主办单位:中国气象局沈阳大气环境研究所
国际刊号:ISSN 1673-503X
国内刊号:CN 21-1531/P

气象与环境学报 ›› 2011, Vol. 27 ›› Issue (3): 51-55.

• 简报 • 上一篇    下一篇

基于残差周期修正的辽宁省粮食产量灰色预测

江和文1 赵铭2 郭婷婷3张丽敏1 孙卓4 高井宝5   

  1. 1.葫芦岛市气象局, 辽宁  葫芦岛 125000;2.秦皇岛市气象局,河北 秦皇岛 066000;3.辽宁省气象局,辽宁 沈阳 110003; 4. 盘锦市气象局, 辽宁 盘锦 124010; 5. 大洼县气象局, 辽宁 大洼 124200
  • 收稿日期:2010-12-02 修回日期:2011-02-28 出版日期:2011-06-30 发布日期:2011-02-28

Grey prediction of grain yield based on period residual modification method in Liaoning province

JIANG He-wen1 ZHAO Ming2 GUO Ting-ting3 ZHANG Li-min1 SUN Zhuo4 GAO Jing-bao5   

  1. 1. Huludao Meteorological Service, Huludao 125000, China;2. Qinhuangdao Meteorological Service, Hebei 125000, China; 3. Liaoning Meteorological Service, Shenyang 110003, China; 4. Panjin Meteorological Service, Panjin 124010, China; 5. Dawa Meteorological Service, Dawa 124200, China
  • Received:2010-12-02 Revised:2011-02-28 Online:2011-06-30 Published:2011-02-28

摘要: 基于1971—2005年辽宁省主要粮食作物水稻和玉米的全区平均单产资料,根据产量数据的多周期和波动性特点,采用周期修正残差值的方法,构建了辽宁省粮食产量的灰色预测模型,采用正弦曲线拟合残差序列,对模型的残差进行周期修正,预测精度提高。结果表明:“十二五”期间,辽宁省的水稻在2013年以后出现小幅度下降,而单产的绝对值仍处于一个相对较高的水平。玉米单产至2014年呈持续增产趋势,2015年略有下降,单产波动幅度较小。以2006—2009年实际产量数据对模型进行检验,效果较好。水稻单产预测模型精度为一级,玉米单产预测模型精度为二级。该模型对于大灾年份的粮食产量预测精度较差,但可反映其趋势。

关键词: 水稻, 玉米, 粮食产量, 灰色预测, GM(1,1)模型

Abstract: According to multi cycles and fluctuant features of crop yield data, a grey prediction model of grain yield was established in Liaoning province based on the mean yearly yield per unit area of main crops (maize and rice) in Liaoning province from 1971 to 2005 using a period residual modification method. The sine curve was applied to simulate the residual sequence and the residual was revised, so the prediction precision of the model was improved. The results indicate that rice yield would decline slightly after 2013, while the absolute value of yield per unit area still remains a relatively high level from 2011 to 2015. Maize yield per unit area may increase successively till 2014, then it may declines in 2015. In general, the fluctuation of crop yield per unit area is slight. The model is tested by using the actual yield from 2006 to 2009, and the prediction effect is good. The prediction precision of rice yield per unit area reaches the first grade in terms of this model, and that of maize is the second grade. The prediction precision of crop yield in severe disaster year is low, but the trend could be predicted according to the model.

Key words: Rice, Maize, Grain yield, Grey prediction, GM (1, 1) model

中图分类号: