高月嬌，黃生志，王韓葉，王志霞，郭雯雯，穆振俠，陳 剛，黃 強(qiáng)

考慮非一致性的黃土高原區(qū)旱澇復(fù)合事件的演變特征及其動(dòng)態(tài)變化

高月嬌1，黃生志1※，王韓葉2，王志霞1，郭雯雯1，穆振俠3，陳 剛2，黃 強(qiáng)1

（1. 西安理工大學(xué) 西北旱區(qū)生態(tài)水利國(guó)家重點(diǎn)實(shí)驗(yàn)室，西安 710048；2. 云南省水利水電勘測(cè)設(shè)計(jì)院，昆明 650021；3. 新疆農(nóng)業(yè)大學(xué)水利與土木工程學(xué)院，烏魯木齊 830052）

受氣候變化和人類活動(dòng)的雙重影響，傳統(tǒng)水文序列的一致性假設(shè)受到破壞，在考慮非一致性的條件下探究相鄰季節(jié)間旱澇復(fù)合事件的動(dòng)態(tài)變化及主導(dǎo)因子，對(duì)區(qū)域的糧食安全與旱澇災(zāi)害防御意義重大。為探究非一致性條件下旱澇復(fù)合事件的動(dòng)態(tài)演變特征及其主導(dǎo)因子，該研究以黃土高原為研究對(duì)象，基于廣義可加模型擬合單季節(jié)標(biāo)準(zhǔn)化降水指數(shù)的邊緣分布，構(gòu)建二維Copula模型分析旱澇復(fù)合事件（中、重和極端情景下）的發(fā)生概率，并利用變量投影重要性準(zhǔn)則探究復(fù)合事件動(dòng)態(tài)變化的主導(dǎo)因子。結(jié)果表明：1）1982—2015年間正常轉(zhuǎn)旱、旱轉(zhuǎn)正常、正常轉(zhuǎn)澇和澇轉(zhuǎn)正常事件分布廣泛且發(fā)生頻次較高（高于22次）；2）春-夏內(nèi)蒙古持續(xù)干旱、夏-秋青海持續(xù)干旱、秋-冬寧夏持續(xù)干旱、冬-春山西持續(xù)干旱、夏-秋陜西持續(xù)洪澇、夏-秋甘肅持續(xù)洪澇事件的發(fā)生概率較大；3）春-夏由旱轉(zhuǎn)澇、夏-秋持續(xù)洪澇、秋-冬由澇轉(zhuǎn)旱、秋-冬持續(xù)干旱和冬-春季持續(xù)干旱事件的發(fā)生概率顯著上升，對(duì)該區(qū)域社會(huì)經(jīng)濟(jì)與生態(tài)將產(chǎn)生不利影響；4）復(fù)合事件發(fā)生概率動(dòng)態(tài)變化的主導(dǎo)因素為北極濤動(dòng)指數(shù)和太陽(yáng)黑子指數(shù)。研究成果將為黃土高原地區(qū)旱澇復(fù)合事件的精準(zhǔn)防御提供科技支撐。

干旱；洪澇；模型；旱澇復(fù)合事件；非一致性；動(dòng)態(tài)變化；黃土高原

0 引 言

全球持續(xù)變暖改變了氣候系統(tǒng)的熱力環(huán)境，影響了全球的水循環(huán)過(guò)程，進(jìn)而導(dǎo)致極端事件頻發(fā)[1]。干旱和洪澇是較典型的極端事件[2]，具有影響范圍廣、發(fā)生頻率高的特點(diǎn)，易對(duì)生態(tài)環(huán)境、糧食產(chǎn)量和社會(huì)生產(chǎn)活動(dòng)產(chǎn)生不利影響。干旱和洪澇災(zāi)害在全世界的發(fā)生頻率增加，強(qiáng)度加大[3]，每年造成的經(jīng)濟(jì)損失分別超過(guò)80億美元和300億美元。雖然干旱和洪澇災(zāi)害幾乎不會(huì)同時(shí)發(fā)生，但有時(shí)空相關(guān)性的災(zāi)害會(huì)相互作用發(fā)展成為旱澇復(fù)合型災(zāi)害[4-5]，從而增加災(zāi)害的影響范圍和影響強(qiáng)度，造成更嚴(yán)重的損失，例如澳大利亞、英國(guó)和秘魯?shù)葒?guó)發(fā)生的由旱轉(zhuǎn)澇災(zāi)害均對(duì)經(jīng)濟(jì)、環(huán)境和糧食安全造成嚴(yán)重影響[6-8]。

在全球變暖的背景下，季節(jié)間持續(xù)干旱（洪澇）和季節(jié)性旱澇交替等現(xiàn)象更加頻繁[9]，其帶來(lái)的負(fù)面影響呈現(xiàn)出多維和多層次性的特點(diǎn)[10]。相鄰季節(jié)間正常轉(zhuǎn)旱、旱轉(zhuǎn)正常、正常轉(zhuǎn)澇、澇轉(zhuǎn)正常、持續(xù)干旱、持續(xù)洪澇、由旱轉(zhuǎn)澇和由澇轉(zhuǎn)旱事件統(tǒng)稱為旱澇復(fù)合事件[11]，其中，持續(xù)干旱（洪澇）事件不僅對(duì)工農(nóng)業(yè)生產(chǎn)、糧食安全和生態(tài)系統(tǒng)產(chǎn)生嚴(yán)重的負(fù)面影響[11-13]，更會(huì)進(jìn)一步加劇災(zāi)害的影響時(shí)間、加大災(zāi)害的破壞程度[14]；而旱澇交替現(xiàn)象是干旱和洪澇災(zāi)害在短時(shí)間內(nèi)的快速轉(zhuǎn)變過(guò)程[15]，具有突變性的特點(diǎn)，可能會(huì)引發(fā)湖泊干涸、城市缺水、泥石流和山洪災(zāi)害頻發(fā)等一系列問(wèn)題[16-17]。有研究表明，連續(xù)旱澇災(zāi)害在未來(lái)可能成為一種常見(jiàn)的現(xiàn)象[18]。

國(guó)外主要側(cè)重于單一干旱和洪澇災(zāi)害的研究[19-20]，而對(duì)旱澇復(fù)合事件的研究較少，如：HE等[21]發(fā)現(xiàn)全球約5.9%和7.6%的陸地分別發(fā)生了春-夏和秋-冬季由旱轉(zhuǎn)澇事件；MARENGO等[22]發(fā)現(xiàn)“拉尼娜”現(xiàn)象是南美洲發(fā)生旱澇急轉(zhuǎn)現(xiàn)象的主要原因；ESPINOZA等[23]發(fā)現(xiàn)亞馬遜流域的由旱轉(zhuǎn)澇事件主要受厄爾尼諾現(xiàn)象的影響。國(guó)內(nèi)對(duì)旱澇組合事件的研究主要集中在旱澇交替現(xiàn)象，其研究的時(shí)間尺度集中在夏季或汛期[24]，研究?jī)?nèi)容涵蓋旱澇急轉(zhuǎn)事件的成因、演變規(guī)律、預(yù)測(cè)、應(yīng)對(duì)方法和對(duì)農(nóng)業(yè)的影響等[24-26]。持續(xù)干旱（洪澇）事件也開(kāi)展了部分研究，例如：張冬冬等[27]研究云南省季節(jié)連旱的概率特征，發(fā)現(xiàn)云南北部在春-夏、夏-秋和冬-春季發(fā)生持續(xù)干旱事件的概率較大；劉宇峰等[28]發(fā)現(xiàn)黃土高原的持續(xù)干旱事件增多；SHI等[29]發(fā)現(xiàn)黃河流域春-夏和夏-秋季傾向于發(fā)生持續(xù)干旱（持續(xù)洪澇）事件；楊志勇等[10]發(fā)現(xiàn)灤河流域在夏-秋季易發(fā)生旱澇復(fù)合事件。

綜上所述，以往研究雖涉及單一旱災(zāi)、單一澇災(zāi)、旱澇交替現(xiàn)象和持續(xù)干旱（洪澇）事件，但未系統(tǒng)揭示相鄰季節(jié)間旱澇復(fù)合事件的演變機(jī)理，尚不明確旱澇復(fù)合事件的驅(qū)動(dòng)因子；此外，以往研究均在一致性的條件下基于氣象站點(diǎn)數(shù)據(jù)分析旱澇復(fù)合事件的演變特征，忽略了氣候變化的影響和小地理尺度上的水文變化特征。因此，本文以旱澇災(zāi)害頻發(fā)的黃土高原為研究對(duì)象，在考慮非一致性的條件下，基于高精度的格點(diǎn)數(shù)據(jù)開(kāi)展相鄰季節(jié)間旱澇復(fù)合事件演變特征與影響因子研究，以期為黃土高原旱澇復(fù)合災(zāi)害的精準(zhǔn)防御提供科學(xué)依據(jù)。

1 數(shù)據(jù)與方法

1.1 研究區(qū)概況

黃土高原（圖1）位于黃河中上游地區(qū)，東起太行山，西至烏鞘嶺，南連秦嶺，北抵長(zhǎng)城，是中國(guó)北方與西北地區(qū)的交界處。流域總面積為6.2×105km2，地處34?41'～41?16'N，100?52'～114?33'E，橫跨中國(guó)7個(gè)省份（包括山西、陜西、甘肅、內(nèi)蒙古、寧夏、青海和河南）。流域?qū)儆诖箨懶约撅L(fēng)氣候區(qū)，區(qū)域內(nèi)降水年際變化大，年內(nèi)分布不均勻，降水主要發(fā)生在夏季，多年平均降雨量為466 mm，自東南向西北遞減，具有明顯的梯度變化特征；多年平均氣溫在-4.0～13.0℃之間，由北到南逐漸升高。

圖1 黃土高原分區(qū)及多年平均降水量示意圖

1.2 數(shù)據(jù)來(lái)源

黃土高原氣象數(shù)據(jù)來(lái)自全球陸地?cái)?shù)據(jù)同化系統(tǒng)生成的（GLDAS-Noah）降水產(chǎn)品（https://search.earthdata. nasa.gov/search?q=GLDAS/），研究所用的時(shí)間范圍從1981年到2015年，以月為時(shí)間尺度，空間分辨率為0.25°×0.25°，由于選取的GLDAS-V2.0時(shí)間僅到2014年，故本文用GLDAS-V2.1降水資料補(bǔ)齊2015年的降水?dāng)?shù)據(jù)[30]。此外，研究所用數(shù)據(jù)還有同期大氣環(huán)流因子，包括太陽(yáng)黑子指數(shù)（Sunspots）、厄爾尼諾南方濤動(dòng)指數(shù)（El Ni?o-Southern Oscillation，ENSO3.4）、北極濤動(dòng)指數(shù)（Arctic Oscillation，AO）和太平洋十年濤動(dòng)指數(shù)（Pacific Decadal Oscillation，PDO），其中，Sunspots來(lái)自比利時(shí)皇家天文臺(tái)（http://sidc.oma.be/silso/dayssplot）；ENSO3.4、AO和PDO均來(lái)自NOAA地球系統(tǒng)研究實(shí)驗(yàn)室（https://www.esrl.noaa.gov/psd/data/climateindices/list/）。

1.3 研究方法

1.3.1 旱澇等級(jí)的劃分

目前，國(guó)內(nèi)常用的旱澇指標(biāo)包括降水距平百分率、Z指數(shù)、標(biāo)準(zhǔn)化降水蒸散指數(shù)和標(biāo)準(zhǔn)化降水指數(shù)（standardized precipitation index，SPI）等[31-33]。SPI指數(shù)具有計(jì)算簡(jiǎn)單、多時(shí)間尺度和穩(wěn)定性好等特點(diǎn)，故本文選取SPI指數(shù)作為季尺度旱澇等級(jí)劃分的依據(jù)，同時(shí)參照國(guó)家規(guī)范GB/T 20481—2017《氣象干旱等級(jí)》，最終確定的旱澇等級(jí)劃分標(biāo)準(zhǔn)見(jiàn)表1。

表1 旱澇等級(jí)劃分標(biāo)準(zhǔn)

1.3.2 非一致性檢驗(yàn)

水文序列非一致性檢驗(yàn)包括趨勢(shì)、周期和突變檢驗(yàn)。檢驗(yàn)方法包括Mann-Kendall（M-K）檢驗(yàn)法、雙累積曲線法、有序聚類法和Pettitt檢驗(yàn)法等[34-35]。本文用M-K檢驗(yàn)單季節(jié)SPI序列的變化趨勢(shì)，同時(shí)采用Pettitt檢驗(yàn)法分析其突變情況，具體計(jì)算過(guò)程可參考文獻(xiàn)[35]。其中，當(dāng)<0.05時(shí)，序列存在有效突變點(diǎn)，說(shuō)明序列的一致性遭到破壞，需要在非一致性的條件下進(jìn)行頻率分析。

1.3.3 GAMLSS模型

GAMLSS（generalized additive models for location，scale and shape）模型最早由RIGBY和STASINOPOULOS[36]于2005年提出，是一種基于位置、尺度和形狀的半?yún)?shù)廣義可加模型。該模型是時(shí)變矩模型的進(jìn)一步發(fā)展，能夠靈活地描述統(tǒng)計(jì)參數(shù)與解釋變量之間的關(guān)系，且比時(shí)變矩法更便捷，極大地方便了非一致性分析工作[37-39]，目前已廣泛用于經(jīng)濟(jì)學(xué)、醫(yī)學(xué)和水文研究等領(lǐng)域[40]。因此，本文基于該模型擬合單季節(jié)SPI序列的非一致性邊緣分布，并在此基礎(chǔ)上探究旱澇復(fù)合事件的演變特征。

模型內(nèi)含有諸多分布函數(shù)，但由于SPI在計(jì)算過(guò)程中已經(jīng)標(biāo)準(zhǔn)化，同時(shí)考慮SPI序列的取值范圍，本文僅考慮用正態(tài)分布進(jìn)行擬合。同時(shí)，選擇冪次函數(shù)（bfp）和三次樣條函數(shù)（cs）作為參數(shù)和解釋變量之間的連接函數(shù)，考慮到冪次函數(shù)的冪與三次樣條函數(shù)的自由度過(guò)高會(huì)存在過(guò)度擬合現(xiàn)象，故本文僅選取bfp(t,1)、bfp(t,2)、bfp(t,3)、cs(t,0)、cs(t,1)、cs(t,2)及cs(t,3)進(jìn)行模型連接。

1.3.4 Copula函數(shù)

Copula函數(shù)能夠有效刻畫(huà)變量間的相依性，同時(shí)能夠靈活構(gòu)造多變量聯(lián)合分布，目前已廣泛應(yīng)用于干旱、洪水、泥沙等水文事件的研究中[41-42]。因此，本研究利用該函數(shù)構(gòu)造非一致性/一致性條件下的二維聯(lián)合分布模型，定量描述旱澇復(fù)合事件的發(fā)生概率。

本文依據(jù)表1定義了中度、重度和極端情景。為計(jì)算不同情景下復(fù)合事件的發(fā)生概率（表2），參照文獻(xiàn)[43-44]推導(dǎo)出由旱轉(zhuǎn)澇、由澇轉(zhuǎn)旱、持續(xù)干旱、持續(xù)洪澇、正常轉(zhuǎn)旱、旱轉(zhuǎn)正常、正常轉(zhuǎn)澇、澇轉(zhuǎn)正常事件的發(fā)生概率計(jì)算式分別如下（以中度情景為例）：

表2 不同情景下復(fù)合事件的發(fā)生概率

注：復(fù)合事件表示當(dāng)季出現(xiàn)一種降水情形的條件下，后季出現(xiàn)不同或相似的降水情形，例如由旱轉(zhuǎn)澇表示當(dāng)季發(fā)生干旱而后季則發(fā)生洪澇。

Note: Compound events mean that under the condition of one precipitation situation in the current season, different or similar precipitation situations occur in the later season. For example, the change from drought to waterlogging means that drought occurs in the current season and then floods occur in the next season.

1.3.5 變量投影重要性準(zhǔn)則

變量投影重要性準(zhǔn)則（variable importance in projection，VIP）指自變量對(duì)因變量影響的重要程度。若自變量的VIP值大于1，表明自變量對(duì)因變量的影響較為重要；若VIP值介于0.5～1，表明重要性一般；若VIP值小于0.5，則表明自變量對(duì)因變量基本沒(méi)有影響，具體計(jì)算過(guò)程可參考文獻(xiàn)[45]。本文用VIP準(zhǔn)則來(lái)反映大氣環(huán)流因子對(duì)旱澇復(fù)合事件動(dòng)態(tài)變化的影響情況，并選擇VIP值解釋度最大的因子為復(fù)合事件動(dòng)態(tài)變化的主導(dǎo)因子。

2 結(jié)果與分析

2.1 相鄰季節(jié)間復(fù)合事件的時(shí)空分布特征

季節(jié)按常規(guī)劃分為：春-夏（3—8月）、夏-秋（6—11月）、秋-冬（9—2月）和冬-春季（12—5月）。依據(jù)表 2統(tǒng)計(jì)各像元在不同情景下黃土高原復(fù)合事件的發(fā)生頻次。計(jì)算相鄰季節(jié)間各像元的平均發(fā)生次數(shù)，發(fā)現(xiàn)秋-冬季最易發(fā)生旱澇復(fù)合事件，發(fā)生次數(shù)為28.88次，隨后依次是冬-春季（27.40次）和夏-秋季（26.42次）春-夏季發(fā)生旱澇復(fù)合事件的頻次較少，為25.05次，（圖 2a）。計(jì)算各像元由旱轉(zhuǎn)澇、由澇轉(zhuǎn)旱、持續(xù)干旱、持續(xù)洪澇、正常轉(zhuǎn)旱、旱轉(zhuǎn)正常、正常轉(zhuǎn)澇和澇轉(zhuǎn)正常事件的平均發(fā)生頻次，發(fā)現(xiàn)正常轉(zhuǎn)旱、旱轉(zhuǎn)正常、正常轉(zhuǎn)澇和澇轉(zhuǎn)正常事件的頻次較高，分別為22.15、22.83、22.81和22.42次；此外，就持續(xù)性旱澇事件與旱澇交替現(xiàn)象而言，持續(xù)干旱事件的發(fā)生頻次最高，為5.46次，其次為持續(xù)洪澇事件（4.70）和由澇轉(zhuǎn)旱事件（3.97次），最后為由旱轉(zhuǎn)澇事件（圖 2b）。

空間上，由旱轉(zhuǎn)澇易發(fā)生在陜西與山西地區(qū)，由澇轉(zhuǎn)旱事件主要發(fā)生在山西地區(qū)，持續(xù)干旱與持續(xù)洪澇事件則易發(fā)生在內(nèi)蒙古地區(qū)，而正常轉(zhuǎn)旱、旱轉(zhuǎn)正常、正常轉(zhuǎn)澇與澇轉(zhuǎn)正常事件則廣泛分布在整個(gè)流域上。

2.2 單季節(jié)SPI序列的非一致性檢驗(yàn)

用M-K檢驗(yàn)法分析黃土高原地區(qū)各季節(jié)SPI序列的變化趨勢(shì)，由圖3可知：流域大部分地區(qū)春季和冬季SPI呈下降趨勢(shì)，表明春季和冬季的降水減少，干旱化趨勢(shì)突出；而夏季與秋季SPI呈上升趨勢(shì)，表明夏季與秋季的降水增多，濕潤(rùn)化趨勢(shì)突出。

采用Pettitt檢驗(yàn)法探究單季節(jié)SPI序列的突變情況，發(fā)現(xiàn)流域內(nèi)春季的SPI序列未發(fā)生突變，夏、秋和冬季的SPI序列均發(fā)生突變，其突變的發(fā)生區(qū)域分別在青海、山西北部和甘肅東部地區(qū)。

2.3 旱澇復(fù)合事件的演變特征及其主導(dǎo)因子

2.3.1 邊緣分布與聯(lián)合分布

由于SPI在計(jì)算過(guò)程中已經(jīng)標(biāo)準(zhǔn)化，同時(shí)考慮SPI序列的取值范圍，本文僅用正態(tài)分布擬合單季節(jié)SPI序列的邊緣分布。若單季節(jié)SPI序列發(fā)生突變，基于GAMLSS模型擬合該季節(jié)SPI序列的邊緣分布，同時(shí)利用赤池信息準(zhǔn)則（akaike information criterion，AIC）與貝葉斯信息準(zhǔn)則（schwarz bayesian criterion，SBC）篩選出最優(yōu)連接方式，得出對(duì)應(yīng)的位置參數(shù)和尺度參數(shù)，并基于此得到該SPI序列的最優(yōu)邊緣分布；若SPI序列不發(fā)生突變，則在一致性條件下用正態(tài)分布擬合得出最優(yōu)邊緣分布。

用均方根誤差（root mean square error，RMSE）和AIC準(zhǔn)則從Clayton-Copula、Frank-Copula、Gumbel- Copula、Gaussian-Copula和t-Copula函數(shù)中選取相鄰季節(jié)間SPI序列的最優(yōu)Copula函數(shù)。

圖3 單季節(jié)標(biāo)準(zhǔn)化降水指數(shù)的變化趨勢(shì)

2.3.2 旱澇復(fù)合事件的發(fā)生概率

根據(jù)優(yōu)選出的Copula函數(shù)及其對(duì)應(yīng)的相關(guān)參數(shù)，計(jì)算中度、重度和極端情景下相鄰季節(jié)間旱澇復(fù)合事件的發(fā)生概率、動(dòng)態(tài)變化和主導(dǎo)因子，由于在中度、重度和極端情景下相鄰季節(jié)間旱澇復(fù)合事件的演變特征基本一致，因此以中度情景為例進(jìn)行分析，下同。

時(shí)間上：春-夏季易發(fā)生正常轉(zhuǎn)澇和澇轉(zhuǎn)正常事件，發(fā)生概率均為11%；夏-秋季和冬-春季易發(fā)生正常轉(zhuǎn)旱（旱轉(zhuǎn)正常）事件，其發(fā)生概率分別為16%和15%；而秋-冬季易發(fā)生正常轉(zhuǎn)旱和正常轉(zhuǎn)澇事件。

空間上：正常轉(zhuǎn)旱、旱轉(zhuǎn)正常、正常轉(zhuǎn)澇和澇轉(zhuǎn)正常事件在流域上廣泛分布；就持續(xù)性旱澇事件與旱澇交替現(xiàn)象而言，內(nèi)蒙古地區(qū)易在春-夏季發(fā)生持續(xù)干旱事件，山西北部與河南地區(qū)易發(fā)生冬-春季持續(xù)干旱事件，寧夏地易發(fā)生秋-冬季持續(xù)干旱事件，青海地區(qū)易發(fā)生夏-秋季持續(xù)干旱事件，陜西南部與甘肅地區(qū)分別易在夏-秋季與秋-冬季發(fā)生持續(xù)洪澇事件（圖4）。此外，持續(xù)干旱（洪澇）事件的發(fā)生概率比旱澇交替事件（由旱轉(zhuǎn)澇、由澇轉(zhuǎn)旱事件）大，與2.1節(jié)的頻次統(tǒng)計(jì)結(jié)果一致，可以相互印證結(jié)果的合理性。

2.3.3 旱澇復(fù)合事件的動(dòng)態(tài)變化

探究旱澇復(fù)合事件的動(dòng)態(tài)變化特征，可以為預(yù)防旱澇復(fù)合事件的發(fā)生提供一定依據(jù)。本節(jié)以5 a時(shí)間序列為滑動(dòng)窗口[46]探究復(fù)合事件發(fā)生概率的變化趨勢(shì)，并用M-K趨勢(shì)法進(jìn)一步分析復(fù)合事件發(fā)生概率的非參數(shù)變化趨勢(shì)。由圖5可知：流域內(nèi)大部分地區(qū)的春-夏由旱轉(zhuǎn)澇、春-夏正常轉(zhuǎn)澇、夏-秋持續(xù)洪澇、夏-秋正常轉(zhuǎn)旱、夏-秋旱轉(zhuǎn)正常、秋-冬由澇轉(zhuǎn)旱、秋-冬持續(xù)干旱、秋-冬正常轉(zhuǎn)澇、秋-冬澇轉(zhuǎn)正常、冬-春持續(xù)干旱、冬-春正常轉(zhuǎn)旱與冬-春正常轉(zhuǎn)澇事件的發(fā)生概率普遍呈上升趨勢(shì)；而春-夏由澇轉(zhuǎn)旱、春-夏正常轉(zhuǎn)旱、夏-秋由旱轉(zhuǎn)澇、夏-秋正常轉(zhuǎn)澇、秋-冬由旱轉(zhuǎn)澇、秋-冬旱轉(zhuǎn)正常、冬-春持續(xù)洪澇和冬-春澇轉(zhuǎn)正常事件的發(fā)生概率則呈下降趨勢(shì)。

發(fā)生概率較大的旱澇復(fù)合事件中，春-夏內(nèi)蒙古持續(xù)干旱、秋-冬寧夏持續(xù)干旱、冬-春山西持續(xù)干旱、夏-秋陜西持續(xù)洪澇與夏-秋甘肅持續(xù)洪澇事件均呈上升的趨勢(shì)，而夏-秋青海持續(xù)干旱事件則呈下降趨勢(shì)。頻繁發(fā)生的正常轉(zhuǎn)旱、旱轉(zhuǎn)正常、正常轉(zhuǎn)澇和澇轉(zhuǎn)正常事件中，除春-夏正常轉(zhuǎn)旱、秋-冬旱轉(zhuǎn)正常、夏-秋正常轉(zhuǎn)澇與冬-春澇轉(zhuǎn)正常事件呈下降趨勢(shì)外，其余復(fù)合事件均普遍上升。

圖4 中度情景下復(fù)合事件的發(fā)生概率

2.3.4 旱澇復(fù)合事件動(dòng)態(tài)演變的主導(dǎo)因素

大量研究表明，復(fù)合事件與大氣環(huán)流異常密切相關(guān)[47]。為了進(jìn)一步揭示變化環(huán)境下旱澇復(fù)合事件動(dòng)態(tài)變化的主導(dǎo)因子，本節(jié)用VIP準(zhǔn)則分析各大氣環(huán)流因子對(duì)復(fù)合事件動(dòng)態(tài)變化的影響情況，基于此得到旱澇復(fù)合事件動(dòng)態(tài)演變的主導(dǎo)因素。

圖5 中度情景下旱澇復(fù)合事件發(fā)生概率的MK趨勢(shì)檢驗(yàn)圖

由圖6可知：流域內(nèi)呈上升趨勢(shì)的復(fù)合事件中，夏-秋持續(xù)洪澇、夏-秋正常轉(zhuǎn)旱和秋-冬持續(xù)干旱事件主要由北極濤動(dòng)（AO）主導(dǎo)，秋-冬正常轉(zhuǎn)澇、秋-冬澇轉(zhuǎn)正常、冬-春持續(xù)干旱和冬-春正常轉(zhuǎn)旱事件主要受太陽(yáng)黑子（Sunspots）的影響，春-夏由旱轉(zhuǎn)澇、春-夏正常轉(zhuǎn)澇、夏-秋旱轉(zhuǎn)正常和冬-春正常轉(zhuǎn)澇事件則由Sunspots與AO共同主導(dǎo)；呈下降趨勢(shì)的復(fù)合事件中，春-夏由澇轉(zhuǎn)旱、春-夏正常轉(zhuǎn)旱和夏-秋由旱轉(zhuǎn)澇事件主要受AO的影響，秋-冬旱轉(zhuǎn)正常、冬-春持續(xù)洪澇和冬-春澇轉(zhuǎn)正常事件由Sunspots主導(dǎo)，而夏-秋正常轉(zhuǎn)澇和秋-冬由旱轉(zhuǎn)澇事件則主要受Sunspots與AO的影響。此外，發(fā)生概率較大的春-夏內(nèi)蒙古持續(xù)干旱事件由Sunspots主導(dǎo)，秋-冬寧夏持續(xù)干旱、夏-秋陜西持續(xù)洪澇、夏-秋甘肅持續(xù)洪澇、夏-秋青海持續(xù)干旱事件主要由AO主導(dǎo)，而冬-春山西持續(xù)干旱事件則由Sunspots與AO共同主導(dǎo)。

注：ENSO3.4、AO、 PDO、Sunspots分別為厄爾尼諾南方濤動(dòng)指數(shù)、北極濤動(dòng)指數(shù)、太平洋十年濤動(dòng)指數(shù)和太陽(yáng)黑子指數(shù)。

總體而言，流域內(nèi)大部分地區(qū)的春-夏正常轉(zhuǎn)旱、夏-秋由旱轉(zhuǎn)澇、夏-秋持續(xù)干旱、夏-秋正常轉(zhuǎn)旱、夏-秋澇轉(zhuǎn)正常、秋-冬由澇轉(zhuǎn)旱事件的動(dòng)態(tài)變化主要受AO影響；而春-夏澇轉(zhuǎn)正常、秋-冬正常轉(zhuǎn)旱、秋-冬旱轉(zhuǎn)正常、秋-冬正常轉(zhuǎn)澇、冬-春持續(xù)干旱、冬-春持續(xù)洪澇和冬-春正常轉(zhuǎn)旱事件的動(dòng)態(tài)變化由Sunspots主導(dǎo)。綜合分析復(fù)合事件動(dòng)態(tài)變化的主導(dǎo)因素，可發(fā)現(xiàn)流域的復(fù)合事件主要受AO和Sunspots的影響。張永瑞等[48]研究發(fā)現(xiàn)AO與降水在黃土高原地區(qū)密切相關(guān)；竇睿音[49]發(fā)現(xiàn)黃土高原地區(qū)的干旱和洪澇災(zāi)害基本隨著太陽(yáng)黑子的升降而升降。

3 討 論

有關(guān)復(fù)合事件的文獻(xiàn)報(bào)道，其研究對(duì)象多集中于夏季（汛期）的旱澇急轉(zhuǎn)事件或相鄰季節(jié)的旱澇交替（由旱轉(zhuǎn)澇與由澇轉(zhuǎn)旱）事件或持續(xù)干旱（洪澇）事件，少有研究考慮相鄰季節(jié)中降水正常的情況（主要包括正常轉(zhuǎn)旱、旱轉(zhuǎn)正常、正常轉(zhuǎn)澇和澇轉(zhuǎn)正常事件），且相關(guān)研究均在一致性的條件下探究復(fù)合事件的發(fā)生概率、演變規(guī)律、對(duì)農(nóng)業(yè)的影響以及預(yù)測(cè)方法等。但在氣候變化的影響下，水文序列的一致性假設(shè)遭到破壞，在一致性條件下探究復(fù)合事件演變規(guī)律的結(jié)果可能與實(shí)際不符。因此，文章在考慮非一致性的條件下分析相鄰季節(jié)間由旱轉(zhuǎn)澇、由澇轉(zhuǎn)旱、持續(xù)干旱、持續(xù)洪澇、正常轉(zhuǎn)旱、旱轉(zhuǎn)正常、正常轉(zhuǎn)澇與澇轉(zhuǎn)正常事件的發(fā)生概率、演變規(guī)律及其影響因子。研究發(fā)現(xiàn)黃土高原的旱澇災(zāi)害與北極濤動(dòng)指數(shù)（太陽(yáng)黑子指數(shù)）密切相關(guān)，此結(jié)論與張永瑞等[48-49]的結(jié)論基本一致；此外，劉宇峰等[28]發(fā)現(xiàn)黃土高原的持續(xù)干旱事件增多的現(xiàn)象也與本文秋-冬、冬-春季持續(xù)干旱事件呈上升趨勢(shì)的結(jié)論一致；而SHI等[29]發(fā)現(xiàn)發(fā)生頻次較高的旱澇復(fù)合事件的風(fēng)險(xiǎn)降低，與本文研究結(jié)論（春-夏內(nèi)蒙古持續(xù)干旱、秋-冬寧夏持續(xù)干旱、冬-春山西持續(xù)干旱、夏-秋陜西持續(xù)洪澇與夏-秋甘肅持續(xù)洪澇事件的發(fā)生頻率較高且呈上升趨勢(shì)）不一致，這可能與文章考慮了非一致性有關(guān)。由于非一致性條件下的邊緣分布函數(shù)以時(shí)間為解釋變量，更能反映變化環(huán)境下水文序列變異性，因而與一致性條件下的邊緣分布函數(shù)存在差異，故導(dǎo)致兩種條件下復(fù)合事件的動(dòng)態(tài)演變規(guī)律存在一定的差異。

復(fù)合事件的發(fā)生與降水變化過(guò)程關(guān)系密切，而降水的變化過(guò)程受氣溫、對(duì)流有效位能（convective available potential energy，CAPE）和對(duì)流抑制位能（convective inhibition，CIN）等未來(lái)大氣熱力狀況的影響[50]。隨著氣溫的升高，弱CAPE和（或）CIN事件減少導(dǎo)致與之對(duì)應(yīng)的弱-中等強(qiáng)度的降水減少；而中等-強(qiáng)CAPE和（或）CIN事件增多，其引起的強(qiáng)降水事件有所增多?？偠灾?，CAPE和CIN的平均強(qiáng)度普遍增加，與之對(duì)應(yīng)的由旱轉(zhuǎn)澇事件在未來(lái)的發(fā)生頻率將會(huì)增加，強(qiáng)度也會(huì)隨之增大[18-19]。在進(jìn)一步的研究中，可深入分析在氣候變暖背景下，隨著氣溫升高，CAPE與CIN的氣候態(tài)分布與變化特征，運(yùn)用全球氣候模式對(duì)未來(lái)CAPE與CIN的變化特征進(jìn)行模擬，并在此基礎(chǔ)上分析CAPE和CIN對(duì)未來(lái)降水的抑制/促進(jìn)作用，進(jìn)而分析CAPE和CIN對(duì)復(fù)合事件的作用機(jī)理，以期揭示復(fù)合事件形成的物理機(jī)制。

4 結(jié) 論

本文以黃土高原為研究對(duì)象，考慮單季節(jié)SPI序列的非一致性，基于GAMLSS模型擬合單季節(jié)SPI序列的邊緣分布，同時(shí)采用Copula函數(shù)構(gòu)建聯(lián)合分布模型分析相鄰季節(jié)旱澇復(fù)合事件的演變特征及其動(dòng)態(tài)變化，同時(shí)探究大氣環(huán)流因子對(duì)復(fù)合事件動(dòng)態(tài)變化的影響情況，得出以下結(jié)論：

1）時(shí)間上，春-夏季易發(fā)生正常轉(zhuǎn)澇與澇轉(zhuǎn)正常事件，發(fā)生概率為11%；夏-秋季與冬-春季則易發(fā)生正常轉(zhuǎn)旱與旱轉(zhuǎn)正常事件，發(fā)生概率分別為15%和16%；而秋冬季發(fā)生正常轉(zhuǎn)旱（正常轉(zhuǎn)澇）事件的頻率較高。

2）空間上，1981—2015年間正常轉(zhuǎn)旱、旱轉(zhuǎn)正常、正常轉(zhuǎn)澇與澇轉(zhuǎn)正常事件在流域上分布廣泛且發(fā)生頻次較多（大于22次），持續(xù)性旱澇事件比旱澇交替事件更為頻繁；就持續(xù)性旱澇事件與旱澇交替現(xiàn)象而言，內(nèi)蒙古地區(qū)、青海、山西北部、河南和寧夏地區(qū)易發(fā)生持續(xù)干旱事件，而陜西南部和甘肅地區(qū)易發(fā)生持續(xù)洪澇事件。

3）流域大部分地區(qū)的春-夏由旱轉(zhuǎn)澇、春-夏正常轉(zhuǎn)澇、夏-秋持續(xù)洪澇、夏-秋正常轉(zhuǎn)旱、夏-秋旱轉(zhuǎn)正常、秋-冬由澇轉(zhuǎn)旱、秋-冬持續(xù)干旱、秋-冬正常轉(zhuǎn)澇、秋-冬澇轉(zhuǎn)正常、冬-春持續(xù)干旱、冬-春正常轉(zhuǎn)旱與冬-春正常轉(zhuǎn)澇事件的發(fā)生概率普遍上升；此外，發(fā)生概率較大的復(fù)合事件的發(fā)生概率亦普遍上升。

4）復(fù)合事件動(dòng)態(tài)變化的主導(dǎo)因子為AO和Sunspots。其中，發(fā)生概率呈上升趨勢(shì)的夏-秋持續(xù)洪澇、夏-秋正常轉(zhuǎn)旱和秋-冬持續(xù)干旱事件主要由AO主導(dǎo)，秋-冬正常轉(zhuǎn)澇、秋-冬澇轉(zhuǎn)正常、冬-春持續(xù)干旱和冬-春正常轉(zhuǎn)旱事件主要受Sunspots的影響。

本文在考慮非一致性的條件下開(kāi)展了相鄰季節(jié)間復(fù)合事件的演變規(guī)律及其動(dòng)態(tài)變化研究，并揭示復(fù)合事件動(dòng)態(tài)變化的影響因子，對(duì)變化環(huán)境下黃土高原地區(qū)復(fù)合事件的精準(zhǔn)防御提供依據(jù)，且該研究框架可應(yīng)用于非一致性條件下全球其他區(qū)域復(fù)合事件的演變特征分析。

[1] 尹家波，郭生練，顧磊，等. 中國(guó)極端降水對(duì)氣候變化的熱力學(xué)響應(yīng)機(jī)理及洪水效應(yīng)[J]. 科學(xué)通報(bào)，2021，66(33)：4315-4325. YIN Jiabo, GUO Shenglian, GU Lei, et al. Thermodynamic response of precipitation extremes to climate change and its impacts on floods over China [J]. Chinese Science Bulletin, 2021, 66(33): 4315-4325. (in Chinese with English abstract)

[2] 夏軍，佘敦先，杜鴻. 氣候變化影響下極端水文事件的多變量統(tǒng)計(jì)模型研究[J]. 氣候變化研究進(jìn)展，2012，8(6)：397-402. XIA Jun, SHE Dunxian, DU Hong. The multi-variable statistical models of extreme hydrological events under climate change[J]. Climate Change Research, 2012, 8(6): 397-402. (in Chinese with English abstract)

[3] 胡毅鴻，李景保. 1951—2015年洞庭湖區(qū)旱澇演變及典型年份旱澇急轉(zhuǎn)特征分析[J]. 農(nóng)業(yè)工程學(xué)報(bào)，2017，33(7)：107-115. HU Yihong, LI Jinbao. Analysis on evolution of drought-flood and its abrupt alternation in typical year from 1951 to 2015 in Dongting Lake area[J]. Transactions of the Chinese Society of Agricultural Engineering (Transactions of the CSAE), 2017, 33(7): 107-115. (in Chinese with English abstract)

[4] MEEHL G A, TEBALDI C. More intense, more frequent, and longer lasting heat waves in the 21st century[J]. Science, 2004. 305(5686): 994-997.

[5] 方偉. 多變量視角下珠江流域洪旱災(zāi)害時(shí)變風(fēng)險(xiǎn)研究[D]. 西安：西安理工大學(xué)，2020. FANG Wei. Assessing Time-Varying Risk of Drought and Flood from A Multivariate Perspective in the Pearl River Basin, China[D]. Xi’an: Xi’an University of Technology, 2020. (in Chinese with English abstract)

[6] BEARD G, CHANDLER E, WATKINS A B, et al. How does the 2010-11 La Nina compare with past La Nina events[J]. Australian Meteorological and Oceanographic Journal, 2011, 24: 17-20.

[7] PARRY S, MARSH T, KENDON M. 2012: From drought to floods in England and Wales[J]. Weather, 2013, 68(10): 268-274.

[8] SON R, WANG S Y, TSENG W L, et al. Climate diagnostics of the extreme floods in Peru during early 2017[J]. Climate Dynamics. 2019, 54(1): 935-945.

[9] MU W, YU F, XIE Y, et al. The copula function-based probability characteristics analysis on seasonal drought & flood combination events on the North China Plain[J]. Atmosphere, 2014, 5(4): 847-869.

[10] 楊志勇，袁喆，方宏陽(yáng)，等. 基于Copula函數(shù)的灤河流域旱澇組合事件概率特征分析[J]. 水利學(xué)報(bào)，2013，44(5)：556-561. YANG Zhiyong, YUAN Zhe, FANG Hongyang, et al. Study on the characteristic of multiply events of drought and flood probability in Luanhe River Basin based on Copula[J]. Journal of Hydraulic Engineering, 2013, 44(5): 556-561. (in Chinese with English abstract)

[11] HRDINKA T, NOVICKY O, HANSLIK E, et al. Possible impacts of floods and droughts on water quality[J]. Journal of Hydro-environment Research, 2012, 6(2): 145-150.

[12] LEIGH C, BUSH A, HARRISON E T, et al. Ecological effects of extreme climatic events on riverine ecosystems: insights from Australia[J]. Freshwater Biology, 2014, 60(12): 648-657.

[13] NING Z, QIAN H, ROEDENBECK C, et al. Impact of 1998-2002 midlatitude drought and warming on terrestrial ecosystem and the global carbon cycle[J]. Geophysical Research Letters, 2005, 32(22): 45-81.

[14] 張利平，秦琳琳，張迪，等. 南水北調(diào)中線水源區(qū)與海河受水區(qū)旱澇遭遇研究[J]. 長(zhǎng)江流域資源與環(huán)境，2010，19(8)：940-945. ZHANG Liping, QIN Linlin, ZHANG Di, et al. Drought-waterlogging encounter probability research between the water source area and water receiving areas in the middle route of South-to-North Water Transfer Project[J]. Resources and Environment in the Yangtze Basin, 2010, 19(8): 940-945. (in Chinese with English abstract)

[15] 涂新軍，龐萬(wàn)寧，陳曉宏，等. 傳統(tǒng)旱澇急轉(zhuǎn)評(píng)估指數(shù)的局限和改進(jìn)[J]. 水科學(xué)進(jìn)展，2022，33(4)：592-601. TU Xinjun, PANG Wanning, CHEN Xiaohong, et al. Limitations and improvement of the traditional assessment index for drought-wetness abrupt alternation[J]. Advances in Water Science, 2022, 33(4): 592-601. (in Chinese with English abstract)

[16] 付文藝. 旱澇急轉(zhuǎn)現(xiàn)狀及水利設(shè)施發(fā)展對(duì)策[J]. 現(xiàn)代農(nóng)業(yè)科技，2014(15)：224-225.

[17] 吳紹飛. 基于Copula函數(shù)的水環(huán)境多變量概率分布及其應(yīng)用研究[D]. 武漢：武漢大學(xué)，2017. WU Shaofei. Copula-Based Multivariate Probability Distribution of Water Environmental Variables and its Applications [D]. Wuhan: Wuhan University, 2017. (in Chinese with English abstract)

[18] CHEN J, DAI A, ZHANG Y. Projected changes in daily variability and seasonal cycle of near-surface air temperature over the globe during the 21st century[J]. Journal of Climate, 2019, 32(24): 8537-8561.

[19] QUESADA-MONTANO B, BALDASSARRE G D, RANGECROFT S, et al. Hydrological change: Towards a consistent approach to assess changes on both floods and droughts[J]. Advances in Water Resources, 2017, 111: 31- 35.

[20] CAVAZOS T. Large-scale circulation anomalies conducive to extreme precipitation events and derivation of daily rainfall in Northeastern Mexico and Southeastern Texas[J]. Journal of Climate, 1999, 12(12): 1506-1523.

[21] HE X, SHEFFIELD J. Lagged compound occurrence of droughts and pluvials globally over the past seven decades[J]. Geophysical Research Letters, 2020, 47(14): 1-14.

[22] MARENGO J A, ALVES L M, SOARES W R, et al. Two contrasting severe seasonal extremes in Tropical South America in 2012: Flood in Amazonia and Drought in Northeast Brazil[J]. Journal of Climate, 2013, 26(22): 9137-9154.

[23] ESPINOZA J C, RONCHAIL J, GUYOT J L, et al. From drought to flooding: Understanding the abrupt 2010-11 hydrological annual cycle in the Amazonas River and tributaries[J]. Environmental Research Letters, 2012, 7(2): 024008.

[24] 劉煒，趙艷麗，馮曉晶. 內(nèi)蒙古地區(qū)夏季旱澇急轉(zhuǎn)環(huán)流異常特征及其預(yù)測(cè)[J]. 干旱氣象，2021，39(2)：203-214. LIU Wei, ZHAO Yanli, FENG Xiaojing. Circulation anomaly characteristics and prediction of drought and flood abrupt alternations in summer in Inner Mongolia[J]. Journal of Arid Meteorology, 2021, 39(2): 203-214. (in Chinese with English abstract)

[25] FENG M A, AY A, JY B, et al. 2015-16 floods and droughts in China, and its response to the strong EI Nino[J]. Science of the Total Environment, 2018, 627: 1473- 1484.

[26] 高蕓，胡鐵松，袁宏偉，等. 淮北平原旱澇急轉(zhuǎn)條件下水稻減產(chǎn)規(guī)律分析[J]. 農(nóng)業(yè)工程學(xué)報(bào)，2017，33(21)：128-136. GAO Yun, HU Tiesong, YUAN Hongwei, et al. Analysis on yield reduced law of rice in Huaibei plain under drought-flood abrupt alternation[J]. Transactions of the Chinese Society of Agricultural Engineering (Transactions of the CSAE), 2017, 33(21): 128-136. (in Chinese with English abstract)

[27] 張冬冬，魯帆，嚴(yán)登華，等. 云南省干干旱時(shí)空演變規(guī)律及季節(jié)連旱的概率特征分析[J]. 應(yīng)用基礎(chǔ)與工程科學(xué)學(xué)報(bào)，2014，22(4)：705-717. ZHANG Dongdong, LU Fan, YAN Denghua, et al. Spatio-temporal analysis of droughts and the characteristic of continuous seasonal droughts probability in Yunnan Province[J]. Journal of Basic Science and Engineering, 2014, 22(4): 705-717. (in Chinese with English abstract)

[28] 劉宇峰，原志華，李文正，等. 1961-2013年黃土高原地區(qū)旱澇特征及極端和持續(xù)性分析[J]. 地理研究，2017，36(2)：345-360. LIU Yufeng, YUAN Zhihua, LI Wenzheng, et al. Extreme and persistent analysis of drought-flood variation in the Loess Plateau during 1961-2013[J]. Geographical Research, 2017, 36(2): 345-360. (in Chinese with English abstract)

[29] SHI W, HUANG S, LIU D, at al. Dry and wet combination dynamics and their possible driving forces in a changing environment[J]. Journal of Hydrology, 2020, 589(5), 125211.

[30] 宋海清，朱仲元，李云鵬. 陸面同化及再分析降水資料在內(nèi)蒙古地區(qū)的適用性[J]. 干旱區(qū)研究，2021，38(6)：1624-1636. SONG Haiqing, ZHU Zhongyuan, LI Yunpeng. Validation of land data assimilation and reanalysis precipitation datasets over Inner Mongolia[J]. Arid Zone Research, 2021, 38(6): 1624-1636. (in Chinese with English abstract)

[31] 張尚印，姚佩珍，吳虹，等. 我國(guó)北方旱澇指標(biāo)的確定及旱澇分布狀況[J]. 自然災(zāi)害學(xué)報(bào)，1998(2)：25-31. ZHANG Shangyin, YAO Peizhen, WU Hong, et al. Determination of drought-flood index and distribution of drought-flood in the north of China[J]. Journal of Natural Disasters, 1998(2): 25-31. (in Chinese with English abstract)

[32] 黃晚華，楊曉光，李茂松，等. 基于標(biāo)準(zhǔn)化降水指數(shù)的中國(guó)南方季節(jié)性干旱近58a演變特征[J]. 農(nóng)業(yè)工程學(xué)報(bào)，2010，26(7)：50-59. HUANG Wanhua, YANG Xiaoguang, LI Maosong, et al. Evolution characteristics of seasonal drought in the south of China during the past 58 years based on standardized precipitation index[J]. Transactions of the Chinese Society of Agricultural Engineering (Transactions of the CSAE), 2010, 26(7): 50-59. (in Chinese with English abstract)

[33] 楊云川，張會(huì)婭，程禹灝，等. 基于DSSAT-Canegro模型的廣西來(lái)賓市甘蔗生長(zhǎng)對(duì)氣象干旱的響應(yīng)[J]. 農(nóng)業(yè)工程學(xué)報(bào)，2022，38(2)：119-130. YANG Yunchuan, ZHANG Huiya, CHENG Yuhao, at al. Effects of meteorological drought on sugarcane growth using DSSAT-Canegro model in Laibin, Guangxi of China[J]. Transactions of the Chinese Society of Agricultural Engineering (Transactions of the CSAE), 2022, 38(2): 119-130. (in Chinese with English abstract)

[34] 黃強(qiáng)，孔波，樊晶晶. 水文要素變異綜合診斷[J]. 人民黃河，2016，38(10)：18-23. HUANG Qiang, KONG Bo, FAN Jingjing. Hydrological elements comprehensive detecting variation[J]. Yellow River, 2016, 38(10): 18-23. (in Chinese with English abstract)

[35] 隆院男，唐蓉，蔣昌波，等. 近60年湘江流域水沙特性及其對(duì)人類活動(dòng)的響應(yīng)[J]. 農(nóng)業(yè)工程學(xué)報(bào)，2018，34(24)：132-143. LONG Yuannan, TANG Rong, JIANG Changbo, et al. Variability characteristics of runoff-sediment discharge and their response to human activities in Xiang River basin in recent 60 years[J]. Transactions of the Chinese Society of Agricultural Engineering (Transactions of the CSAE), 2018, 34(24): 132-143. (in Chinese with English abstract)

[36] RIGBY R A, STASINOPOULOS D M. Generalized additive models for location, scale and shape[J]. Journal of the Royal Statistical Society, 2005, 54(3): 507-554.

[37] 溫天福，熊立華，江聰，等. 基于時(shí)變矩BMA方法的贛江流域年輸沙量變化歸因分析[J]. 農(nóng)業(yè)工程學(xué)報(bào)，2021，37(7)：140-149. WEN Tianfu, XIONG Lihua, JIANG Cong, et al. Attribution analysis of annual sediment load of Ganjiang River Basin using BMA based on time-varying moment models[J]. Transactions of the Chinese Society of Agricultural Engineering (Transactions of the CSAE), 2021, 37(7): 140-149. (in Chinese with English abstract)

[38] VILLARINIA G, SERINALDIB F. Development of statistical models for at-site probabilistic seasonal rainfall forecast[J]. International Journal of Climatology, 2012, 32(14): 2197-2212.

[39] GU X. ZHANG Q, LI J, et al. Impact of urbanization on nonstationarity of annual and seasonal precipitation extremes in China[J]. Journal of Hydrology, 2019, 575: 638-655.

[40] 陳伏龍，楊寬，蔡文靜，等. 基于GAMLSS模型的水文干旱指數(shù)研究—以瑪納斯河流域?yàn)槔齕J]. 地理研究，2021，40(9)：2670-2683. CHEN Fulong, YANG Kuan, CAI Wenjing, et al. Study on hydrological drought index based on GAMLSS: Taking Manas River Basin as an example[J]. Geographical Research, 2021, 40(9): 2670-2683. (in Chinese with English abstract)

[41] Xiao M, Yu Z, Zhu Y. Copula-based frequency analysis of drought with identified characteristics in space and time: a case study in Huai River basin, China[J]. Theoretical and Applied Climatology, 2019, 137(3): 2865-2875.

[42] 賈路，任宗萍，李占斌，等. 基于耦合協(xié)調(diào)度的大理河流域徑流和輸沙關(guān)系分析[J]. 農(nóng)業(yè)工程學(xué)報(bào)，2020，36(11)：86-94. JIA Lu, REN Zongping, LI Zhanbin, et al. Relationship between runoff and sediment load in Dali River Basin based on coupling coordination degree[J]. Transactions of the Chinese Society of Agricultural Engineering (Transactions of the CSAE), 2020, 36(11): 86-94. (in Chinese with English abstract)

[43] 張倩. 基于Copula函數(shù)的黃河中下游干支流多庫(kù)來(lái)水豐枯遭遇分析[D]. 鄭州：鄭州大學(xué)，2019. ZHANG Qian. Analysis of Water Encounters of Multi-reservoirs in the Middle and Lower Reaches of the Yellow River Based on Copula Function [D]. Zhengzhou: Zhengzhou University, 2019. (in Chinese with English abstract)

[44] 王飛宇，張彥，王偲，等. 基于Copula函數(shù)的漢江流域降水徑流豐枯遭遇研究[J]. 灌溉排水學(xué)報(bào)，2022，41(8)：95-105. WANG Feiyu, ZHANG Yan, WANG Cai, et al. Using copula model to analyze consecutive wetting-drying occurrence in rainfall-runoff in Hanjiang Basin[J]. Journal of Irrigation and Drainage, 2022, 41(8): 95-105. (in Chinese with English abstract)

[45] 楊福芹，戴華陽(yáng)，馮海寬，等. 基于赤池信息準(zhǔn)則的冬小麥植株氮含量高光譜估算[J]. 農(nóng)業(yè)工程學(xué)報(bào)，2016，32(23)：161-167. YANG Fuqin, DAI Huayang, FENG Haikuan, et al. Hyperspectral estimation of plant nitrogen content based on Akaike's information criterion[J]. Transactions of the Chinese Society of Agricultural Engineering (Transactions of the CSAE), 2016, 32(23): 161-167. (in Chinese with English abstract)

[46] 黃錦濤. 基于文本數(shù)據(jù)的水旱災(zāi)害風(fēng)險(xiǎn)評(píng)估—以河南省為例[D]. 鄭州：華北水利水電大學(xué)，2020. HUANG Jintao. Flood and Drought Disaster Risk Assessment Based on Text Data-A Case Study of Henan Province[D]. Zhengzhou: North China University of Water Resources and Electric Power, 2020. (in Chinese with English abstract)

[47] 王兆禮，李軍，黃澤勤，等. 基于改進(jìn)帕默爾干旱指數(shù)的中國(guó)氣象干旱時(shí)空演變分析[J]. 農(nóng)業(yè)工程學(xué)報(bào)，2016，32(2)：161-168. WANG Zhaoli, LI Jun, Huang Zeqin, et al. Spatiotemporal variations analysis of meteorological drought in China based on scPDSI[J]. Transactions of the Chinese Society of Agricultural Engineering (Transactions of the CSAE), 2016, 32(2): 161-168. (in Chinese with English abstract)

[48] 張永瑞，張?jiān)儡? 北極濤動(dòng)對(duì)黃土高原氣溫和降水的影響研究[A]. 中國(guó)氣象學(xué)會(huì). 第35屆中國(guó)氣象學(xué)會(huì)年會(huì) S7 東亞氣候、極端氣候事件變異機(jī)理及氣候預(yù)測(cè)[C]. 中國(guó)氣象學(xué)會(huì)：中國(guó)氣象學(xué)會(huì)，2018：493-497.

[49] 竇睿音，延軍平. 關(guān)中平原太陽(yáng)黑子活動(dòng)周期與旱澇災(zāi)害的相關(guān)性分析[J]. 干旱區(qū)資源與環(huán)境，2013，27(8)：76-82. DOU Ruiyin, YAN Junping. Relationships between drought and flood disasters in Guanzhong plain and the activities of sunspot[J]. Journal of Arid Land Resources and Environment, 2013, 27(8): 76-82. (in Chinese with English abstract)

[50] MYOUNG B, NIELSEN-GAMMON J W. Sensitivity of monthly convective precipitation to environmental conditions[J]. Journal of Climate, 2010, 23(1): 166-188.

Evolution characteristics and dynamic changes of drought-flood complex events on Loess Plateau in terms of non-consistency

GAO Yuejiao1, HUANG Shengzhi1※, WANG Hanye2, WANG Zhixia1, GUO Wenwen1, MU Zhenxia3, CHEN Gang2, HUANG Qiang1

(1.-,,710048,; 2.,650021,; 3.,,830052,)

The consistency hypothesis cannot fully meet the current hydrological series in recent years, due to the dual impacts of climate change and human activities. The hydrological frequency has been also questioned under the consistency condition. Much effort has been made into the spatiotemporal evolution characteristics and leading factors of drought-flood complex events between adjacent seasons on the Loess Plateau. Therefore, it is of great significance to regional food security and the prevention of drought and flood disasters. Taking the Loess Plateau as the research object, this study aims to determine the evolution characteristics and dynamic changes of drought-flood complex events, in terms of non-consistency. Firstly, the non-consistencies of the single-season standardized precipitation index (SPI) were diagnosed to construct the two-dimensional joint distribution model of SPI sequences between adjacent seasons using the generalized additive model (GAMLSS model).Secondly, eight events were defined, including from the drought to the flood, from the flood to the drought, persistent drought, persistent flood, from the normal to the drought, from the drought to the normal, from the normal to the flood, and from the flood to the normal. The moderate, severe, and extreme scenarios were identified, according to the classification criteria of drought and flood. A systematic analysis was implemented on the spatiotemporal distribution of the combined events. Thirdly, the occurrence probability of combined drought-flood events under different scenarios was calculated using the consistent/inconsistent two-dimensional joint distribution model. Finally, the 5 year sliding window was combined with the Mann-Kendall test to explore the dynamic evolution characteristics of drought-flood complex events. The important criterion of variable projection was then used to reveal the leading factors of dynamic changes in complex events. The results showed as follows: 1) The occurrence frequencies of the normal to the drought, the drought to the normal, the normal to the flood, and the flood to the normal events were higher than 22 times between adjacent seasons. In addition, the occurrence frequencies of persistent drought (persistent flood) events were greater than that of alternating drought and flood events. 2) The combined events of drought and flood were more likely to occur, with a frequency of 28.88 and 27.40, respectively, from autumn to winter, and from winter to spring. To be specific, the events of the normal to the flood, and the flood to the normal were tended to occur in spring and summer. The events of the normal to the drought, and the normal to the drought were more likely to occur from summer to autumn, and from winter to spring. The probability of the normal drought (flood) events was higher in autumn-winter. 3) Spatially, the events of the normal to the drought, the drought to the normal, the normal to the flood, and the flood to the normal were evenly distributed over the whole basin. In addition, Inner Mongolia, Qinghai, Ningxia, and Shanxi regions tended to sustain the drought events, while Shaanxi and Gansu regions tended to the flood events. 4) There was a significant increase in the occurrence probability of spring-summer drought to flood, summer-autumn sustained flood, autumn-winter from flood to drought, autumn-winter sustained drought, and winter-spring sustained drought. At the same time, an increasing trend was found in the occurrence probability of spring-summer sustained drought in Inner Mongolia, summer-autumn sustained drought in Qinghai, autumn-winter sustained drought in Ningxia, winter-spring sustained drought in Shanxi, and summer-autumn sustained flood in Shaanxi (Gansu), indicating the adverse effects on the social economy and ecology in the region. 5) The leading factors of dynamic change in the occurrence probability of composite events were determined as the Arctic oscillation and sunspot index. The finding can provide scientific and technological support for the precise prevention of drought-flood complex events in the Loess Plateau.

drought; flood; models; combined events; non-consistency; dynamic change; The Loess Plateau

2022-10-12

2022-12-29

國(guó)家自然科學(xué)基金項(xiàng)目（5227090529）；黑土地保護(hù)與利用科技創(chuàng)新工程專項(xiàng)（XDA28060100）

高月嬌，研究方向?yàn)樗呐c水資源。Email：1624508340@qq.com

黃生志，教授，博士生導(dǎo)師，研究方向?yàn)楦珊敌纬杉皞鞑C(jī)理。Email：huangshengzhi7788@126.com

10.11975/j.issn.1002-6819.202210090

P426.616

A

1002-6819(2023)-08-0133-11

高月嬌，黃生志，王韓葉，等. 考慮非一致性的黃土高原區(qū)旱澇復(fù)合事件的演變特征及其動(dòng)態(tài)變化[J]. 農(nóng)業(yè)工程學(xué)報(bào)，2023，39(8)：133-143. doi：10.11975/j.issn.1002-6819.202210090 http://www.tcsae.org

GAO Yuejiao, HUANG Shengzhi, WANG Hanye, et al. Evolution characteristics and dynamic changes of drought-flood complex events on Loess Plateau in terms of non-consistency[J]. Transactions of the Chinese Society of Agricultural Engineering (Transactions of the CSAE), 2023, 39(8): 133-143. (in Chinese with English abstract) doi：10.11975/j.issn.1002-6819.202210090 http://www.tcsae.org