期刊名称：ISPRS Annals of the Photogrammetry, Remote Sensing and Spatial Information Sciences
印刷版ISSN：2194-9042
电子版ISSN：2194-9050
出版年度：2007
卷号：XXXVI-7/C50
出版社：Copernicus Publications
摘要：In this investigation the capability of Landsat ETM+ data for estimation of forest stand density and canopy area in National Park of Khabr in Kerman province, Iran in different density classes was evaluated. In a preliminary study, density mapping in the study area was qualitatively implemented and very thin, thin and semi-dense classes were concluded. Measurements made in sample plots in each density class were considered as a reference. Orthorectified satellite data were used to produce different ratios, vegetation indices (VIs), VIs for soil influence decreasing, components of Principal Component Analysis (PCA) and Tasselled Cap transformation features to apply in further analyses together with the original bands. Simple regression between stand parameters and digital values of each synthetic and original band was separately investigated. Ratio3 ((ETM3 – ETM1)/ (ETM3 + ETM1)) showed the highest correlation (R = 0.75 and Adj. R 2 = 0.54) with stand density measurements in sample plots of semi-dense class. Through multiple regression the results of explanation of variability of dependent variables were improved. Utilizing Ratio3 together with Ratio5 ((ETM4 – ETM5)/ (ETM4 + ETM5)) showed the highest ability to model stand density (Adj. R 2 = 0.75) in semi-dense class. Near results achieved from some other bands in this class and the other ones, too. Regression validation showed that the whole models were statistically valid. Higher spatial resolution satellites data such as SPOT5 and IRS are advised to be examined to improve the results
