南方农业学报2026，Vol.57Issue(2)：398-408,11.DOI:10.3969/j.issn.2095-1191.2026.02.009
基于气象因子的荔枝成花枝率预测模型构建
Establishment of predictive model of litchi flowering rate based on meteorological factors
摘要
Abstract
[Objective]This study aimed to develop a prediction model of litchi flowering rate and a system of warmth damage indicators,thereby providing a scientific basis for flowering quantity estimation,meteorological forecasting and warning,as well as plan making of disaster prevention and mitigation.[Method]Taking the main variety Feizixiao of Hainan litchi as research objects and based on phenological materials and meteorological observation of 8 litchi demonstration or-chards from 2011 to 2024,a flowering rate prediction model of Hainan litchi Feizixiao was established using a second-order polynomial surface fitting method,and the spatiotemporal distribution characteristics of warmth damage during litchi flowering induction period in the study area were analyzed.[Result]During the flower bud induction period of litchi Feizixiao,low and high temperatures fluctuated obviously,as chill accumulation(≤15℃)and heat accumulation(≥28℃)were key meteorological factors affecting the flowering rate.The established second-order polynomial predic-tion model was expressed as z=-0.53x2+0.81x+0.23xy+0.04y2-0.19y+0.69,with a coefficient of determination(R2)of 0.86,a correlation coefficient between predicted and observed values of 0.951,and the root mean square error and standard deviation of 0.038 and 0.095 respectively,indicating a high prediction accuracy and generalization ability.In terms of the spatiotemporal distribution of warmth damage,higher chill accumulations were found in nor-thern and northwestern cities(counties)such as Baisha,Haikou,Danzhou,Chengmai,and Dingan,where the average chill accumulation>100.0 ℃·d,indicating the sufficient chilling units,adequate flower bud induction,and low frequency of warm damage,with the average flowering rate exceeding 90%.Lingshui at the southern region had the lowest average chill accumulation(38.9 ℃·d),while Ledong had the highest average heat accumulation(29.2 ℃·d);the southern regions were more frequently af-fected by warmth damage,with risks of moderate to severe events,and in 2016 and 2024,it was hit by relatively severe warmth damage.[Conclusion]The established prediction model of litchi flowering rate can provide technical support for monitoring litchi flowering and early warning of warmth damage.As the climate continues to warm in the future,the flow-ering promotion and shoot growth control of litchi face increasing challenges,so optimizing cultivar composition and strengthening research on climate adaptability are essential to improving yield and quality of litchi.关键词
荔枝/成花枝率/二阶多项式拟合/气象因子/预测模型/海南省Key words
litchi/flowering rate/second-order polynomial fitting/meteorological factor/prediction model/Hainan分类
农业科技引用本文复制引用
侯伟,陈哲,王祥和,陈小敏,李伟光,张慧,王秀珍..基于气象因子的荔枝成花枝率预测模型构建[J].南方农业学报,2026,57(2):398-408,11.基金项目
海南省"南海新星"科技创新人才平台项目(NHXXRCXM202355) Hainan"New Star of the South China Sea"Science and Technology Innovation Talent Platform Project(NHXXRCXM202355) （NHXXRCXM202355）
