Aerial view of Warra 5D 23.2.2004. Image: Leigh Edwards

This work presents results from the landscape-level carbon-sequestration-forecasting model CAR4D, by way of visualisation. We start by comparing the standing sequestered carbon, calculated by different means, for three similar oldgrowth forests of relatively high biomass. We then proceed to long-term steady states under different management options, landscape-level differences, fire scenarios, fly-throughs and 3D animation of growth. Carbon sequestered in live biomass of the eucalypts/total-forest were: Eucalyptus regnans(Tasmania): 600/850 t-C/ha; Eucalyptus obliqua (Tasmania) 732/760 t-C/ha; and E. regnans (Victoria, Van Pelt et al., 2004): 914/953 t-C/ha. The differences in these values of sequestration for these sites were not necessarily matched to individual trees’ yields based on site index, but were related more to stand-level effects: the rate of self thinning, height growth due to competition and possibly higher than expected error margins in allometrics. This unexpected result indicated a need for models that are more comprehensive at various levels. In this study, the appearance of mature trees on hilly terrain varied greatly depending on the direction and amount of the spur development in the buttress and on local topography. No definitive directional preference for spur development was found, but instead a diversity of buttress shapes and micro-habitats on steep slopes. Taper equations are the most scientifically representative method for rendering the trunks of mathematically derived forests. Five taper equations were tested on various sizes ofEucalyptus regnans. Two new equations allowed better fits for individual trees. The overall best fit equation was based on first principles of tree growth plus an extra parameter for buttress shape—totalling six parameters. That equation was combined with realistic cross-sections to calculate whole-trunk 3D VRML models as a function of DBH. A complex adjustment to trunk shape was included for uneven, sloping ground. Animation of trunk growth was created by general expansion of the perimeter of cross-sections combined with preferential expansion corresponding to spurs. For more mature trees the preferential expansion was reduced, corresponding to achievement of structural stability. Tree trunks and animations were rendered in VRML format. The result allowed within-stand models, including growth, to be rendered and visualised by still frames or by fly-throughs.