基于光周期和温度的物候模型对华北平原刺槐始花期的模拟  

Modeling the first flowering date of Robinia pseudoacacia L.based on photoperiod and temperature in the North China Plain

作　　者：林楠 刘冬 王雨欣[1,3] 王姝纯 卢凡青 王淼 张煦庭[3,7] 李秋月 徐琳 LIN Nan;LIU Dong;WANG Yuxin;WANG Shuchun;LU Fanqing;WANG Miao;ZHANG Xuting;LI Qiuyue;XU Lin(College of Resources and Environmental Sciences,China Agricultural University,Beijing 100193,China;Huafeng Meteorological Media Group,Beijing 100081,China;CMA-CAU Joint Laboratory of Agriculture Addressing Climate Change,Beijing 100193,China;Sichuan Meteorological Service Center,Chengdu 610072,China;Kaifeng Meteorological Bureau,Kaifeng 475000,China;College of Water Resources and Civil Engineering,China Agricultural University,Beijing 100083,China;Shaanxi Meteorological Service Center of Agricultural Remote Sensing and Economic Crops,Xi′an 710016,China;Beijing Municipal Climate Centre,Beijing 100089,China)

机构地区：[1]中国农业大学资源与环境学院,北京100193 [2]华风气象传媒集团有限责任公司,北京100081 [3]中国气象局-中国农业大学农业应对气候变化联合实验室,北京100193 [4]四川省气象服务中心,成都610072 [5]开封市气象局,开封475000 [6]中国农业大学水利与土木工程学院,北京100083 [7]陕西省农业遥感与经济作物气象服务中心,西安710016 [8]北京市气候中心,北京100089

出　　处：《生态学报》2024年第9期3745-3758,共14页Acta Ecologica Sinica

基　　金：国家重点研发计划项目(2019YFA0607402);国家重点研发计划项目(2018YFA0606102);北京市自然科学基金面上项目(8232034)。

摘　　要：光周期和温度是影响木本植物开花的两大关键气象因素。基于我国华北平原1963—2018年57个站点的刺槐始花期实测数据,构建了6种基于气温或光周期驱动的春季物候过程模型(积温模型、光周期模型、光周期-温度顺序模型、温度-光周期顺序模型、光周期-温度平行模型、光周期-温度乘法模型),并进行了参数率定和模型比较优选,以期明确刺槐开花的主导驱动因子以及潜在的光温作用机制。根据内部模拟的赤池信息准则(AIC)判定光周期-温度顺序模型的表现最佳,模拟值与观测值的相关系数(r)为0.86,均方根误差(RMSE)为4.81d,纳什效率系数(NSE)达到0.74,说明刺槐始花期的发生同时受到光周期和温度的作用。同时,在华北平原14个代表性站点上,光周期-温度顺序模型的表现普遍优于均值模型。基于光周期-温度顺序模型的参数率定结果,以下限温度6.5℃和下限日长4.5h来统计刺槐开花的实际有效积温量和实际有效日长累积量的年际变化情况,结果表明56年来刺槐开花的实际有效积温量呈现显著递增趋势,平均每10年增加4.5℃·d(P<0.05),而实际有效日长累积量却呈现极显著递减趋势,平均每10年减少23.9h·d(P<0.01),这说明在气候变暖背景下,刺槐开花的热量需求会更快速完成,引起刺槐开花日期的提前,但自然光周期的年内变化只与地理位置有关,日长累积需求则需要更长的时间完成,因此会在一定程度上抑制春季升温引起的开花日期提前。Photoperiod and temperature are regarded as two key meteorological factors affecting flowering of woody plants.In this study,we constructed six spring phenological mechanism models driven by photoperiod or temperature(TT model,PP model,TPa model,TPb model,TPc model,and TPd model)derived from the observations of the first flowering date of Robinia pseudoacacia L.at 57 phenological observation stations in the North China Plain from 1963 to 2018.Based on these data,we firstly calibrated the parameters of each model by the means of simulated annealing algorithm and then selected the optimal one on the basis of indicators.Our purpose is to identify the dominant drivers of the first flowering date of Robinia pseudoacacia L.and try to explain the potential mechanisms of photoperiod and temperature.According to the internal validation indicators,TPa model was considered to be optimal among the six phenological models with the lowest Akaike Information Criterion(AIC)2372.3;the correlation coefficient(r)between the simulated and observed Day of Year was 0.86,the Root Mean Squared Error(RMSE)was 4.81 days,and the Nash⁃Sutcliffe Efficiency(NSE)reached 0.74.The internal validation results indicated that the first flowering date of Robinia pseudoacacia L.was influenced by both photoperiod and temperature.Besides,TPa model generally outperformed the mean value model at 14 representative phenological observation stations in the North China Plain.In addition,based on the parameterization results of TPa model,we set 6.5℃as the base temperature and 4.5h as the base daylength to separately calculate the actual effective accumulated temperature and daylength of the first flowering date of Robinia pseudoacacia L.in the past 56 years.The results showed that there was a significant increasing trend of the actual effective accumulated temperature,with an average increase rate of 4.5℃·d per decade(P<0.05),while the actual effective accumulated daylength showed an extremely significant decreasing trend,with an average decrease rate

关 键 词：光周期 温度 刺槐 始花期 物候模型 

分 类 号：S792.27[农业科学—林木遗传育种]