西藏林芝苹果花期与气象条件的关系及预测模型

Relationship between apple’s flowering phase and meteorological conditions and prediction model in Nyingchi of Tibet

【目的】西藏是中国气候资源最多样与果树资源最丰富的区域之一,分析西藏最适宜栽培区——林芝的苹果花期与气象条件的关系及其对气候变化的响应,选取主导影响气候因子,构建花期预测回归模型,为当地应对气候变化和开展苹果花期气象服务提供科技支撑。【方法】利用2001—2021年林芝苹果花期物候和逐日平均气温(Tm)、最高气温(Tmax)、最低气温(Tmin)、相对湿度(RH)、降水量(Pr)、日照时数(S)和≥0℃积温(∑T0)等数据,采用线性倾向估计法、Pearson相关系数和逐步回归方法等,分析苹果花期物候及其长度的变化趋势,以及气候因子、晚霜冻对苹果花期的影响,并开展苹果花期预测,构建花期预测回归模型。【结果】2001—2021年林芝苹果各花期以0.884~1.132d·a-1的速度呈推迟趋势,以花芽膨大期推迟的最多,开花始期次之,为1.058d·a-1。花期长度为显著缩短趋势,平均每年缩短0.471d。花期生长期内Tm、Tmax、Tmin均趋于升高,Tmin升幅大于Tmax升幅;DTR、S、RH趋于减少。始花期之前,Pr为增加趋势,之后Pr趋于减少。花期长度生长期内,只有Tm、Tmin表现为上升趋势,其他要素均趋于下降,以∑T0减幅最为明显。【结论】上年10月Pr是影响林芝苹果始花期、盛花期变化的主导气候因子,若上年10月Pr每增加(减少)10mm,苹果始花期和盛花期将分别提前(推迟)2.47d和2.17d。花期晚霜冻日趋于提前,且低温降低,但开花期明显推迟,花期霜冻的气候风险变小。基于旬、月气候要素的统计模型,预测苹果花期精准度为95.8%~97.1%。

【Objective】Tibet has one of the most diverse climatic resources and the richest fruit tree re-sources in China.Nyingchi is the most suitable cultivation area for the growth of fruit trees.This study aimed to establish a regression model for flowering prediction with the leading climatic factors influenc-ing apple growth in Nyingchi,through analysis of the relationship between the apple flowering phases and meteorological conditions and its response to climate change,in order to provide scientific and tech-nological support for adaptation of apple production to climate change,together with prediction servic-es for apple flowering phases over the Tibetan Plateau.【Methods】Both phenology and meteorological data were collected in Nyingchi station from 2001 to 2021.The apple’s flowering phases included the flower bud enlargement date(FED),flower bud opening date(FOD),initial flowering date(IFD),full blooming date(FBD)and terminal flowering date(TFD),while the meteorological data included daily average temperature(Tm),maximum temperature(Tmax),minimum temperature(Tmin),relative humidity(RH),precipitation(Pr),sunshine hours(S)and≥0℃cumulative temperature(∑T0)etc..The trends of apple flowering and the influences of climatic factors on the flowering phases were revealed with rel-evant statistical methods,including the linear propensity estimation method,the Pearson correlation co-efficient,and the stepwise regression method,and the apple flowering prediction were carried out ac-cordingly.【Results】(1)The apple flowering phases had been delayed at a rate of 0.884-1.132 d·a-1 in Nyingchi of Tibet from 2001 to 2021,among them the flower bud enlargement dates were delayed the most,followed by IFD at a rate of 1.058 d·a-1(p<0.01),and TFD the least.The length of flowering pe-riod was shortened significantly by a rate of 0.471 d d·a-1(p<0.05).While the IFD advanced signifi-cantly in the main apple producing areas in China such as Xinjiang,Loess Plateau and Bohai Gulf,the IFD postponed prominently in Nyingchi of Tibet.The main reason for this difference was that apple re-quirement for cold temperature in winter could always be met even under warming conditions,while the increased temperature could accelerate the apple growth rate in the other regions.In contrast,apple requirement for cold temperature in winter would not be met successfully over the Tibetan Plateau un-der the warming context,meaning that the additional temperatures lower than 7.2℃in spring had to be used for accumulation of low temperature to meet the chilling requirement,which inevitably led to the postponed flowering phase in Nyingchi of Tibet.(2)During the flowering periods,all temperature items,including Tm,Tmax,and Tmin,presented the increasing trends by a rate of 0.050-0.125℃·a-1,among them Tmin exhibited a greater increasing rate than Tmax.In contrast,all the other meteorological factors such as DTR,S,and RH decreased with a rate of 0.024-0.125℃·a-1,0.099-3.495 h·a-1,and 0.014-0.153%·a-1,respectively.(3)The Pr before IFD showed an increasing trend with a rate of 0.045-0.131 mm·a-1,and tended to decrease afterwards by a rate of 0.094-0.627 mm·a-1.(4)The∑T0 tended to increase from flower bud opening date to IFD and from IFD to FBD,but decreased from FBD to TFD with a rate of-4.695℃·d·a-1.(5)During the flowering length growth period,only Tm and Tmin showed an increasing trend,while the other elements exhibited the decreasing trends with the most ob-vious in∑T0 by a rate of-3.748℃·d·a-1.(6)The RH,S and∑T0 during the flowering length growth period had fluctuated greatly in the last 21 years.These fluctuations showed a three-curve trend,the RH presented a general“down-up-down”pattern,and the change rate in the phases of 2001 to 2006,2006 to 2017 and 2017 to 2021 was-1.12%·a-1,1.42%·a-1(p<0.01)and-4.25%·a-1,respectively.In con-trast,the S and∑T0 showed an“increase-decrease-increase”trend,among them the S changed with a rate of 21.66 h·a-1(p<0.05),-11.14 h·a-1(p<0.05)and 7.98 h·a-1 in 2001 to 2006,2006 to 2017 and 2017 to 2021,and the∑T0 changed with a rate of 35.78,-17.25 and 17.20℃·d·a-1in the same period,respectively.The Pr showed a quadratic curve variation of increasing and then decreased,with an in-creasing trend of 2.91 mm·10 a-1 in the first ten years of the 21st century and an decreasing rate of-1.53 mm·10 a-1 during 2011 to 2021.【Conclusion】(1)The temperature,especially in winter time,has in-creased significantly in Nyingchi of Tibet in the recent 60 years,which resulted in the insufficient cold demand for apples in the winter and delaying the process of breaking of bud dormancy,and delaying the dates.(2)The Pr in October of the previous year was identified as the dominant climatic factor influ-encing the change of IFD and FBD.In addition,the average ground temperature under 10 cm in March mainly affected the TFD.(3)Only individual flowering phase and growing period lengths were nega-tively correlated with the Tm significantly,while apple flowering phase was positively correlated with the ten-day Tg(5,10,and 20 cm),and the contribution of ground temperature was greater than that of air temperature in the flowering prediction regression model.(4)Both the and RH had significant ef-fects on apple flowering phases,mainly negative effects.The also showed a negative correlation with the flowering phase,but the effect of the S on the flowering phase was less than that of Pr.(5)Although the late frost tended to advance during apple flowering phase and its Tmin decreased,the flowering were significantly delayed,so the climatic risk of the frost for apple’s FD would become less.(6)Based on the climatic elements at decadal and monthly scales,the established statistical model could achieve an accuracy ranging from 95.8%to 97.1%,which seems to be possible to provide practical information on the prediction of the apple flowering phase in Nyingchi of Tibet.