1 氣候預(yù)測理論與方法

1.1 南亞高壓強(qiáng)度和位置變化的2個主導(dǎo)模態(tài)及其形成機(jī)制

揭示了5月南亞高壓強(qiáng)度和位置變化的2個主導(dǎo)模態(tài)，提出了ENSO的持續(xù)性和印度洋的海溫作用對上述2種模態(tài)的影響和機(jī)理。5月南亞高壓的形態(tài)變異將亞洲夏季風(fēng)爆發(fā)的各個階段連接成一個完整接續(xù)的動力過程。本工作根據(jù)觀測資料定義了5月南亞高壓的強(qiáng)度、東西伸脊點(diǎn)和脊線位置，并利用主成分分析（PCA）方法提取了5月南亞高壓的主導(dǎo)模態(tài)。結(jié)果表明，5月南亞高壓的年際變化具有兩種主導(dǎo)模態(tài)：一種是強(qiáng)度模態(tài)，表現(xiàn)為南亞高壓的強(qiáng)度和緯向跨度的一致變化，另一種是經(jīng)向位置模態(tài)，表現(xiàn)為南亞高壓脊線經(jīng)向位置的年際差異。

雖然5月南亞高壓的強(qiáng)度和經(jīng)向位置模態(tài)都受前冬ENSO事件影響，但ENSO事件影響南亞高壓各主導(dǎo)模態(tài)的物理過程卻并不相同。盡管前冬ENSO事件在春季衰減，其信號仍能通過“大氣橋”轉(zhuǎn)移到熱帶印度洋地區(qū)，并通過印度洋海溫異常（SSTA）來影響周邊的大氣環(huán)流異常。觀測和數(shù)值模擬結(jié)果表明，印度洋和太平洋SSTA影響南亞高壓主導(dǎo)模態(tài)的相對貢獻(xiàn)卻與ENSO事件在春季的衰減快慢有關(guān)。當(dāng)前冬ENSO事件在春季衰減較快時，其引起的5月印度洋SSTA能夠改變亞洲南部局地對流強(qiáng)度，其激發(fā)的Rossby波響應(yīng)位于印度西南部上空，令南亞高壓強(qiáng)度模態(tài)發(fā)生變化；當(dāng)前冬ENSO事件衰減較慢時，5月印度洋SSTA始終受ENSO引起的大氣環(huán)流異常影響，因此來自熱帶太平洋的SSTA通過改變亞洲南部地區(qū)的垂直運(yùn)動，進(jìn)而調(diào)整局地經(jīng)向溫度梯度，并影響南亞高壓的經(jīng)向位置模態(tài)。本研究結(jié)果說明，除ENSO事件的冷、暖位相外，它對亞洲夏季風(fēng)高空環(huán)流的影響還隨其在春季的演變過程不同而改變（圖1）。（劉伯奇，祝從文）

1.2 2015/2016年超強(qiáng)El Ni?o事件次年盛夏西太平洋副高異常偏弱的成因

以往研究認(rèn)為，當(dāng)前冬發(fā)生了El Ni?o事件后，次年夏季西太平洋副高將異常偏強(qiáng)，這主要和El Ni?o引起的西北太平洋海—?dú)庀嗷プ饔煤蜔釒в《妊蠛Ｅ璩叨犬惓Ｔ雠嘘P(guān)。但最新的觀測指出，在2015/2016年冬季超強(qiáng)El Ni?o事件發(fā)生后，2016年7、8月西太平洋副高異常偏弱，這與已有的ENSO-西太平洋副高關(guān)系不符。本研究通過資料分析和數(shù)值試驗(yàn)表明，與歷史事件相比，2015/2016年冬季超強(qiáng)El Ni?o事件衰減更快，令次年夏季印度洋無法充分加熱，因此印度洋暖SSTA無法在盛夏維持，從而削弱了印度洋暖SSTA通過局地緯向環(huán)流異常來加強(qiáng)西太平洋副高的作用。同時，來自中緯度的對流層中上部波列在2016年夏季能夠南傳到達(dá)西太平洋地區(qū)，進(jìn)一步削弱了西太平洋副高的強(qiáng)度。（劉伯奇，祝從文，蘇京志，華麗娟）

1.3 全球變暖已由趨緩期轉(zhuǎn)為加速期

自1999年至2012年間，全球平均表面溫度（GMST）升溫速率較先前幾十年偏低，此即“全球變暖停滯/趨緩”現(xiàn)象。但“變暖趨緩”何時結(jié)束，此前存在較大爭議。然而在2014年、2015年和2016年GMST相繼創(chuàng)下歷史最高記錄。我們研究表明，GMST這種“三級跳”式破紀(jì)錄攀升現(xiàn)象尚屬首次，這標(biāo)志著全球變暖由“趨緩期”轉(zhuǎn)為“加速期”。對排名前十的年平均高溫記錄在各年代的發(fā)生頻次（NT10）統(tǒng)計顯示，NT10在1960年代后節(jié)節(jié)攀升，即便是在“趨緩期”也顯著增加，這表明全球變暖的持續(xù)增溫趨勢始終沒有改變。就機(jī)理而言，太平洋年代際振蕩（PDO）自2013年開始由負(fù)位相轉(zhuǎn)為正位相，這有利于全球表面溫度升溫速率重新加速。變暖趨緩期間，海洋吸收了更多的熱量，這些熱量的釋放將導(dǎo)致氣溫升高。2014/2015年超強(qiáng)厄爾尼諾進(jìn)一步加劇了全球溫度升高。厄爾尼諾對滯后3個月GMST的影響最顯著，因此冬季年（6月至次年5月）平均的GMST年際變率與NINO3區(qū)溫度有著很好的關(guān)聯(lián)。然而，導(dǎo)致這幾次破紀(jì)錄高溫的最根本的驅(qū)動力，仍來源于全球變暖持續(xù)升溫的背景趨勢。對未來幾十年情景預(yù)估表明，全球氣溫將會持續(xù)維持歷史高位，破紀(jì)錄歷史高溫將頻繁出現(xiàn)，并形成一種新常態(tài)（圖2）。（蘇京志）

1.4 天山中部地區(qū)夏季降水量變化的氣候特征

研究了天山中部地區(qū)夏季降水量日變化的氣候特征?；谥鹦r降雨量數(shù)據(jù)和衛(wèi)星觀測所得對流指數(shù)的分析表明：此區(qū)域有3個顯著特征，即南部山區(qū)的清晨高峰值，山區(qū)傍晚的高峰值，以及北部山區(qū)夜間的高峰值。進(jìn)一步分析了這些日變化特征之間的關(guān)系。通過定義區(qū)域降水事件（RRE），記錄每個RRE的初始位置。在南部盆地南部地區(qū)，山脈南部的早期降雨是由局地所引起的。山區(qū)傍晚的降水峰值和北部山區(qū)的夜間降水峰值都受到了山區(qū)降雨事件的影響。這些降水事件在下午出現(xiàn)在山區(qū)，其中一些會向北移動，導(dǎo)致北部盆地的夜間降雨。山區(qū)下午對流觸發(fā)以及在南部盆地的清晨降水都與山脈周圍的晝夜變化的風(fēng)和熱力學(xué)條件有關(guān)。在山脈（南部盆地）的對流形成之前的中午（晚上），熱力不穩(wěn)定相伴隨的低層輻合對流出現(xiàn)（圖3）。（李建）

1.5 青藏高原低渦對西南渦生成的影響

西南渦產(chǎn)生于青藏高原東側(cè)的我國西南地區(qū)，定義在700 hPa上，是影響我國西南乃至東部廣大地區(qū)降水的重要天氣系統(tǒng)。青藏高原低渦是產(chǎn)生于青藏高原主體上的500 hPa天氣系統(tǒng)，它們的移出能夠?qū)ξ髂蠝u的活動產(chǎn)生影響。然而，目前關(guān)于移出的青藏高原低渦對西南渦生成的影響的研究很少。本文基于FNL資料對2000—2011年移出型高原低渦對西南渦的生成的影響進(jìn)行了研究。研究選擇了3類過程：第1類，有高原低渦伴隨下的西南渦生成過程；第2類，僅有高原低渦，沒有西南渦生成過程；第3類，沒有高原渦伴隨下的西南渦生成過程。每個過程包含9個個例，研究分別對這3個過程進(jìn)行了分析，并進(jìn)一步通過對比指出了移出型高原低渦在西南渦生成過程中的作用。結(jié)果指出，凝結(jié)潛熱加熱對西南渦是西南渦生成的決定性因素，而移出型高原低渦能夠在西南地區(qū)創(chuàng)造更加有利的降水條件，從而有利于西南渦的生成。有高原低渦伴隨下的西南渦生成后，強(qiáng)度更大，維持時間更長。但是高原低渦并非西南渦產(chǎn)生的充分條件，強(qiáng)烈的西南氣流及相應(yīng)的水汽輸送對西南渦對的產(chǎn)生起到了關(guān)鍵作用。（李論）

1.6 中國夏季極端暖事件的類型分析

根據(jù)白天最高溫度和夜間最低溫度的極端性及二者的組合情況，將中國夏季的暖事件細(xì)分為3種互不重疊的子類，即獨(dú)立白日型、獨(dú)立夜間型及混合型。這些子類的變化與傳統(tǒng)定義中的暖日、暖夜事件的變化在趨勢的顯著性、強(qiáng)度甚至是符號正負(fù)等方面均存在明顯差異，這些差異在包含極端白天溫度的子類中尤為明顯。按照上述分類，一些在傳統(tǒng)定義中被忽視的顯著變化被成功挖掘出來。如，獨(dú)立白日型的支配地位正在逐漸減弱，而混合型事件和獨(dú)立夜間型事件逐漸演化成主要的暖事件類型。這兩類暖事件具體表現(xiàn)為頻次顯著增長、持續(xù)時間顯著延長、影響范圍顯著擴(kuò)張，并且強(qiáng)度越強(qiáng)的事件增強(qiáng)的幅度越大。（陳陽，翟盤茂）

1.7 近20年中國東部地區(qū)的持久且較強(qiáng)的變暖減緩現(xiàn)象

在1997年以來，中國東部地區(qū)經(jīng)歷了一次變暖減緩過程，主要表現(xiàn)為早冬的最低溫度的顯著變冷。通過隨機(jī)組合起始年份，發(fā)現(xiàn)中國東部的“優(yōu)勢變暖減緩期”發(fā)生在1998—2013間，這段時間內(nèi)的區(qū)域平均最低溫度變冷的趨勢最強(qiáng)并且達(dá)到顯著變冷的站點(diǎn)數(shù)最多。對變暖減緩最為敏感的區(qū)域主要位于華北、江淮地區(qū)和華南，這些敏感區(qū)的顯著變冷趨勢一直持續(xù)到了2016年。這種持續(xù)的變暖減緩使得嚴(yán)重的冷事件在上述地區(qū)頻發(fā)。達(dá)到如此強(qiáng)度且持久的變暖減緩期在過去50年間都不曾發(fā)生，因此可以被視為氣候變暖大背景下的一個“奇異值”（圖4）。（陳陽，翟盤茂）

1.8 全球增暖導(dǎo)致盛夏華東高溫事件的發(fā)生概率增大

包括熱浪在內(nèi)的極端氣候事件的頻率、強(qiáng)度和持續(xù)時間正在發(fā)生著變化。極端氣候事件極容易引發(fā)災(zāi)害，對極端氣候事件的歸因是氣候?qū)W界關(guān)注的熱點(diǎn)方向，這其中的一個挑戰(zhàn)性問題是對單一極端氣候事件的歸因，原因在于單一極端事件既可能受到氣候變化的影響，又包含自然變率的信號。2013年7—8月，我國華東地區(qū)經(jīng)歷了1951年以來持續(xù)時間最長（長于40天）、強(qiáng)度最高（區(qū)域最高溫度44.1℃）的熱浪事件。圍繞著這次高溫事件的歸因問題，基于超級集合歸因模擬試驗(yàn)和耦合模式集合模擬結(jié)果，研究了氣候變化和自然內(nèi)部變率對該極端事件的影響，研究發(fā)現(xiàn)大氣自然變率和人類活動對這次熱浪事件均有貢獻(xiàn)；與這次熱浪事件直接相關(guān)的大氣環(huán)流型，表現(xiàn)為華東地區(qū)上空的異常正高壓，這種高壓型自身因大氣內(nèi)部變率而產(chǎn)生；人類活動影響顯著增加了類似2013年盛夏極端高溫事件的發(fā)生概率（圖5）。（馬雙梅）

2 氣候系統(tǒng)模式研發(fā)

2.1 氣候系統(tǒng)模式簡介

中國氣象科學(xué)研究院氣候系統(tǒng)模式（CAMS-CSM）得到進(jìn)一步發(fā)展完善?，F(xiàn)有版本CAMS-CSM包含了先進(jìn)的大氣模式ECHAM5（v5.4）、海洋模式MOM4、海冰模式SIS、陸地表面模式CoLM和FMS耦合器等組成分量。在大氣模式ECHAM5中引入了歐拉型“兩步保形平流方案”（TSPAS），并采用了跳點(diǎn)差分算法，模擬結(jié)果顯示新方案對東亞地區(qū)降水特別是青藏高原大地形周邊降水有明顯改善，顯著降低了陡峭地形區(qū)高海拔區(qū)域的降水。同時將大氣模式的輻射模塊替換為BCC_RAD輻射方案，模擬結(jié)果顯示新方案顯著改善了大氣頂?shù)妮椛淦胶?，同時明顯改善了對東亞地區(qū)短波云輻射狀況的模擬。在各模塊通量耦合過程中，針對大氣模式中半隱式垂直擴(kuò)散方案發(fā)展了一套新的守恒的通量計算和插值算法，保持了?！?dú)夂捅獨(dú)饨缑娴哪芰渴睾阋约澳Ｊ降姆€(wěn)定積分。通過改進(jìn)?！戇吔绺顸c(diǎn)的通量算法，保證了?！戇吔绺顸c(diǎn)的通量守恒，模式積分過程中海溫和海冰的長期變化趨勢明顯減弱。利用該版本CAMS-CSM，對工業(yè)革命前控制試驗(yàn)和歷史試驗(yàn)進(jìn)行積分模擬。通過分析上述試驗(yàn)結(jié)果，發(fā)現(xiàn)該版本CAMS-CSM很好地再現(xiàn)了氣候平均狀態(tài)和主要?dú)夂蛳到y(tǒng)的季節(jié)周期，包括海表溫度、降水、海冰范圍和海洋溫躍層等要素特征，并能合理再現(xiàn)主要的氣候變率模態(tài)，如Madden-Julian振蕩（MJO）、ENSO、東亞夏季風(fēng)（EASM）以及太平洋年代際振蕩（PDO）。特別是，該模型在模擬東亞夏季風(fēng)（EASM）變異性和ENSO-EASM關(guān)系上顯示了明顯優(yōu)勢。模擬結(jié)果中也存在幾種偏差，例如年平均降水形態(tài)中的雙ITCZ特征，高估了ENSO振幅，并且與ENSO相關(guān)的Bjerkness反饋機(jī)制強(qiáng)度較弱。

2.2 降水分布

總體上，模式模擬的年平均降水量的總體水平分布與觀測結(jié)果相符。在主要的降水中心，如赤道輻合帶（ITCZ）、南太平洋輻合帶（SPCZ）以及熱帶印度洋和熱帶大西洋的熱帶地區(qū)，模擬結(jié)果均得到很好體現(xiàn)。然而，與觀測值相比，這些降水中心的強(qiáng)度通常更大。另外，從季節(jié)循環(huán)來看，觀測的主要雨帶以及熱帶輻合帶（ITCZ）和南太平洋輻合帶（SPCZ）的季節(jié)性遷移都能被合理地模擬出來（圖略）。模擬的ITCZ向最北的位置移動，在7—8月達(dá)到了峰值。而SPCZ在2—3月的最南端位置是最強(qiáng)的，與觀測結(jié)果一致。在5—12月期間，ITCZ控制了熱帶地區(qū)的降水，而SPCZ在1—4月，這一季節(jié)性的時間特征在模型和觀測之間是一致的。

2.3 熱帶太平洋特征

ENSO是在年際時間尺度上最主要的氣候變率模態(tài)。在ENSO事件演變過程中，大氣和海洋均發(fā)生顯著異常變化，包括SST、降水、緯向風(fēng)以及溫躍層。溫躍層深度的變化會導(dǎo)致赤道東太平洋海溫的上升和混合，稱為溫躍層反饋機(jī)制，這在ENSO動力學(xué)中起著重要的作用。溫躍層反饋過程在赤道中東太平洋最為顯著，因?yàn)榇颂帨剀S層深度較淺。因此，溫躍層及其季節(jié)循環(huán)的特征對ENSO模擬至關(guān)重要。與觀測結(jié)果相比，該模型能很好地再現(xiàn)溫躍層深度。模擬的溫躍層在西太平洋較深，東太平洋較淺，表明模型中溫躍層的帶狀斜坡略強(qiáng)。ENSO事件通常在北半球冬季成熟，即季節(jié)鎖相性。已有研究指出，氣候狀態(tài)的季節(jié)變化對ENSO鎖相的作用至關(guān)重要，其中一個重要指標(biāo)即是赤道SST的季節(jié)周期。模式所模擬的赤道SST季節(jié)周期與觀測相一致。赤道東太平洋地區(qū)的年循環(huán)以及赤道西太平洋地區(qū)的半年循環(huán)都能夠得到很好的模擬。模擬的偏差主要位于東太平洋附近。例如，模擬的暖位相強(qiáng)度比觀測偏弱，而模擬的冷位相比觀測偏強(qiáng)。其次，模擬的冷位相季節(jié)演變超前觀測結(jié)果1～2個月（圖6）。

2.4 ENSO的空間形態(tài)

作為熱帶氣候系統(tǒng)最顯著的年際時間尺度變率信號，ENSO能夠影響整個熱帶太平洋以及印度洋等區(qū)域。從Nino3.4指數(shù)與SST異常之間的回歸形態(tài)來看（圖7a、c），與ENSO相關(guān)的SST空間形態(tài)能夠很好地被模擬出來。與HadISST觀測結(jié)果相比，由于模式中冷舌過度西伸，赤道中東太平洋SST正異常在子午線方向上過于狹窄，更向西延伸。在南中國海和熱帶印度洋的正相關(guān)區(qū)都能被模擬出來，并且澳大利亞西海岸的弱負(fù)相關(guān)區(qū)也能反映出來。Nino3.4指數(shù)與降水和850 hPa風(fēng)異常的回歸場（圖7b、d）進(jìn)一步顯示，該模式能夠充分模擬出與ENSO相關(guān)的降水和風(fēng)異常的空間形態(tài)。例如，在赤道太平洋中部的降水增加，這是由于暖SSTA導(dǎo)致熱帶對流增強(qiáng)所致。同時，異常西風(fēng)占據(jù)了赤道中東太平洋（150°E～120°W），表明大氣與海洋的反饋機(jī)制。降水負(fù)異常分布在東印度洋和冷海溫區(qū)域，此處對流活動被抑制。上述太平洋和印度洋上的降水異常中心和低空風(fēng)異常的量值幅度都能被模式所刻畫。例如，模式能夠模擬出降水異常和低層風(fēng)異常的觀測所得的不對稱特征：降水異常和低層風(fēng)異常通常在赤道以南最大。

2.5 東亞夏季風(fēng)變率

通常用降水和風(fēng)場來指示東亞夏季風(fēng)（EASM）的變率。這里評估主要集中在與EASM變異性相關(guān)的降水和風(fēng)場變化方面。在此采用Wang and Fan（1999）的季風(fēng)指數(shù)（WFI），這是從850 hPa緯向風(fēng)切變來定義的：WFI=U850（5°～ 15°N，90°～ 130°E）?U850（22.5°～ 32.5°N，110°～ 140°E）。圖8給出負(fù)WFI指數(shù)與850 hPa風(fēng)異常的回歸場。觀測結(jié)果中，環(huán)流形態(tài)為具有典型的反氣旋特征，即在中國東南部和沿長江和日本東南部為西南風(fēng)異常，以及從西太平洋延伸至南中國海的東風(fēng)異常。降水異常增強(qiáng)主要位于東亞副熱帶鋒，降水異常減弱主要位于反氣旋的南支附近。整體來說，模擬結(jié)果能夠很好地體現(xiàn)出觀測的形態(tài)，在反氣旋的北部（南部）降水增強(qiáng)（減弱）的降水形態(tài)能夠模擬出來。特別地，模式能夠模擬出與梅雨帶變化相關(guān)的異常降水中心。然而，模擬結(jié)果也存在一些不足之處。例如，降水增強(qiáng)中心的強(qiáng)度通常比GPCP的結(jié)果偏弱，而西太平洋的降水減弱中心相比觀測結(jié)果偏強(qiáng)。此外，南中國海的負(fù)降水中心模擬結(jié)果偏弱，中南半島的負(fù)降水中心在模擬結(jié)果中也偏移到更南方。

ENSO對東亞夏季風(fēng)的影響方面，該模式的模擬性能有著顯著優(yōu)勢。厄爾尼諾現(xiàn)象在衰減的夏季期間，西北太平洋通常為反氣旋的特征（圖9），即為西太平洋反氣旋（WNPAC）。WNPAC起源于厄爾尼諾發(fā)展階段的秋季，并可能維持至厄爾尼諾事件衰減年的夏季，這是ENSO影響中國夏季降水的一種重要機(jī)制。雖然WNPAC持續(xù)到厄爾尼諾事件次年的夏季的機(jī)制存在爭議，但它可能與西北太平洋、印度洋以及熱帶北大西洋等相關(guān)聯(lián)。在模式模擬結(jié)果中，ENSO對東亞夏季風(fēng)的影響形態(tài)與觀測分析結(jié)果相符，這表明EASM變率的主模式與ENSO密切相關(guān)。該模式成功地再現(xiàn)了降水和850 hPa風(fēng)場與ENSO相關(guān)形態(tài)。從模擬結(jié)果可以清楚地看出，WNP存在反氣旋環(huán)流。與GPCP數(shù)據(jù)類似，模擬的梅雨帶從中國東部一直延伸到日本東南部，這表明該模式在模擬ENSO-EASM關(guān)系方面具有相當(dāng)好的能力。（容新堯，李建，陳昊明，辛羽飛，蘇京志，華麗娟，齊艷軍等）

3 極地研究

3.1 基于EAGLE氣象站自動監(jiān)測記錄和雪層定年的南極內(nèi)陸地區(qū)風(fēng)對降雪、物質(zhì)積累等過程的影響

結(jié)合氣象站4 m高度氣溫記錄，我們計算了Eagle雪坑各雪層沉降時的氣溫，并對比了同期氧同位素比率記錄。結(jié)果發(fā)現(xiàn)，雖然δ18O和氣溫總趨勢一致，上部雪層的δ18O值和氣溫的變化規(guī)律非常相似，但在50 cm深度以下，定年結(jié)果和實(shí)際降雪日期有約6個月的誤差，這可能是由雪層混合作用引起的。對同期δ18O值和氣溫做相關(guān)分析，發(fā)現(xiàn)其相關(guān)系數(shù)只有0.33，相關(guān)較低的值主要集中在2006 年間，這可能是在2006年的強(qiáng)風(fēng)背景下，雪層受到再搬運(yùn)作用的影響較強(qiáng)（搬運(yùn)過來的雪的沉降環(huán)境無法確定）。由此可以看出，風(fēng)在雪積累過程中具有極其重要的作用，其搬運(yùn)、堆積以及加速升華等作用可以造成雪層的丟失或者異常積累，并在雪坑化學(xué)記錄中留下干擾信號，影響了我們的判斷，即短期的雪層化學(xué)記錄可能受到了沉降后過程的影響，可信度有限。（丁明虎）

3.2 FELIX-S模型的構(gòu)建及其與MISMIP3D計劃的模式性能的比較

與美國CESM地球系統(tǒng)模式下的LIWG（陸冰模式課題組）積極開展相關(guān)合作，參與其中的full-Stokes模型（FELIX-S）開發(fā)工作，構(gòu)建了三維冰蓋和冰架系統(tǒng)的full-Stokes冰流模式，包括接地線動力學(xué)過程、網(wǎng)格刻畫等。該模式是世界率先融合冰蓋和冰架2個冰凍圈重要分量于一體的模型，突破了以往將冰蓋模型和冰架模型分別描述的局限，是冰體動力學(xué)描述方法的重要突破。本研究同時將其與參加MISMIP3D計劃的所有模式在接地線和冰架底部過程的模擬性能進(jìn)行了對比和適用性討論。（張通）

3.3 南極衛(wèi)星臭氧總量產(chǎn)品的評估

根據(jù)1993—2015年南極中山站Brewer光譜儀臭氧總量測值， 分析比較不同時期衛(wèi)星探測反演的大氣臭氧總量誤差特征。結(jié)果表明，衛(wèi)星測值總體偏高， 這與以南、北半球中緯度為主的全球比對結(jié)果（衛(wèi)星測值總體偏低）不同， 但誤差沒有超過4%。對同一顆衛(wèi)星一天多次過境測值的選取中， 注意到太陽天頂角（SZA）最低時的測值與地基一致性最好（平均誤差為?0.02%～1.15%）。TOMS算法反演的臭氧總量（含SBUV、TOMS-EP、OMI-TOMS）與地基測值最接近， 其次是GOD-FIT法（以GOME-2A為代表）和DOAS-TOGOM法（含GOME、SCIAMACHY和OMI-DOAS）。衛(wèi)星臭氧總量誤差對SZA均有一定的依賴性∶ 當(dāng)SZA在60°～70°以上時DOAS、GOME-2A的臭氧總量誤差呈增加趨勢而TOMS則下降， 但在80°～85°時GOME-2A下降，衛(wèi)星測值在地基臭氧總量為300～350 DU時與地基測值最接近，DOAS-TOGOM和GOME-2A的誤差在300 DU以下時隨臭氧總量降低而呈增加趨勢。衛(wèi)星臭氧總量誤差對衛(wèi)星與地基在觀測時間上的差異呈一定的統(tǒng)計特點(diǎn)∶ 當(dāng)時間差別在4 h以上時誤差呈上升趨勢;在8 h時OMI-TOMS的誤差＞10%， 而9 h時DOAS-TOGOM誤差可達(dá)＞15%，但GOME-2A沒有超過10%。當(dāng)衛(wèi)星過境點(diǎn)與地基測點(diǎn)的距離在100 km以上時， 衛(wèi)星臭氧總量誤差可達(dá)-5%; 而當(dāng)TOMS-EP或OMI-TOMS的過境位置在中山站東南方的南極大陸上空時， 其臭氧總量總體偏低， 而在中山站西北方的海洋上空則相反， 可能反映了地表反射率差異對TOMS算法反演的影響?！俺粞醵础逼陂g衛(wèi)星臭氧總量與地基測值的一致性較非“臭氧洞”期間明顯降低，TOMS算法的衛(wèi)星臭氧總量誤差變化未超過1%/10a。1996—2015年中山站SBUV和Brewer的臭氧總量月距平變化趨勢分別為1%/10a和0.9%/10a， 表明臭氧層較一致的微弱恢復(fù)態(tài)勢。（張雷）

3.4 南極羅斯海的海冰變化趨勢

通過去趨勢方法分析了1979—2015年南極羅斯海、別林斯高晉海-阿蒙森海、威德爾海、印度洋和太平洋5個扇區(qū)的海冰范圍變化，發(fā)現(xiàn)僅有羅斯海的海冰范圍增長具有顯著性趨勢，其他4個海區(qū)的海冰范圍增長/減小均不顯著。此研究說明氣候變暖背景下的南極海冰變化仍然存在很大不確定性。（袁乃明，丁明虎）

3.5 2012年1—2月亞洲嚴(yán)寒事件及其與北極海冰減少的關(guān)系

通過對再分析資料的分析和數(shù)值模擬試驗(yàn)，揭示了2011年由夏季到冬季，持續(xù)性北極海冰異常以及北極夏季大氣環(huán)流對后期冬季亞洲大陸極端嚴(yán)寒事件的可能影響。研究發(fā)現(xiàn)，太平洋-阿留申區(qū)域以及歐亞大陸中部是兩個關(guān)鍵區(qū)域，這兩個區(qū)域大氣環(huán)流演變對發(fā)生在2012年1月中、下旬（2012年1月17日至2月1日）的亞洲大陸極端嚴(yán)寒事件有重要貢獻(xiàn)。在本次極端嚴(yán)寒事件爆發(fā)前期，阿留申區(qū)域海平面氣壓持續(xù)快速升高，當(dāng)該區(qū)域海平面氣壓開始減弱時，伴隨著極地阻塞高壓的出現(xiàn)以及西伯利亞高壓迅速加強(qiáng)，從而導(dǎo)致北極冷空氣在亞洲大陸的向南爆發(fā)。因此，阿留申區(qū)域大氣環(huán)流對本次極端嚴(yán)寒事件的影響即大氣環(huán)流的下游效應(yīng)起關(guān)鍵作用。數(shù)值模擬試驗(yàn)證明，2011年夏季北極大氣環(huán)流狀況，顯著地加強(qiáng)了北極海冰異常偏少對冬季阿留申區(qū)域和歐亞大陸中部大氣環(huán)流的負(fù)反饋?zhàn)饔?，?dǎo)致有利于2012年1月中、下旬極端嚴(yán)寒事件出現(xiàn)的大氣環(huán)流異常。研究表明，阿留申區(qū)域以及東北太平洋中緯度區(qū)域的擾動可能提供了前兆信號，該信號可以增加對歐亞大陸極端嚴(yán)寒事件季節(jié)內(nèi)演變的預(yù)測技巧。（武炳義）

3.6 冬奧會服務(wù)

自2015年11月起，針對冬奧會滑雪賽道保障及儲雪評估兩大難點(diǎn)，在河北崇禮和北京延慶賽區(qū)開展積雪觀測，并在人工造雪演變特征、賽道雪冰監(jiān)測方法等方面取得初步成果。針對賽道雪冰質(zhì)量核心問題（造雪、賽道雪坡制作、賽道雪冰監(jiān)測到賽道雪冰質(zhì)量預(yù)測）之一賽道雪坡準(zhǔn)備，通過文獻(xiàn)調(diào)研、冬奧組委座談、國際雪聯(lián)專家咨詢、實(shí)地調(diào)查等方式初步建立了可量化的科學(xué)解決方案。結(jié)合國外經(jīng)驗(yàn)及極地雪冰氣象監(jiān)測方案，提出了賽道雪冰質(zhì)量監(jiān)測方案，并在崇禮萬龍雪場開展屬地試驗(yàn)，進(jìn)展良好。結(jié)合極地冰—?dú)庀嗷プ饔媚Ｐ秃头e雪演變模型，初步建立了賽道雪冰質(zhì)量預(yù)測模型，并撰寫了運(yùn)行手冊。

3.7 極地業(yè)務(wù)工作

圓滿完成2017年度南極中山氣象臺和長城站氣象站地面氣象業(yè)務(wù)觀測任務(wù)，全年無錯報漏報；完成中山氣象臺臭氧總量監(jiān)測任務(wù)，并協(xié)助發(fā)布《南極臭氧公報》。超低溫自動氣象站研發(fā)取得新進(jìn)展，重建了LGB69、Dome A超低溫自動氣象站，獲取了2017年度連續(xù)監(jiān)測數(shù)據(jù)。

圖1 基于南亞高壓強(qiáng)度模態(tài)指數(shù)（a，b）和經(jīng)向位置模態(tài)指數(shù)（c，d）的5月150 hPa位勢高度合成場（gpm，黑色粗實(shí)線和陰影分別表示合成和氣候平均14270 gpm等高線，黑色虛線表示氣候平均高壓脊線，紅色和藍(lán)色虛線分別表示強(qiáng)度模態(tài)偏強(qiáng)和偏弱時的高壓脊線位置）Fig.1 Composites of 150 hPa geopotential height (solid contours, gpm) and ridgeline (dashed curves) of the South Asian High(SAH) with respect to (a, b) the intensity and (c, d) meridional position mode of the SAH on the interannual timescale.The bold solid contours are the 14,270 gpm geopotential height.The climatological 14,270 gpm geopotential height and SAH ridgeline are represented by gray shading and the black dashed curves, respectively

圖2 排名前十的年平均高溫記錄的發(fā)生頻次（NT10）在各年代的分布(每個年代的El Ni?o、La Ni?a和正常年份期間的NT10由不同顏色表示；細(xì)線為年平均GMST時間序列，粗線為其7年滑動平均)Fig.2 Relative frequency of the top 10 record (NT10) of annual mean temperatures for each decade.The NT10 is divided into three subgroups (El Ni?o, La Ni?a, and neutral years) as indicated as the legend.The NT10 frequency in the 2010s is weighted by 10/6.The number of El Ni?o, La Ni?a, and neutral conditions during a decade is indicated by the thick line in each bar.The annual mean GMST is plotted in thin black, and its seven-year moving average is plotted in thick black

圖3 垂直縱剖面的異常垂直運(yùn)動(顏色陰影；10-2 m/s)及相應(yīng)的垂直環(huán)流(經(jīng)向風(fēng)分量和垂直速度的100倍，m/s)在86o和87.5oE經(jīng)向平均值：(a)12∶00;(b)18∶00;(c)00∶00;(d)06∶00Fig.3 Vertical profile of the anomalous vertical motion (color shading; 10-2 m s-1) and the corresponding vertical circulation(Vectors having components of the meridional wind and 100 the vertical velocity; m s-1) longitudinally averaged between 86o E and 87.5o E at different times: (a) 12:00; (b) 18:00; (c) 00:00; (d) 06:00

圖4 （a）不同起始年份的區(qū)域平均最低溫度的趨勢（℃/10a）和(b)所有短期（10～20年）線性趨勢的概率密度函數(shù)（圖a中黑色正方形表明該趨勢至少在0.1的水平上顯著；圖b中紅色梯形為起始年份在1996年以前的線性趨勢的概率分布，藍(lán)色線為起始年份在1997年以后的線性趨勢的概率分布，3條垂直的虛線為所有早期（開始于1996年以前）的短期趨勢中的第10百分位，第5百分位和第1百分位）Fig.4 (a) Historical trends (℃ decade?1) for domain-averaged minimum temperature (Tmin) with different starting years and interval lengths.All trends significant at the 90% conf i dence level at least are enclosed by black squares.(b) PDFs of all shortterm (10?20 years) trends, which began before 1996 (red histograms) and after 1997 (blue curves).Three vertical dashed lines locate the lower 10th, 5th and 1st percentiles of all short-term trends during earlier period (starting before 1996)

圖5 2013年盛夏地表氣溫（SAT，℃，填色） 和500 hPa位勢高度異常（Z500, m，等值線）空間分布(a)； CMIP5(b)、CAM5.1(c)和MIROC5(d)模式模擬的中國中東部盛夏平均SAT異常的直方圖（柱狀圖）和PDF（曲線）（b～d中的紫色線對應(yīng)2013年盛夏中國中東部SAT異常的觀測值）Fig.5 (a) 2013 July–August mean surface air temperature (SAT) anomalies (℃, shaded) and geopotential height anomalies at 500 hPa (Z500, units in m, contours).Histogram (bars) and probability density functions (PDFs, curves) of July–August mean SAT anomalies averaged over Central and Eastern China derived from (b) CMIP5, (c) CAM5.1 and (d) MIROC5 simulations.The vertical purple lines in (b)–(d) are the observed 2013 July–August SAT anomaly

圖6 HadISST（a）和CAMS-CSM(b)歷史模擬的赤道海溫（℃，5°S～5°N平均）氣候平均年循環(huán)Fig.6 Annual cycle of equatorial SST (℃, averaged over 5°S?5°N) from (a) HadISST and (b) CAMS-CSM historical simulation

圖7 與Nino3.4指數(shù)回歸所得的冬季(DJF)平均的SST(℃，左圖)和降水(陰影，mm /d)和850 hPa風(fēng)異常(向量，m/s)(右圖)(上圖為觀察結(jié)果，下圖為CAMS-CSM歷史試驗(yàn)?zāi)M結(jié)果，其中SST、降水和850 hPa風(fēng)場分別來自于HadISST、GPCP和NCEP2數(shù)據(jù)(1980-2013))Fig.7 Winter (DJF) SST (℃, left panels), precipitation (shaded, mm day?1) and 850 hPa winds anomalies (vector, m/s) (right panels) regressed on the Nino-3.4 index from observation (upper panels) and CAMS-CSM piControl simulation (bottom panels).The SST, precipitation and 850 hPa observations are derived from the HadISST, GPCP and NCEP2 data (1980?2013),respectively

圖8 由負(fù)WFI指數(shù)回歸所得夏季降水(陰影，mm/d)和850 hPa風(fēng)異常(矢量，m/s)：(a)為觀測結(jié)果；(b)為CAMS-CSM歷史試驗(yàn)?zāi)M結(jié)果Fig.8 Summer precipitation(shaded, mm/day) and 850 hPa winds anomalies (vector, m s-1) regressed on the negative WFI index from (a) observation and (b) CAMS-CSM (PI) Control simulation

圖9 冬季Nino3指數(shù)(DJF0)與次年夏季(JJA1)降水(陰影，mm/d)和850 hPa風(fēng)(矢，m/s)異常的相關(guān)系數(shù)場：(a)為觀測結(jié)果；(b)為CAMS-CSM歷史試驗(yàn)?zāi)M結(jié)果（顯示超過95%的信度水平）Fig.9 Correlation between the winter Nino3 index (DJF0) and the following summer (JJA1) precipitation (shaded, mm/day)and 850 hPa wind (vector, m s?1) anomalies from (a) observationand (b) CAMS-CSM (PI) Control simulation.Conf i dence levels above 95% are shown

Progress in Climate System and Climate Change Research

1 Theory and methodology of climate prediction

1.1 Dominant modes of the South Asian High and their mechanism

We have found that two dominant modes exist in the interannual variation of the South Asian High(SAH) in May, in terms of its intensity and meridional position mode, respectively.Both of them are affected by the preceding ENSO events.The pattern variation of SAH in May integrates all onset processes of the Asian summer monsoon as a whole.We have objectively defined the indices of the SAH intensity, western and eastern extension and its meridional position.Two dominant modes were extracted using the principle component analysis (PCA) on the above SAH indices.One is the SAH intensity mode, which presents the uniform variation of the SAH strength and zonal expansion.The other is the SAH meridional position mode featured by the interannual changes of the SAH meridional position with little variation of the SAH intensity.

Although these two modes are affected by the preceding ENSO events, the physical mechanisms are distinct.In boreal spring, the ENSO event is decaying, but its influences on the atmospheric circulation can be memorized by the SSTA in the tropical Indian Ocean (TIO) via the “atmospheric bridge”.Both observation and numerical sensitivity experiments suggest that the relative contribution of SSTA in the tropical Pacific and TIO to the two modes depends on the decaying rate of ENSO event in boreal spring.On the one hand, when the ENSO damps faster, its resultant TIO SSTA can alter the tropical convection over South Asia to modify the SAH intensity mode by stimulating a Rossby wave response over the southwestern India.On the other hand, the SAH meridional position mode is regulated by the SSTA in the tropical Pacific when the ENSO event persists in boreal spring.In particular, the zonal SSTA gradient in the Indo-Pacific Ocean modulates the vertical motion over South Asia, leading to the variation of the meridional temperature gradient in situ.The present study indicates that the interannual variation of the SAH not only depends on the ENSO phase, but also on its temporal evolution in boreal spring (Fig.1).(Liu Boqi, Zhu Congwen)

1.2 Reason for the weaker WPSH in boreal late summer after the 2015/2016 super El Ni?o.

Previous studies have proposed that the summer WPSH should be enhanced after an El Ni?o event occurred in the preceding winter.This is partly due to the air-sea interaction over the western North Pacific(WNP), and partly associated with the warming SSTA in the tropical Indian Ocean (TIO) under the influences of El Ni?o.However, the latest observational studies indicate that the WPSH became weaker in late summer of 2016 after the 2015/2016 super El Ni?o, which contradicts the existing conceptual model.Compared with the situation in 1983 and 1998, both data analysis and numerical experiment showed that the contrast is primarily attributed to the much faster decaying of the first super El Ni?o in the 21st century.The fast-decaying El Ni?o prevented the warming of the TIO in the subsequent summer, which attenuated the anomalous zonal circulation originating from the TIO.Thus, the descending anomaly over the WNP weakened, corresponding to the weaker WPSH.In the meantime, TIO SSTA can alter the mid-latitudinal wave guide in the upper troposphere by changing the meridional temperature gradient.The colder SSTA in the TIO facilitated the southward intrusion of the upper-level wave train into the WNP, which further weakened the WPSH in late summer of 2016.(Liu Boqi, Zhu Congwen, Su Jingzhi, Hua Lijuan)

1.3 Consecutive record-breaking high temperatures marked the handover from hiatus to accelerated warming

Following the recent warming hiatus period, two astonishing high temperature records reached in 2014 and 2015 consecutively.To investigate the occurrence features of record-breaking high temperatures in recent years, a new index focusing the frequency of the top 10 high annual mean temperatures was defined in this study.Analyses based on this index showed that record-breaking high temperatures occurred over most regions of the globe with a salient increasing trend after 1960s, even during the so-called hiatus period.Overlapped on the ongoing background warming trend and the interdecadal climate variabilities, the El Ni?o events,particularly the strong ones, can make a significant contribution to the occurrence of high temperatures on interannual timescale.High temperatures associated with El Ni?o events mainly occurred during the winter.As the Pacific Decadal Oscillation (PDO) struggled back to its positive phase since 2014, the global warming returned back to a new accelerated warming period, marked by the record-breaking high temperatures in 2014.Intensified by the super strong El Ni?o, successive high records occurred in 2015 and 2016 also.Higher frequencies of high temperatures would occur in the near future because the PDO tends to maintain a continuously positive phase (Fig.2).(Su Jingzhi)

1.4 Diurnal variation of summer precipitation across the central Tian Shan Mountains

The climatic features of the diurnally varying summer precipitation over and around the central Tian Shan Mountains are investigated.Both the hourly rainfall data observed at eight stations along a transect across the mountains and the convective index derived from the satellite data show that there are three distinct regimes:the early morning peak at stations to the south of the mountains, the late afternoon peak at stations over the mountains, and the night peak at stations to the north of the mountains.The relationship between regimes of the diurnal variation was analyzed.By defining the regional rainfall event (RRE), the initial stations of each RRE were recorded.The early morning rainfall in the southern periphery of the mountains is triggered locally in the southern basin.Both the late afternoon peak over the mountains and the night peak in the northern periphery are influenced by mountain-originated rainfall events.These rainfall events appear over the mountains in the afternoon, and some of them move northward and lead to the nocturnal rainfall in the northern basin.The triggering of convection in the afternoon over the mountains and that in the early morning in the southern basin are related to the diurnally varying wind and thermodynamic conditions over and around the mountains.Lowlevel convergence with thermodynamic instability appears at noon (night) over the mountains (in the southern basin) just before the start of convection (Fig.3).(Li Jian)

1.5 Genesis of southwest vortices and its relation to Tibetan Plateau vortices

Southwest vortices (SWVs) at 700 hPa occurring over the eastern flanks of the Tibetan Plateau are important summer rain-producing systems, often leading to heavy rainfall over southwestern China and even wider areas in eastern China when they move eastward.Tibetan Plateau vortices (TPVs) are the mesoscale systems forming over the Tibetan Plateau defined at 500 hPa, which have an important influence on SWVs when they move off the Tibetan Plateau.However, only a few studies discussed the effects of TPVs on the genesis of SWVs.The present work compares three situations including 9 cases, using reanalysis data to investigate the role played by TPVs in the genesis process of SWVs.The genesis mechanisms of SWVs accompanied by the moving-off TPVs (Situation A) are explored, and then the mechanisms are further verified by comparison with situations of moving-off TPVs that are unaccompanied by the generation of SWVs(Situation B) and genesis process of SWVs without the moving-off TPVs (Situation C).It is revealed that the TPVs moving-off the Tibetan Plateau (moving-off TPVs) can exert significant effects on the genesis of SWVs through both dynamical and thermodynamic processes.The moving-off TPVs are favorable for the generation of SWVs through strengthening the cyclonic vorticity, convergence and ascending motion.Diagnoses of the potential vorticity budgets reveal that the condensational latent heat has the greatest contribution to the generation of SWVs.The SWVs under the influence of TPVs (Situation A) are stronger and have longer lifespans.Analysis of the water vapor budget indicates that the water vapor is mainly transported from south of the genesis region of SWVs associated with strong southerlies.It is demonstrated that the southerlies and the associated water vapor transport are another prominent factor affecting the genesis of the SWVs.(Li Lun)

1.6 Revisiting summertime hot extremes in China during 1961?2015

Summertime hot extremes in China are categorized into three distinct types, i.e.independent hot days,independent hot nights, and compound events, based on differing configurations between daily maximum and minimum temperatures.Linear trends for multiple indictors of these subtypes and traditionally-defined hot days/nights exhibited remarkable differences in significance, magnitude, and even sign, especially for events involving daytime extremes.Thus, some significant changes masked in conventional analyses are successfully uncovered.Particularly, the dominance of independent hot days has decayed significantly, accompanied by a rapid boom of compound events and/or independent hot nights in different regions.These nighttimeaccentuated hot extremes have exhibited significant increases in duration, intensity and spatial extent, with much stronger trends detected in severest events.(Chen Yang, Zhai Panmao)

1.7 Persisting and strong warming hiatus over eastern China during the past two decades

During the past two decades since 1997, eastern China has experienced a warming hiatus punctuated by significant cooling in minimum temperature (Tmin), particularly during early-mid winter.By arbitrarily configuring start and end years, a “vantage hiatus period” in eastern China is detected over 1998-2013, during which the domain-averagedTminexhibited the strongest cooling trend and the number of significant cooling stations peaked.Regions most susceptible to the warming hiatus are located in North China, the Yangtze-Huai River Valley and South China, where significant cooling inTminpersisted through 2016.This sustained warming hiatus gave rise to increasingly frequent and severe cold extremes there.Concerning its prolonged persistency and great cooling rate, the recent warming hiatus over eastern China deviates greatly from most historical short-term trends during the past five decades, and thus could be viewed as an outlier against the prevalent warming context (Fig.4).(Chen Yang, Zhai Panmao)

1.8 Attribution of the July?August 2013 heat event in Central and East China to anthropogenic greenhouse gas emissions

The observed and simulated frequency, intensity, and duration of some extreme weather and climate events have changed as the climate system has been warming.Extreme event attribution has drawn the interest of the public and been a topic of climate change research because of their severely devastating impacts.An individual extreme weather or climate event can occur because of natural internal variability, and can also be influenced by anthropogenic factors, along with the rare sample of extreme events.Therefore, it’s difficult to determine the extent to which climate change influences individual extreme events.In the midsummer of 2013, Central and East China (CEC) was hit by an extraordinary heat event, with the region experiencing the warmest July–August on record.To explore how human-induced greenhouse gas emissions and natural internal variability contributed to this heat event, we compared observed July–August mean surface air temperature (SAT) with that simulated by climate models.It is found that both atmospheric natural variability and anthropogenic factors contributed to this heat event.This extreme warm midsummer was associated with a positive high-pressure anomaly that was closely related to the stochastic behavior of the atmospheric circulation.Diagnosis of CMIP5 models and large ensembles of two atmospheric models indicates that the human influence has substantially increased the chance of the extreme warm midsummers such as that in 2013 in CEC, although the exact estimated increase depends on the selection of climate models (Fig.5).(Ma Shuangmei)

2 Development of the climate system model

2.1 CAMS-CSM Model description

A new coupled climate model has been developed at the Chinese Academy of Meteorological Sciences(CAMS-CSM) by employing several start-of-the-art component models.The coupled model consists of the atmospheric model ECHAM5, the ocean model MOM4, the sea ice model SIS, the land surface model CoLM,as well as the FMS coupler.The atmospheric component is a modified version of the atmospheric general circulation model ECHAM5 (v5.4) developed at the Max-Planck-Institute for Meteorology (MPI-Met).ECHAM5 is a spectral atmospheric model with a triangular truncation.The major differences between the CAMS-CSM version and the standard ECHAM5 model include: (1) A Two-step Shape Preserving Advection Scheme (TSPAS) is used for the passive tracer transport, which has been shown the capability of reducing the overestimation of precipitation over the steep edges of the southern Tibetan Plateau; (2) A k-distribution scheme developed by Zhang et al.(2006a, 2006b) is adopted for shortwave and longwave radiation transfer calculations.The Tiedtke (1989) massflux scheme with modifications for penetrative convection according to Nordeng (1994) is applied for cumulus convection parameterization.For stability considerations, time-stepping of vertical diffusion is treated implicitly in the ECHAM5 model.The implicit algorithm results in a tridiagonal system for momentum, temperature and water vapor tendency, in which the surface fluxes are relying on the states of next time step and thus can only be obtained after solving the tridiagonal equations in the atmospheric model.The model exhibits capability of reproducing the climatological mean states and seasonal cycle of major quantities of the climate system, including sea surface temperature, precipitation, sea ice extent, as well as thermocline.The major modes of climate variability are also reasonably captured by the model, such as the Madden-Julian Oscillation (MJO), ENSO, the East Asian Summer Monsoon (EASM), as well as the Pacific Decadal Oscillation (PDO).In particular, the model displays promising advantage in simulating the East Asian Summer Monsoon (EASM) variability and the ENSO-EASM relationship.Several biases exist in the model:the double-ITCZ in the annual mean precipitation map, overestimated ENSO amplitude, and too weaker Bjerkness feedback associated with ENSO.

2.2 Precipitation

Overall, the general spatial pattern of the annual mean precipitation in the model resembles that from the observations.The major rainfall center, such as ITCZ, SPCZ, and those over the tropical Indian Ocean as well as the tropical Atlantic are clearly seen in the model.The major rainbelts, as well as the seasonal migration of the intertropical convergence zone (ITCZ) and the south Pacific convergence zone (SPCZ) are reasonably captured by the model.The simulated ITCZ moves to its northernmost position with precipitation peak in July–August., while SPCZ is strongest and at its southernmost position in February–March., in good agreement with observations.During May to December., the ITCZ dominates precipitation over the tropical region, while the SPCZ is prevailing from January to April, such a seasonal timing feature is consistent between the model simulation and observations.

2.3 SST seasonal cycle and thermocline

ENSO is the most dominant climate mode on interannual time scales in the climate system.Evident anomalies of the atmosphere and ocean have been observed during the ENSO cycles, including SST,precipitation, zonal wind, as well as the thermocline.The change in thermocline depth can lead to fluctuation in SST in the eastern equatorial Pacific by upwelling and mixing, referred as the thermocline feedback, playing an important role in ENSO dynamics.The thermocline feedback is most effective in the central-eastern equatorial Pacific because of the shallow thermocline there.Therefore, realistically representing the mean thermocline and its seasonal cycle is crucial for the simulation of ENSO.Compared with the observation, the thermocline depth is well reproduced by the model.The simulated thermocline is somewhat deeper in the western Pacific and shallower in the eastern Pacific, indicating a slightly stronger zonal slope of thermocline in the model simulation.A distinctive feature of ENSO is its phase-locking to winter.Previous studies proposed that the seasonal variations of climatological states are crucial for the ENSO phase-locking.One of the essential metrics to measure the seasonal variation of tropical climatology is the seasonal cycle of the equatorial SST.Qualitatively, the model does a good job in reproducing the observed seasonal cycle of equatorial SST.The annual-cycle in the eastern equatorial Pacific as well as the semi-annual-cycle in the western equatorial Pacific are captured by the model.There are some discrepancies in the model, primarily occurring in the eastern Pacific.First, the simulated magnitude of the warm phase is weaker than the observed values, whereas that of the cold phase is stronger.Second, the cold phase in the model tends to peak earlier than that in observation by 1–2 months (Fig.6).

2.4 ENSO

The regression pattern between the Nino3.4 index and SST anomalies in the tropical Pacific and Indian Ocean are shown in Figs.7a and 7c.Overall, the anomalous SST pattern during El Ni?o events is reasonably captured by the model.Compared with the HadISST SST data, the simulated positive SSTA over the central and eastern Pacific is too narrow in the meridional direction and extends more westward as a result of the excessive westward penetration of the cold tongue in the model.The negative anomalies in the western Pacific exhibit a much zonal orientation on the north lobe, while on the south lobe the strength appears to be weaker than that in observations.The spatial pattern and magnitude of the warm signal over the South China Sea and tropical Indian Ocean are successfully simulated, as well as the small negative area on the west coast of Australia.The regression between Nino3.4 index and precipitation and 850 hPa wind anomalies are further shown in Figs.7b and 7d.The anomalous pattern is characterized by the increased precipitation over the central equatorial Pacific as a result of the enhanced convection in response to the warm SSTA.Meanwhile, anomalous westerlies extend from 150°E to 120°W over the equatorial Pacific Ocean, indicating the atmospheric aspect of the Bjerkness feedback.Negative precipitation anomalies spread over the eastern Indian Ocean and the regions of cold SSTAs, due to the weakened Walker circulation and suppressed convection over there.The model shows its ability to capture the spatial pattern of precipitation and wind anomalies associated with ENSO.The anomalous precipitation centers and low-level winds in both Pacific and Indian Oceans are reproduced with comparable magnitudes relative to those in observations.For example, the model captures the observed asymmetrical feature of precipitation and low-level wind anomalies: both precipitation and low-level wind anomalies tend to maximize south of equator.

2.5 East Asian summer monsoon (EASM) variability

As precipitation and winds are usually used to represent the EASM, in this study the evaluation mainly focused on the precipitation and wind variations associated with the EASM variability.While a number of indices have been proposed to measure the strength of the EASM (Wang 2008), here the monsoon index by Wang and Fan (1999) is adopted, which is defined as the 850 hPa zonal wind shear: WFI=U850(5°?15°N,90°?130°E) ?U850(22.5°?32.5°N, 110°?140°E).

Fig.8 shows the regression pattern between the negative WFI and the 850 hPa wind and precipitation anomalies from observations and the model.The observed circulation pattern is characterized by an anticyclone with anomalous southwesterly winds over southeastern China and westerly along the Yangtze River (YRV)and the southeast of Japan, as well as the increased easterly stretching from the western Pacific to the South China Sea.The precipitation pattern displays enhanced rainbelt spanning along the East Asian subtropical front and suppressed rainfall spreading in the southern wing of the anticyclone, with the anomalous precipitation centers close to the climatological rainfall centers.Overall, the model does a fairly good job in reproducing the observed pattern.The enhanced/deficit rainbelt over the northern/southern wing of the anticyclone is captured by the model.In particular, the model is able to capture the anomalous rainfall centers associated with the variations of Meiyu/Baiu/Changma rainbelt, which remains poorly simulated by most of the present-day climate models.

The distribution of winds and precipitation in Fig.8 reflects the anomalous pattern during the decaying summer of El Ni?o event, which also features an anticyclone over the northwestern Pacific and appears to be the dominant mode of the EASM interannual variability.This anticyclone, referred to as the western North Pacific anticyclone (WNPAC), originates in the autumn of the developing phase and sustains to the decaying summer of El Ni?o event, providing a mechanism connecting the summer precipitation of China and ENSO.While the mechanism of sustaining the WNPAC through the decayed summer is controversial, it might link to the combined forcing from northwestern Pacific, the Indian Ocean, as well as the tropical North Atlantic Ocean.To examine how well the model captures this dominant pattern in the ENSO-EASM relationship,the correlation between winter Ni?o3 index and the following summer precipitation and 850 hPa wind anomalies are calculated with the results shown in Fig.9.Indeed, the correlation pattern resembles that in Fig.8, suggesting that the dominant mode of the EASM is closely related to ENSO.The model successfully reproduces the ENSO related pattern for both precipitation and 850 hPa wind anomalies.The anomalous anticyclonic circulation over the WNP can be clearly observed from the simulated pattern.Analogous to that in the GPCP data, the simulated anomalous Meiyu/Baiu/Changma rainbelt stretches from the eastern China to the southeast of Japan, suggesting that the model has a fairly good capability in simulating the ENSO-EASM relationship.(Rong Xinyao, Li Jian, Chen Haoming, Xin Yufei, Su Jingzhi, Hua Lijuan, Qi Yanjun, et al.)

3 Polar research

3.1 Snowdrift effect on snow deposition: Insights from a comparison of a snow pit profile and meteorological observations in east Antarctica

A high-frequency, precise ultrasonic sounder was used to monitor precipitated/deposited and drift snow events over a 3-year period (17 January 2005 to 4 January 2008) at the Eagle automatic weather station site,inland Antarctica.Ion species and oxygen isotope ratios were also generated from a snow pit below the sensor.These accumulation and snowdrift events were used to examine the synchronism with seasonal variations of δ18O and ion species, providing an opportunity to assess the snowdrift effect under typical Antarctic inland conditions.There were up to 1-year differences for this 3-year-long snow pit between the traditional dating method and ultrasonic records.This difference implies that in areas with low accumulation or high wind, the snowdrift effect can induce abnormal disturbances on snow deposition.The snowdrift effect should be seriously taken into account for high-resolution dating of ice cores and estimation of surface mass balance, especially when the morphology of most Antarctic inland areas is similar to that of the Eagle site.(Ding Minghu)

3.2 A comparison of two Stokes ice sheet models applied to the Marine Ice Sheet Model Intercomparison Project for plan view models (MISMIP3D)

We presented a comparison of the numerical and simulation results for two “full” Stokes ice sheet models,FELIX-S and Elmer/Ice.The models were applied to the Marine Ice Sheet Model Intercomparison Project for plan view models (MISMIP3D).For the diagnostic experiment (P75D) the two models gave similar results (with＜ 2% difference in terms of along-flow velocities) when using identical geometries and computational meshes,which we interpreted as an indication of inherent consistencies and similarities between the two models.For the standard (Stnd), P75S, and P75R prognostic experiments, we found that FELIX-S (Elmer/Ice) grounding lines were relatively more retreated (advanced) with the results that were consistent with minor differences in the diagnostic experiment results.We showed that this is due to different choices in the implementation of basal boundary conditions in the two models.While we were unable to argue for the relative favorability of either implementation, we did show that these differences decreased with increasing horizontal (i.e., both along- and across-flow) grid resolution and that grounding-line positions for FELIX-S and Elmer/Ice converged within the estimated truncation error for Elmer/Ice.Stokes model solutions are often treated as an accuracy metric in model intercomparison experiments, but computational cost may not always allow for the use of model resolution within the regime of asymptotic convergence.In this case, we proposed that an alternative estimate for the uncertainty in the grounding-line position is the span of grounding-line positions predicted by multiple Stokes models.(Zhang Tong)

3.3 Comparison of long-term total ozone observations from space- and ground-based methods at Zhongshan Station, Antarctica

Total ozone errors for satellite observations at Zhongshan Station in Antarctica were characterized using their relative difference (RD) from ground-based Brewer observations during 1993–2015.All satellite total ozone observations slightly overestimated ground-based ones (with RD less than 4%).This is in contrast to the conclusions drawn from global-scale validation studies, where main ground-based reference stations are located in middle latitudes.Given multiple total ozone data per day at Zhongshan Station, observed by a sun synchronous orbit satellite, measurements at the lowest solar zenith angle (SZA) showed greatest consistency with Brewer ones, having an overall RD of ?0.02%–1.15%.Algorithm-retrieved total ozone data from the total ozone mapping spectrometer (TOMS), including solar backscatter ultra violet (SBUV), TOMS-Earth probe (EP), ozone monitoring instrument (OMI)-TOMS, showed the best agreement with ground-based values,followed by the global ozone measurement experiment-type direct fi tting (GOD-FIT) algorithm for the GOME-2A, and finally the differential optical absorption spectroscopy (DOAS) —Algorithm retrieved products for satellites-detectors of global ozone measurement experiment (GOME), scanning imaging absorption spectrometr for atmospheric chartography (SCIAMACHY), and OMI.Satellite total ozone RD presented some statistical characteristics, but no specific trends.Values for DOAS and GOME-2A algorithms significantly increase when the SZA was above 60°–70°, whereas values for GOME-2A decrease when the SZA is 80°–85°.Satellite total ozone RD is the minimum when the Brewer total ozone is 300–350 DU, with an obvious increase in RD values for DOAS- and GOME-2A when the Brewer total ozone is 150–300 DU.Satellite total ozone RD obviously increases as the time difference between satellite overpasses and Brewer measurements grows.Specifically, RD rises as the absolute time difference increases to more than 4 h, yielding an OMI-TOMS RD of more than 10% as this difference increases to 8 h.The DOAS- RD may be up to 15%, while GOME-2A RD does not exceed 10%.The satellite total ozone RD may reach ?5%, as the distance between the satellite overpass pixel and the station become more than 100 km.Possibly because of the discrepancy in surface albedo, the TOMS-algorithm retrieved total ozone is underestimated when the pixel on the south-east side of the station (the Antarctica continent) is used but overestimated on the north-west side of the station (the Indian Ocean).Consistency between space- and ground-based total ozone data is least for the “ozone hole”.Typically,the RD of TOMS-algorithm retrieved total ozone is within 1%/10yr.Thus, the SBUV and Brewer monthly averaged total ozone anomalies from 1996 to 2015 were 1%/10yr and 0.9%/10yr, respectively.Both indicate a weak but consistent ozone layer recovery.(Zhang Lei)

3.4 Significant increase of the Antarctic sea ice extent in the Ross Sea

In the context of global warming, the question of why Antarctic sea ice extent (SIE) has increased is one of the most fundamental unsolved mysteries.Although many mechanisms have been proposed, it is still unclear whether the increasing trend is anthropogenically originated or only caused by internal natural variability.In this study, we employed a new method where the underlying natural persistence in the Antarctic SIE can be correctly accounted for.It is found that the Antarctic SIE is not simply short-term persistent as assumed in the standard significance analysis, but actually characterized by a combination of both short- and long-term persistence.By generating surrogate data with the same persistence properties, the SIE trends over Antarctica (as well as five sub-regions) are evaluated using Monte-Carlo simulations.It is found that the SIE trends over most sub-regions of Antarctica are not statistically significant.Only the SIE over Ross Sea has experienced a highly significant increasing trend (p=0.008) which cannot be explained by natural variability.Influenced by the positive SIE trend over Ross Sea, the SIE over the entire Antarctica has also increased over the past decades,but the trend is only marginally significant (p=0.034).(Yuan Naiming, Ding Minghu)

3.5 A cold event in Asia during January?February 2012 and its possible association with Arctic seaice loss

Through both observational analyses and simulation experiments, the intraseasonal evolution of atmospheric circulation anomalies associated with a persistent cold event in the Asian continent during late January to early February 2012, and the possible association with Arctic sea-ice loss were investigated.The results suggest that the northeastern Pacific-Aleutian region and central Eurasia are two critical areas where the atmospheric circulation evolution contributed to the development of this cold event.A persistent increase in sea level pressure (SLP) over the Aleutian region was a predominant feature prior to the cold event, and then decreasing SLP over this region was concurrent with both occurrence of a polar blocking high aloft and the rapid strengthening of the Siberian High, triggering outbreaks of Arctic air over the Asian continent.Consequently, the influence of the Aleutian region on this cold event, i.e., the downstream effect of the atmospheric circulation, played a critical role.Results from simulation experiments demonstrate that Arctic atmospheric circulation conditions in the summer of 2011 significantly enhanced a negative feedback of Arctic sea-ice loss on SLP over the Aleutian region and central Eurasia during the ensuing wintertime, which may be a major reason for the development of this cold event.This finding also implies that the Aleutian Low and disturbances in the mid-latitudes over the northeastern Pacific may provide precursors to increase skills in predicting the intraseasonal evolution of extreme cold events over Eurasia.(Wu Bingyi)

3.6 Winter Olympics services

Since November 2015, according to the two difficulties that guarantee of pistes and the snow storage evaluation in the Winter Olympic Games, the observation of snow was carried out in Chongli and Yanqing,and the preliminary results were obtained in terms of artificial snow evolution and snow and ice monitoring of track.In view of preparing a ski racing track, one of the core issues of the racing track quality, we have initially established a quantif i able scientific solution through the literature research, the Winter Olympic Games Group Committee discussion, the International Snow Union Expert consultation, the field investigation and so on.Combined with oversea experience and meteorological monitoring scheme of snow and ice in polar, the project puts forward the quality monitoring scheme of snow and ice on pistes and carries out the field experiment in Wanlong ski resort in Chongli, which is in progress well.Based on the polar ice-gas interaction model and the Snow Evolution model, a forecast model for the quality of snow and ice on pistes was established, and a running manual was written.

3.7 Polar meteorological service

We have fulfilled meteorological observations at Zhongshan station and Great Wall station in Antarctic.We also have fulfilled ozone observations at Zhongshan station and helped to release “Antarctic Ozone Bulletin”.We have made new progress in Ultralow temperature AWS research and development.We have rebuilt LGB69 and Dome A Ultralow temperature AWS and acquired continuous observational data in 2017.