摘要
利用新疆艾比湖农业灌溉地区MODIS EVI遥感影像数据和同期野外调查获得的75个样方生物量数据,对比分析了EVI与地表生物量多个回归方程的相关系数,进而建立了EVI与农作物生物量的多元回归模型。地表农作物生物量与EVI拟合方程相关系数大小依次为:幂函数>指数方程>三次多项式方程>一元线性拟合方程,同时采用13个独立样方采样数据进行误差分析,证明幂函数模型拟合精度最高。因此采用幂函数模型对研究区农作物生物量进行估算,结果表明,在人工灌溉地区,作物生长茂盛,生物量分布于1～10 kg/m2区间;而在非浇灌地区,地表植被稀疏,多为耐旱耐盐碱植物,地表生物量多在1 kg/m2以下。西北农业灌溉地区地表生物量与土壤水分密切相关,人工灌溉是影响地表农作物生物量变化的主要因素。
Based on field survey data, MODIS Enhanced Vegetation Index（EVI） image acquired in 2005 was applied for crop biomass estimation in Ebinur lake area. A total of 75 plots were investigated and the relationship between biomass and vegetation indices was clarified. A series of regression models comparing biomass and EVI were established. The sequence of correlation coefficients from high to low was power regression〉exponential regression〉polynomial regression〉linear regression. The tested result by expressing with power regression model gave more accurate estimation（r=0.78）. A crop biomass distribution map was derived, the crop biomass was higher in irrigated areas （between 1 kg/m^2 and 10 kg/m^2） and lower in non-irrigated areas （less than 1 kg/m^2 ）, indicating that the local crop biomass is mainly influenced by agricultural irrigation conditions.
出处
《农业工程学报》
EI
CAS
CSCD
北大核心
2008年第10期141-144,314,共5页
Transactions of the Chinese Society of Agricultural Engineering
基金
中国科学院知识创新重大方向性项目(KZCX3-SW-334)
广西科学研究与技术开发计划(0663020)
关键词
农业灌溉区
植被指数
生物量
EVI
回归分析
irrigation farming area, vegetation index, biomass, EVI, regression analysis
