Journal
CENTRAL EUROPEAN JOURNAL OF GEOSCIENCES
Volume 5, Issue 2, Pages 208-222Publisher
VERSITA
DOI: 10.2478/s13533-012-0124-9
Keywords
regression; multispectral imagery; coniferous forest; forest inventory; Rila Mountain
Categories
Funding
- European Social Fund through contract with the Ministry of Education of Bulgaria [BG051PO001/07/3.3-02/63/170608]
Ask authors/readers for more resources
In this study, regression-based prediction of volume and aboveground biomass (AGB) of coniferous forests in a mountain test site was conducted. Two datasets - one with applied topographic correction and one without applied topographic correction - consisting of four spectral bands and six vegetation indices were generated from SPOT 5 multispectral image. The relationships between these data and ground data from field plots and national forest inventory polygons were examined. Strongest correlations of volume and AGB were observed with the near infrared band, regardless of the topographic correction. The maximal correlation coefficients when using plotwise data were -0.83 and -0.84 for the volume and AGB, respectively. The maximal correlation with standwise data was -0.63 for both parameters. The SCS+C topographic correction did not significantly affect the correlations between spectral data and forest parameters, but visually removed much of the topographically induced shading. Simple linear regression models resulted in relative RMSE of 32-33% using the plotwise data, and 43-45% using the standwise data. The importance of the source and the methodology used to obtain ground data for the successful modelling was pointed out.
Authors
I am an author on this paper
Click your name to claim this paper and add it to your profile.
Reviews
Recommended
No Data Available