生態(tài)與農(nóng)業(yè)氣象研究進(jìn)展

生態(tài)環(huán)境與農(nóng)業(yè)氣象
Ecological Environment and Agrometeorology

生態(tài)與農(nóng)業(yè)氣象研究進(jìn)展

2012年，生態(tài)環(huán)境與農(nóng)業(yè)氣象研究所在生態(tài)與農(nóng)業(yè)氣象理論及應(yīng)用、農(nóng)業(yè)氣象防災(zāi)減災(zāi)、生態(tài)與農(nóng)業(yè)對(duì)氣候變化的響應(yīng)與適應(yīng)3個(gè)領(lǐng)域取得了一系列成果，部分成果得到了推廣應(yīng)用。

1 生態(tài)與農(nóng)業(yè)氣象理論及應(yīng)用技術(shù)

1.1 基于集合卡爾曼濾波的PyWOFOST模型在東北玉米估產(chǎn)中的適用性研究

在基于集合卡爾曼濾波（EnKF）構(gòu)建的遙感信息-作物模型耦合模型（PyWOFOST）的基礎(chǔ)上，建立了以葉面積指數(shù)（LAI）為結(jié)合點(diǎn)、使用MODIS LAI數(shù)據(jù)作為外部同化數(shù)據(jù)的中國(guó)東北玉米同化模擬模型。LAI的模擬結(jié)果表明，同化后的模擬LAI明顯比同化前的模擬LAI更接近實(shí)測(cè)值，特別是同化前模擬較差的部分。產(chǎn)量模擬結(jié)果表明，盡管同化外部觀測(cè)信息后，存在一個(gè)與實(shí)測(cè)產(chǎn)量相對(duì)接近的同化模擬產(chǎn)量，但由于WOFOST模型本身無(wú)法準(zhǔn)確地模擬嚴(yán)重災(zāi)害條件下的作物生長(zhǎng)發(fā)育情況，因此即使同化后的模擬產(chǎn)量較同化前有所改進(jìn)，但仍與實(shí)測(cè)產(chǎn)量存在一定誤差。（趙艷霞）

1.2 華北地區(qū)冬小麥潛在產(chǎn)量和雨養(yǎng)產(chǎn)量模擬研究

自主研發(fā)了適合華北地區(qū)冬小麥研究的區(qū)域模擬系統(tǒng)。收集整理了華北平原40個(gè)站點(diǎn)的模型所需數(shù)據(jù)，驅(qū)動(dòng)區(qū)域模擬系統(tǒng)，模擬了2003—2007年冬小麥生長(zhǎng)狀況。冬小麥潛在產(chǎn)量高值區(qū)集中在華北平原中部一帶，南北兩端較低，其中北部又比南部稍高；平均雨養(yǎng)產(chǎn)量呈南北高、中間低的趨勢(shì)。北部產(chǎn)量高的原因是潛在產(chǎn)量較高，且生育期內(nèi)的降水相對(duì)較多。平原中部產(chǎn)量低的原因是生育期內(nèi)降水較少。南部產(chǎn)量相對(duì)較高的原因是該區(qū)域是降水量最高的區(qū)域。從模擬結(jié)果看，潛在產(chǎn)量和雨養(yǎng)產(chǎn)量的模擬結(jié)果基本合理地反映了生產(chǎn)潛力的實(shí)際，表明所建立的模型能較好地模擬區(qū)域冬小麥的生長(zhǎng)發(fā)育以及產(chǎn)量形成，適用于華北地區(qū)作物的模擬研究。(鄔定榮)

1.3 青藏高原探空大氣水汽總量誤差及其訂正方法

以地基GPS遙感的大氣水汽總量（GPS_PW）為參照標(biāo)準(zhǔn)，對(duì)拉薩（1999—2010年）和那曲（2003年）探空觀測(cè)的大氣水汽總量（RS_PW）進(jìn)行對(duì)比分析和大氣水汽總量（PW）偏差（GPS_PW-RS_PW）訂正。結(jié)果表明，近10多年拉薩站RS_PW比GPS_PW明顯偏小，偏小程度隨使用不同的探空儀而異。近10多年大氣水汽總量偏差呈減少趨勢(shì)，這與探空儀性能改進(jìn)有關(guān)。分析發(fā)現(xiàn)，青藏高原大氣水汽總量偏差具有明顯季節(jié)變化和日變化特征，夏季比冬季明顯，12∶00比00∶00（世界時(shí)）明顯。太陽(yáng)輻射加熱與氣溫的日變化和季節(jié)變化是造成高原大氣水汽總量偏差日變化和季節(jié)變化的重要原因。據(jù)此，提出了高原大氣水汽總量偏差訂正方法，并以拉薩和那曲站為例進(jìn)行大氣水汽總量偏差訂正，訂正后的大氣水汽總量系統(tǒng)偏差顯著減小，隨機(jī)誤差也相應(yīng)得到改善。（梁宏）

1.4 華北夏玉米生長(zhǎng)模型(GCGA Maize)的初步構(gòu)建

積溫法（TSUM）、熱量單位法（TCHU）和熱量累積法（THU）等幾種常用發(fā)育模式的比較結(jié)果顯示，THU的方差和變異系數(shù)均最小，穩(wěn)定性最高。進(jìn)一步構(gòu)建了發(fā)育階段溫度強(qiáng)度和土壤水分對(duì)發(fā)育進(jìn)程的訂正模式，使發(fā)育模式的模擬能力得到改善?；贔arquhar等的光合生化理論構(gòu)建了夏玉米葉片光合生化模式，并與簡(jiǎn)單的負(fù)指數(shù)模式進(jìn)行了模擬比較，結(jié)果表明前者的總體擬合效果更好。根據(jù)土壤水分運(yùn)動(dòng)方程及其簡(jiǎn)化求解方法建立了農(nóng)田土壤水分運(yùn)移模式，并與土壤水分平衡模式進(jìn)行比較發(fā)現(xiàn)，水分平衡模式的反應(yīng)更靈敏，而水分運(yùn)移模式的模擬值變化較小，但更穩(wěn)定。探討了基于觀測(cè)數(shù)據(jù)同化的作物生長(zhǎng)模型參數(shù)估計(jì)方法。聯(lián)立以上發(fā)育模式、光合生化模式、水分運(yùn)移模式以及參數(shù)估計(jì)方法，初步建成了華北夏玉米生長(zhǎng)模型（GCGA Maize）。（馬玉平）

1.5 基于遙感數(shù)據(jù)與作物生長(zhǎng)模型同化的華北夏玉米長(zhǎng)勢(shì)評(píng)價(jià)

開展了觀測(cè)數(shù)據(jù)與作物生長(zhǎng)模型同化方法的正確性驗(yàn)證，通過(guò)模型狀態(tài)變量對(duì)不同參數(shù)或變量初值的敏感性分析、觀測(cè)數(shù)據(jù)對(duì)敏感參數(shù)的約束性分析，確定了模型的待優(yōu)化參數(shù)，利用優(yōu)化算法實(shí)現(xiàn)遙感數(shù)據(jù)與作物生長(zhǎng)模型的同化。利用遙感作物生長(zhǎng)同化模型模擬的不同年份生物量動(dòng)態(tài)累積過(guò)程及最終生物量與同期平均氣候下生物量的比較，確定了作物長(zhǎng)勢(shì)評(píng)價(jià)指標(biāo)。在作物生育期結(jié)束獲得全生育期氣象資料后，根據(jù)模擬地上總干重的大小進(jìn)行了年度作物長(zhǎng)勢(shì)評(píng)價(jià)。利用各地實(shí)測(cè)作物出苗日期、品種熟性數(shù)據(jù)以及實(shí)時(shí)逐日氣象數(shù)據(jù)驅(qū)動(dòng)遙感作物生長(zhǎng)同化模型，計(jì)算不同年份該日作物累積生物量與同期平均氣候條件下生物量的差值百分比，動(dòng)態(tài)評(píng)價(jià)了生育期內(nèi)各時(shí)段夏玉米長(zhǎng)勢(shì)變化。（馬玉平）

1.6 西北春小麥生育期與氣溫變化的關(guān)系研究

以典型春小麥灌溉耕作區(qū)武威農(nóng)試站1990—2011年和雨養(yǎng)耕作區(qū)定西農(nóng)試站1986—2011年觀測(cè)數(shù)據(jù)為依據(jù)，統(tǒng)計(jì)分析了春小麥發(fā)育期變化與積溫變化的關(guān)系。武威春小麥1990—2011年發(fā)育期縮短了20天，發(fā)育期有顯著縮短趨勢(shì)（P＜0.01，r =0.827），定西春小麥1986—2011年發(fā)育期縮短了13天，發(fā)育期也有顯著縮短趨勢(shì)（P＜0.05，r = 0.463）。研究表明，自苗期開始的積溫持續(xù)增加是導(dǎo)致西北地區(qū)春小麥發(fā)育期提前和生育期縮短的主要原因（圖1）。（俄有浩）

1.7 基于TOPMODEL模型和新安江水文模型構(gòu)建的新水文模型

考慮到TOPMODEL模型刻畫地下水動(dòng)態(tài)變化時(shí)存在與實(shí)際情況偏差的問(wèn)題(平均地下水水位線高于地面)和新安江模型無(wú)法有效表達(dá)地下水位的動(dòng)態(tài)變化，研究提出了降雨-徑流模型XXT。該模型的一個(gè)非常重要的特征是把土壤蓄水容量曲線和地下水水位線緊密關(guān)聯(lián)，形成新的蓄水容量曲線概念，并與TOPMODEL模型土壤分層結(jié)構(gòu)及地下產(chǎn)流方程結(jié)合，通過(guò)嚴(yán)密地?cái)?shù)學(xué)推導(dǎo)形成了新產(chǎn)流方程和新的水量平衡方程，在此基礎(chǔ)上構(gòu)建了新型降雨-徑流模型XXT（第1個(gè)X代表新安江，第2個(gè)X借鑒了農(nóng)學(xué)中的“雜交”符號(hào)，T代表TOPMODEL），該模型主要由4個(gè)模塊構(gòu)成，即數(shù)據(jù)前處理模塊、土壤濕度空間分布可視化模塊、產(chǎn)流模塊及匯流模塊。（趙俊芳）

1.8 水熱影響的植物展葉期模型

大量研究表明，水分是影響植物展葉期的重要因子，特別是前一年降水對(duì)次年的返青期影響顯著。通過(guò)深入分析不同時(shí)期降水對(duì)植物展葉期的作用，構(gòu)建了基于降水的水分因子分模型，結(jié)合積溫模型，發(fā)展了一個(gè)普適的水熱物候模型。對(duì)東北地區(qū)植物展葉期的驗(yàn)證表明，模型具有很好的適用性，模擬誤差小于4天，優(yōu)于國(guó)內(nèi)外同類模型。模型定量解釋了植物展葉期對(duì)相互作用的水熱因子響應(yīng)機(jī)制，對(duì)準(zhǔn)確評(píng)估生態(tài)系統(tǒng)生產(chǎn)力與碳收支具有重要意義。（周廣勝）

1.9 草原光能利用率參數(shù)化

基于研究站點(diǎn)數(shù)據(jù)分析，比較了內(nèi)蒙古典型草原與荒漠草原的光能利用率季節(jié)及年際變異。結(jié)果表明，盡管典型草原的總初級(jí)生產(chǎn)力顯著大于荒漠草原，其光能利用率卻顯著小于荒漠草原。水分是內(nèi)蒙古典型草原與荒漠草原光能利用率的主要限制因子，而溫度的影響不顯著。光能利用率對(duì)環(huán)境因子的敏感性(直線回歸的斜率)在兩類草原間存在顯著差異。散射輻射在典型草原中對(duì)光能利用率具有一定的促進(jìn)作用，但在荒漠草原中這種促進(jìn)作用不顯著。（周廣勝）

2 農(nóng)業(yè)氣象防災(zāi)減災(zāi)技術(shù)研究

2.1 不同等級(jí)干熱風(fēng)災(zāi)損評(píng)估模型的初步建立

采用黃淮海地區(qū)逐日氣象資料和小麥發(fā)育期、產(chǎn)量、干熱風(fēng)災(zāi)害等數(shù)據(jù)，基于公認(rèn)的冬小麥干熱風(fēng)災(zāi)害指標(biāo)，計(jì)算了不同等級(jí)干熱風(fēng)危害指數(shù)，構(gòu)建了不同等級(jí)干熱風(fēng)災(zāi)損評(píng)估模型并進(jìn)行了評(píng)估，得到一些初步結(jié)論。黃淮海各地區(qū)冬小麥不同發(fā)育時(shí)段的干熱風(fēng)危害指數(shù)平均為：輕度干熱風(fēng)危害下，各地區(qū)干熱風(fēng)危害指數(shù)在抽穗-開花時(shí)段最大，開花-乳熟時(shí)段居中，乳熟-成熟時(shí)段最小；重度干熱風(fēng)危害下，各地區(qū)干熱風(fēng)危害指數(shù)在乳熟-成熟時(shí)段最大，抽穗-開花時(shí)段居中，開花-乳熟時(shí)段最小。從抽穗到成熟，輕度干熱風(fēng)危害指數(shù)最大，重度干熱風(fēng)危害指數(shù)最小。此外，基于構(gòu)建的不同等級(jí)干熱風(fēng)災(zāi)損評(píng)估模型，計(jì)算了發(fā)生干熱風(fēng)后的實(shí)測(cè)產(chǎn)量比灌漿期未受災(zāi)的正常預(yù)計(jì)產(chǎn)量的減產(chǎn)百分比。結(jié)果顯示，重度干熱風(fēng)危害下，小麥減產(chǎn)率平均為26.1%。（趙艷霞）

2.2 農(nóng)業(yè)干旱災(zāi)害靜態(tài)和動(dòng)態(tài)風(fēng)險(xiǎn)評(píng)估概念模型和指標(biāo)體系

農(nóng)業(yè)干旱災(zāi)害的風(fēng)險(xiǎn)形成是多因素綜合作用的結(jié)果。作物干旱除受氣候條件影響外，土壤性質(zhì)、地貌類型、地下水狀況、作物本身的需水特征以及干旱災(zāi)害管理水平、區(qū)域的抗旱減災(zāi)能力等人為因素都影響其干旱災(zāi)害風(fēng)險(xiǎn)的發(fā)生及強(qiáng)度。因此，基于農(nóng)業(yè)干旱災(zāi)害風(fēng)險(xiǎn)形成機(jī)理，從農(nóng)業(yè)干旱災(zāi)害發(fā)生學(xué)的角度建立農(nóng)業(yè)干旱災(zāi)害風(fēng)險(xiǎn)概念框架，并在此基礎(chǔ)上選取遼西北最主要的玉米作物作為研究示范對(duì)象，建立了農(nóng)業(yè)干旱災(zāi)害風(fēng)險(xiǎn)評(píng)價(jià)的指標(biāo)體系。（王春乙）

2.3 全國(guó)農(nóng)業(yè)洪澇災(zāi)害風(fēng)險(xiǎn)分析與評(píng)估

采用逐日降水量資料、地形數(shù)據(jù)、基礎(chǔ)地理數(shù)據(jù)、公里網(wǎng)格土地利用數(shù)據(jù)、公里網(wǎng)格人口數(shù)據(jù)和公里網(wǎng)格GDP數(shù)據(jù)等，綜合考慮降水、地形、河流與湖泊分布對(duì)洪澇災(zāi)害危險(xiǎn)性的影響，以及耕地面積百分比、人口密度、GDP密度分布對(duì)洪澇災(zāi)害脆弱性的影響，初步構(gòu)建了農(nóng)業(yè)洪澇災(zāi)害危險(xiǎn)性、脆弱性評(píng)價(jià)模型、等級(jí)評(píng)價(jià)指標(biāo)和災(zāi)害風(fēng)險(xiǎn)評(píng)估模型，分8級(jí)進(jìn)行全國(guó)農(nóng)業(yè)洪澇災(zāi)害風(fēng)險(xiǎn)評(píng)估。結(jié)果顯示，中國(guó)中東部地區(qū)以4級(jí)以上風(fēng)險(xiǎn)為主，西部地區(qū)以3級(jí)以下風(fēng)險(xiǎn)為主。8級(jí)以上洪澇災(zāi)害風(fēng)險(xiǎn)主要位于河南省東南部、淮北平原西部、江西鄱陽(yáng)湖周邊、湖南洞庭湖周邊、江漢平原、安徽沿江平原、江蘇中部及沿江地區(qū)、浙江東北部、四川盆地西側(cè)、廣東沿海和珠江三角洲地區(qū)、廣西沿海和中部、海南島中北部沿海地區(qū)等。（霍治國(guó)）

2.4 主要農(nóng)作物病蟲害對(duì)氣候變化的響應(yīng)

針對(duì)全國(guó)農(nóng)作物病蟲害、病害、蟲害以及小麥病蟲害、小麥白粉病，研究并揭示了氣候變化導(dǎo)致的全年及作物全生育期的光、溫、水等氣象因子變化對(duì)不同病蟲害的影響關(guān)系。如冬小麥全生育期平均溫度為10.2 °C，且以0.46 °C/10a的速率升高；平均溫度每升高1 °C，可使冬小麥白粉病發(fā)生面積增加255.3萬(wàn)hm2。平均降水強(qiáng)度為5.6 mm/d，且以0.18 mm/(d 10a)的速率增加；平均降水強(qiáng)度每增加1 mm/d，將使冬小麥白粉病發(fā)生面積增加270.8萬(wàn)hm2。平均日照時(shí)數(shù)為1139.3 h，且以9.90 h/10a的速率減少；平均日照時(shí)數(shù)每減少100 h，將使冬小麥白粉病發(fā)生面積增加269.0萬(wàn)hm2。在氣候變化導(dǎo)致的小麥全生育期溫度、降水、日照因子變化中，溫度增加對(duì)冬小麥白粉病發(fā)生面積增加的影響最為顯著，其次為日照時(shí)數(shù)減少，再次為平均降水強(qiáng)度增大。（霍治國(guó)）

2.5 基于作物生長(zhǎng)模型的淮河流域玉米旱澇損失評(píng)估

利用玉米生長(zhǎng)模型定義了生物量旱澇脅迫指數(shù)（TAGPI）、產(chǎn)量旱澇脅迫指數(shù)（WSOI）以及水分影響指數(shù)（RFWSI）。以WSOI指數(shù)為主要評(píng)價(jià)依據(jù)，其余為輔助評(píng)價(jià)指標(biāo)，利用玉米生長(zhǎng)模型模擬淮河流域多個(gè)站點(diǎn)48年的生物量旱澇脅迫指數(shù)、產(chǎn)量旱澇脅迫指數(shù)和水分影響指數(shù)，根據(jù)其概率分布確定了全生育期和不同生育階段的旱澇指標(biāo)。假定后期氣象條件正常的情況下，利用模擬最終產(chǎn)量結(jié)合旱澇指標(biāo)可以進(jìn)行旱澇災(zāi)害損失的定量評(píng)估。利用不斷更新的實(shí)時(shí)氣象數(shù)據(jù)驅(qū)動(dòng)作物生長(zhǎng)模型，結(jié)合逐日旱澇指標(biāo)，可以開展旱澇災(zāi)害損失的時(shí)間動(dòng)態(tài)評(píng)估。（馬玉平）

2.6 全國(guó)農(nóng)業(yè)臺(tái)風(fēng)和暴雨災(zāi)害綜合風(fēng)險(xiǎn)評(píng)估

針對(duì)全國(guó)農(nóng)業(yè)臺(tái)風(fēng)、暴雨災(zāi)害綜合風(fēng)險(xiǎn)評(píng)估，以降水、大風(fēng)、農(nóng)業(yè)產(chǎn)值、農(nóng)作物播種面積、耕地面積、高程、高程標(biāo)準(zhǔn)差、河網(wǎng)密度、距離水體的遠(yuǎn)近和植被覆蓋度為指標(biāo)，建立了全國(guó)農(nóng)業(yè)臺(tái)風(fēng)災(zāi)害風(fēng)險(xiǎn)指數(shù)模型。以年暴雨日數(shù)、年暴雨累積量、年暴雨過(guò)程次數(shù)、年暴雨最大過(guò)程累積降水量、年極端降水日數(shù)、年極端降水累積量、年極端降水過(guò)程次數(shù)和年極端降水平均集中度為指標(biāo)，建立了全國(guó)農(nóng)業(yè)暴雨綜合氣候風(fēng)險(xiǎn)指數(shù)模型。根據(jù)風(fēng)險(xiǎn)疊加原理，建立了全國(guó)農(nóng)業(yè)臺(tái)風(fēng)和暴雨綜合災(zāi)害風(fēng)險(xiǎn)指數(shù)模型。利用自然斷點(diǎn)分級(jí)法將風(fēng)險(xiǎn)指數(shù)劃分為高風(fēng)險(xiǎn)、次高風(fēng)險(xiǎn)、中等風(fēng)險(xiǎn)、次低風(fēng)險(xiǎn)和低風(fēng)險(xiǎn)5個(gè)等級(jí)，基于GIS繪制了空間分辨率為5 km × 5 km風(fēng)險(xiǎn)區(qū)劃圖(圖2)。（毛飛）

2.7 東北草原蝗蟲氣象監(jiān)測(cè)預(yù)報(bào)與災(zāi)損評(píng)估技術(shù)

為研究東北草原亞洲飛蝗孵化進(jìn)程與熱量條件的關(guān)系，利用人工氣候箱進(jìn)行了平均氣溫、最低氣溫、積溫對(duì)亞洲飛蝗越冬卵孵化影響的試驗(yàn)。結(jié)果表明，出蝻數(shù)、出蝻率、累積出蝻數(shù)、累積出蝻率均隨日平均氣溫升高而增大。日最低氣溫連續(xù)3天穩(wěn)定通過(guò)25 °C時(shí)，蝗蟲開始孵化出蝻；在26 °C左右時(shí)孵化最快，低于25 °C時(shí)，蝗蟲出蝻速率緩慢。25 °C以上有效積溫達(dá)到11.6 °C d，活動(dòng)積溫達(dá)到211.6 °C d時(shí)，蝗卵開始孵化出土。當(dāng)有效積溫增加到20 °C d，活動(dòng)積溫達(dá)320 °Cd后，孵化速度回落，在達(dá)到另一個(gè)小高峰后孵化過(guò)程結(jié)束。(白月明)

3 農(nóng)業(yè)應(yīng)對(duì)氣候變化研究

3.1 氣候變化對(duì)東北地區(qū)熱量資源及玉米溫度適宜度的影響

利用RegCM3 模式輸出的東北3省1951—2100年逐日溫度資料，分析了東北3省熱量資源和玉米不同生育期氣溫適宜度的時(shí)空分布。結(jié)果表明， 1951—2100 年?yáng)|北地區(qū)熱量資源顯著增加，穩(wěn)定通過(guò)10 °C初日不斷提前，初日在4月25日之前的區(qū)域北界向東向北擴(kuò)展；≥10 °C活動(dòng)積溫大于3000 °Cd的區(qū)域面積不斷增加，適宜種植晚熟玉米的區(qū)域面積也不斷增加； 1951—2100年玉米播種-出苗期以及出苗-抽雄期的氣溫適宜度隨時(shí)間逐漸升高，1951—2040年遼寧省玉米抽雄-成熟期以及全生育期的氣溫適宜度較高，而黑龍江省較低；2041—2100年遼寧省玉米抽雄-成熟期以及全生育期的氣溫適宜度逐漸降低，吉林省東部和黑龍江省呈逐漸增加的趨勢(shì)。（郭建平）

3.2 基于實(shí)際生育期的東北地區(qū)玉米氣候生產(chǎn)潛力研究

利用1981—2006年?yáng)|北3省30個(gè)農(nóng)業(yè)氣象觀測(cè)站玉米生育期資料和1961—2006年70個(gè)氣象臺(tái)站的氣象資料，統(tǒng)計(jì)分析了東北3省玉米主要發(fā)育期的變化特點(diǎn)，研究了東北地區(qū)玉米實(shí)際生長(zhǎng)期間的氣候生產(chǎn)潛力的變化特征。結(jié)果表明，東北地區(qū)玉米光合生產(chǎn)潛力、光溫生產(chǎn)潛力均呈現(xiàn)由西向東、由北向南遞減的趨勢(shì)，而氣候生產(chǎn)潛力呈現(xiàn)由西南向東北逐漸遞減的趨勢(shì)。46年來(lái)，東北地區(qū)光合生產(chǎn)潛力、氣候生產(chǎn)潛力呈逐年減少的趨勢(shì)，而光溫生產(chǎn)潛力遼寧、吉林呈現(xiàn)逐年減少的趨勢(shì)，黑龍江呈現(xiàn)增加趨勢(shì)。從年代際來(lái)看，20世紀(jì)60年代最高，進(jìn)入90年代后顯著減少，2001年后進(jìn)入氣候生產(chǎn)潛力的最小時(shí)期。（郭建平）

3.3 中國(guó)北方春小麥種植區(qū)氣候適宜性分布

以Maxent模型為基礎(chǔ)，利用春小麥種植空間分布數(shù)據(jù)和1961—2008年全國(guó)560個(gè)臺(tái)站日值氣象觀測(cè)數(shù)據(jù)的≥0 °C積溫、≥0 °C持續(xù)日數(shù)、年降水量、1月平均氣溫、≥3 °C積溫、≥3 °C持續(xù)日數(shù)、≥3 °C蒸散發(fā)量、≥3 °C總輻射等8個(gè)春小麥空間分布環(huán)境數(shù)據(jù)，模擬了中國(guó)北方春小麥氣候適宜性分布區(qū)域，氣候適宜性分布區(qū)域概率達(dá)到50% ~80%。與當(dāng)前北方春小麥種植區(qū)域?qū)Ρ劝l(fā)現(xiàn)，具有較高的一致性(圖3)。（俄有浩）

3.4 氣候變化背景下黃淮海地區(qū)冬小麥干熱風(fēng)演變趨勢(shì)

基于黃淮海地區(qū)52個(gè)氣象站50年（1961—2010年）5月11日至6月10日逐日最高氣溫、14∶00空氣相對(duì)濕度和風(fēng)速資料，利用氣候趨勢(shì)和相似性分析方法分析了黃淮海地區(qū)冬小麥生長(zhǎng)后期干熱風(fēng)的發(fā)生趨勢(shì)。結(jié)果表明，黃淮海大部分地區(qū)平均日最高氣溫呈現(xiàn)下降趨勢(shì)，與氣候總體變暖的趨勢(shì)相反；14∶00空氣相對(duì)濕度呈升高趨勢(shì)，也與大氣總體變干的趨勢(shì)相反；14∶00風(fēng)速呈現(xiàn)減小趨勢(shì)，與北方地區(qū)植樹造林有較大關(guān)系，從而使得單一要素達(dá)到干熱風(fēng)標(biāo)準(zhǔn)的日數(shù)呈減少趨勢(shì)。說(shuō)明黃淮海地區(qū)出現(xiàn)各級(jí)干熱風(fēng)的日數(shù)有減少趨勢(shì)。（郭建平）

3.5 錦州玉米和盤錦水稻農(nóng)田生態(tài)系統(tǒng)碳水平衡研究

對(duì)遼寧省錦州玉米和盤錦水稻農(nóng)田生態(tài)系統(tǒng)中碳水平衡的研究表明，玉米和水稻田年CO2凈交換量(NEE)分別為280 gC/m2和195 gC/m2，二者均是大氣CO2的匯；生態(tài)系統(tǒng)碳固定的同時(shí)，玉米和水稻田生態(tài)系統(tǒng)的年蒸散量(ET)分別為407 mm和660 mm。以生態(tài)系統(tǒng)總初級(jí)生產(chǎn)力(GPP)與ET的比值(生態(tài)系統(tǒng)通過(guò)光合作用吸收固定單位CO2所消耗水量)來(lái)衡量生態(tài)系統(tǒng)水分利用效率(WUE)，玉米、水稻田生態(tài)系統(tǒng)WUE分別為11.3 g/kg和4.4 g/kg。玉米、水稻田生態(tài)系統(tǒng)總初級(jí)生產(chǎn)力的季節(jié)變化與溫度密切相關(guān)，且玉米田生態(tài)系統(tǒng)總初級(jí)生產(chǎn)力高于水稻田；由于兩類農(nóng)田GPP和ET季節(jié)變化趨勢(shì)的差異，玉米田生態(tài)系統(tǒng)WUE在玉米出苗后呈上升趨勢(shì)，直至玉米成熟期達(dá)到峰值后下降，水稻田生態(tài)系統(tǒng)WUE的峰值則出現(xiàn)在營(yíng)養(yǎng)和生殖生長(zhǎng)并進(jìn)階段，比玉米田生態(tài)系統(tǒng)WUE峰值出現(xiàn)早（圖4）。（周莉）

4 農(nóng)業(yè)氣象科技成果推廣應(yīng)用

4.1 西北氣候脆弱地區(qū)農(nóng)業(yè)干旱及農(nóng)田灌溉預(yù)報(bào)技術(shù)推廣應(yīng)用

以陜西、甘肅、寧夏農(nóng)業(yè)干旱為研究對(duì)象，建立了氣象和土壤數(shù)據(jù)庫(kù)，改進(jìn)了作物干旱模式，并與分布式水文模式相結(jié)合，基于GIS技術(shù)研發(fā)了“西北農(nóng)業(yè)干旱及農(nóng)田灌溉預(yù)報(bào)系統(tǒng)”，實(shí)現(xiàn)了西北農(nóng)業(yè)干旱、農(nóng)田灌溉預(yù)報(bào)及服務(wù)產(chǎn)品的制作功能。對(duì)西北地區(qū)氣象業(yè)務(wù)服務(wù)人員進(jìn)行了培訓(xùn)，并在西北省級(jí)氣象業(yè)務(wù)部門進(jìn)行了推廣應(yīng)用，制作發(fā)布服務(wù)產(chǎn)品100期以上，該項(xiàng)目推廣受益面積超過(guò)6.67萬(wàn)hm2，農(nóng)業(yè)干旱預(yù)報(bào)準(zhǔn)確率總體達(dá)80%以上，取得了較好的社會(huì)和經(jīng)濟(jì)效益。（劉建棟）

4.2 東北地區(qū)玉米低溫冷害動(dòng)態(tài)監(jiān)測(cè)預(yù)警技術(shù)推廣應(yīng)用

在“十一五”科技支撐計(jì)劃課題研究成果的基礎(chǔ)上，對(duì)東北地區(qū)玉米區(qū)域分布進(jìn)行了進(jìn)一步細(xì)分，補(bǔ)充和利用最新的資料對(duì)低溫冷害預(yù)警模式參數(shù)進(jìn)行重新計(jì)算，完善了原有模式。通過(guò)模式的完善已基本完成了區(qū)域和分省低溫冷害監(jiān)測(cè)預(yù)警業(yè)務(wù)系統(tǒng)的研發(fā)。2012年?yáng)|北3省共發(fā)布各類服務(wù)產(chǎn)品10多期，取得了較好的效果。（郭建平）

圖1 武威和定西農(nóng)試站春小麥發(fā)育期變化及趨勢(shì)Fig1 The variability and change trend of spring-wheat growth period

圖2 全國(guó)農(nóng)業(yè)臺(tái)風(fēng)和暴雨災(zāi)害綜合風(fēng)險(xiǎn)區(qū)劃Fig2 Agricultural typhoon and rainstorm disaster comprehensive risk regionalization

圖4 東北農(nóng)田生態(tài)系統(tǒng)碳水通量平衡關(guān)系Fig4 Carbon and water budget of agro-ecosystem in Northeast China

Progress in Ecology and Agrometeorology Research

In 2012, the Institute of Eco-environment and Agrometeorology of CAMS obtained signifcant research progress on agrometeorological theories and application techniques, agrometeorological disaster prevention and mitigation, and response of agriculture to climate change. Some results have been applied in real-time operations.

1 Agrometeorological theories and application techniques

1.1 Application of ensemble Kalman flter to Northeast China maize yield estimation from the PyWOFOST crop model

By coupling with an ensemble Kalman flter (EnKF), the coupled remote sensing information-crop model (PyWOFOST) with leaf area index (LAI) as the coupling point was used to estimate maize yield in Northeast China. The LAI simulation results show that the LAI values from the coupled model with EnKF assimilation are more consistent with the observations than that without assimilation. However, the maize yield simulation results show that there is still a large error in the crop yield compared with the observations when there occurs a severe disaster, although certain improvements have been brought up by the EnKF assimilation. The reasons for the error are being investigated. (Zhao Yanxia)

1.2 Simulation of the potential and rainfed yield of winter wheat in the North China Plain

A regional simulation system was developed. The system is suitable for simulation of winter wheat growth and potential and rainfed yield in the North China Plain (NCP). Data from 40 meteorological stations in the NCP were collected and used as input to the system to simulate the potential and rainfed yield of winter wheat in the period 2003—2007. The results show that potential yield of winter wheat is decreased from the central NCP to the north and south, with higher values in the north than the south, while rainfed yield has the reverse trends. This is because rainfall is the key factor that controls winter wheat rainfed yield in the NCP. Middle plain has the lowest rainfall, thus has the lowest rainfed yield. Rainfed yield is higher in the north than in the middle because north has both higher rainfall and higher potential yield, with the latter being the base of the rainfed yield. Rainfed yield is higher in the south than in the middle because south plain has the highest rainfall. The simulation results are consistent with the observation, showing that the simulation system is capable to simulate winter wheat growth in the NCP, and can be used in the next study. (Wu Dingrong)

1.3 Random errors of the radiosonde precipitable water over the Tibetan Plateau and associated calibration methods

The characteristics of the systematic and random errors of the radiosonde (RS) precipitable water (PW) data at Lhasa during 1999—2010 and at Naqu in 2003, were compared with ground-based GPS measurements. The results show that RS_PW was signifcantly smaller than GPS_PW at Lhasa. Different types of radiosonde humidity sensors showed different magnitudes of the dry bias of PW. Due to the introduction of the high performance humidity sensors (GST-1), the PW bias was apparently being reduced gradually over the past 10 years. The temporal variation characteristics of the RS_PW dry bias were also investigated. The results show that the RS_PW dry bias exhibited pronounced diurnal and annual variations. The dry bias of RS_PW was much larger at 12∶00 UCT than at 00∶00 UTC, and larger in summer than in winter. Additionally, the causes of diurnal and annual variations of the RS_PW dry bias were investigated. The solar radiative heating to the humidity sensors may have played an important role. It can be seen that the diurnal variation of RS_PW drybias was signifcant partly because air temperature was higher at 12∶00 UTC than at 00∶00 UTC. The annual variation of RS_PW dry bias was pronounced also partly because air temperature was higher in summer than in winter. The calibration methods for the RS_PW dry bias were developed and applied to the GZZ-2 and GTS-1 sounding PW datasets at Lhasa and Naqu. The corrections greatly improved the accuracy of the RS_PW data.(Liang Hong)

1.4 Construction of the Grid Crop Growth and Assess model (GCGA Maize)

Comparisons of the accumulated temperature, heat unit, and heat cumulation (THU) methods showed that both the variance and coeffcient of variation of THU were the smallest and its stability was the highest. A correction model that considers the impacts of temperature strength and soil moisture on the crop development process was firstly constructed. A photosynthesis model for summer maize was then constructed based on Farquhar’s photosynthesis biochemistry theory. The ftting effect of this photosynthesis model was better than the simple negative index model. The soil water movement model was established according to the equation of soil water movement and its numerical calculation methods. Compared with the soil water balance model, the soil water movement model was more stable. Moreover, the methods for estimating crop model parameters were discussed based on assimilation of observation data. The Grid Crop Growth and Assess model (GCGA Maize) was fnally set up on the basis of the above models and methods. (Ma Yuping)

1.5 Evaluation of summer maize growth in North China based on the crop growth model with assimilation of remote sensing data

Validation of the assimilation of observation data into the crop growth model was firstly carried out. The optimal and suitable parameters were then obtained based on both the sensitivity analysis of different parameters and initial values and the constraint analysis of sensitive parameters. Assimilation of remote sensing data into crop growth model was fnally achieved by using a derived optimization algorithm. The crop growth index was determined by comparing the dynamic accumulation of biomass simulated by the assimilation model in different years with the biomass under the average climatic conditions. Annual evaluation of maize growth at maturity was carried out based on the total above ground production simulated by the assimilation model. Dynamic evaluation of summer maize growth was then realized by using the assimilation model driven by daily meteorological data in real time. (Ma Yuping)

1.6 Relationship between temperature and spring-wheat growth period in Northwest China

Based on the observation data from Wuwei and Dingxi stations, which represent the irrigation and rainfed crop areas, respectively, the relationship of accumulated temperature and spring-wheat growth period was analyzed. The results show that spring-wheat growth period was reduced by 20 days during 1990—2011 at Wuwei station, and 13 days during 1986—2011 at Dingxi station, implying a signifcant decrease trend (P ＜ 0.01, r = 0.827 for Wuwei and P ＜ 0.05, r = 0.463 for Dingxi). The conclusion is that continuous increase in accumulated temperature occurring from seedling was the main cause of the decrease in the growth period (Fig1). (E Youhao)

1.7 A new hydrological model based on TOPMODEL and Xin'anjiang hydrological model

Considering the shortcomings that TOPMODEL model simulated ground water dynamic changes deviate from the actual situation, i.e., the average ground water level is even above the ground, and Xin'anjiang model can not express the dynamic change of underground water level, we have developed a new model of rainfall-runoff XXT. XXT has an important feature that the soil water storage capacity curve is closely related to ground water level, forming the new concept of water storage capacity curve. Then, XXT was further refned by taking advantage of the TOPMODEL layered soil and underground runoff equations. After rigorous mathematical deduction, new runoff production and water balance equations were obtained. Finally, new rainfall-runoff model XXT was set up, with the frst X representing the Xin'anjiang, the second X the “hybrid”symbol in the agriculture, and T the TOPMODEL. XXT is mainly made up of four modules, namely, the data processing module, the visualization module of spatial distribution of soil moisture, the runoff module, and theconfux module. (Zhao Junfang)

1.8 A temperature-precipitation based leafng model

By analyzing the effects of precipitation in different periods on plant leafng, a precipitation-based submodel was firstly established. Then, a universal temperature-precipitation based leafing model (TP) was developed, combined with the accumulated temperature model. The TP has been proved to be more suitable for simulating leafng of all the plant species in Northeast China than the prior models, with the simulation error less than 4 days. Furthermore, the results indicate that the TP has quantitatively examined the responses of plant leafng to interactive hydrothermic factors and played critical roles in accurately evaluating ecosystem productivity and carbon budget. (Zhou Guangsheng)

1.9 Light use effciency over two temperate steppes

Combining eddy covariance fux data with the fraction of photosynthetically active radiation absorbed by the plant canopy from MODIS, we report the seasonal and interannual variations of light use effciency (LUE) on a typical steppe and a desert steppe in Inner Mongolia, northern China. The results show that both annual average LUE and maximum LUE were higher on the desert steppe than on the typical steppe, despite the higher GPP of the latter. Water availability was the primary limiting factor of LUE at both sites; however, the sensitivity of LUE to water condition differed signifcantly between the two sites. LUE increased with the diffuse radiation ratio on the typical steppe; however, such a trend was not found for the desert steppe.(Zhou Guangsheng)

2 Agrometeorological disaster prevention and mitigation

2.1 Constructing dry hot wind damage evaluation model at different grades

Based on the meteorological data, the developmental period, yield, dry hot wind disasters data of wheat in the Huanghuaihai area, and the recognized indicators of winter wheat dry hot wind disasters, the different grades of dry hot wind hazard index were calculated. Then, the dry hot wind damage evaluation model was constructed at different levels. Preliminary conclusions are as follows. The average hazard index of dry hot wind during different developmental periods of winter wheat is different. Under the light dry hot wind hazard, the average hazard index of dry hot wind is the biggest from tasseling to fowering, smaller from fowering to milking, and least from milking to riping. Under the severe dry hot wind hazard, the average hazard index of dry hot wind is the biggest from milking to riping, smaller from tasseling to fowering, and least from fowering to milking. In addition, based on the built dry hot wind damage evaluation model, the reduction of wheat yield caused by dry hot wind hazard is calculated. The results show that average reduction of wheat yield caused by severe dry hot wind hazard is about 26.1%. (Zhao Yanxia)

2.2 Establishment of the conceptual model and index system for static and dynamic risk assessment

of agro-drought disasters

Agro-drought disaster risk can be attributed to the complexity of multi-factors, including climatic conditions, soil properties, geomorphic type, groundwater status, water requirement of specific crop, and anthropogenic infuences such as the management level of drought disaster, regional capacity of mitigation, prevention to drought disaster, etc. Thereby, a conceptual framework of agro-drought disaster risk was raised in view of agro-drought disaster embryology and the formation mechanism of agro-drought disaster risk. Based on the conceptual model, the indicator system for agro-drought disaster was established and maize grown over the northwestern Liaoning Province was taken as a case study. (Wang Chunyi)

2.3 Analysis and assessment of agricultural food risk in China

Considering the effect of the distribution of precipitation, terrain, rivers, and lakes on food hazard, aswell as the effect of the distribution of arable land percentage, population density, GDP density on flood vulnerability, assessment model of food hazard and vulnerability, level assessment index and risk assessment model were constructed. The index was divided into 8 levels to assess agricultural food risk, based on daily precipitation data, terrain data, basic geographic data, km-grid land use data, km-grid population data, kmgrid GDP data, etc. The results show that the levels of food risk are above 4 in east-central China and below 3 in western China. The food risk of level 8 is mainly located in southeastern Henan, west of Huaibei Plain, areas around Poyang Lake and Dongting Lake, Jianghan Plain, areas along the Yangtze River in Anhui and Jiangsu provinces, central Jiangsu Province, northeastern Zhejiang, west of Sichuan Basin, coastal areas of Guangdong, Pearl River Delta, coastal and central areas of Guangxi Province, northern and central areas of Hainan Province, and coastal areas of Hainan Province. (Huo Zhiguo)

2.4 Effect of climate change on pests and diseases of major crops

The effect of changes of the meteorological factors and crop growth period induced by climate change including sunshine, temperature, precipitation, and so on, on pests and diseases was revealed, in the studies of national crop pests and diseases, pests, diseases, wheat pests and diseases, and powdery mildew. It is found that average temperature is 10.2 °C in the growth period of winter wheat, increasing at the rate of 0.46 °C/10 yr. Powdery mildew occurrence area will increase by 255.3×104hm2when average temperature increases by 1 °C. Average precipitation intensity is 5.6 mm/day, increasing at the rate of 0.18 mm/day per decade. Powdery mildew occurrence area will increase by 270.8×104hm2when average precipitation intensity increases by mm/day. Average sunshine hours are 1139.3 h, decreasing at the rate of 9.9 h per decade. Powdery mildew occurrence area will increase by 269.0×104hm2when average sunshine hour decreases by 100 h. Among the changes of temperature, precipitation and sunshine, the increase of average temperature has the greatest impact on the occurrence of crop diseases, followed by the decrease of sunshine hours and the increase of average precipitation intensity. (Huo Zhiguo)

2.5 Seamless assessment methods of loss from droughts and foods for maize in the Huaihe River Basin based on the crop growth model

The production of drought and food stress index (WSOI), which was the main basis of evaluation, the biomass drought and food stress index (TAGPI), and the water impact index (RFWSI) were defned based on maize growth model. WSOI, TAGPI, and RFWSI were frstly simulated by the maize growth model for 48 years at different stations in the Huaihe River Basin. The thresholds of droughts and foods stress index at different development stages and during the whole growing period were then determined according to probability distribution of the simulation results. Quantitative assessment of losses caused by droughts and floods can be achieved through the combination of simulated final biomass and droughts and floods stress index if the weather conditions are normal. The dynamic assessment of losses can be carried out by means of using the maize growth model driven by daily meteorological data in real time. (Ma Yuping)

2.6 Agricultural typhoon and rainstorm disaster comprehensive risk assessment in China

For the national agricultural typhoon and rainstorm disaster comprehensive risk assessment, the agricultural typhoon disaster risk index models in China were established with the indexes of precipitation, wind, agricultural output value, sown area of crops, cultivated area, elevation, standard deviation of elevation, river density, distance to water and vegetation coverage. The agricultural rainstorm comprehensive climate risk index models in China were also established with the indexes of annual rainstorm days, annual rainstorm accumulation, annual rainstorm process times, accumulated precipitation of the largest annual rainstorm process, annual extreme precipitation days, annual extreme precipitation accumulation, annual extreme precipitation process times and average concentration of annual extreme precipitation. According to the risk superposition principle, the risk index model of agricultural typhoon and rainstorm integrated disaster in China was established. Using natural breaks classifcation, the risk index was divided into 5 classifcations, i.e., highrisk, second high risk, medium risk, second low risk, and low risk. The comprehensive risk zoning map in China was drawn with a spatial resolution of 5 km × 5 km based on GIS (Fig2). (Mao Fei)

2.7 Monitoring, forecasting and loss assessment of Asiatic Migratory Locust in northeast grassland

In order to clarify the relationship of Asiatic Migratory Locust hatching process and heat conditions in northeast grassland, this experiment was conducted in an artificial climate box to study the influence of temperature and effective accumulated temperature on the Asiatic Migratory Locust hatching of the overwintering. The results show that, the hoppers, the rate of hoppers, the accumulation hoppers, and the rate of accumulation hoppers increased with increasing temperature. The locust nymphae began to sprout with the daily minimum temperature stabilizing at 25 °C for 3 days. As the day of temperature increasing, the speed of hatching increased continuously; at about 26 °C, the hatching occurred at the fastest speed. If daily minimum temperature was below 25 °C, the rate of locust hoppers became slow. When effective accumulated temperature reached 11.6 °C·day, the active accumulated temperature reached 211.6 °C·day, locust egg began to hatch. When effective accumulated temperature increased to 20 °C·day and the active accumulated temperature exceeded 320 °C·day, the incubation speed slowed down, then reached the second small peak, and then the incubation process was over. (Bai Yueming)

3 Response of agriculture to climate change

3.1 Impact of climate change on the heat resources and corn suitability in Northeast China

Based on the daily temperature data from RegCM3 model output (1951—2100), the spatial and temporal distributions of heat resources and corn suitability during the different growth stages in Northeast China were analyzed. The results show that∶ from 1951 to 2100, the heat resources in Northeast China signifcantly increase. The initial days passing stably the daily average temperature of 10 °C are also signifcantly prolonged, extending eastward and northward before 25 April. The areas with10 °C active accumulated temperature greater than 3000 °C·day continue to increase. The suitable areas for planting late-maturing corn are also expanding. The temperature suitability of corn during sowing-emergence and emergence-tasseling gradually increases. From 1951 to 2040, the temperature suitability of corn during tasseling-mature and whole growth period is high in Liaoning Province but low in Heilongjiang Province. From 2041 to 2100, the temperature suitability of corn during tasseling-mature and whole growth period is low in Liaoning Province but high in eastern Jilin and Heilongjiang provinces. (Guo Jianping)

3.2 Evaluation of climate potential productivity of corn based on actual growth period in Northeast China

Based on the corn growing data from 30 agro-meteorological observation stations during 1981—2006 and the meteorological data from 70 meteorological stations during 1961—2006 in Northeast China, the characteristics of climate potential productivity of corn were assessed. The photosynthetic production potential and the light and temperature production potential of corn were gradually reducing from west to east and from north to south. However, the climate production potential presented a progressively decreasing trend from southwest to northeast. In the past 46 years, the photosynthetic production potential and the climate production potential in Northeast China showed decreasing trends. In particular, the production potentials of light and temperature in Liaoning and Jilin provinces presented a decreasing trend, while they were increasing in Heilongjiang Province. From decadal perspective, the climate production potential was the highest in the 1960s, signifcantly reduced in the 1990s, and the minimum occurred after 2001. (Guo Jianping)

3.3 Simulated suitability distribution of spring wheat in Northwest China

Based on the Maxent model, the distribution of spring wheat and the weather data, including0 °C accumulated temperature,0 °C interval days, annual rainfall, annual average temperature,3 °C accumulated temperature,3 °Cinterval days,3 °C evapotranspiration,3 °C radiation, were used to simulate the suitability distribution of the spring wheat in Northwest China. The simulation results show that the probability of suitability distribution area is 50%-80%, indicating a better match with current plant area in Northwest China (Fig3). (E Youhao)

3.4 Changes of dry hot wind of winter wheat in Huanghuaihai area under climate change

Based on daily maximum temperature, air relative humidity and wind speed at 14∶00 local time from 52 meteorological stations from 1961 to 2010 in the Huanghuaihai area, combined with climate trends and similarity analysis method, the trend of dry hot wind of winter wheat during the late growth stage was systematically analyzed. From 1961 to 2010, the average daily maximum temperature in most of the Huanghuaihai area showed a downward trend, contrary to the climate warming. The air relative humidity at 14∶00 tended to increase, which was opposite to general atmospheric drying. The wind speed at 14∶00 showed a decreasing trend, which was related to afforestation in the northern part of the area. It is diffcult for a single factor to meet the dry hot wind standard. Thereby, the decreasing trends of dry hot wind of winter wheat in the Huanghuaihai area at different levels were obvious. (Guo Jianping)

3.5 Carbon and water budget of agro-ecosystem in Northeast China

During the study period, both maize and paddy rice in Northeast China behaved as potential CO2sink, and the annual net uptake of CO2were 280 gC/m2and 195 gC/m2respectively. Seasonal variations in gross primary production (GPP) for these two ecosystems were associated with temperature, and the maize ecosystem had a higher value of GPP than the paddy rice. Annual evapotranspiration was 407 mm for maize, and 660 mm for paddy rice. Ecosystem water use effciency (WUE = GPP/ET) of maize (11.3 g CO2/kg H2O) was higher than that of paddy rice (4.4 g CO2/kg H2O). WUE seasonal pattern of maize differed from that of paddy rice, as the maximum WUE appeared earlier in paddy rice than in maize (Fig4). (Zhou Li)

4 Research application

4.1 Application of the prediction technology for agro-drought and irrigation in the climatic vulnerable regions of Northwest China

A prediction system for agro-drought and irrigation in Northwest China (including Shanxi, Gansu and Ningxia provinces) was established to provide detailed information for farmers and the government. Service products such as agro-drought and optimum irrigation were made available by the system, supported by soil and meteorological data and a revised crop-drought model coupled with a distributed hydrological module. More than 100 periodical issues on agro-drought information were released shared, and the software training was conducted on the provincial scale in Northwest China. The objectives of the project have been fullfled, with over 66666.7 hm2crop areas being benefted from the prediction information at the accuracy level of 80% and above. (Liu Jiandong)

4.2 Application of the dynamic monitoring and warning technology of maize chilling in Northeast China

On the basis of the results from the Key Projects in the National Science & Technology Pillar Program during the 11th Five-Year Plan Period, the regional distributions of corn in Northeast China were further subdivided. After supplementing the latest data, the parameters of cold damage warning model were recalculated, and the original warning model was improved. Then, the monitoring and early warning system of chilling damage at regional and provincial levels was completed. In 2012, a total of over 10 service booklet were issued in the three northeastern provinces, achieving good results. (Guo Jianping)