摘要
采用美国ASD公司Fieldspec3光谱仪,在内蒙古呼伦贝尔草原区进行了高光谱遥感地面观测与地上干物质量测定,并获取与试验同期的Terra/MODIS遥感数据。运用回归分析方法,构建实测归一化植被指数(ASD NDVI)与地上干物质量(ANPP)之间的地面光谱模型,分析ASD NDVI与Modis NDVI相关关系,建立基于Modis NDVI预测地上干物质量的Modis光谱模型。研究表明:呼伦贝尔草原的区域最优地面光谱模型为指数函数(n=49,R2=0.7496,P<0.001),ASD NDVI与Modis NDVI为线性相关关系(n=49,R2=0.7939,P<0.001),所建立Modis光谱模型的标准误差SE=62g/m2,平均预测精度为68.86%,该模型为呼伦贝尔草原区区域产草量测定提供了新方法。
During July and August 2009, an ASD Fieldspec3 spectroradiometer was used to measure hyperspectral data; Harvested above -ground net primary productivities（ANPP） were measured, and Terra/MODIS remotely - sensed data corresponded with the field surveys on the Hulunbeier grassland, Inner Mongolia, China were ob- tained. By regression analysis method, the ground spectral models were built for estimating ANPP with Normal- ized Difference Vegetation Index measured in the field （ASD NDVI）. Then the relationship between NDVI got from MODIS/TERRA （MODIS NDVI） and ASD NDVI was analyzed to establish a MODIS Spectral Model to es- timate ANPP with MODIS NDVI. The results showed that the best ground Spectral Model was a exponential equa- tion in study area（ n = 49, R2 = 0. 7496, P 〈 0. 001 ）. The relationship between MODIS NDVI and ASD NDVI was linear equation（n = 49, R2 = 0. 7939, P 〈 0. 001 ）. MODIS Spectral Model established with two equations had standard error SE = 62g/m2 and 68.86% average prediction accuracy. The study provided a new method to pre- dict grassland biomass on the Hulunbeier grassland.
出处
《干旱区资源与环境》
CSSCI
CSCD
北大核心
2012年第5期108-112,共5页
Journal of Arid Land Resources and Environment
基金
环保公益性行业科研专项经费(200809125
200909021)
中国环境科学研究院中央级公益性科研院所基本科研业务专项项目(2009KYYW07)资助
作者简介
陈艳梅（1970-），女，河北邢台人。Email：ehenyml970@sohu．com
