基于经验模态分解的中国粮食单产波动特征及影响因素

探索中国粮食单产的多尺度波动特征及其影响因素,进而有针对性地采取措施来稳定和提高粮食单产对保障中国粮食安全有重要意义。该文对1978-2012年中国粮食及其主要构成作物稻谷、小麦和玉米的单位面积产量,利用经验模态分解方法进行了多尺度波动分解,并分析了改革开放以来中国粮食单产波动的多尺度特征及其主要影响因素。结果表明:1)应用经验模态分解方法可以将1978年以来中国的粮食和三大作物的单产分解为1个趋势项和2个波动项,分别反应了科技、政策和气候对于中国粮食单产波动的影响。其中技术趋势为主导,2个尺度的波动都较小。2)粮食趋势单产年增长率从1997年前的2.28%下降到之后的0.69%,明显放缓。稻谷的趋势单产与粮食走势最为接近,但1997年以后几乎走平。玉米和小麦的趋势单产增速1997年后虽有所下降,但目前上升依然明显。3)中期波动项包括了3个完整的周期和1个进行中的周期,且与中国粮食政策的重大调整及其所产生的后效在时段上具有很好的一致性。总体看粮食生产领域政策的影响大于气候的影响,但不同作物对各种政策的响应也有所不同,土地、税收、补贴、奖励和保险政策的普惠性和同步性较价格政策更强。小麦的中期波动率明显大于其他作物,显示其对政策更加敏感。4)粮食单产的短期波动率为1.80%,其中玉米为3.38%,高于小麦(2.55%)和稻谷(1.06%),显示中国的粮食生产系统有较强的抵御气象灾害的能力,但玉米防灾能力不及稻谷和小麦。5)综合长期趋势和中短期波动来看,3大作物中稻谷最为稳产。稻谷的主要问题是1998年以后单产的趋势产量上升乏力,玉米的主要问题是短期波动较大,受气象灾害的影响大,而小麦的政策波动大于玉米和稻谷。当前中国的农业已经进入了新的发展阶段,农业政策的调整势在必行,该研究结果可为新阶段农业宏观调控政策的制定提供决策依据。

Analysis on characteristics and influential factors of grain yield fluctuation in China based on empirical mode decomposition

Abstract: There is an important significance for China's food security to explore the fluctuation characteristics and influence factors of grain output per sown area (hereinafter referred to as grain yield), and then to take corresponding measures to stabilize and improve the grain yield level. In this paper, the time-series yields of grain and main crops (rice, wheat and maize) from 1978 to 2012 in China were decomposed using empirical mode decomposition (EMD) method, and then multi-scale fluctuation characteristics and its main influence factors of grain yield were analyzed. The results showed: 1) the time series of grain yield could be decomposed into two intrinsic mode functions (IMF) and a residual item. The residual item denotes a slow and steady rising trend in grain yield which is dominated by the technological progress. The IMF1 and IMF2 represent a short-term fluctuation in about 4-yr cycle and a medium-term fluctuation in about 9-yr cycle, which respectively reflect the influence of meteorological and policy factors on grain yields according to the analysis of grain production system. The variance contribution rate of the residual item was greatly larger than that of the IMFs, which means the upward trend in grain yield was more pronounced than fluctuations. 2) Fluctuation of trend yield of grain was distinct before and after 1997: fast rise with an average annual growth rate of 2.28% before 1997 and slow rise with an average annual growth rate of only 0.69% after 1997. The change of trend yield was similar between rice and grain, but the former seemed to rise more slowly after 1997. For wheat and maize, the trend yield was still increased obviously although the growing rate had been significantly slowing down after 1997. 3) The IMF2 included 3 complete cycles (1979-1988, 1989-1993, and 1994-2002) and an ongoing cycle (2003 to present) which coincided with adjustment of Chinese agricultural policy. In general, the impact of policy and market factors on fluctuation of grain yield is greater than that of climate factors, but responses of different crops to different policies varied: policies on land, tax, subsidies, incentives and insurance had positive or negative effect on medium-term fluctuation of main crops synchronously, but price policies was not the case for main crops especially for maize. Medium-term fluctuation rate of grain yield was 2.94%, and that of wheat was 3.83%, significantly higher than that of maize (2.38%) and rice (2.18%). It suggested wheat's vulnerability to policies. 4) The short-term fluctuation rate of grain yield was 1.80%, and that of maize, wheat, and rice was 3.38%, 2.55%, and 1.06%, respectively. It indicated that Chinese grain production system had a strong ability to resist against meteorological disaster. The disaster prevention capabilities of different crops was not the same, and maize was more easily affected by meteorological disasters than wheat and rice. 5) Both considering the short-term and medium-term fluctuation rates, rice was the most stable crop in yield. The main problem of rice production was the slow growth of yield trend after 1998, while maize was strongly influenced by meteorological disaster, and fluctuation of wheat was significantly affected by policy cycles. At present, China's agriculture production has entered a new development stage. It is necessary to adjust the agricultural policy in order to adapt to the new situation. The results here can help with decision-making and adjustment of agricultural policies.