吳 蕾,柏軍華,肖 青,杜永明,柳欽火,徐麗萍
?
作物生長模型與定量遙感參數(shù)結(jié)合研究進展與展望
吳 蕾1,2,柏軍華1,3※,肖 青1,3,杜永明1,3,柳欽火1,3,徐麗萍2
(1. 中國科學(xué)院遙感與數(shù)字地球研究所,北京 100101;2.石河子大學(xué),石河子832000; 3. 中國科學(xué)院遙感與數(shù)字地球研究所遙感科學(xué)國家重點實驗室,北京 100101)
作物生長模型與定量遙感參數(shù)的結(jié)合,不僅滿足前者實現(xiàn)區(qū)域應(yīng)用的需求,也有助于提高后者的反演精度,在生態(tài)、農(nóng)業(yè)、資源調(diào)查與全球氣候變化等研究上意義重大。該文從作物生長模型空間應(yīng)用拓展的角度,對國內(nèi)外主流作物生長模型、定量遙感參數(shù)以及兩者結(jié)合的參數(shù)與方法進行了概述,分析了典型作物生長模型的主要模擬過程及其驅(qū)動、初始化、輸出等參數(shù),總結(jié)了當前定量遙感正反演結(jié)果可為作物生長模型區(qū)域應(yīng)用提供的參數(shù)數(shù)據(jù);建立了作物生長模型模擬過程與定量遙感參數(shù)的對應(yīng)關(guān)系,對比分析了作物生長模型與定量遙感參數(shù)的不同結(jié)合方式?;谝陨蟽?nèi)容,對作物生長模型面應(yīng)用的限制因素及其與定量遙感參數(shù)的關(guān)系、作物生長模型面應(yīng)用時參數(shù)尺度效應(yīng)的影響、作物生長模型與定量遙感參數(shù)耦合方法的發(fā)展3個方面展開了討論,以期為作物生長模型與定量遙感參數(shù)開展更好的結(jié)合研究提供參考。
遙感;模型;植被;作物生長;定量遙感;結(jié)合;發(fā)展
作物生長模型與定量遙感參數(shù)的結(jié)合,最初的出發(fā)點是后者為前者提供準確的數(shù)據(jù)源,達到實現(xiàn)大面積作物生長狀態(tài)及產(chǎn)量的模擬的目的[1-3]。作物生長模型模擬的基礎(chǔ)是物質(zhì)平衡和能量守恒原理,以農(nóng)田的光、溫、水、肥等條件因子為驅(qū)動,模擬作物光合、呼吸、蒸騰等生理過程及其對環(huán)境條件的正反饋結(jié)果[4-5]。雖然作物生長模型是實際作物生長過程的數(shù)字簡化表達,但大部分模型參數(shù)的獲取仍然困難,加上地表特征的異質(zhì)性和氣象條件的多變性,在區(qū)域應(yīng)用中,以點帶面的參數(shù)數(shù)據(jù)應(yīng)用較多,不確定性突出[6-7]。隨著遙感技術(shù)的完善與發(fā)展,使用定量遙感正反演參數(shù)得到的地表理化特性參數(shù)信息,為解決作物生長模型區(qū)域化應(yīng)用的參數(shù)數(shù)據(jù)的需求問題提供了可能[8-10]。
20世紀80年代,美國LACIE(large area crop inventory experiment)研究計劃表明,遙感信息對作物生長模型的應(yīng)用范圍和模擬精度的提高有很大作用[11]。定量遙感模型形成伊始,其主要目的就是為農(nóng)學(xué)研究提供參數(shù)數(shù)據(jù)。1986年,Wiegand等[12]首次提出利用衛(wèi)星數(shù)據(jù)和作物產(chǎn)量估算模型的結(jié)合以解決定量分析的參數(shù)復(fù)雜化問題,隨后許多學(xué)者在兩者結(jié)合方面開展了諸多研究和應(yīng)用。Delecolle等[13-17]先后歸納和討論了作物生長模型與遙感的結(jié)合方法及其在大尺度上的應(yīng)用,內(nèi)容多以兩者結(jié)合方法的分析為主,缺少對兩者結(jié)合的理論基礎(chǔ)、各自發(fā)展需求、當前結(jié)合參數(shù)、發(fā)展中可能的科學(xué)問題以及各自研究領(lǐng)域進步對結(jié)合的可能推動等內(nèi)容[6-9,18-19]的系統(tǒng)梳理和闡述。為了兩者結(jié)合更好的繼往開來,本文從作物生長模型與定量遙感參數(shù)相結(jié)合出發(fā),綜合分析了作物生長模型的發(fā)展和定量遙感參數(shù)正反演的現(xiàn)狀,對以下4個問題進行了探討:1)主流作物生長模型的基礎(chǔ)原理、特征、輸入和輸出參數(shù)、面應(yīng)用亟需解決的問題是什么?2)國內(nèi)外主要定量遙感產(chǎn)品在時空分辨率方面的表現(xiàn)及作物生長模擬過程與定量遙感參數(shù)對應(yīng)關(guān)系是怎樣的?3)作物生長模型與定量遙感參數(shù)結(jié)合時,可選的鏈接參數(shù)及方法有哪些?4)在總結(jié)和討論了作物生長模型與定量遙感參數(shù)結(jié)合的困難與發(fā)展方向的基礎(chǔ)上,如何更好地發(fā)揮定量遙感在作物生長模型面應(yīng)用中的作用以及作物生長模型對定量遙感反演的促進作用呢?
1.1 作物生長模型的發(fā)展
作物生長模型是將作物、環(huán)境、栽培技術(shù)作為一個整體,應(yīng)用系統(tǒng)分析原理和方法,對作物生長發(fā)育、光合生產(chǎn)、器官生成和產(chǎn)量形成等生理過程及其對環(huán)境和技術(shù)的反饋關(guān)系進行理論概括和數(shù)量分析,繼而建立相應(yīng)數(shù)學(xué)模型,進行作物生長過程的動態(tài)定量模擬[20]。從作物生長模型空間應(yīng)用拓展角度,將作物生長模型的發(fā)展分為機理性模型構(gòu)建、模型點應(yīng)用和模型面應(yīng)用3個階段(表1)。
表1 作物生長模型的發(fā)展
1)機理性模型構(gòu)建
20世紀60年代,隨著對作物生理變化機理的深入認識和計算機技術(shù)的發(fā)展,采用數(shù)學(xué)公式描述植物生理過程,成為作物生長動態(tài)模擬模型研制的起步[21-22]。
荷蘭最早開始作物生長模型的研制。1965年,Wit[23]發(fā)表的文章“葉冠層的光合作用”,奠定了作物生長機理性模型的理論基礎(chǔ),并于1970年發(fā)表了第1個作物生長過程碳素平衡模擬模型ELCROS(elementary crop simulator)[24];1982年,荷蘭瓦赫寧根大學(xué)研究中心在ELCROS的基礎(chǔ)上,通過增加呼吸作用及作物微氣象呼吸過程,改進了碳素的平衡模擬,形成了相對復(fù)雜的綜合模型BACROS(basic crop simulation)[25]。但這2個模型只能模擬潛在生產(chǎn)力的作物生長發(fā)育。出于對作物實際生長狀態(tài)的模擬要求,水分[26-29]和營養(yǎng)限制[30-32]下的作物生長模型的研制得以開展。1981年,Keulen等[33]發(fā)表了可以模擬施肥牧草生長與水分利用的ARID CROP模型。
美國作物機理性模型的研究略晚,但很快在玉米生長發(fā)育涉及到的光合、呼吸、水循環(huán)與吸收、營養(yǎng)循環(huán)與吸收等方面的模擬迅速發(fā)展。1968年,Duncan等[34]以“葉冠層的光合作用”理論為基礎(chǔ),通過建立冠層分層的光強、CO2濃度、光合速率、干物質(zhì)、籽粒生長模擬以及土壤濕度模擬算法,形成了玉米模型SIMAIZ(simulation maize);1969年,Chen等[35]將植物光合產(chǎn)物同化過程分解為光合作用和呼吸作用,并于1971年與Curry[36]研制了包括形態(tài)發(fā)育、光合作用及其物質(zhì)積累和各器官分配等過程的玉米生長模型。同年,Stapleton等[37]通過引進一些常數(shù),簡化光合產(chǎn)物累積過程中光能轉(zhuǎn)換成化學(xué)能的算法,并增加氮素平衡和根系生長子模塊,形成了SIMCOT(simulation cotton)模型。
2)模型的點應(yīng)用
初期研制的機理性模型,主要目的是準確模擬作物的生理和生長過程。上世紀80年代,模型逐步趨于模塊化和通用化,標志著機理性模型開始向點應(yīng)用模型轉(zhuǎn)變。
1989年,荷蘭瓦赫寧根大學(xué)的科學(xué)家在ELCROS模型基礎(chǔ)上,開發(fā)了一年生的作物概要性模型MACROS(modules for annual crop simulation)[38];Spitters等[39]開發(fā)了SUCROS(simple and universal crop simulator)模型,因其普適性得到廣泛應(yīng)用;1990年,SARP(simulation and systems analysis for rice production)計劃在MACROS和SUCROS基礎(chǔ)上,開發(fā)了一系列針對應(yīng)用的水稻模型,統(tǒng)稱為ORYZA模型[40]。1994年,世界糧食研究中心在SUCROS模型的基礎(chǔ)上開發(fā)了WOFOST(world food studies)模型[41]。
20世紀80年代初,美國以DSSAT(decision support system for agro-technology transfer)系統(tǒng)[42]為推動,集成了29種不同的作物模擬模型,包括CERES(crop environment resource synthesis)系列模型(6種)、LEGUMES豆類作物模型(7種)、ROOT CROPS模型(4種)、OIL CROP向日葵模型、VEGETABLES模型(5種)、FIBER模型、FORAGES模型(4種)、SUGARCANE模型以及FRUIT CROPS模型。Williams等[43]以一個基本生長函數(shù)為主算法建立了EPIC模型,通過調(diào)整作物參數(shù),模擬了20多種作物和牧草的生長。GOSSYM(gossypium simulation models)模型[44]和GLYCIM(dynamic simulation of soybean crop growth)模型[45]也是模型點應(yīng)用階段的重要組成部分。
這個階段,作物生長模型的應(yīng)用前景引起了其他國家的重視。1987年,加拿大的Place和Brown[46]研制了SIMCOY(simulation of corn yield)模型,用以模擬玉米形態(tài)發(fā)育、根系生長、葉面積指數(shù)變化、水分平衡等過程。1995年,澳大利亞農(nóng)業(yè)生產(chǎn)系統(tǒng)研究組(APSRU)、聯(lián)邦科工組織及昆士蘭政府等單位通過將不同作物生長模型集成到一個公用平臺,建成了作物生產(chǎn)潛力模型APSIM(agricultural production system simulation)[47]。中國在作物生長模型的研究上起步較晚。1993年,高亮之和金之慶[48]提出并建立了“水稻鐘模型”RCSODS(rice cultivation simulation optimization decision making system);1996年,潘學(xué)標等[49]研制了棉花生長發(fā)育模擬模型COTGROW(cotton growth and development simulation model)。
3)模型的面應(yīng)用
20世紀90年代,作物生長模型在農(nóng)業(yè)生態(tài)區(qū)劃、土地質(zhì)量評價和作物估產(chǎn)等領(lǐng)域的應(yīng)用逐漸開展,由于包括遙感在內(nèi)的地理信息技術(shù)的快速發(fā)展,作物生長模型大范圍的空間應(yīng)用得以展開。此時,面應(yīng)用模型的參數(shù)獲取困難問題凸顯,最有效的解決方式有2種:①對點應(yīng)用模型的進一步改進,代表性的是荷蘭LINTUL(light interception and utilization)模型[50],避開了機理性程度較高的植物光合作用過程模擬,改為應(yīng)用冠層輻射截獲量和光能利用率間接決定作物光合產(chǎn)物積累;②保留原有模型,通過結(jié)合GIS技術(shù)實現(xiàn)模型的面應(yīng)用需求。如美國在DSSAT系統(tǒng)基礎(chǔ)上,開發(fā)了GIS輔助的GIS-DSSAT,可對參數(shù)的空間變異進行分析[51];日本科學(xué)家[52-53]將EPIC模型與GIS相結(jié)合,建立了Spatial EPIC系統(tǒng),可在國家尺度上評價氣候變化對主要禾谷作物產(chǎn)量的影響;辛紅敏[54]基于APSIM模型與GIS的結(jié)合分析了小麥-玉米增產(chǎn)潛力和地域差異性;潘學(xué)標等[49]的COTGROW模型也實現(xiàn)了與GIS技術(shù)的結(jié)合,可對不同田間管理情況下的棉花生長進行模擬。
1.2 典型作物生長模型模擬過程和參數(shù)
作物生長模型發(fā)展過程中,階段性的標志模型有WOFOST、ORYZA、EPIC、CERES、APSIM和RCSODS等,據(jù)其模擬的作物種類,分為單一作物模型和通用模型。從表2可以看出,所有通用模型共同的過程模擬有光截獲和利用、光合物質(zhì)生產(chǎn)和分配、形態(tài)發(fā)育、蒸騰、水分平衡、養(yǎng)分平衡,由于各模型目的性側(cè)重點不同,不同通用模型也有其獨特的模擬過程:WOFOST、EPIC、CERES模型注重建立脅迫狀態(tài)模擬;APSIM模型注重根生長、殘茬分解模擬;ORYZA和RCSODS模型有水稻的分蘗模擬。多數(shù)模型可實現(xiàn)水肥限制條件下的模擬,但APSIM模型和EPIC模型,只能實現(xiàn)潛在模擬和水分限制的模擬。表3為典型作物生長模型的主要輸入及輸出參數(shù)。
表2 典型作物生長模型的主要模擬過程
表3 典型作物生長模型的主要輸入/輸出參數(shù)
注:ü表示模型有該參數(shù)的輸入/輸出;×表示模型無該參數(shù)的輸入/輸出。
Note: ü indicates that the model has input / output of this parameter;× indicates that the model does not have input / output of this parameter.
由表3可知,作物生長模型模擬結(jié)果主要表現(xiàn)作物的形態(tài)發(fā)育、冠層結(jié)構(gòu)特征、光合產(chǎn)物積累及分布特征。模擬主要輸出參數(shù)包括狀態(tài)參數(shù)和結(jié)果輸出參數(shù),其中狀態(tài)參數(shù)有葉面積指數(shù)(LAI)、生物量、生育時期、冠層高度、葉齡,各器官比例、水肥限制因子,結(jié)果輸出參數(shù)有潛在產(chǎn)量和限制產(chǎn)量。不同限制條件下,生物量和LAI等狀態(tài)參數(shù)及潛在產(chǎn)量是每個模型都可實現(xiàn)的模擬參數(shù),除CERES和APSIM外,其他模型并未將根系生長作為模型主要模擬過程。涉及的主要輸入?yún)?shù)可分為4類:1)氣象參數(shù):主要有太陽輻射、溫度、濕度、氣壓、降水量和風(fēng)速,多作為驅(qū)動參數(shù)。對所有典型作物生長模型來說,降水、濕度、風(fēng)速主要參與蒸散、土壤水分平衡過程模擬。太陽輻射、最高、最低氣溫和降水是驅(qū)動作物生長的主要動力,與光的截獲和利用、光合產(chǎn)物生產(chǎn)和分配和形態(tài)發(fā)育等過程都密切相關(guān),是要求輸入的最小數(shù)據(jù)集。2)土壤參數(shù):反映土壤理化特性的數(shù)據(jù),多作為初始化參數(shù)。其中土壤物理參數(shù)用于計算土壤的保水性,土壤化學(xué)參數(shù)用于計算土壤基礎(chǔ)肥力,土壤基礎(chǔ)肥力是作物連續(xù)生長過程的水分和養(yǎng)分限制因子載體,與蒸騰、水分平衡、養(yǎng)分平衡、環(huán)境脅迫等過程密切相關(guān);3)管理參數(shù):為模型提供模擬起始、終止及生長過程肥水供應(yīng)條件,參與蒸騰、為初始化或者驅(qū)動參數(shù),水分平衡、養(yǎng)分平衡、環(huán)境脅迫等過程計算;4)遺傳參數(shù):反應(yīng)作物品種特征,多作為初始化參數(shù),是影響物形態(tài)發(fā)育過程的主要參數(shù)之一。
隨著遙感對地觀測能力不斷提高,將遙感信息嵌入作物生長模型或者校正有關(guān)參數(shù),有助于解決作物生長模擬模型由點向面空間應(yīng)用擴展中的參數(shù)數(shù)據(jù)難以獲取的問題[55]。
2.1 可結(jié)合的主要定量遙感參數(shù)
經(jīng)過近幾十年的發(fā)展,遙感能夠提供給作物生長模型的數(shù)據(jù)發(fā)生了巨大變化,從單純提供影像轉(zhuǎn)變到可以提供地學(xué)或生物學(xué)定量遙感參數(shù)數(shù)據(jù);目前,基于AVHRR、MODIS、MSG、VEGETATION等傳感器數(shù)據(jù),可以生產(chǎn)區(qū)域乃至全球范圍的多個地學(xué)特征參數(shù)產(chǎn)品[56]。其中與作物生長相關(guān)的包括LAI、太陽光合有效輻射(photosynthetically active radiation, PAR)、吸收性太陽光合有效輻射(fraction of photosynthetically active radiation, FPAR)、總初級生產(chǎn)力(gross primary productivity, GPP)、凈初級生產(chǎn)力(net primary productivity, NPP)、植被覆蓋度(fractional vegetation coverage, FVC)、物候(phenophase, PHE)、葉綠素(chlorophyll, CHL)、陸表溫度(land surface temperature, LST)、土壤水分(soil moisture, SM)、植被指數(shù)(vegetation index, VI)、土地利用/覆被變化(land-use and land-cover change, LUCC)、反照率(albedo)、蒸騰蒸發(fā)(evapotranspiration, ET)等。
為提高遙感數(shù)據(jù)質(zhì)量,生產(chǎn)出更高時空連續(xù)性的序列數(shù)據(jù)集,多源遙感數(shù)據(jù)被用作計算同一參數(shù)產(chǎn)品。以LAI為例,由于其在生態(tài)環(huán)境和農(nóng)業(yè)生產(chǎn)中的重要指示作用,多數(shù)以陸地特征為主要觀測目標的傳感器,都可對其進行定量反演。加拿大遙感中心采用光譜植被指數(shù)建立LAI的計算模型,利用AVHRR數(shù)據(jù)和VEGETATION數(shù)據(jù)[57]生成了連續(xù)時間序列的LAI遙感產(chǎn)品(時間分辨率為旬,空間分辨率可達1 km)。美國MODIS的LAI/FPAR模型可以反演4和8 d時間頻率的LAI/FPAR數(shù)據(jù)產(chǎn)品[58]。中國的GLASS LAI產(chǎn)品的時間分辨率為8 d,空間分辨率可達1和5 km。
然而,作物生長模擬對數(shù)據(jù)有更高時、空分辨率并存的要求,以上公開發(fā)布的定量遙感正反演產(chǎn)品,由于多關(guān)注全球變化,數(shù)據(jù)在空間上多為幾百米到十幾千米空間分辨率,在時間上多為幾天到幾十天的時間分辨率。從目前對地觀測發(fā)展分析,美國EOS對地觀測衛(wèi)星系列、法國SPOT衛(wèi)星系列、加拿大RADARSAT雷達衛(wèi)星系列、印度INSAT衛(wèi)星系列以及中國的ZY資源衛(wèi)星系列、HJ環(huán)境系列和GF高分衛(wèi)星系列等中高分辨率衛(wèi)星傳感器,可以提供多波段、多時相、多角度的對地觀測數(shù)據(jù),空間分辨率可達10 m以內(nèi),時間分辨率可達1 d,為獲得更高時、空分辨率遙感正反演參數(shù)數(shù)據(jù),進行作物生長模型面應(yīng)用的模擬提供了可能。
2.2 作物生長模擬過程與定量遙感參數(shù)的對應(yīng)關(guān)系
在區(qū)域乃至更大范圍上,定量遙感可獲取的陸地地表主要參數(shù)包括土地利用、植被理化和冠層結(jié)構(gòu)3類,其中多數(shù)參數(shù)與作物生長模型模擬過程關(guān)系密切。由表4可知,PAR、LST、SM直接影響作物形態(tài)發(fā)育、光合物質(zhì)生產(chǎn)和蒸散等過程,是觸發(fā)和推動作物生長的動力和物質(zhì)條件;FPAR、CHL、Albedo影響作物光合作用強度;LAI不僅影響和體現(xiàn)作物群體光合作用強弱,也影響作物生長大多數(shù)過程;GPP/NPP是作物同化過程中不同階段物質(zhì)累計狀態(tài),是器官形成的最直接物質(zhì)基礎(chǔ);PHE反映作物不同生育階段;VI綜合表現(xiàn)作物群體光合強度、能力以及光合產(chǎn)物累積,根據(jù)不同算法可間接估算與作物生長密切相關(guān)的LAI、GPP、NPP、FPAR等參數(shù);ET和Albedo參與水分和能量平衡過程,間接影響作物光合強度;LUCC可在模型模擬前,提供作物空間分布信息。目前,PAR、VI、LST和Albedo為遙感通過反射推導(dǎo)出的參數(shù),不確定性相對較小,在作物生長模型與定量遙感參數(shù)結(jié)合中可作為首選參數(shù)。
2.3 作物生長模型與定量遙感信息結(jié)合時參數(shù)選擇
因作物生長模型與定量遙感相結(jié)合時,研究目的不同,導(dǎo)致結(jié)合參數(shù)的選擇有差異,本文根據(jù)兩者結(jié)合時的參數(shù)數(shù)量,將其分為單參數(shù)和多參數(shù)結(jié)合。
單參數(shù)結(jié)合是指結(jié)合參數(shù)只有一個。LAI[59-60]、物候期[61]、冠層反射率[62]、冠層溫度[63]、土壤濕度[64]、植被指數(shù)等均可成為結(jié)合參數(shù)。最早和最常見結(jié)合的單參數(shù)是LAI,冠層反射率也是單參數(shù)結(jié)合時選擇較多的對象,不同之處在于作物生長模型需與冠層輻射傳輸模型相結(jié)合模擬反射率,然后進行冠層反射率或者VI指數(shù)的同化。Delecolle等[65]使用SPOT/HRV數(shù)據(jù)反演多個時期小麥LAI,輸入ARCWHEAT模型,預(yù)測大面積小麥產(chǎn)量。Maas等[66]、Clevers等[67]、趙艷霞[68]、Dente等[69]、Curnel等[70]、黃健熙等[71-74]分別通過Landsat/TM、航空遙感和MODIS等反演的LAI,與SUCROS、CERES-Wheat、WOFOST等作物生長模型相鏈接。另外,馬玉平等[75-81]雖然也是以LAI作為結(jié)合參數(shù),特殊點是作物生長模型與冠層輻射傳輸模型直接通過鏈接參數(shù)結(jié)合后,模擬植被冠層反射率,然后將模擬的冠層反射率或者VI與觀測值進行同化。總之,以LAI、冠層反射率或其他參數(shù)作為同化參數(shù)時,優(yōu)化作物生長模型初始化或者狀態(tài)參數(shù)均為研究的最終目標,可優(yōu)化的參數(shù)包括LAI、生長率、光能利用效率、最大葉面積指數(shù)、播種日期和種植密度等。
表4 作物生長模型主要模擬過程與定量遙感參數(shù)對應(yīng)關(guān)系
注:+表示定量遙感參數(shù)可以參與作物生長模型模擬過程計算;++表示定量遙感參數(shù)等同于作物生長模型的輸出參數(shù)。
Note: +indicates that parameters derived from quantitative remote sensing can participate in simulation process calculation of crop growth model; ++ indicates that parameters derived from quantitative remote sensing are similar to output parameters of crop growth model.
多參數(shù)結(jié)合是指結(jié)合參數(shù)數(shù)量大于一個。Maas[82]將遙感反演的LAI和表面溫度作為作物生長模型輸入?yún)?shù),用以改善玉米地上生物量的模擬結(jié)果。郭建茂[83]利用遙感反演的LAI和ET初始化/參數(shù)化WOFOST模型,以提高冬小麥生長過程的模擬精度。包姍寧等[84]在作物水分脅迫模式下,同化ET和LAI雙變量,其結(jié)果優(yōu)于單變量對作物產(chǎn)量的模擬。Weiss等[85]通過植被冠層遙感輻射SAIL模型與作物生長模型STICS,同化LAI、葉綠素含量、生物量和相對水分含量,以提高模型模擬的冠層反射率與SPOT的吻合度。黃健熙等[86-87]通過同化冬小麥的遙感反演和SWAP 作物生長模擬的LAI、ET,優(yōu)化SWAP模型中的出苗日期和灌溉量2個參數(shù)。李衛(wèi)國等[88]用遙感反演得到的LAI和生物量校正作物模型運行軌跡加強對模型對產(chǎn)量的預(yù)測精度。朱元勵等[89-90]利用優(yōu)化算法以LAI和葉片氮積累量共同作為結(jié)合點和更新點,建立了基于遙感信息與作物生長模型結(jié)合的作物生長監(jiān)測與產(chǎn)量預(yù)測技術(shù)??偠灾?,LAI與表面溫度、ET、葉綠素含量、生物量、相對水分含量、葉片氮積累量等均可組合作為鏈接參數(shù)優(yōu)化模型,優(yōu)化后的模型模擬結(jié)果相對以單變量作為鏈接參數(shù)時更加準確[84]。
2.4 作物生長模型與定量遙感參數(shù)結(jié)合方法的對比
作物生長模型與定量遙感參數(shù)結(jié)合研究早期,主要的結(jié)合方法是驅(qū)動法,即利用遙感“觀測值”直接作為作物生長模型的輸入?yún)?shù)數(shù)據(jù)源,參與區(qū)域作物狀態(tài)參數(shù)的模擬,將點上的作物生長模型拓展到區(qū)域乃至更大范圍,驅(qū)動法結(jié)合參數(shù)可以很多,表3中出現(xiàn)的參數(shù)均可驅(qū)動作物模型的運轉(zhuǎn)。隨后,以遙感“觀測值”或反演參數(shù)數(shù)據(jù)作為標準參考值,通過同化法實現(xiàn)作物生長模型初始化參數(shù)或者狀態(tài)參數(shù)的優(yōu)化,提高作物生長模型面應(yīng)用能力[9]。本文根據(jù)執(zhí)行優(yōu)化時參數(shù)選擇上的差異進一步將同化法分為:參數(shù)化法Ⅰ和參數(shù)法Ⅱ。參數(shù)化法Ⅰ是以定量遙感反演模型與作物生長模型模擬的形式結(jié)合,同化LAI、葉綠素、冠層溫度、ET這類狀態(tài)參數(shù),參數(shù)化法Ⅱ則是以作物生長模型與遙感模型通過共同參數(shù)結(jié)合后,同化參數(shù)定量遙感正演模型的是反射率或者VI(圖1)。
a)如果Label[j]>0且medarray[Label(j)]<1,則表示該標號的塊連通域首次被掃描到,則Label_index=Label_index+1;medarray[Label(j)]=Label_index;Label[j]=Label_index。
Figure 1 Three methods of combining crop growth model with parameters derived from quantitative remote sensing
本文首先從作物生長模型空間應(yīng)用范圍拓展的角度,綜述了以荷蘭和美國為代表的作物生長模型研究進展,從中可以看出,作物生長模型從機理性模型構(gòu)建開始,逐步發(fā)展到模型的面應(yīng)用,既體現(xiàn)了模型的發(fā)展過程,也指明了模型的發(fā)展方向,作物生長模型能夠在生態(tài)、農(nóng)業(yè)和氣候變化等領(lǐng)域發(fā)揮重要作用。
目前,國內(nèi)外在小麥、玉米、水稻3個主要糧食作物以及甜菜、苜蓿、甘蔗等小作物上,都開展了作物生長模型與定量遙感參數(shù)結(jié)合的研究[90-91],最初目的是實現(xiàn)作物生長模型面應(yīng)用的需求,并多采用驅(qū)動法。隨著研究的深入,同化的方法和參數(shù)趨于多元化,發(fā)展到現(xiàn)在的驅(qū)動法、參數(shù)化法Ⅰ和參數(shù)化法Ⅱ,同化參數(shù)也從LAI逐步增加到冠層反射率、土壤濕度、冠層溫度、ET、LNA、葉綠素等或多個參數(shù)的組合。因此,在當前定量遙感參數(shù)獲取能力加快發(fā)展時期,對定量遙感與作物生長模擬兩者的關(guān)系、點模型的升尺度適用性以及結(jié)合方法發(fā)展3個方面進行總結(jié)和討論,將有助于提高兩者結(jié)合應(yīng)用的廣度和深度。
3.1 作物生長模型面拓展應(yīng)用的限制因素及其與定量遙感參數(shù)的關(guān)系
作物生長模型模擬過程主要有形態(tài)發(fā)育、光截獲與利用、光合物質(zhì)生產(chǎn)與分配以及水分平衡和養(yǎng)分平衡,作物生長模型運行需要的驅(qū)動和初始化參數(shù),主要包括氣象、土壤和田間管理等共計40余個。在作物生長模型空間應(yīng)用范圍逐漸擴展過程中,實現(xiàn)區(qū)域乃至全球尺度所有參數(shù)高時空分辨率的數(shù)據(jù)獲取很有挑戰(zhàn)。
面對這種挑戰(zhàn),通過遙感等對地觀測方法進行反演可以解決其中的部分難題。目前,能夠反演的陸地定量遙感參數(shù)達10多種,公開發(fā)布的有土壤水分、溫度、LAI、覆蓋度、太陽光合有效輻射、植被指數(shù)、反照率等[92-93]。這些產(chǎn)品可作為輸入數(shù)據(jù)源驅(qū)動模型,達到對區(qū)域乃至全球不同空間范圍的模擬。但是,當前的定量遙感參數(shù)正反演結(jié)果還存在較大的不確定性,部分參數(shù)精度較低。以參數(shù)化法Ⅱ優(yōu)化作物生長模型和輻射傳輸模型中敏感參數(shù),在一定程度上是誤差傳遞較小的一種情況。但在使用驅(qū)動法時,如果不能正確選擇鏈接參數(shù)和結(jié)合方法,誤差層層傳遞,將增加作物生長模型模擬結(jié)果的不確定性,降低模擬精度。
總的來看,當前作物生長模型與定量遙感參數(shù)的結(jié)合,是以實現(xiàn)作物生長模型面應(yīng)用需求為主要目的,隨著區(qū)域性作物生長模型模擬精度的提高,生長期等信息可準確提供,可為定量遙感正反演參數(shù)(如FPAR、ET、GPP、NPP等)精度的提高提供支持[94]。因此,作物生長模型與定量遙感參數(shù)不僅是上下游的數(shù)據(jù)生產(chǎn)與應(yīng)用關(guān)系,也是相互支持共同提高地物信息獲取能力的方式。
3.2 作物生長模型面拓展應(yīng)用時參數(shù)尺度效應(yīng)的影響
作物生長模型面應(yīng)用,是在模型點應(yīng)用基礎(chǔ)上發(fā)展而來,當前的研究主要關(guān)注模型面應(yīng)用可利用的驅(qū)動數(shù)據(jù)和初始化數(shù)據(jù)不足的問題,對數(shù)據(jù)的時間和空間適應(yīng)性考慮較少,這導(dǎo)致了兩者結(jié)合對植被及其環(huán)境參數(shù)監(jiān)測的不確定性提高。在時間尺度上,有研究表明,提高作物關(guān)鍵物候期的 LAI 精度比提高遙感觀測的時間分辨率對農(nóng)業(yè)數(shù)據(jù)同化系統(tǒng)更加有效[61]。植被冠層部分理化參數(shù)特征隨其生長發(fā)育階段變化顯著,正確考慮時間尺度有益于提高定量遙感對植被及其生長環(huán)境信息的獲取,促進定量遙感與作物生長模型更好的結(jié)合。在考慮空間尺度效應(yīng)時,需比較相同名稱參數(shù)在定量遙感正反演模型與作物生長模擬模型中的物理意義,主要原因就是混合像元的存在會造成作物生長模型與遙感正反演模型的結(jié)合從點尺度上升到面尺度的不適應(yīng)。在空間尺度上的通過資源環(huán)境衛(wèi)星等得到的定量遙感產(chǎn)品,空間分辨率為30~1 000 m,不同空間分辨率的地表覆被理化屬性異質(zhì)性特征明顯。假設(shè)有3塊試驗小區(qū),分別代表大小不同的作物群體,相應(yīng)的作物產(chǎn)量也有差異,并有高、低兩種空間分辨率遙感影像,對較高空間分辨率遙感數(shù)據(jù),一個像元對應(yīng)一塊試驗小區(qū),對較低空間分辨率遙感數(shù)據(jù),一個像元對應(yīng)整個3塊試驗小區(qū),接下來使用以上兩種分辨率的定量遙感產(chǎn)品參與作物生長模型模擬,比較兩者模擬的LAI、生物量和產(chǎn)量等結(jié)果:高空間分辨率數(shù)據(jù)驅(qū)動模擬下的三個像元值與低分辨率下一個像元值的關(guān)系是怎樣的?異質(zhì)性地表特征對作物生長模型在尺度上推應(yīng)用時,初始化參數(shù)有怎樣的變化特征,怎樣影響作物生長模擬過程及結(jié)果?從這個角度看,作物生長模型在后期的面應(yīng)用研究中,不僅要通過遙感等手段解決驅(qū)動和初始化參數(shù)難以獲取的問題,也要分析升尺度對作物生長模型初始化參數(shù)的影響。
通過植被空間理化參數(shù)與冠層入射和透射光的關(guān)系,定量遙感正演模型可模擬植被冠層太陽光合有效輻射的吸收率、透射率和反射率,而當冠層太陽光合有效輻射的吸收率、透射率和反射率已知時,反過來利用計算算法,可實現(xiàn)LAI等參數(shù)的定量遙感反演。作物生長過程模擬是通過比爾定理將消光系數(shù)與LAI相結(jié)合,計算植被冠層太陽光合有效輻射透過率,然后加上經(jīng)驗性的冠層反射率信息得到冠層光吸收率,并以此作為光合物質(zhì)生產(chǎn)和分配的能量基礎(chǔ)。冠層對光的吸收、透射和反射是作物生長過程模擬和定量遙感正反演的基礎(chǔ),兩者最大結(jié)合之處在于對光的截獲。
當前多數(shù)研究致力于通過同化法優(yōu)化狀態(tài)參數(shù)或初始化參數(shù),在這方面已經(jīng)積累了大量研究成果基礎(chǔ)上,是否可以嘗試使用定量遙感正演算法,引入遙感獲取的植被冠層反射率信息,計算作物生長模型中的吸收性太陽光合有效輻射,改進當前消光系數(shù)經(jīng)驗性取值,提高作物生長模型光截獲計算的機理性和準確性,為定量遙感參數(shù)與作物生長模型算法結(jié)合的奠定理論基礎(chǔ)。同時,作物生長模擬的氮素、纖維素、木質(zhì)素等積累量,以及葉片、莖、穗、株高等狀態(tài)量,都是定量遙感正反演模型需要的初始化變量,精確獲取這些參數(shù)的值有助于定量遙感正反演精度的提高。由此可以看出,定量遙感參數(shù)與作物生長模型的結(jié)合方法和參數(shù)還需進一步的探索和豐富。
拓展更大空間的應(yīng)用能力是作物生長模型發(fā)展的方向,參數(shù)的適用尺度是作物生長模型模擬精度的重要影響因素;全面把握遙感對大空間尺度空間信息觀測現(xiàn)狀和能力,建立作物生長模型與定量遙感參數(shù)更為緊密的結(jié)合方式,不僅可以通過2個模型的同化,降低過程參數(shù)的誤差傳遞與不確定性,也可以嘗試算法的融合,以達到提高對目標參數(shù)的作物生長模型模擬和遙感模型正反演精度的目的,發(fā)揮定量遙感觀測在農(nóng)業(yè)資源環(huán)境監(jiān)測方面更大的應(yīng)用價值和作物生長精確模擬對提高定量遙感正反演精度的改進價值。
[1] Ma Yuping, Wang Shili, Zhang Li, et al. Monitoring winter wheat growth in North China by combining a crop model and remote sensing data[J]. International Journal of Applied Earth Observation and Geoinformation, 2008, 10(4): 426-437.
[2] Li Tao, Feng Yongsheng, Li Xiaomei. Predicting crop growth under different cropping and fertilizing management practices[J]. Agricultural and Forest Meteorology, 2009, 149(6/7): 985-998.
[3] 張峰,吳炳方,劉成林,等. 區(qū)域作物生長過程的遙感提取方法[J]. 遙感學(xué)報,2004,8(6):515-528.
Zhang Feng, Wu Bingfang, Liu Chenglin, et al. A method to extract regional crop growth profile with time series of NDVI data[J]. Journal of Remote Sensing, 2004, 8(6): 515-528. (in Chinese with English abstract)
[4] 宇振榮. 作物生長模擬模型研究和應(yīng)用[J]. 生態(tài)學(xué)雜志,1994,13(1):69-73.
Yu Zhenrong. Studies on crop growth modelling and its application[J]. Chinese Journal of Ecology, 1994, 13(1): 69-73. (in Chinese with English abstract)
[5] 曹宏鑫,趙鎖勞,葛道闊,等. 作物模型發(fā)展探討[J]. 中國農(nóng)業(yè)科學(xué),2011,44(17):3520-3528.
Cao Hongxin, Zhao Suolao, Ge Daokuo, et al. Discussion on development of crop models[J]. Scientia Agricultura Sinica, 2011, 44(17): 3520-3528. (in Chinese with English abstract)
[6] 邢雅娟,劉東升,王鵬新. 遙感信息與作物生長模型的耦合應(yīng)用研究進展[J]. 地球科學(xué)進展,2009,24(4):444-451.
Xing Yajuan, Liu Dongsheng, Wang Pengxin. Advances of the coupling application of remote sensing information and crop growth model[J]. Advances in Earth Science, 2009, 24(4): 444-451. (in Chinese with English abstract)
[7] 張佳寶,張均華,劉建立. 作物模型研究進展[J]. 土壤,2012,44(1):1-9. Zhang Jiabao, Zhang Junhua, Liu Jianli. Advances on Crop Models[J]. Soils, 2012, 44(1): 1-9. (in Chinese with English abstract)
[8] 蒙繼華,吳炳方,杜鑫,等. 遙感在精準農(nóng)業(yè)中的應(yīng)用進展及展望[J]. 國土資源遙感,2011,90(3):1-7.
Meng Jihua, Wu Bingfang, Du Xin, et al. A review and outlook of applying remote sensing to precision agriculture[J]. Remote Sensing for Land & Resources, 2011, 90(3): 1-7. (in Chinese with English abstract)
[9] 李存軍,王紀華,王嫻,等. 遙感數(shù)據(jù)和作物模型集成方法與應(yīng)用前景[J]. 農(nóng)業(yè)工程學(xué)報,2008,24(11):295-301.
Li Cunjun, Wang Jihua, Wang Xian, et al. Methods for integration of remote sensing data and crop model and their prospects in agricultural application[J]. Transactions of the Chinese Society of Agricultural Engineering (Transactions of the CSAE), 2008, 24(11): 295-301. (in Chinese with English abstract)
[10] 殷飛,金世佳. 遙感在農(nóng)業(yè)旱情監(jiān)測中的應(yīng)用現(xiàn)狀與展望[J]. 干旱環(huán)境監(jiān)測,2015,29(2):87-92.
Yin Fei, Jin Shijia. Application status and prospect on agricultural drought monitoring based on Remote sensing[J]. Arid Environmental Monitoring, 2015, 29(2): 87-92. (in Chinese with English abstract)
[11] Macdonald R B, Hall F G, Erb R B. The Large Area Crop Inventory Experiment (LACIE). Part 1: The use of LANDSAT data in LACIE[J]. LARS Symposia, 1975, 6(3/4/5): 1-46.
[12] Wiegand C L, Richardson A J, Jackson R D, et al. Development of agrometeorological crop model inputs from remotely sensed information[J]. IEEE Transactions on Geoscience & Remote Sensing, 1986, 24(1): 90-98.
[13] Delécolle R, Maas S J, Guérif M, et al. Remote sensing and crop production models: present trends[J]. Isprs Journal of Photogrammetry & Remote Sensing, 1992, 47(2): 145-161.
[14] Moulin S, Bondeau A, Delecolle R. Combining agricultural crop models and satellite observations: From field to regional scales[J]. International Journal of Remote Sensing, 1998, 19(6): 1021-1036.
[15] Fischer A, Kergoat L, Dedieu G. Coupling satellite data with vegetation functional models: Review of different approaches and perspectives suggested by the assimilation strategy[J]. Remote Sensing Reviews, 2009, 15(1/2/3/4): 283-303.
[16] 馬玉平,王石立,張黎,等. 基于遙感信息的作物模型重新初始化/參數(shù)化方法研究初探[J]. 植物生態(tài)學(xué)報,2005,29(6):918-926.
Ma Yuping, Wang Shili, Zhang Li, et al. A preliminary study on the re-initialization、re-parameterization of a crop model based on remote sensing data[J].Acta Phytoecologica Sinica, 2005, 29(6): 918-926. (in Chinese with English abstract)
[17] 閆巖. 基于定量遙感產(chǎn)品和作物生長模型同化的農(nóng)作物長勢監(jiān)測與估產(chǎn)方法研究[D]. 北京:中國科學(xué)院研究生院(遙感應(yīng)用研究所),2006.
Yan Yan. Study on Crop Growth Monitoring and Yield Prediction by Assimilation of Quantitative Remote Sensing Product and Crop Growth Model[D]. Beijing: Graduate University of Chinese Academy of Sciences (Institute of Remote Sensing Applications), 2006. (in Chinese with English abstract)
[18] 趙艷霞,周秀驥,梁順林. 遙感信息與作物生長模式的結(jié)合方法和應(yīng)用:研究進展[J]. 自然災(zāi)害學(xué)報,2005,14(1):103-109.
Zhao Yanxia, Zhou Xiuji, Liang Shunlin. Methods and application of coupling remote sensing data and crop growth models:advance in research[J]. Journal of Natural Disasters, 2005, 14(1): 103-109. (in Chinese with English abstract)
[19] 王靜,李新. 基于作物生長模型和多源數(shù)據(jù)的融合技術(shù)研究進展[J]. 遙感技術(shù)與應(yīng)用,2015,30(2):209-219.
Wang Jing, Li Xin. Progress on data fusion technology of crop growth model and Multi-source observation information[J]. Remote Sensing Technology and Application, 2015, 30(2): 209-219. (in Chinese with English abstract)
[20] Edwards D, Hamson M. Example Models[M]. United Kingdom : Macmillan Education, 1989.
[21] Duncan W G, Loomis R S, Williams W A, et al. A model for simulating photosynthesis in plant communities[J]. Hilgardia, 1967, 38(4): 181-205.
[22] 李軍. 作物生長模擬模型的開發(fā)應(yīng)用進展[J]. 西北農(nóng)業(yè)大學(xué)學(xué)報,1997,25(4):102-107.
Jun Li. The Progress in the Development and Application of Crop Growth Simulation Models[J]. Acta Univ. Agric. Boreali-occidentalis, 1997, 25(4): 102-107. (in Chinese with English abstract)
[23] Wit C T D. Photosynthesis of leaf canopies[J]. Agricultural Research Report, 1965, 663.
[24] Wit C T, Setlík I. DE Dynamic concepts in biology. Prediction and Measurement of Photosynthetic Productivity. Proceedings of the IBP/PP Technical Meeting[C] Wageningen: Centre for Agricultural Publishing, 1970.
[25] Penning D V F W, Laar H H V. Simulation of Plant Growth and Crop Production[M]. Wageningen: Centre for Agricultural Publishing, 1982.
[26] Gebremichael M, Vivoni E R. Spatial sampling uncertainty in SMEX04 soil moisture fields: A data-based resampling experiment[J]. Remote Sensing of Environment, 2008, 112(2): 326-336.
[27] 張栩然,宮阿都,李京,等. 水分脅迫條件下玉米生產(chǎn)潛力遙感評估模型:以重慶市長壽區(qū)為例[J]. 遙感信息,2015,30(1):37-42.
Zhang Xuran, Gong A'dou, Li Jing, et al. Productivity evaluating model of maize under water stress based on remote sensing: A case study of Changshou, Chongqing[J]. Remote Sensing Information, 2015, 30(1): 37-42. (in Chinese with English abstract)
[28] Abduwasit G, Li Zhaoliang, Qin Qimin, et al. Estimating crop water stress with ETM+ NIR and SWIR data[J]. Agricultural and Forest Meteorology, 2008, 148(11): 1679-1695.
[29] 張黎,王石立,何延波,等. 遙感信息應(yīng)用于水分脅迫條件下的華北冬小麥生長模擬研究[J]. 作物學(xué)報,2007,33(3):401-410.
Zhang Li, Wang Shili, He Yanbo, et al. Winter wheat growth simulation under water stress by remote sensing in north China[J]. Acta Agronomica sinica, 2007, 33(3): 401-410. (in Chinese with English abstract)
[30] Inoue Y, Sakaiya E, Zhu Y, et al. Diagnostic mapping of canopy nitrogen content in rice based on hyperspectral measurements[J]. Remote Sensing of Environment, 2012, 126(6829): 210-221.
[31] Xu Xingang, Zhao Chunjiang, Wang Jihua, et al. Using optimal combination method and in situ hyperspectral measurements to estimate leaf nitrogen concentration in barley[J]. Precision Agriculture, 2014, 15(2): 227-240.
[32] Croft H, Chen JingM, Zhang Yongqin, et al. Evaluating leaf chlorophyll content prediction from multispectral remote sensing data within a physically-based modelling framework[J]. ISPRS Journal of Photogrammetry and Remote Sensing, 2015, 102(5): 85-95.
[33] Keulen H V, Seligman N G, Benjamin R W. Simulation of water use and herbage growth in arid regions—A re-evaluation and further development of the model ‘a(chǎn)rid crop’[J]. Agricultural Systems, 1981, 6(3): 159-193.
[34] Duncan W G, Hesketh J D. Net photosynthetic rates, relative leaf growth rates, and leaf numbers of 22 races of maize grown at eight temperatures[J]. Crop Science, 1968, 8(6): 670-674.
[35] Chen L H, Huang B K, Splinter W E. Developing a physical-chemical model for a plant growth system[J]. Amer Soc Agr Eng Trans Asae, 1969, 12: 698-702.
[36] Curry R B. Dynamic simulation of plant growth part I. development of a model[J]. Transactions of the Asabe, 1971, 14(5): 946-949.
[37] Stapleton H N, Meyers R P. Modeling subsystems for the cotton plant simulation[J]. Amer Soc Agr Eng Trans Asae, 1971(5): 950-953.
[38] Penning D V F W, Jansen D M, Berge H F M T, et al. Simulation of Ecophysiological Processes of Growth in Several Annual Crops[M]. Wageningen: Center for Agricultural Publishing and Documentation (Pudoc), 1989.
[39] Spitters C J T, Keulen H V, Kraalingen D W G V. A simple and universal crop growth simulator: SUCROS87[J]. Simulation & Systems Management in Crop Protection, 1989,147-181.
[40] Landsberg J J, Wit C T D. Simulation of Assimilation, Respiration and Transpiration of Crops[J/OL]. Journal of Ecology, 1978, 68(1): 333.
[41] Hijmans R J, Guiking Lens I M, Diepen V C A. WOFOST 6.0: User’s guide for the WOFOST 6.0 crop growth simulation model[M]. Technical document 12, DLOW inand Staring Centre, 1994.
[42] Berge H F M T. Building Capacity for Systems Research at National Agricultural Research Centres: SARP's Experience[M]. Springer Netherlands: Systems approaches for agricultural development, 1993: 515-538.
[43] Williams J R, Jones C A, Dyke P T. A modeling approach to determining the relationship between erosion and soil productivity[J]. Transactions of the ASAE, 1984: 27(1):129-144.
[44] Lemmon H. Comax: An expert system for cotton crop management[J]. Science, 1986, 233(4759): 29-33.
[45] Reddy V R, Acock B, Whisler F D. Crop management and input optimization with GLYCIM: Differing cultivars[J]. Computers & Electronics in Agriculture, 1995, 13(1): 37-50.
[46] Place R E, Brown D M. Modelling corn yields from soil moisture estimates: Description, sensitivity analysis and validation[J]. Agricultural & Forest Meteorology, 1987, 41(41): 31-56.
[47] Mccown R L, Hammer G L, Hargreaves J N G, et al. APSIM: An agricultural production system simulation model for operational research[J]. Mathematics & Computers in Simulation, 1995, 39(3): 225-231.
[48] 高亮之, 金之慶. RCSODS—水稻栽培計算機模擬優(yōu)化決策系統(tǒng)[J]. 計算機農(nóng)業(yè)應(yīng)用,1993(3):14-20.
Gao Liangzhi, Jin Zhiqing. RCSODS - Rice Cultivation Computer Simulation Optimization Decision System[J]. Computer Agriculture Applications, 1993(3): 14-20. (in Chinese with English abstract)
[49] 潘學(xué)標, 韓湘玲, 石元春. 一個可用于栽培管理的棉花生長發(fā)育模擬模型: COTGROW[J]. 中國農(nóng)業(yè)科學(xué), 1996, 29(2): 94.
Pan Xuebiao, Han Xiangling, Shi Yuanchun. A cotton growth and development simulation model for culture management - COTGROW[J]. Scientia Agricultural Sinica, 1996, 29(2): 94. (in Chinese with English abstract)
[50] Spitters C J T, Schapendonk A H C M. Evaluation of breeding strategies for drought tolerance in potato by means of crop growth simulation[J]. Plant & Soil, 1990, 123(2): 193-203.
[51] Thornton P K, Booltink H W G, Stoorvogel J J. A computer program for geostatistical and spatial analysis of crop model outputs[J]. Agronomy Journal, 1997, 89(4): 620-627.
[52] Priya S, Shibasaki R. National spatial crop yield simulation using GIS-based crop production model[J]. Ecological Modelling, 2001, 136(2/3): 113-129.
[53] Tan G X, Shibasaki R. Global estimation of crop productivity and the impacts of global warming by GIS and EPIC integration[J]. Ecological Modelling, 2003,168(3): 357-370.
[54] 辛紅敏. 基于APSIM模擬與GIS分析的小麥-玉米增產(chǎn)潛力及其地域差異研究[D]. 保定:河北農(nóng)業(yè)大學(xué),2013.
Xin Hongmin. An APSIM Simulation and GIS Analysis Based Study on Yield Increase Potential and Its Spatial Variation of the Wheat-Maize System[D]. Baoding: Agricultural university of Hebei, 2013. (in Chinese with English abstract)
[55] 徐冠華,柳欽火,陳良富,等. 遙感與中國可持續(xù)發(fā)展:機遇和挑戰(zhàn)[J]. 遙感學(xué)報,2016,20(5):679-688.
Xu Guanhua, Liu Qinhuo, Chen Liangfu, et al. Remote sensing for China’s sustainable development: opportunities and challenges[J]. Journal of Remote Sensing, 2016, 20(5): 679-688. (in Chinese with English abstract)
[56] Xiao Zhiqiang, Liang Shunlin, Wang Jindi, et al. Use of general regression neural networks for generating the GLASS leaf area index product from time-series MODIS surface reflectance[J]. IEEE Transactions on Geoscience and Remote Sensing, 2014, 52(1): 209-223.
[57] Chen J M, Pavlic G, Brown L, et al. Derivation and validation of Canada-wide coarse-resolution leaf area index maps using high-resolution satellite imagery and ground measurements[J]. Remote Sensing of Environment, 2002, 80(1): 165-184.
[58] Myneni R B, Hoffman S, Knyazikhin Y, et al. Global products of vegetation leaf area and fraction absorbed PAR from year one of MODIS data[J]. Remote Sensing of Environment, 2002, 83(1/2): 214-231.
[59] Launay M, Guerif M. Assimilating remote sensing data into a crop model to improve predictive performance for spatial applications[J]. Agriculture, Ecosystems & Environment, 2005, 111(1/2/3/4): 321-339.
[60] 楊鵬,吳文斌,周清波,等. 基于作物模型與葉面積指數(shù)遙感影像同化的區(qū)域單產(chǎn)估測研究[J]. 農(nóng)業(yè)工程學(xué)報,2007,23(9):130-136.
Yang Peng, Wu Wenbin, Zhou Qingbo, et al. Assimilating remotely sensed LAI into GIS-based EPIC model for yield assessment on regional scale[J]. Transactions of the Chinese Society of Agricultural Engineering (Transactions of the CSAE), 2007, 23(9): 130-136. (in Chinese with English abstract)
[61] 姜浩. 基于作物模型同化遙感物候信息的冬小麥估產(chǎn)方法研究[D]. 成都:電子科技大學(xué),2011.
Jiang Hao. Study on Estimation of Yield of Winter Wheat Based on Crop Model[D]. Chengdu: University of Electronic Science and Technology of China, 2011. (in Chinese with English abstract)
[62] Bouman B A M. Linking Physical Remote Sensing Models with Crop Growth Simulation Models, Applied for Sugar Beet[J]. International Journal of Remote Sensing, 1992, 13(14): 2565-2581.
[63] 王純枝,宇振榮,辛景峰,等. 基于遙感和作物生長模型的作物產(chǎn)量差估測[J]. 農(nóng)業(yè)工程學(xué)報,2005,21(7):84-89.
Wang Chunzhi, Yu Zhenrong, Xin Jingfeng, et al. Yield gap estimation by combining remote sensing and crop growth model[J]. Transactions of the Chinese Society of Agricultural Engineering (Transactions of the CSAE), 2005, 21(7): 84-89. (in Chinese with English abstract)
[64] de Wit A J W, Van D C A. Crop model data assimilation with the Ensemble Kalman filter for improving regional crop yield forecasts[J]. Agricultural and Forest Meteorology, 2007, 146(1/2): 38-56.
[65] Delecolle R, Guerif M. Introducing spectral data into a plant process model for improving its prediction ability[J]. Spectral Signatures of Objects in Remote Sensing, 1988, 127(1): 125.
[66] Maas S J. Using satellite data to improve model estimates of crop yield[J]. Agronomy Journal, 1988(80): 655-662.
[67] Clevers J G P W, Leeuwen H J C V. Combined use of optical and microwave remote sensing data for crop growth monitoring[J]. Remote Sensing of Environment, 1996, 56(1): 42-51.
[68] 趙艷霞. 遙感數(shù)據(jù)和作物模型結(jié)合方法及初步應(yīng)用[D]. 北京:北京大學(xué),2005.
Zhao Yanxia. Combination of Remote Sensing Data and Crop Model and Its Preliminary Application[D]. Beijing: Beijing University, 2005. (in Chinese with English abstract)
[69] Dente L, Satalino G, Mattia F, et al. Assimilation of leaf area index derived from ASAR and MERIS data into CERES-Wheat model to map wheat yield[J]. Remote Sensing of Environment, 2008, 112(4): 1395-1407.
[70] Curnel Y, Ajw D W, Duveiller G, et al. Potential performances of remotely sensed LAI assimilation in WOFOST model based on an OSS experiment[J]. Agricultural & Forest Meteorology, 2012, 151(12): 1843-1855.
[71] 黃健熙,武思杰,劉興權(quán),等. 基于遙感信息與作物模型集合卡爾曼濾波同化的區(qū)域冬小麥產(chǎn)量預(yù)測[J]. 農(nóng)業(yè)工程學(xué)報,2012,28(4):142-148.
Huang Jianxi, Wu Sijie, Liu Xingquan, et al. Regional winter wheat yield forecasting based on assimilation of remote sensing data and crop growth model with Ensemble Kalman method[J]. Transactions of the Chinese Society of Agricultural Engineering (Transactions of the CSAE), 2012, 28(4): 142-148. (in Chinese with English abstract)
[72] Huang Jianxi, Tian Liyan, Liang Shunlin, et al. Improving winter wheat yield estimation by assimilation of the leaf area index from Landsat TM and MODIS data into the WOFOST model[J]. Agricultural & Forest Meteorology, 2015, 204: 106-121.
[73] Ma Guannan, Huang Jianxi, Wu Wenbin, et al. Assimilation of MODIS-LAI into the WOFOST model for forecasting regional winter wheat yield[J]. Mathematical & Computer Modelling, 2013, 58(3/4): 634-643.
[74] Huang Jianxi, Sedano F, Huang Yanbo, et al. Assimilating a synthetic Kalman filter leaf area index series into the WOFOST model to improve regional winter wheat yield estimation[J]. Agricultural & Forest Meteorology, 2016, 216: 188-202.
[75] 馬玉平,王石立,張黎,等. 基于遙感信息的華北冬小麥區(qū)域生長模型及模擬研究[J]. 氣象學(xué)報,2005,63(2):204-215.
Ma Yuping, Wang Shili, Zhang Li, et al. Study on winter wheat regional simulation model based on remete sensing data and its simulations in north China[J]. Acta Meteorologica sinica, 2005, 63(2): 204-215. (in Chinese with English abstract)
[76] Guerif M., Duke C. Calibration of the SUCROS emergence and early growth module for sugar beet using optical remote sensing data assimilation[J]. European Journal of Agronomy, 1998, 9: 127-136.
[77] Moulin S, Fischer A, Dedieu G, et al. Temporal variations in satellite reflectances at field and regional scales compared with values simulated by linking crop growth and SAIL models[J]. Remote Sensing of Environment, 1995, 54(3): 261-272.
[78] Moulin S, Fischer A, Dedieu G. Assimilation of Shortwave Remote Sensing Observations Within an Agrometeorological Model-Crop Production Estimation[C]. Geoscience and Remote Sensing Symposium, 1996. IGARSS '96. 'Remote Sensing for a Sustainable Future, International. IEEE Xplore, 1996(4): 2366-2368.
[79] Laurent K, Fischer A, Moulin S, et al. Satellite measurements as a constraint on estimates of vegetation carbon budget[J]. Tellus Series B-chemical & Physical Meteorology, 1995, 47(1/2): 251-263.
[80] 楊沈斌,申雙和,李秉柏,等. ASAR數(shù)據(jù)與水稻作物模型同化制作水稻產(chǎn)量分布圖[J]. 遙感學(xué)報,2009,13(2):282-290.
Yang Shenbin, Shen Shuanghe, Li Bingbo, et al. Mapping rice yield based on assimilation of ASAR data with rice growth model[J]. Journal of Remote Sensing, 2009, 13(2): 282-290. (in English with Chinese abstract)
[81] 王東偉,王錦地,梁順林. 作物生長模型同化MODIS反射率方法提取作物葉面積指數(shù)[J]. 中國科學(xué):地球科學(xué),2010,40(1):73-83.
Wang Dongwei, Wang Jindi, Liang Shunlin. Retrieving crop leaf area index by assimilation of MODIS data into crop growth model[J]. Sci China Earth Sci, 2010, 40(1): 73-83. (in Chinese with English abstract)
[82] Maas S J. Use of remotely-sensed information in agricultural crop growth models[J]. Ecological Modelling, 1988, 41(3): 247-268.
[83] 郭建茂. 基于遙感與作物生長模型的冬小麥生長模擬研究[D]. 南京:南京信息工程大學(xué),2007.
Guo Jianmao. Simulation of Regional Winter Wheat Growth Using Remote Sensing Data and Crop Growth Model[D]. Nanjing: Nanjing University of Information Science & Technology, 2007. (in Chinese with English abstract)
[84] 包姍寧,曹春香,黃健熙,等. 同化葉面積指數(shù)和蒸散發(fā)雙變量的冬小麥產(chǎn)量估測方法[J]. 地球信息科學(xué)學(xué)報,2015,17(7):871-882.
Bao Shanning, Cao Chunxiang, Huang Jianxi, et al. Research on winter wheat yield estimation based on assimilation of leaf area index and evapotranspiration data[J]. Journal of Geo-information Science, 2015, 17(7): 871-882. (in Chinese with English abstract)
[85] Weiss M, Troufleau D, Baret F, et al. Coupling canopy functioning and radiative transfer models for remote sensing data assimilation[J]. Agricultural & Forest Meteorology, 2001, 108(2): 113-128.
[86] 黃健熙,馬鴻元,田麗燕,等. 基于時間序列LAI和ET同化的冬小麥遙感估產(chǎn)方法比較[J]. 農(nóng)業(yè)工程學(xué)報,2015,31(4):197-203.
Huang Jianxi, Ma Hongyuan, Tian Liyan, et al. Comparison of remote sensing yield estimation methods for winter wheat based on assimilating time-sequence LAI and ET [J]. Transactions of the Chinese Society of Agricultural Engineering (Transactions of the CSAE), 2015, 31(4): 197-203. (in Chinese with English abstract)
[87] Huang J, Ma H, Su W, et al. Jointly Assimilating MODIS LAI and ET products into the SWAP model for winter wheat yield estimation[J]. IEEE Journal of Selected Topics in Applied Earth Observations & Remote Sensing, 2015, 8(8): 4060-4071.
[88] 李衛(wèi)國,王紀華,趙春江,等. 基于定量遙感反演與生長模型耦合的水稻產(chǎn)量估測研究[J]. 農(nóng)業(yè)工程學(xué)報,2008,24(7):128-131.
Li Weiguo, Wang Jihua, Zhao Chunjiang, et al. Estimating rice yield based on quantitative remote sensing inversion and growth model coupling[J]. Transactions of the Chinese Society of Agricultural Engineering (Transactions of the CSAE), 2008, 24(7): 128-131. (in Chinese with English abstract)
[89] 朱元勵. 基于遙感和模型同化的水稻生長監(jiān)測預(yù)測技術(shù)研究[D]. 南京:南京農(nóng)業(yè)大學(xué),2009.
Zhu Yuanli. Study on Growth Monitoring and Prediction Technique Based on Assimilation of Remote Sensing and Modeling in Rice[D]. Nanjing: Agricultural University Of Nanjing, 2009. (in Chinese with English abstract)
[90] 王航,朱艷,馬孟莉,等. 基于更新和同化策略相結(jié)合的遙感信息與水稻生長模型耦合技術(shù)的研究[J]. 生態(tài)學(xué)報,2012,32(14):4505-4515.
Wang Hang, Zhu Yan, Ma Mengli, et al. Coupling remotely sensed information with a rice growth model by combining updating and assimilation strategies[J]. Acta Ecologica Sinica, 2012, 32(14): 4505-4515. (in Chinese with English abstract)
[91] 靳華安,王錦地,柏延臣,等. 基于作物生長模型和遙感數(shù)據(jù)同化的區(qū)域玉米產(chǎn)量估算[J]. 農(nóng)業(yè)工程學(xué)報,2012(6):162-173.
Jin Hua’an, Wang Jindi, Bo Yanchen, et al. Estimation on regional maize yield based on assimilation of remote sensing data and crop growth model[J]. Transactions of the Chinese Society of Agricultural Engineering (Transactions of the CSAE), 2012(6): 162-173. (in Chinese with English abstract)
[92] 陳洪萍,賈根鎖,馮錦明,等. 氣候模式中關(guān)鍵陸面植被參量遙感估算的研究進展[J]. 地球科學(xué)進展,2014,29(1):56-67.
Chen Hongping, Jia Gensuo, Feng Jinming, et al. Remote sensing estimates of key land surface vegetation variables used in climate model: A review[J]. Advances in Earth Science, 2014, 29(1): 56-67. (in Chinese with English abstract)
[93] 馬明國,宋怡,王旭峰,等. AVHRR、VEGETATION和MODIS時間系列遙感數(shù)據(jù)產(chǎn)品現(xiàn)狀與應(yīng)用研究進展[J]. 遙感技術(shù)與應(yīng)用,2012,27(5):663-670.
Ma Mingguo, Song Yi, Wang Xufeng, et al. Development status and application research of time series remote sensing data products based on AVHRR, VEGETATION and MODIS[J]. Remote Sensing Technology and Application, 2012, 27(5): 663-670. (in Chinese with English abstract)
[94] 馬玉平,王石立. 利用遙感技術(shù)實現(xiàn)作物模擬模型區(qū)域應(yīng)用的研究進展[J]. 應(yīng)用生態(tài)學(xué)報,2004,15(9):1655-1661.
Ma Yuping, Wang Shili. A review on research of implement application for crop simulation model in regional scale by using remote sensing technology[J]. Chinese Journal of Applied Ecology, 2004, 15(9): 1655-1661. (in Chinese with English abstract)
Research progress and prospect on combining crop growth models with parameters derived from quantitative remote sensing
Wu Lei1,2, Bai Junhua1,3※, Xiao Qing1,3, Du Yongming1,3, Liu Qinhuo1,3, Xu Liping2
(1.100101; 2.832000;3.100101)
Combining the crop growth models with the remote sensing parameters, is important to realize the applications in the large spatial scale for the former, and also to improve the rationality and accuracy of inversion theory for the latter. Some research fields, such as the ecology, agriculture, resource investigation and global climate change, would use the data derived from the combination form. The overview includes 3 parts, i.e. the international crop growth model, the quantitative remote sensing parameters and the parametric methods. From the view of the spatial expansion for the application of the crop growth model, the development duration of crop growth models was divided into 3 stages: The construction of the mechanism models, the application in the point scale, and the application in the regional and global scale. In order to understand the situation and foundation of the inter-discipline combination from the crop growth simulation and the quantitative remote sensing, the paper describes 3 important contents. The first is overviewing the main simulation processes and the input and output parameters for the typical crop growth models. The second is summarizing the remote sensing inversion parameters which can be used as the initialization and driving data for the application of crop growth simulation models in the regional and global scale, establishing the corresponding relation between the simulation process of crop growth model and the parameters from the quantitative remote sensing. And the third is comparing 3 kinds of combination methods between the crop growth model and the parameters derived from the quantitative remote sensing, and emphasizing the differences, advantages and disadvantages for the 3 combination methods. Based on the contents mentioned above, 3 topics for discussion are proposed. The first topic is the application limitations of crop growth models in the large spatial scale and its relationship with quantitative remote sensing parameters. The second one is the influence of the scale effect from the input parameters when the crop growth model is used to simulate the crop growth in the larger region. And the third one is to discuss the development direction of combination methods. It is hopeful to provide a kind of thinking for combining the crop growth models with the parameters from the quantitative remote sensing through the overview, summary and discussion. And it is clearly concluded that the data from the quantitative remote sensing can provide initialization data for crop growth models to some extent in the regional and global scale, and the application in a large space scale is the direction of crop growth model. The conclusion shows further that it is important to pay attention to the scale problem of the model parameters, and that the data dis-matching for the same parameter from the crop growth and remote sensing can result in the huge error of estimation on the output data due to the difference of physical meaning from the 2 disciplines. Understanding that the data from the quantitative remote sensing could enhance the ability of simulating the crop conditions and yield at the large scale was also helpful to understand that the remote sensing had the ability of deriving the biochemical and biophysical information from the ground surface exactly. And furthermore, it is expected that the correct combination parameters should be chosen to deduce the propagation of error and uncertainty, the assimilation methods would still preserve the mainstream style for the combination, and following the increasing accuracy of data from the remote sensing and crop model, the fusion model for the model of crop growth and remote sensing can be constructed to play a greater application value in the environmental monitoring and agricultural production.
remote sensing; models; vegetation; crop growth; quantitative remote sensing; combination; development
10.11975/j.issn.1002-6819.2017.09.020
TP79; S31
A
1002-6819(2017)-09-0155-12
2016-11-08
2017-04-01
國家重點基礎(chǔ)研究發(fā)展計劃(973計劃)課題“復(fù)雜地表遙感輻射/散射機理及動態(tài)建?!保?013CB733401);國家自然基金“多尺度觀測與作物生長模型相結(jié)合的輻射傳輸一體化模擬研究”(41671366);國家自然基金“基于能量平衡的作物冠層熱輻射方向性模型”(41571359);國家科技基礎(chǔ)平臺建設(shè)項目“測繪地物波譜本底數(shù)據(jù)庫建設(shè)”(2014FY210800)
吳 蕾,女,新疆哈密人,研究方向為全球變化與生態(tài)響應(yīng)。石河子石河子大學(xué),832000。Email:wl992917@sina.com
柏軍華,男,重慶人,博士,助理研究員,研究方向為作物生長模型定量遙感、地面試驗遙感。北京中國科學(xué)院遙感與數(shù)字地球研究所,100101。Email:bjh0902@163.com
吳 蕾,柏軍華,肖 青,杜永明,柳欽火,徐麗萍. 作物生長模型與定量遙感參數(shù)結(jié)合研究進展與展望[J]. 農(nóng)業(yè)工程學(xué)報,2017,33(9):155-166. doi:10.11975/j.issn.1002-6819.2017.09.020 http://www.tcsae.org
Wu Lei, Bai Junhua, Xiao Qing, Du Yongming, Liu Qinhuo, Xu Liping. Research progress and prospect on combining crop growth models with parameters derived from quantitative remote sensing[J]. Transactions of the Chinese Society of Agricultural Engineering (Transactions of the CSAE), 2017, 33(9): 155-166. (in Chinese with English abstract) doi:10.11975/j.issn.1002-6819.2017.09.020 http://www.tcsae.org