遥感技术在新乡市人民公园绿化调查中的应用

Application of remote sensing technique to the investigation of plants in People's Park of Xinxiang

对2011年8月覆盖新乡市人民公园的SPOT5遥感影像进行预处理提取植被信息,再用代表性样地法采集植被信息,选定归一化植被指数(NDVI)和比值植被指数(RVI)作为自变量,以实测样本数据(LVV)作为应变量,采用多元回归分析法建立基于遥感影像的新乡市人民公园植被遥感模型.结果表明:LVV与VI呈极显著的相关关系,其相关系数多以相对均质植被高于植被总体.每种植被样方优化出一个模型,即针阔混交林:LVV=16.216RVI_(TOA)+19.698RVI_(DN)-9.112(R^2=0.866,RMSE=0.289);阔叶林;LVV=8.111RVI_(PAC)-3.142(R^2=0.795,RMSE=0.512);灌木:LVV=313.621NDVI_(DN)~3-19.118NDVI_(DN)~2+2.612(R^2=0.812,RMSE=0.714);草地:LVV=3.121RVI_(TON)+1.992RVI_(DN)-4.002(R^2=0.892,RMSE=0.547);总体植被:LVV=2.231RVI_(PAC)-7.112NDVI_(SR)+5.122NDVI_(PAC)+9.982NDVI_(DN)-1.417(R^2=0.796,RMSE=0.712).这些优选模型在新乡市人民公园的植被调查中具有一定的应用价值.

Extracting vegetation information from SPOT5 remote sensing image and representative sample method to collect the vegetation information of the People＇s Park in Xinxiang City were used to derive two vegetation indices,i. e.,normalized difference vegetation index（NDVI）,and ratio vegetation index （RVI）,to establish the vegetation remote sensing investigation model using multiple regression analysis.The results showed that LVV was significantly correlated with VI.LVV-VI correlation coefficients of relatively ＇pure＇ vegetation are higher than those of total vegetation.One ＇best＇ model was selected for each of the vegetation quadrates,i.e.,broad-conifer leaf mixed forest： LVV=16.216RVITOA ＋19.698 RVIDN-9.112 （RZ=0.866,RMSE=0.289）,broad-leaf forest ：LVV=8.111RVPAC-3.142 （RZ=0.795,RMSE=0.512）,shrub：LVV=313.621NDVIDN3-19.118NDVIDN2＋2.612 （R2=0.812,RMSE=0.714）,grass ：LVV= 3.12 IR V In, ＋ 1.992R V IDN-4.002 （ R2 =0.892,RMSE =0.547）,and total vegetation ： L V V =2.231R V Ip3C-7.112NDVIsR ＋ 5.122NDVIPAC＋9.982NDVIDN-1.417 （R2=0.796,RMSE=0.712）.The optimization model has certain application value in the People＇s Park vegetation survey in Xinxiang City.