自然资源学报 ›› 2005, Vol. 20 ›› Issue (5): 745-751.doi: 10.11849/zrzyxb.2005.05.014

• 研究方法 • 上一篇    下一篇

基于EMD的我国粮食产量波动及其成因多尺度分析

刘会玉1, 林振山1, 张明阳2   

  1. 1. 南京师范大学地理科学学院, 南京 210097;
    2. 中国科学院亚热带农业生态研究所, 长沙 410125
  • 收稿日期:2005-01-21 修回日期:2005-03-29 出版日期:2005-10-25 发布日期:2005-10-25
  • 作者简介:刘会玉(1978~):女:湖南辰溪人:博士生:主要从事资源生态的研究.E-mai:foozle@163.com
  • 基金资助:

    国家自然科学基金项目(40371044);国家“211”二期工程重大项目“不同时空尺度的环境演变和生态建设”.

Analysis on the Fluctuation of Grain Output in China and Its Causes at Multi-time Scale Based on Empirical Mode Decomposition Method

LIU Hui-yu1, LIN Zhen-shan1, ZHANG Ming-yang2   

  1. 1. Geography Science College, Nanjing Normal University, Nanjing 210097, China;
    2. Institute of Subtropical Agriculture, CAS, Changsha 410125, China
  • Received:2005-01-21 Revised:2005-03-29 Online:2005-10-25 Published:2005-10-25

摘要: 论文利用经验模态分解(EMD)方法对1949年以来我国粮食产量波动及其成因进行了多时间尺度的分析。研究结果表明:我国粮食产量存在3年和9年左右的准周期波动,并以9年左右的波动为主。从粮食产量波动的趋势量分布来看,1949年以来我国粮食产量不断增长,但是20世纪90年代后粮食增长停滞;在对粮食作物播种面积单产和粮食播种面积两个影响因子进行EMD分解时发现,粮食作物播种面积单产是粮食产量3年左右波动的控制因子,而粮食播种面积却是粮食产量9年左右波动的主要控制因子。对两个尺度的周期性波动分析发现,我国粮食产量波动幅度近年来不断增加,这将给我国已经相当紧张的粮食供需关系带来巨大的隐患。由于3年和9年左右的两个准周期波动在2005~2006年期间都将处于波谷,论文预测2005或2006年的粮食将严重短缺,并且短期内我国粮食产量仍会下降,但是较长期内粮食产量将回升。

关键词: 粮食产量, 波动, 经验模态分解, 多尺度, 本征模函数

Abstract: Grain security is a very important issue which has a significant influence on the social stability of both a country and a region.Meanwhile,the fluctuation of grain output has a close relation with grain security.Thereby,the researches on the periods,amplitudes and characteristics of the fluctuation of grain output and its causes are propitious to reduce the intensity of the fluctuation,to eliminate unfavorable factors,and further to keep the stability of grain output to insure grain security.Empirical Mode Decomposition(EMD) is a powerful method for analyzing the nonlinear and non-stationary time series.Different from the traditional method in doing integral transformation to signal,it decomposes signal into several intrinsic mode functions (IMFs),which contain and extrude the local characters of signal.So the characteristic information of the original signal can be well held by analyzing the IMFs.Therefore,EMD method was introduced to study the fluctuation of grain production in China and its causes at multi-time scale,and the data have been decomposed into two IMFS and a residual trend term.The conclusions are drawn :1)The fluctuations of grain output have 3-year and 9-year time scales.Hereinto,fluctuation of the 9-year time scale is more prominent.Shown from the residual trend term,the grain output has increased continually since 1949,but stagnated after 1996. 2)We have decomposed the data of per capita grain yield and cultivated land by EMD.We found that per capita grain yield is the control factor of the 3-year fluctuation of grain output while cultivated land is the main influencing factor of the 9-year fluctuation of grain yield,moreover,the national policy and natural disasters also influence the 9-year fluctuation of grain yield to a certain degree.Therefore,in order to eliminate the short-term unfavorable factors,we must improve the foundation of grain production,add the investment in science and technology and enhance the ability of disaster prevention so as to stabilize per capita grain yield.With regarded the elimination of the long-term unfavorable factors,we should control the decrease in cultivated land.3)Analysis on the periodic fluctuation of the two time scales revealed that the grain output will decrease in the recent 2~3 years,but it will soon reverse to increase for a long time.Meanwhile,the amplitudes of the oscillation of grain output is increasing,which indicates that it is possible for the grain output to increase or decrease with big amplitudes.So it is urgent to take measures to guarantee the grain security in the case of crop failure.

Key words: grain output, fluctuation, empirical mode decomposition, intrinsic mode function

中图分类号: 

  • F307.11