基于两步剔除趋势法的中国粮食单产波动特征分析  

Grain yield fluctuation in China using the two-stage de-trending method

作　　者：刘忠[1] 黄峰[1] 李保国[1] 

机构地区：[1]中国农业大学资源与环境学院农业部华北耕地保育重点实验室,北京100193

出　　处：《资源科学》2015年第6期1279-1286,共8页Resources Science

基　　金：中央高校基本科研业务费专项(2014JD071);"十二五"国家科技支撑计划项目(2012BAD05B02);国家自然科学基金项目(41271532);长江学者和创新团队发展计划(IRT0412)

摘　　要：粮食单产的增加是改革开放以来中国粮食总产上升的主要原因。进行粮食单产波动分解,并分析其多尺度波动特征及主导因素,从而揭示粮食单产变化的规律,对于今后中国制定相关政策,稳定粮食生产,保障粮食安全具有重要的意义。本文基于粮食生产的技术-经济复合系统分析,利用1978-2011年粮食生产统计数据,通过多年滑动平均和曲线拟合两级趋势剔除,进行了单产波动分解,并分析了粮食单产的多尺度波动特征及主导因素。结果表明:1应用两级趋势去除的方法可以将粮食单产分解为长期趋势、中期波动和短期波动3个项,其中短期波动主要受年际气候变化影响,中期波动曲线主要反映了研究期社会经济和粮食政策的变化及其对单产的影响,而长期趋势则主要反映了技术进步和农作制度对于粮食单产的作用;2长期来看,研究期粮食单产以上升趋势为主,总体波动不大,但个别时段波动较大;3中期波动显著大于短期波动,同时,短期波动率在逐渐收窄,而中期波动率没有明显的趋势变化,显示粮食单产波动的主导因素为社会经济条件和宏观政策。研究结果可为今后农业政策的调整提供决策依据。Grain yield,as measured by grain output per unit harvested area,may be responsible for the improvement of gross grain output in China since 1978. Identifying major factors that account for such an increase and the fluctuation of grain yield and patterns of change may contribute to modification of food policy and food security. Starting from the rational-farmer hypothesis and technology-economy system analysis,we decomposed time-series(1978-2011)grain yield into different components using two-stage de-trending method. In addition,multi-scale fluctuation and major influencing factors of grain yield were also examined. We found that time-series grain yield could be decomposed into three parts,including a short-term fluctuation obtained by de-trending moving averages from time-series grain yield,long-term trend by curve fitting of moving averages and a medium-term fluctuation by subtracting the previous two. Short-term fluctuation was primarily determined by climatic factors;medium fluctuation by social-economic and policy cycles;and long-term fluctuation by technology and farming systems. Despite volatile time intervals,grain yield featured on overall long-term up-ward trend with little fluctuation. The degree of fluctuation as measured by maximum,minimum and range of fluctuation rates in the medium-term was larger than that over the short-term. Meanwhile,the degree of short-term fluctuation seemed to decline over the study period,and a similar trend was not observed for medium-term fluctuation. We conclude that medium-term or social-economic and policy cycles may be the largest contributor to grain yield fluctuation and this finding has significant scientific and policy implications for China's food policy.

关 键 词：粮食 单产 波动 趋势去除 多尺度 中国 

分 类 号：F326.11[经济管理—产业经济]