不同利用状况下草原遥感估产模型  

Yield estimation model for different utilization status grassland based on remote sensing data

作　　者：陶伟国[1] 徐斌[1] 刘丽军[2] 杨秀春[1] 覃志豪[1] 

机构地区：[1]农业部资源遥感与数字农业重点开放实验室,北京100081 [2]中国农业科学院农业资源与农业区划研究所,北京100081

出　　处：《生态学杂志》2007年第3期332-337,共6页Chinese Journal of Ecology

基　　金：国家高技术研究发展专项(2006AA10Z242);农业部全国草原遥感监测资助项目

摘　　要：在内蒙古自治区的温性草甸草原、温性草原、温性荒漠草原、温性草原化荒漠和温性荒漠5种具有代表性的草地类型区地面观测数据的基础上,考虑草地的利用状况信息,分类建立了产草量与4种植被指数(由MODIS数据计算得到)的回归估产模型,将利用状况这一定性变量作为虚拟变量与遥感估算模型相结合,找出了3种利用状况下的鲜草产量最优混合估算模型和估算指数。结果表明:(1)EVI是反映产草量变化的最好指标,分类构建的模型平均测产精度达到了80%;(2)将利用状况作为虚拟变量考虑之后建立的混合测算模型精度达到了79%,明显高于利用状况未知时构建的混合模型的精度,比分类建模应用更简洁方便。Based on the observation data obtained from five representative steppes, i. e. , temperate meadow steppe, temperate steppe, temperate desert steppe, temperate steppe-desert and temperate desert in Inner Mongolia, regression models were established to express the relations between each grassland biomass and four vegetation indices, with the information about grassland utilization status considered. The qualitative variable utilization status was introduced into remote sensing estimation models as a dummy variable, and the optimal mixed estimation models and vegetation indices for three types of utilization status were selected. The results showed that EVI was the best index reflecting grassland yield variation, and the average estimation precision of the models was 80%. The precision of the mixed models introduced with utilization status （mixed model Ⅱ ） arrived to 79% , being higher than that of the models without the information of utilization status. Comparing with the models established for different utilization status, mixed model Ⅱ had the advantages of concision and convenience.

关 键 词：草原估产 利用状况 植被指数 MODIS 虚拟变量 

分 类 号：S812[农业科学—草业科学]