摘要
该文以我国云南省西双版纳的热带森林植被为例,对热带森林植被生物量的遥感地学估算进行初步探讨。研究表明,只用LANDSATTM的波段数据建立生物量的回归估算模型时,其模型的复相关系数只有0 118,而利用其主成分或植被指数建立回归估算模型时,其模型的复相关系数不仅没有提高,反而在一定程度上有所降低。原始波段与其主成分结合能提高估算模型的复相关系数,其效果比原始波段与植被指数结合的效果要好。模型中如果考虑气象因子中的年平均降雨量,能在很大程度上提高其复相关系数。在此基础上引入地形因子,未能显著提高模型的复相关系数。有气象因子参与建立的遥感气象模型和遥感地学模型在0 01水平上回归显著。
The biomass of the tropical forest vegetation was estimated by using remote sensing data,meteorological data and topographical data in Xishuangbanna,Yunnan Province.It is shown as follows: the correlation coefficient of the biomass estimation model formed by LANDSAT TM data is 0 344,which is higher than that of the model formed by principal components of LANDSAT TM or vegetation index.The correlation coefficient of the model formed by LANDSAT TM data and principal component data is higher than that of the model formed by LANDSAT TM data and vegetation index.When the average annual rainfall data is taken into the model,the correlation coefficient can be improved,which reaches 0 588.When the topographical data is taken into the model,the correlation coefficient can only be less improved.The regression is significant for the model with the consideration of meteorological data or the model with consideration of meteorological data and topographic data
出处
《地理与地理信息科学》
CSSCI
CSCD
北大核心
2004年第6期22-25,共4页
Geography and Geo-Information Science
基金
中国科学院知识创新项目(CX10G-E01-02-03)
国家自然科学基金项目(40161007)
四川省青年基金项目(03ZQ026-032)
