基于关键气象因子的油茶种仁含油率预测模型构建

prediction Model of seed oil content of camellia oleifolia Based on key Meteorological Factors

摘要:

含油率是油茶重要的经济性状,构建油茶含油率预测模型对于茶油产量预报有重要的意义 。本研究以普通油茶 为研究对象,通过分析油茶种仁含油率与不同气象因子的关系,确定影响油茶种仁含油率的关键气象因子,然后用回归分 析法构建基于关键气象因子的油茶含油率预测模型,并用独立资料检验预测模型效果 。结果表明:8月平均气温、8 月 最高 气温、9月最高气温、油脂转化积累高峰期最高气温与含油率呈显著的负相关关系 。通过逐步回归分析法拟合得到了 3个 油茶含油率预测模型,用独立资料检验后,发现基于 9月 降水量(儿1) 、9月最高气温(儿2)和油脂转化积累高峰期最长连续无 降水天数(儿3)的油茶种仁含油率(y)预测模型(y= 79. 46- 0. 03儿1 - 0. 86x2 - 0. 30儿3)效果好,平均相对误差为 4. 6%,可应 用于普通油茶种仁含油率预测。

Abstract:

oil content is an important economic character of Camellia oleifera . It is of great significance to construct a prediction model of oil content in seed kernel for C. oleifera . In this study, common C. oleifera was selected as the re- search object. By analyzing the relationship between oil content of C. oleifolia and different meteorological factors , the key meteorological factors affecting oil content of C. oleifolia were determined. Then the prediction model of oil content of C. oleifera based on key meteorological factors was constructed by regression analysis method and the effect of predic- tion model was tested by independent data . The results showed that the average temperature in August, the maximum tem- perature in August, the maximum temperature in september and the maximum temperature in the peak period of oil con- version and accumulation were significantly negatively correlated. Three oil content prediction models of C. oleifera were fitted by stepwise regression analysis . After testing with independent data, it was found that the oil content prediction model based on precipitation in september, maximum temperature in september and longest consecutive days without pre- cipitation during the peak period of oil conversion and accumulation had the best effect, and could be applied to the oil content prediction of C. oleifera .