Using forest growth functions in satellite remote sensing
2002
Lang, M. | Nilson, T. | Luekk, T. (Estonian Agricultural University, Tartu (Estonia). Department of Forest Management)
Efficient use of remote sensing tools, especially satellite images, in sustainable forest management requires good knowledge of the processes influencing the spectral characteristics of forest stands. In this work, a forest growth model and a forest reflectance model were used to interpret the spectral changes in remotely sensed signals from forest stands. A set of spruce stands as well as fertile-site and infertile-site birch stands were selected from database of Jaervselja Training and Experimental Forest District, and their reflectance factors were measured from the atmospherically corrected Landsat 7 ETM image of 10 July 1999. Using the forest growth model and the forest reflectance model, the respective theoretical age-dependent reflectance factors were simulated for each ETM channel and forest type. The leaf area index and the canopy closure were estimated using the regression functions derived from the output of the forest growth model. The results showed that most of the variations in age-dependent forest reflectance curves can be explained with forest growth and forest reflectance models. A forest stand is usually described using lots of well-correlated parameters, such as average tree DBH, tree height, etc. These correlation functions, however, are subject to constant change due to forest growth. To be able to make decisions on sustainable forest management, we need growth functions for covariance matrices or the principal components of forest parameters concerning stand age, which would allow a more efficient integration of forest growth and reflectance models, the existing databases and the data obtained by remote sensing
Mostrar más [+] Menos [-]Palabras clave de AGROVOC
Información bibliográfica
Este registro bibliográfico ha sido proporcionado por Estonian University of Life Sciences