中国陆地生态系统总初级生产力VPM遥感模型估算  

ESTIMATION OF GROSS PRIMARY PRODUCTIVITY IN CHINESE TERRESTRIAL ECOSYSTEMS BY USING VPM MODEL

作　　者：陈静清 闫慧敏[1] 王绍强[1] 高艳妮[1] 黄玫[1] 王军邦[1] 肖向明 

机构地区：[1]中国科学院地理科学与资源研究所,北京100101 [2]中国科学院大学,北京100049 [3]Department of Microbiology and Plant Biology,Center for Spatial Analysis,University of Oklahoma

出　　处：《第四纪研究》2014年第4期732-742,共11页Quaternary Sciences

基　　金：中国科学院战略性先导科技专项项目(批准号:XDA05050602);中国科学院重点部署项目(批准号:KSZD-EW-Z-021-02)共同资助

摘　　要：陆地生态系统总初级生产力（Gross Primary Productivity,简称GPP）时空格局及其变化动态的准确监测是区域碳收支研究的核心问题之一,遥感模型正在为区域碳通量监测提供更为实时、准确的模拟数据.基于中分辨率成像光谱仪（MODerate-resolution Imaging Spectroradiometer,简称MODIS）遥感数据和涡度相关碳通量观测数据发展而来的VPM模型经过10年的努力,目前已经在全球涵盖十类生态系统的21个站点上开展模型的校验与验证研究,为区域GPP的准确估算与监测奠定了方法基础.本研究构建了评估GPP区域格局的VPM模型区域模式,以空间分辨率500m、时间步长8天的MODIS卫星影像数据以及相同时空分辨率的温度与光合有效辐射数据为模型输入数据,模拟估算我国2006～2008年GPP及其空间分布格局.VPM模型模拟的中国陆地生态系统GPP年总量平均值为5.0Pg C/a,其中森林、草地、农田和灌丛生态系统分别占34％,17％,37％和12％.本研究模拟的全国GPP总量与多模型模拟的平均结果（5.40 Pg C/a）相当,但不同模型估算的各类生态系统GPP存在较大差异.本研究通过利用遥感数据对VPM模型中的关键参数（最大光能利用率）进行参数空间化,表达同一土地覆被类型内部光能利用率的空间异质性;应用目前我国准确性最为可靠的土地利用与植被数据提取土地覆被数据,首次将农业多熟种植作为单独的植被类型引入模型中,模型参数与输入数据的精度保障了模型模拟结果的可靠性.Remote sensing model provides more real-time and more accurate simulation data for regional carbon flux observing.Accurate simulation needs sufficient validations of every ecosystem at the ecosystem scale by using flux observation data.More and more studies have shown that current Gross Primary Productivity （GPP） simulation still has great uncertainty,for example,MODIS GPP product （MOD17） doesn＇t take multiple cropping and C4 plant as separate vegetation types,therefore,the simulation results of such ecological zones have large underestimation.VPM model has been developed based on MODIS data and flux observation data.After 10 years of effort,a series of studies on model calibration and validation has been carried out in 21 sites covering ten types of typical ecosystems throughout the world,including Haibei alpine shrub and alpine wetland,temperate semi-arid grasslands of Inner Mongolia,the Changbai Mountain conifer,winter wheat-maize rotation cropland of Yucheng flux station and Tongyu cropland （corn） station,which have shown good simulation capabilities and laid a scientific basis for the estimation of regional GPP.In this study,we build the VPM model for regional pattern evaluation of GPP in terrestrial ecosystems.Then we estimate the spatial pattern and temporal variation of GPP in China during 2006～2008 using MODIS data,rasterized temperature and photosynthetically active radiation （PAR） data with a resolution of 500m at 8-day intervals as input data.The total annual GPP of China＇s terrestrial ecosystems is 5.0Pg C/a during the period of 2006～2008.Forests,grassland,cropland and shrub account for 34％,17％,37％,and 12％,respectively.These vegetation types can be arranged by GPP values in descending order as follows：evergreen broadleaf forest,deciduous broadleaf forest,mixed forest,triple-cropping cropland,double-cropping cropland,single-cropping cropland,evergreen needleleaf forest,shrub,deciduous needleleaf forest,grassland,alpine meadow and alpine grassland.The total annual GPP

关 键 词：VPM模型 总初级生产力 最大光能利用率 MODIS 

分 类 号：Q148[生物学—生态学] TP79[生物学—普通生物学]