行星波对北半球夏季空间复合热浪的影响：物理机制和S2S预报误差来源分析

【作者】 雷蕾；

【导师】 徐邦琪；

【作者基本信息】 南京信息工程大学 ， 气象学， 2024， 硕士

【摘要】 空间复合热浪事件指的是同时在多个地区发生热浪,相比于单一区域极端天气气候事件,空间复合事件往往造成更大的灾害。为了更好地理解和预报此类事件,本研究基于欧洲中期天气预报中心(ECMWF)的最新再分析数据(ERA5)和参与次季节–季节(Subseasonal to Seasonal,S2S)预测项目的回报数据集,研究北半球空间复合热浪事件的机理和次季节预测的机会窗口。本研究通过定义阈值和利用区域依赖性检验方法实现对空间复合热浪事件的客观识别,分析诊断了中纬度次季节尺度行星波与热浪事件之间的联系,并系统性评估了S2S业务模式对热浪事件的预测技巧,明确其次季节预报误差来源。主要结论如下:(1)考虑了区域气候特征、行星波的空间尺度以及高温热浪事件的高发热点地区后,将北半球中高纬度地区划分为9个区域,分别为北美西部、北美东部、格陵兰岛-冰岛、欧洲、俄罗斯-北极、亚洲西部、亚洲中部、东亚、俄罗斯-远东地区。基于以上分区方案,本研究依据百分位阈值法以及热浪范围阈值定义了区域热浪,并将空间复合热浪事件定义为同一天至少有2个及以上区域发生热浪,实现了在逐日尺度上对事件的客观识别。1979–2022年夏季,空间复合热浪日共有2145天,共有325种不同的热浪发生区域组合,经过区域依赖性检验后辨识出两种典型的区域热浪组合,分别为北美西部-欧洲的区域复合、欧洲-东亚的区域复合,一共出现249天、252天,均呈现显著增加趋势。(2)空间复合热浪事件与中纬度次季节行星波的活动高度相关,其中次季节5–7波为热浪事件发生期间的主导波动。合成分析和温度方程诊断的结果表明,次季节5–7波使同一纬度带上多个区域受异常高压环流控制,而与异常高压环流有关的下沉绝热增温、辐射和感热的非绝热加热过程,以及高压西侧的南风输送的暖平流,是导致多个地区同时升温的重要物理机制。波动振幅越强时,区域发生热浪的概率越高;当波动移速越慢,也越有利于事件维持。(3)评估ECMWF S2S模式对空间复合热浪事件和次季节行星波的预测技巧,发现ECMWF S2S模式能提前15–20天有效预报空间复合热浪事件,其对次季节5–7波的预报技巧在20天左右。提前1–6候预报时,模式能否成功预报两种典型区域热浪组合事件发生与模式对5–7波的预报性能密切相关,模式在正确预报事件发生时预报波动的误差小于漏报个例,证实了5–7波是事件预报误差的来源之一。然而受限于模式对行星波的预测技巧,约20天之后模式对事件的预报准确度较低。在对两种典型区域组合事件的预报中,多数情况下振幅对热浪事件的预报更重要,位相决定了发生热浪的区域,而振幅的强度决定了事件能否发生。更多还原

【Abstract】 Spatially concurrent heatwaves refer to heatwaves occurring simultaneously in multiple regions,which can cause greater disasters than regional extreme weather and climate events.In order to better understand and forecast such heatwaves,based on the latest reanalysis data(ERA5)of the European Centre for Medium-Range Weather Forecasts(ECMWF)and its operational model’s reforecast dataset as part of the subseasonal-to-seasonal prediction(S2S)project,we investigate the mechanism of the heatwaves,as well as the ‘windows of opportunity’for subseasonal forecasting in the Northern Hemisphere.The objective identification of spatial compound heatwaves is achieved by defining thresholds and using the region-dependence test method.We analyze not only the relationship between subseasonal mid-latitude waves and heatwaves,but also systematically evaluate the subseasonal prediction of the S2 S model for those heatwaves for clarifying the sources of the subseasonal forecasting errors.The main conclusions are as follows:(1)By focusing on the regional climate characteristics,the spatial scale of subseasonal planetary waves and the hotspots of heatwaves,we divided the land in the mid to high latitudes of the Northern Hemisphere into 9 regions: Western North America,Eastern North America,Greenland-Iceland,Europe,Russia-Arctic,Western Asia,Central Asia,East Asia,and RussiaFar East.The regional heatwaves were identified using the percentile threshold method and a heatwave range threshold,then,spatially concurrent heatwaves where at least two or more areas experience heatwaves on the same day were defined.This approach allows for the objective identification of events on a daily scale.During the summers of 1979–2022,there were 2145 days with spatially concurrent heatwaves,resulting in 325 distinct combinations of regions experiencing heatwaves.Two typical pairs of spatially compound heatwaves(involving two key regions experiencing heatwaves concurrently)passed the region-dependence test.These pairs are namely ‘Western North America-Europe’ pair and ‘Europe-East Asia’ concurrent heatwave cases,with occurrences accumulating 249 days and 252 days respectively.Both pairs show a significant increasing trend.(2)Spatially concurrent heatwaves are highly correlated with the activity of mid-latitude planetary waves,particularly those subseasonal zonal wavenumbers 5–7(referred to as Waves5–7),which are predominant during these heatwave events.Composite analysis and temperature equation diagnosis show that Waves 5–7 can cause multiple regions along the same latitude to be controlled by abnormal high-pressure circulation,leading to synchronous heating in multiple regions due to related sinking adiabatic warming,radiant and sensible heating processes,and warm advection transported by the southerly wind.The stronger the wave amplitude,the higher the probability of regional heatwaves occurrence.Moreover,the slower the phase-speed of waves,the more favorable they are for the persistence of these events.(3)Through the evaluation of the prediction skills for spatially concurrent heatwaves and planetary waves using the reforecast data of ECMWF S2 S model,it was found that the ECMWF S2 S model can effectively predict events and subseasonal Waves 5–7 at the lead times of 15–20 days and 20 days,respectively.When the model predicts 1–6 pentads in advance,its ability to successfully predict the occurrence of two typical regional heatwaves pairs is closely related to the prediction errors of subseasonal Waves 5–7.In cases when the model correctly predicts the events,the prediction errors of waves are smaller than in cases when the events are missed,indicating that subseasonal Waves 5–7 are one of the sources of forecasting errors for event prediction.However,due to the limited forecast skills for planetary waves at the lead time of20 days or longer,the model has poor skills to predict the events.In the forecast of events involving the aforementioned two typical pairs,the amplitude of the wave is mostly more important for accurately predicting the occurrence of event.The phase of the wave determines the area where the heatwaves occur,while the magnitude of the amplitude determines the likelihood of the events’ occurrence.更多还原