Estimation of forest stand parameters using digital orthorectified aerial photographs
Ground based forest inventory surveys can provide highly accurate measurements of tree and stand characteristics, but are time-consuming and costly, and therefore typically limited to number of sample plots. Estimating tree and stand characteristics from digitised aerial photographs can provide measurements from the whole stand, but is less accurate. The goal of this study was to evaluate the application of scanned, stereo aerial photography and digital photogrammetry in combination with tree crown delineation techniques to measure tree and stand characteristics in two sites in Scotland, Rosarie and Leanachan forests. Existing medium-resolution scanned images of true colour aerial photographs (1:10,000) were used to derive Digital Elevation Models (DEMs) of the forest canopy and digital orthophotographs. Tree crown delineation techniques were used on the derived digital orthophotographs and tree crown measurements including crown area and coordinates of each crown were derived. The DEMs in combination with Digital Terrain Models (DTMs) derived from digital contour maps, were used for the estimation of tree and stand heights. Equations derived from regression analysis of individual tree measurements on the ground, and the orthophotographs from Rosarie forest, were used for the estimation of tree and stand characteristics of both sites. For Rosarie forest the estimations of stand top height, basal area, stand volume stand biomass and stand density (~23.7%) were similar with the ground measured stand characteristics (±10%), where as for Leanachan forest the estimations were less accurate due to the non-optimum illumination conditions during the acquisition of the aerial photographs. The level of accuracy achieved in this study is adequate for measuring tree and stand characteristics, if the acquisition conditions of aerial photographs are optimal. Higher level of accuracy may be possible, but requires more accurate DTMs, possibly derived using active airborne remote sensing sensors.