Abstract:

Based on the Zaozhuang pomegranate yield and the Yicheng national meteorological station data from 2006 to 2020, the pomegranate yield was separated by the sliding average, HP filter and Logistic curve fitting method, and the key meteorological factors were selected to establish a simulation model of pomegranate yield, with a view to providing scientific and technological support for enhancing the Zaozhuang pomegranate brand benefits and helping rural revitalization. The results showed that: (1) the Logistic curve fitting method was the most suitable in separating the yield of Zaozhuang pomegranate from 2006 to 2020, with determination coefficient (R²) of 0.998, the mean relative error (MRE) of 0.389%, the mean absolute error (MAE) of 26.296 kg·ha⁻¹, and the root mean squared error (RMSE) of 29.825 kg·ha⁻¹, the HP filter was second best simulation effect. (2) Zaozhuang pomegranate meteorological yield was positively correlated with the average temperature in early June, the longest consecutive precipitation days in July, the precipitation in mid−August, the longest consecutive precipitation days and the precipitation in August, the sunshine hours and the temperature diurnal variation in late September, which indicated that the high temperature in early June, the sufficient precipitation in July and August, the sufficient sunshine and the larger temperature diurnal variation in late September were all favorable to the improvement of the pomegranate meteorological yield. The negative correlation between the maximum temperature in August, the average relative humidity in September, the average relative humidity in late September and the meteorological yield of pomegranate, which indicated that the hot weather in August and the high humidity in September were unfavorable to the growth of pomegranate. (3) After selection, a simulation model of Zaozhuang pomegranate yield was established using years and seven meteorological factors: the average temperature in early June, the longest continuous precipitation days in July, the maximum temperature and the precipitation in August, the average relative humidity in late September, the sunshine hours and the temperature diurnal variance in late September, with R² of 0.984, MRE of 1.18%, MAE of 82.917kg·ha⁻¹, and an RMSE of 197.770kg·ha⁻¹. The yield simulation model was validated with the pomegranate yield data in 2013, 2017 and 2020, with R² of 0.858, MRE of 8.13%, MAE of 566.983kg·ha⁻¹, and RMSE of 882.967kg·ha⁻¹, indicating that the model can be used for the simulation of Zaozhuang pomegranate yield.

Key words: Pomegranate, Meteorological, Trend yield, Logistic curve fitting, Yield simulation