| 森林生物量与遥感信息的相关性 |
| |
| 引用本文： | 仝慧杰,冯仲科,罗旭,张彦林.森林生物量与遥感信息的相关性[J].北京林业大学学报,2007(Z2). |
| |
| 作者姓名： | 仝慧杰 冯仲科 罗旭 张彦林 |
| |
| 作者单位： | 北京林业大学测绘与3S技术中心，北京林业大学测绘与3S技术中心，中国计量科学研究院，北京林业大学测绘与3S技术中心 中国国土资源部航空物探遥感中心，甘肃林业职业技术学院 |
| |
| 基金项目： | 国家自然科学基金项目(90302014) |
| |
| 摘 要： | 利用遥感数据研究森林的生物量,建立遥感信息模型,首先要分析各波段与生物量的相关性.通过建立甘肃省小陇山党川林场中幼林典型样地,并伐树称量,建立模型计算出样地的生物量.对试验区的TM影像进行校正,对应每块样地中心点的GPS测量坐标,获取了样地像元各波段的灰度值,并计算各种植被指数.利用MATLAB软件计算了样地生物量与遥感影像各波段的灰度值、各种植被指数的相关系数.在P<0.05水平上,生物量与TM1、TM2、TM3、TM6成显著的负相关.与归一化植被指数、比值植被指数和绿度成显著的正相关.建立遥感生物量模型应用这些呈显著正、负相关的波段和派生数据.采用逐步回归的方法建立了生物量与植被指数的统计方程,以及生物量与遥感TM影像的各波段灰度值、植被指数的统计方程.但在P<0.05水平上,只有NDVI、RVI两个因子复相关系数不高.
|
| 关 键 词： | 遥感信息 生物量模型 相关系数 中幼林 植被指数 逐步回归 |
Correlations between forest biomass and remote sensing informaton. |
| |
| TONG Hui-jie.Correlations between forest biomass and remote sensing informaton.[J].Journal of Beijing Forestry University,2007(Z2). |
| |
| Authors: | TONG Hui-jie |
| |
| Institution: | TONG Hui-jie~ |
| |
| Abstract: | The research aims to find the correlations between forest biomass and remote sensing information to build the remote sensing model for forest biomass estimating.The used 40 pieces of plots,in which the trees were in young or middle age,were built by the typical sampling method in Dangchuan Forest Management Station in Gansu Province.The plots were positioned by GPS and calculated using none-real-time difference adjustment method.The biomass of trees in every species cut down was weighted to build the biomass model for each tree species and the biomass of each plot acquired by the model.The TM image of research area captured during the growing season of trees was rectified so that the accuracy was less than one pixel.By intersecting the plot center point shape file to image using ERDAS software,the grey datum of each plot in every band was obtained,and the vegetation indices,bright,green and wet were computed.Then the correlation between plot forest biomass and the remote sensing data was computed.At P<0.005 level,the correlation coefficients between biomass and band 1,band 2,band 3,as well as band 6 were obvious negative and signifcant positive with vegetation indices and green index.The research indicated that the forest biomass model should use the band 1,band 2,band 3,band 6,NDVI,RVI and GREEN.By step regression,the equation used to estimate the biomass of the region was constructed between the biomass of plots and remote sensing data.But at the significant level P<0.05,only two factors,i.e.NDVI and RVI in the equation.So it is necessary to compose more vegetation indices deriving from TM other bands data,and to find the forest spectrum characteristics further. |
| |
| Keywords: | remote sensing information forest biomass model correlation coefficient young and middle aged forests vegetation index step regression |
| 本文献已被 CNKI 等数据库收录！ |
|
