Abstract:

To accurately predict the blooming period of rape and carry out the meteorological service more effectively, the authors selected the temperature, precipitation, sunshine and other observation data of rape from 1990 to 2019 in Xuancheng of Anhui Province, and used SPSS and Excel software based on the Mann-Kendall method, to analyze the relationship between common flowering characteristics and period with meteorological factors. The results showed that the average blooming period of rape in the past 30 years was March 20 in Xuancheng, the interannual variation changed significantly, the difference between the earliest and the latest date was more than one month, and the overall trend was ahead of time with a rate of 2.7 d/10 a. The start of the mutation was around 2013. The blooming period of rape in Xuancheng was significantly and negatively correlated with the average temperature in January, February and the whole winter, as well as the active accumulated temperature ≥3℃, but it had little correlation with precipitation and sunshine. Based on the study, the prediction model of rape blooming period was established by stepwise regression. After a retrospective test, the minimum error of prediction value was 0 d, and the maximum error was 11 d, so the effect was generally good.

Key words: Rape, Blooming Period, Feature Analysis, Prediction, Xuancheng