Construction of high and low temperature disaster prediction model based on microclimate characteristics of kiwifruit orchard

Abstract:【Objective】Kiwifruit is one of the fruit trees with the largest planting area and output in Shaanxi and even in China. It is vulnerable to extreme temperature waterlogging and drought. Meteoro- logical disasters such as high temperature stress in summer, freezing injury in dormancy and germination could seriously affect its normal physiological activities, yield, and quality. Therefore, analyzing the temperature characteristics in kiwifruit orchard and establishing the high and low temperature prediction models are of great significance to prevent meteorological disasters, guide the fine management and elevate the quality and efficiency of kiwifruit industry.【Methods】The microclimate observation system was used to systematically and comprehensively observe the temperature at different heights of the canopy in Shaanxi Guanzhong kiwifruit orchard. The temperature variation characteristics of kiwifruit at different vertical heights and canopy temperature under different weather conditions were ana- lyzed. The prediction model of high and low temperature of orchard canopy was established by using the statistical methods.【Results】The diurnal variation trend of kiwifruit canopy temperature in differ- ent vertical heights during fruit expansion and dormancy and different weather was the same. It was a single peak curve which was generally consistent with other orchards. The daily maximum temperature  in the upper canopy was higher than that in the lower canopy, which appears at 14:00. During the dormant period, the daily minimum temperature in canopy was lower than that in the upper and lower canopy, which appeared at 7:00. And the canopy temperature was lower than the upper canopy and lower canopy from 5: 00 to 18: 00. During the fruit expansion period of kiwifruit, the daily maximum temperature of canopy was the highest on sunny days, followed by cloudy days and the lowest on rainy days. On sunny and cloudy days, the maximum temperature appears at 15:00, and on rainy days it appears at 14:00. During the dormant period, the daily minimum temperature of canopy appeared at 7:00, and was the lowest in sunny days, followed by cloudy days, and the highest in rainy days. The diurnal changes of daily maximum temperature and daily minimum temperature in kiwifruit orchard canopy were con- sistent with those in weather stations, and the daily maximum temperature and daily minimum temperature in orchard canopy are higher than those in weather stations in most periods. The maximum daily temperature of canopy reached the index of high-temperature stress mainly in July and the first half of August, and the minimum daily temperature reached the index of low-temperature freezing damage mainly in late January. High-temperature heat damage was most likely to occur in July and low-tempera- ture freezing damage was most likely to occur in January. The daily dynamic changes of daily maxi- mum temperature and daily minimum temperature in kiwifruit canopy during fruit expansion and dor- mancy are consistent with the trend of weather stations, and canopy temperature is higher than weather stations in most periods. Most of the time when the daily maximum temperature of canopy reaches the index of high temperature and heat damage is concentrated from early July to mid-August, and the time when the dormancy period reaches low temperature and freezing damage is concentrated in late Janu- ary. Based on the weather station temperature, the linear prediction models of the highest temperature and the lowest temperature in three weather types, namely clear, cloudy and overcast or rainy, were established, and the linear prediction models of high and low temperature in the whole growth stage of kiwifruit were established regardless of the weather types. The prediction model equation determination coefficient (R2) of low temperature in dormant period and high and low temperature in different weather types is above 0.90, and the prediction model equation determination coefficient (R2) of high tempera- ture in three weather types in fruit expansion period is above 0.70. The low temperature linear model of kiwifruit dormant period can well predict and calculate the lowest canopy temperature of kiwifruit or- chard, and the absolute error is mostly within 2 °C; The absolute error of the linear model of high tem- perature in kiwifruit expansion period in estimating the highest temperature in orchard canopy is large, which needs to be further revised.【Conclusion】The diurnal variation of temperature in kiwifruit orchard at different heights and different weather types presented a single peak curve, and the canopy temperature was always lower than the upper canopy and lower than the lower canopy most of the time. The daily dynamic change of canopy temperature was consistent with weather stations temperature, and is higher than that in most periods. On the daily scale, the occurrence time of high temperature stress was concentrated from 15: 00 to 16: 00, and the occurrence time of chilling damage was concentrated at 7: 00. On the monthly scale, the high temperature stress occurred from early July to mid-August, and the chilling damage occurred in January. The low temperature prediction model of kiwifruit canopy dormant period and the prediction model of kiwifruit canopy regardless of weather type can well predict the temperature of canopy height. The linear model of high temperature in kiwifruit expanding period needs further model revision.