基于GM(1,1)的茶园晚霜发生灰预测模型

A GM(1, 1)-based Grey Prediction Model of Late Frost Occurrence in Tea Fields

为准确预报早春晚霜发生状况,从而为茶园防霜调控提供决策支持,从宏观时间尺度,建立了晚霜发生的预测模型。对江苏省镇江地区1979~2008年3,4月份晚霜发生的历史数据进行统计分析,基于灰色系统理论的季节灾变灰预测方法,建立晚霜发生的初霜日和末霜日预测模型。结果表明：GM（1,1）灰模型的后验差的比值小于0.35,且模型精度大于0.95,预测精度为优;经验证,预测初霜日3月1日、末霜日4月2日的发生年份与实际基本相符。因此,采用该预测模型进行霜冻预报,对于防霜控制、避免和减轻霜冻害损失,具有重要的实践意义。

A prediction model was established on a long time series scale to forecast the late frost occurrence in tea fields accurately, which would support the decision making for frost protection control. Historical data of late frost occurrence in Zhenjiang between March and April from1979 to 2008 were statistically analyzed, and then the prediction models of the first and the last date of frost occurrences were established based on the seasonal calamities grey method. The results showed that GM（1, 1） grey prediction model has the excellent accuracy with the ratio between posteriori errors less than 0.35 and the model accuracy above 0.95. The predicted years in which the first and the last frost occurred on March 1st and April 2nd respectively are approximately correct. Thus, the above model has the potential to forecast frost damage and is of practical significance for frost protection control and disaster avoidance or reduction.