Net primary productivity (NPP) is a key ecological parameter that is important in estimating carbon stocks in large forested areas. NPP is estimated using models of which leaf area index (LAI) is a key input. This research computes a variety of ground-based and remote sensing LAI estimation approaches and examines the impact of these estimates on modeled NPP. A relative comparison of ground-based LAI estimates from optical and allometric techniques showed that the integrated LAI-2000 and TRAC method was preferred. Spectral mixture analysis (SMA), accounting for subpixel influences on reflectance, outperformed vegetation indices in LAI prediction from remote sensing. LAI was shown to be the most important variable in modeled NPP in the Kananaskis, Alberta region compared to soil water content (SWC) and climate inputs. The variability in LAI and NPP estimates were not proportional, from which a threshold was suggested where first LAI is limiting than water availability. / xii, 181 leaves : ill. ; 28 cm.

Leaf area index (LAI) provides forestry information that is important for regional scale ecological models and in studies of global change. This research examines the effects of mountainous terrain on the radiometric properties of multispectral CASI imagery in estimating ground-based optical measurements of LAI, obtained using the TRAC and LAI- 2000 systems. Field and image data were acquired summer 1998 in Kananaskis, Alberta, Canada. To account for the influence of terrain a new modified approach using the Li and Strahler Geometric Optical Mutual Shadowing (GOMS) model in 'multiple forward mode' (MFM) was developed. This new methodology was evaluated against four traditional radiometric corrections used in comination with spectral mixture analysis (SMA) and NDVI. The MFM approach provided the best overall predictions of LAI measured with ground-based optical instruments, followed by terrain normalized SMA, SMA without terrain normalization and NDVI. / xiv, 151 leaves : ill. (some col.), map ; 29 cm.