Abstract: Based on Landsat TM images and field survey data, and by the aids of 3S techniques, this paper studied the dynamic changes of forest landscape pattern in Taishan Mountain, and their relations with altitude variation. Eight topographical factors were adopted to define the topographical characteristics, and three radiant transects consisting of standard sequential sampling plots were applied to collect information on remote sensing images. The multi-scale pattern of forest landscape differentiation and its correlation with topographical parameters were analyzed by multivariate regression and main factor analyses. The results showed that topographical orientation (position), elevation, and mountain land type were the critical factors at landscape level, slope exposure and gradient were important factors, while slope shape and position were not the significant ones. In the analyses of forest landscape differentiation and gradient, the key elements were elevation exposure, slope shape and position, and elevation, with the consideration of topographical position. There were significant positive correlations among the main topographical factors, which could better explain the multi-scale pattern of forest landscape differentiation. The contribution rate was about 70% at any scale, and the explanatory percentage was increased with increasing scale while the explanatory ability was in adverse, which revealed that the effect of topographical factors on forest landscape pattern is prominent, and scale effect remains an important problem to be explored.

Key words: Biogenic volatile organic compounds(BVOCs), Elevated atmospheric CO₂, O₃