 034016.
[19] Tian H, Wang P, Tansey K, et al. Agricultural and Forest Meteorology, 2021, 310: 108629.
[20] Zhang H, Kang J, Xu X, et al. Computers and Electronics in Agriculture, 2020, 176: 105618.
[21] Ren J Q, Chen Z X, Zhou Q B, et al. Journal of Remote Sensing, 2015, 19(4): 568.
[22] Shanahan J F, Schepers J S, Francis D D, et al. Agronomy Journal, 2021, 93(3): 583.
[23] AN Qin, CHEN Sheng-bo, SUN Shi-chao（安秦, 陈圣波, 孙士超）. Geospatial Information（地理空间信息），2018, 16(3): 14.
[24] Cao Q, Miao Y, Shen J, et al. Agronomy Journal, 2016, 17(4): 136.
[25] KANG Jun-feng, HUANG Lie-xing, ZHANG Chun-yan, et al（康俊锋, 黄烈星, 张春艳, 等）. China Environmental Science（中国环境科学）, 2020, 40(5): 1895.
[26] WANG Teng-jun, FANG Ke, YANG Yun, et al（王腾军, 方珂, 杨耘, 等）. Bulletin of Surveying and Mapping（测绘通报）, 2021，(11): 92.
[27] Saeys W, Mouazen A M, Ramon H. Biosystems Engineering, 2005, 91(4): 393.
[28] WANG Hai-jiang, JIANG Tian-chi, Yunger John A, et al（王海江, 蒋天池, Yunger John A, 等）. Journal of Agricultural Machinery（农业机械学报）, 2018, 49(5): 263.
[29] HAN Wen-ting, PENG Xing-shuo, ZHANG Li-yuan, et al（韩文霆, 彭星硕, 张立元, 等）. Journal of Agricultural Machinery（农业机械学报）, 2020, 51(1): 148.
[30] Li L, Wang B, Feng P, et al. Agricultural and Forest Meteorology, 2021, 308-309(4): 108558.
[31] WANG Lai-gang, XU Jian-hua, HE Jia, et al（王来刚, 徐建华, 贺佳, 等）. Journal of Maize Science（玉米科学），2020, 28(6): 88.