近30 a甘肃省河东地区极端气温指数时空变化特征及趋势预测

基于甘肃省河东地区61个气象站点1988—2017年逐日气温数据,利用Mann-Kendall检验,Sen’s斜率估计方法分析甘肃省河东地区极端气温指数的时空变化趋势,并探讨极端气温指数与其影响因素之间的关系,最后利用NAR神经网络结合Hurst指数对甘肃省河东地区极端气温指数变化进行预测分析。结果表明:(1)从时间上看,冷极值相对指数呈下降趋势,冷极值绝对指数、暖极值以及气温日较差、作物生长期呈上升趋势。(2)从空间上看,对冷极值变化反应最为敏感的是高寒湿润区,对暖极值变化反应最为敏感的是温带半湿润区和北亚热带湿润区,除北亚热带湿润区外各区域作物生长期的变化都达到了显著水平,而气温日较差仅在温带半湿润区达到了显著水平。(3)多数极端气温指数与经纬度、海拔之间有显著相关性,但受区域自然特点影响,经度与海拔对其影响实为一类。(4)亚洲区极涡强度、北半球极涡强度以及青藏高原指数B与极端气温指数变化有密切关系,而太阳黑子等只与个别指数之间存在显著的相关性。(5)预测出的极端气温指数冷极值相对指数仍呈现下降趋势,冷极值的绝对指数、暖极值以及气温日较差、作物生长期仍然呈现增加趋势,但大多数指数与...

Quantifying and predicting spatial and temporalvariations in extreme temperatures since 1990 in Gansu Province,China

We assembled a database of daily temperature data from61 meteorological stations in the Hedong region of Gansu Province,China from1988 to 2017.We used the Mann Kendall test and the Sen’s slope estimation method to analyze the patterns of spatiotemporalchanges in extreme temperature indices in the Hedong region of Gansu.Inaddition,we explored the relationship between extreme temperatures and theirinfluencing factors.Finally,we applied the NAR neural network and Hurst indexto predict and analyze extreme temperature index changes in Hedong.Our results yieldedsix major findings:(1) the relative index of cold extreme values has declinedover time.The absolute indices of cold and warm extreme values are greater inmagnitude,and the temperature index is worse.As a result,the crop growth periodis increasing.(2) The areas that are most sensitive to cold extreme changes arethe alpine humid regions,while temperate semi humidzones and northern subtropical regions are the most sensitive to warm extremechanges.With the exception of the northern subtropical humid region,the cropgrowth season changes were significant in all areas.However,diurnal temperaturedifferences reached significance only in temperate semi humid regions.(3) The most extreme temperature indices weresignificantly correlated with longitude,latitude,and altitude,but were alsoaffected by regional environmental characteristics.(4) The intensity of thepolar vortex in Asia,the northern hemisphere polar vortex intensity,and theTibetan Plateau Index B were allclosely related to changes in the extreme temperature index.(5) Our predictedextreme temperature index matched the observed downward trend for the relativeindex of cold extreme values.It also predicted worsening of extreme heat andcold,and temperatures.Finally,it correctly showed lengthened growingseasons.However,most of the indices deviated slightly from the 1988-2017observational dataset.(6) Compared with other regions,most of the temperatureindex changes in Hedong were at intermediate levels,reflecting locally diverseresponses in the region’s various climate zones.