Here's a summary of the thesis:

Forest covers 40 % of terrestrial areas and Canada holds 10 % of it alone, mostly in a boreal setting. The need for planning and supervising implied by this reality couldn't be filled entirely by traditional means (forest inventories), especially for wood volume, an essential measurement for ecology researchers and forest engineers. A technology named "laser altimetry", which has been in development for several years, seems promising in that regard. It is based on geometric properties: the reflection of a laser beam by any object intercepted. The time elapsed between emission and reception gives the distance between the instrument and the object and, by way of calculation, the elevation of surfaces. Combining a high-resolution scanning laser with a higher sampling intensity, a precise positioning by INS-GPS and an adequate volume model should resolve any remaining shortcomings of that technology so far.

The experimental method was applied mostly on stands which contained two species, a deciduous tree (
Populus tremuloides Michx.) and a coniferous one (Picea glauca [Moench] Voss), both abundant in the Conservation Zone of "Forêt d'Enseignement et de Recherche du Lac Duparquet" (FERLD), in Abitibi, Québec. Volume derived from a sample inventory (38 square plots of 400 m2) done in the summer of 1999 were compared with calculations obtained from the data of an airborne LIDAR (Optech ALTM1020), taken from three flights in the summer of 1998 (400 000 ground points and two million vegetation points). Many of the one-variable relationships tried out proved significant in explaining the ground-truth volume (r2 = 0,52 to 0,77); but the statistical tests linking each of the 8 laser pseudo-volumes to ground wood volumes by a quadratic model gave better results, especially in the case of a laser interpolation were values less than the mean height were left out. Combining that calculation with the cover ratio gives even better results, at least for aspen (r2 = 0,88, good up to 50 m3/ha of variation). If small volumes would have been taken into account (less than 200 m3/ha), results would have been even better still. With the help of stratified classes (density, height and species groupings), a practical cartography of the study zone could be achieved (digital terrain models and vegetation surface models).

You might also be interested in taking a look at the section of the site named "Cartes" (Maps).

Keywords: altimetry, laser, LIDAR, height, volume, forest, FERLD, Duparquet, Abitibi



Conception: Alain Renaud, août 2000