大氣探測(cè)與雷電防護(hù)
Atmospheric Sounding and Lightning Protection

大氣探測(cè)與雷電研究進(jìn)展

2014年大氣探測(cè)研究所以理論研究為基礎(chǔ)，緊密聯(lián)系實(shí)際應(yīng)用，圍繞野外觀測(cè)試驗(yàn)和理論研究的重點(diǎn)方向，開展了大量卓有成效的觀測(cè)和研究工作，并取得了豐碩的成果。2014年共發(fā)表期刊論文13篇，其中SCI論文4篇；申請(qǐng)國內(nèi)發(fā)明專利3項(xiàng)、國際發(fā)明專利1項(xiàng)、實(shí)用新型專利2項(xiàng)；獲得軟件著作權(quán)授權(quán)1項(xiàng)；1人獲得全國優(yōu)秀青年氣象科技工作者稱號(hào)。具體成果主要體現(xiàn)在以下幾個(gè)方面。

1 野外觀測(cè)試驗(yàn)和閃電物理過程研究

1.1 雷電野外觀測(cè)試驗(yàn)

2014年大氣探測(cè)研究所在廣州、重慶以及拉薩分別開展了雷電外場(chǎng)試驗(yàn)。5月6日到8月28日在廣州從化的廣州野外雷電試驗(yàn)基地開展的 “廣東閃電綜合觀測(cè)試驗(yàn)（GCOELD）” 是自2006年以來的第9年度外場(chǎng)試驗(yàn)。本年度的觸發(fā)閃電次數(shù)達(dá)到歷年最高值，實(shí)現(xiàn)了共計(jì)15次包括多回?fù)粼趦?nèi)的人工觸發(fā)閃電，總回?fù)舸螖?shù)達(dá)到了45次，并獲得了上述觸發(fā)閃電的電磁場(chǎng)、光輻射、高速攝像、電流的直接測(cè)量資料。在自然閃電觀測(cè)方面進(jìn)一步發(fā)展了多頻段同步信號(hào)的自動(dòng)化觀測(cè)技術(shù)，完善了低頻電磁場(chǎng)、光輻射、寬帶輻射場(chǎng)和高速攝像的多頻段（低頻-高頻-甚高頻-光信號(hào)）的同步綜合觀測(cè)平臺(tái)，改善了觸發(fā)記錄方式，實(shí)現(xiàn)了自動(dòng)化測(cè)量，獲得了幾十例閃電信號(hào)多頻段觀測(cè)數(shù)據(jù)，進(jìn)一步深化了對(duì)閃電放電過程的研究。同時(shí)，新發(fā)展和研制的高靈敏度閃電電場(chǎng)變化測(cè)量系統(tǒng)已經(jīng)架設(shè)并穩(wěn)定運(yùn)行，獲取了大量的閃電放電的連續(xù)記錄數(shù)據(jù)。在高建筑物閃電觀測(cè)方面，增設(shè)了2套大量程的快慢電場(chǎng)變化測(cè)量儀以及2套全視野閃電通道成像儀，全年共獲取綜合觀測(cè)個(gè)例79個(gè)，積累了更多的高建筑物自然閃電同步觀測(cè)資料（圖1～2）。

雷電防護(hù)技術(shù)測(cè)試獲得了高壓架空線在近距離閃電回?fù)舻碾姶怒h(huán)境中的感應(yīng)情況，檢驗(yàn)了架空線的電磁兼容性能，研究了真實(shí)雷電環(huán)境下電磁脈沖在架空線的耦合機(jī)理，成功獲得了數(shù)十次的架空線感應(yīng)電壓特征，為架空線雷電防護(hù)方案的優(yōu)化提供了重要的基礎(chǔ)數(shù)據(jù)。

重慶野外試驗(yàn)初步建成由9個(gè)子站和1個(gè)閃電數(shù)據(jù)處理中心站構(gòu)成的地基高精度3D閃電定位系統(tǒng)區(qū)域示范網(wǎng)絡(luò)，并完成基于時(shí)差定位技術(shù)的高精度閃電3D定位算法和閃電類型識(shí)別軟件。拉薩外場(chǎng)試驗(yàn)主要針對(duì)“布達(dá)拉宮雷電災(zāi)害防御基礎(chǔ)研究”項(xiàng)目的任務(wù)，按照計(jì)劃完成了雷電光學(xué)圖像和電磁信號(hào)觀測(cè)設(shè)備的架設(shè)。（張陽，呂偉濤）

1.2 雷電先導(dǎo)3D發(fā)展特征的觀測(cè)和分析

閃電通道2D圖像不能反映閃電通道的真實(shí)分布情況，難以滿足準(zhǔn)確分析閃電通道3D發(fā)展特征和閃電先導(dǎo)相互作用的需求。本研究建立了一種利用雙站攝像資料重建閃電3D通道的方法，從2幅在不同位置拍攝到的閃電2D圖像上分別獲取閃電通道的坐標(biāo)序列，利用空間立體幾何原理，尋找2組序列中相互匹配的坐標(biāo)，再由匹配的坐標(biāo)序列重建出最終的閃電3D通道。

利用建立的閃電通道3D重建方法，對(duì)2012年在廣州觀測(cè)到的6次自然負(fù)極性閃電連接過程中上行連接先導(dǎo)（UCL）的通道進(jìn)行了3D重建，并對(duì)比分析了利用單站高速攝像資料和雙站光學(xué)觀測(cè)3D重建資料得到的（UCL）的2D和3D發(fā)展特征。主要結(jié)論如下：重建得到的6個(gè)UCL通道3D長度的范圍為180～818 m，平均值為426 m；計(jì)算得到了38個(gè)3D發(fā)展速率，其變化范圍為（0.8～14.3）×105m/s，平均值為4.7×105m/s，其中4個(gè)（11%）處于106m/s量級(jí)；對(duì)應(yīng)的2D長度和發(fā)展速率的變化范圍分別為147～610 m和（0.3～10.6）×105m/s；平均來看，UCL的3D發(fā)展速率為2D的1.3倍；UCL的發(fā)展速率隨時(shí)間呈現(xiàn)逐漸增加的趨勢(shì)，4個(gè)106m/s量級(jí)的3D速率都出現(xiàn)在回?fù)羟暗?.2 ms內(nèi)；對(duì)于3D長度短于300 m的UCL，77%（20/26）的3D速率小于5×105m/s，而對(duì)于3D長度超過300 m或者頭部高度超過650 m的UCL，其3D速率均高于5×105m/s（圖3）。（呂偉濤，馬穎）

1.3 人工觸發(fā)閃電連續(xù)電流過程與M分量特征

在廣州野外雷電試驗(yàn)基地，對(duì)2008年和2011年夏季人工觸發(fā)閃電回?fù)糁蟮?4個(gè)連續(xù)電流過程和43個(gè)M分量的通道底部電流、電場(chǎng)變化和通道亮度進(jìn)行了同步測(cè)量和分析。結(jié)果表明：M分量的電流、快慢電場(chǎng)變化和亮度變化波形均近似對(duì)稱；觸發(fā)閃電連續(xù)電流過程的持續(xù)時(shí)間、轉(zhuǎn)移電荷量、電流平均值的幾何平均值分別為22 ms，6.0 C和273 A；M分量的幅度、轉(zhuǎn)移電荷量、半峰值寬度、上升時(shí)間、持續(xù)時(shí)間的幾何平均值分別為409 A，205 mC，520 μs，305 μs和1.6 ms；連續(xù)電流持續(xù)時(shí)間與M分量的個(gè)數(shù)、相鄰M分量之間的時(shí)間間隔均存在顯著的正相關(guān)關(guān)系（圖4）。（鄭棟，張陽）

1.4 閃電不規(guī)則脈沖簇事件的發(fā)生規(guī)律

分析了2012 年廣州地區(qū)6 次雷暴過程的負(fù)地閃中不規(guī)則脈沖簇（CPT）放電事件的發(fā)生規(guī)律。結(jié)果表明，CPT 能夠發(fā)生在首次回?fù)糁啊⒏鱾€(gè)回?fù)糸g以及最后一次回?fù)糁?，是?fù)地閃放電過程中普遍存在的一種放電現(xiàn)象。在323次負(fù)地閃中有243次出現(xiàn)了CPT放電事件，比例達(dá)到了75.2%，并且能夠發(fā)生在負(fù)地閃的首次回?fù)糁?、回?fù)糁g以及最后一次回?fù)糁螅渲杏?6.7%的繼后回?fù)糁昂?1.5%的最后一次回?fù)糁蟀l(fā)生了CPT。研究還發(fā)現(xiàn)，CPT在繼后回?fù)糁坝?種分布類型，即單獨(dú)出現(xiàn)一次CPT-c（與繼后回?fù)粝噙B的CPT)、單獨(dú)出現(xiàn)一次CPT-i（與繼后回?fù)粲幸欢〞r(shí)間間隔的CPT)、同時(shí)出現(xiàn)CPT-c 和CPT-i 以及同時(shí)出現(xiàn)多次CPT-i。其中，單獨(dú)出現(xiàn)一次CPT-c類型最為多見，39.4%的繼后回?fù)糁盀榇祟愋停诘?次和第2次繼后回?fù)糁案菀壮霈F(xiàn)多次CPT的現(xiàn)象，分別占到對(duì)應(yīng)次序回?fù)艨倲?shù)的9.4%和7.7%。而隨著繼后回?fù)舸涡虻脑黾?，其之前發(fā)生CPT的概率呈減小的趨勢(shì)，并且除了第1次和第2次繼后回?fù)粢酝?，其他繼后回?fù)糁癈PT-c的出現(xiàn)頻次大于CPT-i。另外，CPT-i和最后一次回?fù)糁蟮腃PT均疊加在K變化上，部分回?fù)糸g的CPT也能夠發(fā)生在J過程中（圖5）。（張陽）

1.5 閃電不規(guī)則先導(dǎo)的多尺度熵分析

針對(duì)不規(guī)則脈沖簇難以判別問題，將多尺度熵應(yīng)用于不規(guī)則先導(dǎo)分析中，給出了一種多尺度熵計(jì)算方法，并通過實(shí)例說明多尺度熵在表征復(fù)雜信號(hào)方面的能力，來區(qū)分不規(guī)則先導(dǎo)與直竄先導(dǎo)及梯級(jí)先導(dǎo)的不同。探討了閃電信號(hào)不規(guī)則脈沖分析應(yīng)用中多尺度熵關(guān)鍵參量的選擇方法。在此基礎(chǔ)上，將不規(guī)則先導(dǎo)與直竄先導(dǎo)及梯級(jí)先導(dǎo)閃電信號(hào)的多尺度熵進(jìn)行比較。統(tǒng)計(jì)分析表明：不規(guī)則先導(dǎo)和直竄先導(dǎo)熵值隨尺度先增加后趨于平穩(wěn)，但熵值有很大差異；梯級(jí)先導(dǎo)熵值隨尺度變化不明顯，整體呈增長趨勢(shì)，與不規(guī)則先導(dǎo)的熵值在大于3的尺度上也有所差異，因此當(dāng)尺度大于3時(shí)可將熵值大于1.5的先導(dǎo)歸類為不規(guī)則先導(dǎo)，熵值小于1.5的先導(dǎo)歸類為梯級(jí)先導(dǎo)或直竄先導(dǎo)。不規(guī)則先導(dǎo)的特征熵平均值為2～2.1，最大值范圍為2.6～2.8，最小值范圍為1.5～1.51（圖6）。（張陽）

1.6 閃電電場(chǎng)變化信號(hào)測(cè)量系統(tǒng)及方法

為了解決閃電電場(chǎng)近距離測(cè)量的飽和問題并兼顧其測(cè)量的動(dòng)態(tài)范圍，發(fā)明了一種閃電電場(chǎng)變化信號(hào)測(cè)量系統(tǒng)。該系統(tǒng)包括探測(cè)天線、電場(chǎng)快變化接收機(jī)、信號(hào)處理模塊：電場(chǎng)快變化接收機(jī)分別連接探測(cè)天線和信號(hào)處理模塊；探測(cè)天線用于將感應(yīng)到的閃電產(chǎn)生的電場(chǎng)變化信號(hào)傳輸至電場(chǎng)快變化接收機(jī)；電場(chǎng)快變化接收機(jī)用于處理電場(chǎng)變化信號(hào)，以得到閃電電場(chǎng)快變化信號(hào)，并將閃電電場(chǎng)快變化信號(hào)傳輸至信號(hào)處理模塊；信號(hào)處理模塊用于采集并處理閃電電場(chǎng)快變化信號(hào)，以得到閃電電場(chǎng)慢變化信號(hào)。同時(shí)發(fā)明了一種與上述測(cè)量系統(tǒng)相對(duì)應(yīng)的閃電電場(chǎng)變化信號(hào)測(cè)量方法。所發(fā)明的閃電電場(chǎng)變化測(cè)量系統(tǒng)和方法，可通過電場(chǎng)快變化接收機(jī)獲取閃電電場(chǎng)快變化信號(hào)，并對(duì)閃電電場(chǎng)快變化信號(hào)處理得到閃電電場(chǎng)慢變化信號(hào)，實(shí)現(xiàn)了對(duì)閃電電場(chǎng)快、慢變化信號(hào)的集成化測(cè)量，有效解決了近距離閃電信號(hào)過強(qiáng)導(dǎo)致的閃電電場(chǎng)慢變化探測(cè)輸出飽和的問題（圖7）。（張陽，孟青）

2 雷暴中閃電活動(dòng)特征及其與天氣現(xiàn)象關(guān)系研究

2.1 基于大氣層結(jié)和雷暴演變的閃電和降水關(guān)系

選取2006—2008年發(fā)生在北京及其周邊地區(qū)的28次雷暴過程，基于大氣不穩(wěn)定度參數(shù)和雷達(dá)參量對(duì)雷暴過程進(jìn)行分類，分析了不同分類條件下的總閃電活動(dòng)（SAFIR3000 3D閃電定位系統(tǒng)觀測(cè)）和對(duì)流降水（雷達(dá)反演）的關(guān)系。結(jié)果表明，總閃對(duì)應(yīng)降水量的平均值為1.92×107kg/fl。依據(jù)對(duì)流有效位能和抬升指數(shù)對(duì)雷暴進(jìn)行分類的分析表明，較強(qiáng)的不穩(wěn)定狀態(tài)對(duì)應(yīng)了較小的總閃對(duì)應(yīng)降水量，同時(shí)總閃頻次與對(duì)流降水量的相關(guān)性更好?；诶走_(dá)特征參數(shù)的分類分析表明，總閃對(duì)應(yīng)降水量在對(duì)流運(yùn)動(dòng)較弱情況下最小，其次是對(duì)流運(yùn)動(dòng)較強(qiáng)的情況下，而對(duì)流運(yùn)動(dòng)適中時(shí)最大。（鄭棟）

2.2 上海及周邊地區(qū)地閃活動(dòng)特征及海陸差異

利用LS8000閃電定位系統(tǒng)2009—2011年的地閃觀測(cè)資料對(duì)上海及周邊地區(qū)（120°～122.5°E，30°～32°N）的地閃活動(dòng)特征進(jìn)行了研究。結(jié)果表明，分析區(qū)域內(nèi)正地閃的比例約占8.5%，大電流地閃（電流絕對(duì)值大于50 kA）的比例約為5.6%。地閃活動(dòng)主要集中在6—9月，峰值出現(xiàn)在8月；日間12∶00—19∶00閃電活動(dòng)最為活躍，峰值出現(xiàn)在14∶00，凌晨閃電活動(dòng)最弱。從日變化上來說，正地閃和大電流地閃比例在地閃活動(dòng)較強(qiáng)時(shí)段低于地閃活動(dòng)較弱時(shí)段；在月分布上，在地閃活動(dòng)最強(qiáng)的夏季，正地閃比例普遍在10%以下，在地閃活動(dòng)較弱的春、秋、冬季，正地閃比例普遍在10%以上。以31°N為界，分析區(qū)域北部地閃密度基本在6～12次／（km2·a），南部基本在2.4～4.8次／（km2·a）。同時(shí)陸地上的地閃密度要顯著高于湖泊和海洋上的地閃密度，而海洋上的正地閃比例和大電流地閃比例要顯著高于陸地。閃電空間分布的時(shí)間變化說明，下午地閃活動(dòng)主要出現(xiàn)在陸地，而凌晨地閃主要出現(xiàn)在水體附近，其他時(shí)段則表現(xiàn)出過渡特征，這與下墊面的加熱作用緊密相關(guān)。（鄭棟）

3 利用WRF-Electric模式模擬熱帶氣旋電荷結(jié)構(gòu)的演變特征

以WRF-ARW模式為基礎(chǔ)，在Milbrandt和Morrison2個(gè)雙參數(shù)微物理方案中分別耦合了感應(yīng)和非感應(yīng)的起電機(jī)制，同時(shí)引入了放電參數(shù)化方案，從而構(gòu)建起一個(gè)完整的中尺度起電放電模式WRFElectric。模式不僅能夠模擬風(fēng)暴內(nèi)電荷結(jié)構(gòu)的演變，同時(shí)還具有區(qū)域閃電活動(dòng)分布的預(yù)報(bào)能力。在非感應(yīng)起電機(jī)制方面，不僅引入了基于液態(tài)水含量的TGZ機(jī)制和GZ機(jī)制，還將基于霰粒子結(jié)淞率的SP98機(jī)制和RR機(jī)制引入到了數(shù)值模式。圖8所示為中尺度起電放電模式的架構(gòu)。利用建立的中尺度起電放電模式，對(duì)于一個(gè)理想的熱帶氣旋起電的演變進(jìn)行了模擬。研究熱帶氣旋的起電及電荷結(jié)構(gòu)，有利于更進(jìn)一步地理解熱帶氣旋中的閃電活動(dòng)。本研究從數(shù)值模擬的角度嘗試描述熱帶氣旋電荷結(jié)構(gòu)的演變特征，已有的大部分觀測(cè)事實(shí)和模擬結(jié)果從不同方面支持了本研究的結(jié)論。

模擬結(jié)果表明（圖9），眼壁區(qū)對(duì)流一般表現(xiàn)出負(fù)的偶極性結(jié)構(gòu)，有一個(gè)負(fù)電荷區(qū)在正電荷區(qū)之上。在加強(qiáng)階段，眼壁區(qū)伴隨的強(qiáng)烈上升氣流的極端強(qiáng)對(duì)流呈現(xiàn)正常的3極性電荷結(jié)構(gòu)，有一個(gè)主負(fù)電荷區(qū)夾在2個(gè)正電荷區(qū)之間。外螺旋雨帶對(duì)流的電荷結(jié)構(gòu)在不同階段均呈現(xiàn)正常的偶極性電荷結(jié)構(gòu)。進(jìn)一步的分析表明，不同的電荷結(jié)構(gòu)主要是由上升氣流和粒子分布的差異造成的。在眼壁區(qū)對(duì)流的上升氣流一般較弱，不同粒子的混合區(qū)域主要分布在較低的層次，導(dǎo)致起電過程主要發(fā)生在霰粒子的正起電區(qū)。在加強(qiáng)階段，眼壁區(qū)爆發(fā)的強(qiáng)對(duì)流具有強(qiáng)的垂直上升氣流，起電在霰粒子的正、負(fù)起電區(qū)同時(shí)進(jìn)行。而在外螺旋雨帶對(duì)流中，霰粒子和冰晶粒子的主要共存區(qū)在云的高層，起電過程也主要發(fā)生在霰粒子的負(fù)起電區(qū)，從而形成正常的偶極性電荷結(jié)構(gòu)。據(jù)此，構(gòu)建了熱帶氣旋電荷結(jié)構(gòu)演變的概念模型（圖8～9）。（王飛，徐良韜）

4 雷電業(yè)務(wù)工作進(jìn)展

4.1 雷電臨近預(yù)警系統(tǒng)的推廣培訓(xùn)

針對(duì)雷電業(yè)務(wù)應(yīng)用系統(tǒng)的需求，2014年9月16—19日，按照“2013年氣象監(jiān)測(cè)與災(zāi)害預(yù)警工程”項(xiàng)目實(shí)施方案的指導(dǎo)精神，中國氣象科學(xué)研究院大氣探測(cè)所承擔(dān)的第2期全國“雷電臨近預(yù)警系統(tǒng)培訓(xùn)班”在中國氣象局氣象干部培訓(xùn)學(xué)院安徽分院舉辦，來自全國30個(gè)?。ㄊ小⒆灾螀^(qū)）氣象臺(tái)、防雷中心以及民航華北空管局氣象中心的技術(shù)人員共65名學(xué)員參加了本次培訓(xùn)班。

大氣探測(cè)所在第1期培訓(xùn)班的基礎(chǔ)上，根據(jù)培訓(xùn)后的反饋精心調(diào)整授課內(nèi)容，更新修訂了教材。培訓(xùn)內(nèi)容涉及了雷電物理、災(zāi)害天氣、雷電探測(cè)、雷電預(yù)警預(yù)報(bào)、防雷工程等多方面知識(shí)，受到了學(xué)員的好評(píng)，進(jìn)一步推動(dòng)了現(xiàn)有科研成果的業(yè)務(wù)轉(zhuǎn)化和應(yīng)用。（姚雯，孟青，張文娟）

4.2 雷電臨近預(yù)警系統(tǒng)推廣應(yīng)用

2014年3月，雷電臨近預(yù)警系統(tǒng)在廣東省防雷中心本地化調(diào)試后，于2014年4月20日投入業(yè)務(wù)試驗(yàn)運(yùn)行。系統(tǒng)運(yùn)行期間，已成功對(duì)廣東省多次雷暴過程提前作出了準(zhǔn)確的雷電預(yù)警，系統(tǒng)POD、FAR以及TS評(píng)分較好。目前，廣東省防雷中心已將雷電臨近預(yù)警系統(tǒng)的業(yè)務(wù)產(chǎn)品應(yīng)用于廣州市超高建筑物、?；穲?chǎng)所等多個(gè)重點(diǎn)單位的雷電臨近預(yù)警服務(wù)，為重點(diǎn)單位合理安排經(jīng)營時(shí)間、規(guī)避雷擊風(fēng)險(xiǎn)以及雷電應(yīng)急安全生產(chǎn)等工作提供了重要的決策參考。（姚雯，孟青）

圖1 人工引雷試驗(yàn)場(chǎng)（左）和一次成功觸發(fā)閃電（右）Fig. 1 Field for artif cially triggered lightning (left) and a successfully triggered lightning (right)

圖2 新架設(shè)的觀測(cè)設(shè)備（左）和一次高建筑閃電（右）Fig. 2 A building observation system (left) and a high-building lightning (right)

圖3 廣州試驗(yàn)觀測(cè)到的一次閃電過程的上行連接先導(dǎo)的3D通道重建結(jié)果（右上為3D通道，左上、左下和右下分別為3D通道在Y-Z平面、X-Y平面和X-Z平面上的投影，顏色指示高度的變化）Fig. 3 3D reconstruction results of the upward connecting leader (UCL) for a lightning f ash observed in Guangzhou. It contains the 3D reconstruction channel of the UCL (top right), the projection of the 3D reconstruction channel in the Y-Z plane (top left), the X-Y plane (bottom left) and the X-Z plane (bottom right). The colors indicate height

圖4 觸發(fā)閃電連續(xù)電流過程觀測(cè)波形：(a)電流波形；(b)快電場(chǎng)變化波形；(c)慢電場(chǎng)變化波形；(d)通道亮度變化波形Fig. 4 Observed waves of continuous current of triggered lightning: (a) current wave; (b) fast electric f eld waveform; (c) slow electric f eld waveform; (d) channel luminosity variation waveform

圖5 不規(guī)則脈沖簇在放電波形中的位置分布Fig. 5 Distribution of chaotic pulses in the discharge waveform

圖6 不同先導(dǎo)的多尺度熵特征Fig. 6 Multi-scale entropy features of different leaders

圖7 閃電電場(chǎng)變化集成測(cè)量系統(tǒng)示意Fig. 7 Integrated measurement system of lightning E-change

圖8 中尺度起電放電模式WRF-Electric的架構(gòu)Fig. 8 Frame of the WRF-Electric model

圖9 熱帶氣旋不同階段的電荷結(jié)構(gòu)演變特征（眼壁區(qū)加強(qiáng)階段(a-b)和準(zhǔn)穩(wěn)態(tài)階段(c-d)電結(jié)構(gòu)；外螺旋云帶加強(qiáng)階段(e)和準(zhǔn)穩(wěn)態(tài)階段(f)電結(jié)構(gòu)）Fig. 9 Evolution of the charge structure of a TC in different stages: Intensif cation stage (a-b) and quasi-steady stage (c-d) of an eyewall; Intensif cation stage (e) and quasi-steady stage (f) of the outer spiral rainband cells

Progress in Atmospheric Sounding and Lightning Research

Based on theoretical research and practical application while focused on key directions of the lightning field experiment and the theoretical research, a lot of observation and research activities were carried out and fruitful results were achieved by the Institute of Atmospheric Sounding (IAS) in 2014. 13 papers were published, 4 of which were collected by SCI/EI; one was requesting an international patent, 3 national patents for invention and 2 patents for practical novelty; one software copyright was authorized; one person obtained the title of “National Outstanding Young Meteorological Science and Technology Worker”. The concrete results are shown as follows.

1 Lightning f eld experiment and physical process research

1.1 Comprehensive observation experiment on lightning discharge

The IAS conducted the comprehensive observation experiment on lightning discharge in 2014. The Guangdong Comprehensive Observation Experiment on Lightning Discharge (GCOELD) carried out in Guangzhou from May 6 to Aug 28 is the ninth f eld experiment since 2006. 15 lightning f ashes, including 45 return strokes, have been triggered successfully, which are the best results since the year of 2006 when GCOELD began. The synchronous observation data, including current, electromagnetic signal, optical radiation and high speed video, have also been acquired. As for the observation of natural lightning, the synchronous auto-observation technology of multiband lightning signals has been further developed, and the multiband observation platform, which detects LF electromagnetic signal, optical radiation, broadband radiation, and high-speed video, has been improved through a f exible triggering method and an auto-observation method. The multiband observation data for tens of f ashes have been recorded, which helps further deepen the research of lightning discharge. At the same time, the novel observation system of electric f eld change and of highsensitivity has been built and operated continuously, and continuous data of lightning discharge have been acquired. As for the observation of lightning on high buildings, we set up some novel observation systems, including two slow/fast antennas with a large detection range and two whole-view imagers of lightning channel, with 79 comprehensive cases of high building lightning recorded.

Additionally, the test of lightning protection technology has been conducted based on the close triggered lighting. The electromagnetic compatibility of high voltage line has been checked and 10 induction signals near the triggered lightning have been recorded, which provide fundamental data for optimizing the lightning protection scheme of high voltage line and establishing the lightning protection level in the future. Further, the coupling mechanism on electromagnetic pulse of a high voltage line has been studied.

The experiment in Chongqing has preliminarily built a local prototype network of three-dimensional (3D) lightning location system of high accuracy, which includes 9 substations and 1 central station. At the same time, the location algorithm of time-of-arrival (TOA) for lightning 3D location of high accuracy and the software for distinguishing lightning types has been developed.

The experiment in Lasa focused on the tasks as described in the project “Fundamental Research for Lightning Protection in the Potala Palace”. The instruments of optical imaging and electromagnetic signal for lightning have been installed and operated (Fig.1–2). (Zhang Yang, LǚWeitao)

1.2 Observation and analysis of the 3D propagation characteristics of lightning leader

A two-dimensional (2D) image of lightning channel, which cannot ref ect the real 3D spatial distribution of lightning channel, is insufficient to meet the needs of an accurate analysis of the 3D lightning channel development characteristics and the interaction between different leaders. A reconstruction method of 3D lightning channel from dual-station optical observation is established. The 2D coordinate sequence of lightning channel is obtained from each 2D lightning image captured at different positions. The principles of 3D spatial geometry are used to match the coordinates from different sequences to reconstruct the 3D lightning channel.

Six downward negative f ashes terminated on tall structures in Guangzhou are analyzed. The 3D lightning channels of the upward connecting leaders (UCL) are reconstructed. For comparison, the corresponding 2D parameters are calculated using the single-station high-speed images. The main conclusions are summarized as follows: the 3D length values of the six UCLs range from 180 to 818 m with an average value being 426 m; 38 3D speed values calculated by combining the 3D UCL channel and the high-speed images for the six UCLs range from 0.8×105to 14.3×105m s-1(average: 4.7×105m s-1) and four of them (11%, 4/38) are on the order of 106m s-1; the corresponding 2D length and 2D speed range from 147 to 610 m and 0.3×105to 10.6×105m s-1, respectively; the average value of the 3D speed is 1.3 times that of the 2D speed; when the time approaches the return stroke, the propagation speed of the UCL increases and all of the four 3D speed values on the order of 106m s-1occur less than 0.2 ms prior to the RS; when the 3D length is shorter than 300 m, 77% (20/26) of the corresponding 3D speed values are smaller than 5×105m s-1, and when the 3D length is longer than 300 m or the UCL tip height is higher than 650 m, all of the corresponding 3D speed values are faster than 5×105m s-1(Fig. 3). (Lǚ Weitao, Ma Ying)

1.3 Characteristics analysis of a continuing current process and M-component in an artificially triggered lightning

The Continuing Current (CC) process of cloud-to-ground lightning is a discharge process in which charges are continuously transferred to ground along the lightning channel after a return stroke. The magnitude of CC is small, but the duration of CC is commonly long. So CC often causes a lightning disaster. It,s very hard to get current data due to the randomness of lightning. An artif cially triggered lightning, in which the location and time of a triggered lightning can be controlled, is an effective way to measure currents of lightning. An artif cially triggered lightning is different from a natural lightning, which only has CC after a return stroke. Yet an artif cially triggered lightning has both CC and Initial Continuous Current (ICC) processes. Only the CC is analyzed using simultaneous observations of current, electric f eld change and channel luminosity by coaxial shunts, fast and slow antennas, and high-speed cameras in Guangzhou Field Experiment Site for Lightning Research and Testing, Conghua, Guangdong, China. Then, photoelectric characteristics and parameters of 14 CC and 43 M-components after return strokes of a triggered lightning observed in summer from 2008 to 2011 are analyzed. The relationships between some characteristic parameters of CC and M-component are analyzed, too.

The current waveforms of CC after return strokes are continuous and change slowly. Usually, there are current pulses ranging in size superimposed on CC waveforms. The slow electric field waveforms of CC are slowly changing, too. The lightning channel below the cloud is always luminescent during CC. The current waveforms, fast and slow electric field waveforms and channel luminosity variation waveforms of M-components are approximately symmetrical. The geometric mean of duration, charge transferred to ground, average current and action integral for CC are 22 ms, 6.0 C, 273 A, 4187 A2s, respectively. The geometric mean of magnitude, charge transferred, half peak width, rise time (10%–90%), duration, preceding CC level, inter-pulse interval, action integral for M-components are 409 A, 205 mC, 520 μs, 305 μs, 1.6 ms, 310 A, 6.5 ms, 465 A2s, respectively. There are remarkable positive correlations between the duration of continuing current and number of M-components, and between the duration of continuing current and inter-pulse interval of M-components. The correlation coeff cients are 0.83 and 0.75, and both pass the signif cant verif cation of 0.01 level (Fig. 4). (Zheng Dong, Zhang Yang)

1.4 Occurrence regularity of CPT discharge events in a negative cloud-to-ground lightning

Occurrence regularity of chaotic pulse trains (CPT) discharge events in a negative cloud-to-ground (CG) lightning during six thunderstorms is analyzed. Results show that CPT is a common phenomenon throughout the negative CG lightning discharge process. 243 times of CPT discharge events occur during 323 negative CG lightning processes with a proportion reaching 75.2%. CPT in negative CG can occur before the first return stroke, between the strokes, and after the last stroke. The proportion of 66.7% of the total subsequent strokes is preceded by CPT, and CPT occurs after 11.5% of the total last strokes. It is also found that there are four distributions prior to subsequent strokes: single CPT-c (CPT connecting with subsequent strokes), single CPT-i (CPT occurring in an interval between CPT and return strokes), CPT-c and CPT-i occur concurrently, and several CPT-is appear concurrently. Single CPT-c is the most common, 39.4% of subsequent strokes are preceded by single CPT-c, and several CPTs are easier to occur before the f rst and second subsequent strokes, which are respectively 9.4% and 7.7% of the total corresponding strokes. With the increasing subsequent strokes, there is a decreasing occurrence of CPT, and in addition to the previous two subsequent strokes, the number of CPT-c is obviously greater than that of CPT-i. In addition, both CPT-i and CPT-c coincide with the negative CG K-change, and some also correspond to the process of J-change in the slow electric f eld wave (Fig. 5). (Zhang Yang)

1.5 The multi-scale entropy feature of a chaotic leader in a CG lightning

To address the difficulty in determining chaotic pulses, we apply the multi-scale entropy method to distinguish chaotic, dart and dart-stepped leaders. The calculation method of multi-scale entropy and the applicability in characterizing complex signals have been investigated. The key parameters of multi-scale entropy during their application to the analysis of lightning chaotic pulses have been researched, and the multiscale entropies for chaotic, dart and dart-stepped leaders have been compared. The results show that although the entropies of chaotic and dart leaders both increase f rstly and then stabilize with the changing scale, their values are obviously different. The entropy of a stepped leader does not change too much with the increasing scale, and its value is different from that of a chaotic leader when the scale is larger than 3. As a result, we suggest that the entropy with a value of ＞1.5 indicates a chaotic leader, in contrast, the entropy with a value of＜ 1.5 corresponds to a stepped-leader or a dart leader. The average value of characteristic entropy for a chaotic leader is 2–2.1, the range of maximum values of it is 2.6–2.8, and the range of minimum values of it is 1.5–1.51 (Fig. 6). (Zhang Yang)

1.6 Lightning electric f eld change signal measuring system and method

To address the saturation during the close measurement of lightning, a lightning electric field change signal measuring system has been invented. The system comprises a detection antenna, an electric f eld rapidchange receiver and a signal processing module, wherein the electric f eld rapid-change receiver is respectively connected with the detection antenna and the signal processing module; the detection antenna is used for transmitting an electric f eld change signal generated by an induced lightning to the electric f eld rapid-change receiver; the electric f eld rapid-change receiver is used for processing the electric f eld change signal to obtain a lightning electric f eld rapid-change signal and transmitting the lightning electric f eld rapid-change signal to the signal processing module; and the signal processing module is used for acquiring and processing the lightning electric f eld rapid-change signal to obtain a lightning electric f eld slow-change signal. The invention also covers a lightning electric f eld change signal measuring method corresponding to the measuring system. The system and method detect the electric f eld change signal by the reception of E-change, and then process the data to get a slow E-change waveform. As a result, it realizes an integrated measurement of lightning electric f eld rapid-change and slow-change signals, and solves the saturation during the close measurement of lightning (Fig. 7). (Zhang Yang, Meng Qing)

2 Characteristics of lightning activity in a thunderstorm and its relationship with weather phenomena

2.1 Relationship between lighting and precipitation based on atmospheric stratification and thunderstorm development

A total of 28 thunderstorms occurring in and around Beijing from 2006 to 2008 are collected to investigate the relationship between total lighting (observed by SAFIR3000) and convective precipitation (by radar inversion). These cases are classif ed according to the parameters of atmospheric stratif cation and the ref ectivity of radar. The quantitative results provide a reference for applications of lightning data to severe weather warning and precipitation estimation. The lightning forecast can also be improved by assimilating the relationship between the hydrometeors and the lighting activities to the numerical prediction models. The analysis can extend the application of the lighting data.

The results show that the average convective rain yield per f ash is 1.92×107kg f-1on the whole, while the linear correlation coefficient between total lightning frequency and convective precipitation is 0.584. Total lightning frequency (expressed by F with the time interval being 6 min) can be used to calculate the amount of convective precipitation with the equation being R = (2.813×108) + (4.570×106) F. A total of 28 thunderstorms are classif ed according to the available potential convective energy (Ecap) and lifting index (LI) of the atmospheric stratif cation. It is found that strong instability of atmospheric stratif cation tends to be associated with smaller precipitation and there is a more pronounced correlation between total lightning and precipitation. The classif ed Ecap (no less than 1600 J·kg-1) has a correlation coeff cient of 0.837. The total lightning frequency can be used to calculate the amount of convective precipitation with the equation being R = (1.620×108) + (5.478×106) F. While the classif ed LI (no less than 4 K) has a correlation coeff cient of 0.853, the total lightning frequency can be used to calculate the area of the amount of convective precipitation with the equation being R = (1.530×108) + (6.276×106) F. Another three parameters are calculated from radar reflectivity, i.e., maximum height of 20 dBz reflectivity, maximum reflectivity at 12 km level, and volume ratio of the ref ectivity larger than 30 dBz above 0℃ to the ref ectivity larger than 40 dBZ above 0℃, in terms of their radar volume scans. The most pronounced relationships between lightning and precipitation occur in the classif cation of H20 dBz＜11.5 km, 25 dBz≤f12 km＜35 dBz, and V40/30＜0.39, when the correlation coeff cients are 0.804, 0.609 and 0.750, respectively. The linear correlation between lightning and precipitation shows obvious differences in different classifications. The fitting equations in different classifications will provide references for the application of relationships between lightning and precipitation according to the characteristics of thunderstorm processes. (Zheng Dong)

2.2 Characteristics of CG lightning in Shanghai and its surrounding regions and land-sea difference

With observations from the LS8000 lightning location system, cloud-to-ground (CG) lightning activities in Shanghai and its surrounding regions are studied. The percentage of positive CG lightning in the total is about 8.5% and that of large current CG lightning (with the absolute value of current larger than 50 kA) is 5.6%. CG lightning activities highly occur in the period from June to September, and peak in August. For the temporal distribution of CG lightning, 12:00 to 19:00 BT is an active period, which peaks at 14:00 BT and bottoms out in the morning. The monthly and hourly distributions of the percentage of the positive cloud-toground (PCG) lightning are seasonal. Setting 31°N as the boundary, the CG lightning density is generally 6 to 12 f ashes/(km2·a) in the northern part and 2.4 to 4.8 f ashes/( km2·a) in the southern part of the analyzed area. At the same time, the CG lightning density on land is higher than those in lake and oceanic areas, while the percentage of PCG lightning and the large current CG lightning in the oceanic area is signif cantly higher than those on land. The time changes of spatial distribution of CG lightning show that CG lightning activities mainly appear on the land in the afternoon and near the water in the morning, while in other periods are transitional, which is closely related to the heating effect of the underlying surface. (Zheng Dong)

3 Simulation of the electrif cation of a tropical cyclone using the WRF-Electric model

Inductive and non-inductive electrif cation schemes and a bulk discharge parameterization are introduced into the Milbrandt and Morrison two-moment microphysical schemes in the WRF-ARW model. The model with electrical processes, referred to as WRF-Electric, is able to simulate charge density and lightning distribution in storms. Four different charge separation schemes (TGZ, GZ, SP98 and RR) are introduced into the two microphysics schemes. Fig. 8 shows the frame of the WRF-Electric model.

The evolution of the electrification of an idealized Tropical Cyclone (TC) is simulated by the WRFElectric model. The characteristics of TC lightning can be further understood by comprehending the electrif cation of TCs. This study makes an attempt to illustrate the evolution of the charge structure of TCs. The results of this study can be supported by most of the previous observations and simulations.

The results indicate that the eyewall generally exhibits an inverted dipole charge structure with negative being above positive. In the intensif cation stage, however, the extremely tall towers of the eyewall may exhibit a normal tripole structure with a main negative region found between two regions of positive charge. The outer spiral rainband cells display a simple normal dipole structure in all stages. Further analyses indicate that the differences in charge structures are associated with different updrafts and particle distributions. Weak updrafts, together with a coexistence region of different particles at lower levels in the eyewall, result in charging processes that occur mainly in the positive graupel charging zone (PGCZ). In the intensification stage, the occurrence of charging processes in both positive and negative graupel charging zones are associated with strong updrafts in the extremely tall towers. In addition, the coexistence region of graupel and ice crystals is mainly situated at upper levels in the outer rainband, so the charging processes mainly occur in the negative graupel charging zone (NGCZ). The conceptual model of the evolution of the charge structure of TCs is built (Fig. 8–9). (Wang Fei, Xu Liangtao)

4 Research working progress about lightning operational application

4.1 Promotion training of the lightning detection, nowcasting and warning system

“2013 Meteorological Monitoring and Disaster Warning Project”, the second stage of the national training on CAMS_LNWS, was held by IAS in Anhui branch of the China Meteorological Administration Training Center during 16–19 September 2014. A total of 65 staff members from meteorological stations and lightning protection centers at national level and from the Meteorological Center of the North China Air Traff c Management Bureau of Civil Aviation Administration of China participated in this course.

The IAS revised the context of the course and updated the teaching materials according to the feedback of the first training course held in 2013. The training course covered lightning physics, lightning disaster, lightning detection, lightning warning, and lightning protection engineering. The course, which was highly praised by the participants, promoted the further application of research outputs to operations and services. (Yao Wen, Meng Qing, Zhang Wenjuan)

4.2 Application of Lightning Nowcasting and Warning System (CAMS_LNWS)

After localization test in the Lightning Protection Center of Guangdong Province, the CAMS_LNWS was put into operation in Guangdong on April 20, 2014. During the operational period, it made several successful early warnings and the evaluation results of POD (Probability of Detection), FAR (False Alarm Rate) and TS (Threat Score) showed that all warnings performed well. Now, the products of CAMS_LNWS have been used in the services for ultra-high buildings, hazardous chemicals sites and other key areas. It can provide an important reference for decisions to be made for work time arrangement, lightning risk avoidance, and safety production.( Yao Wen, Meng Qing)