The impact of spatial resolution on riparian leaf area index modelling using remote sensing

Abstract

This thesis investigated the impact of differing sensor spatial resolutions on leaf area index (LAI) modelling. Airborne images along with ground measurements of LAI were acquired for riparian areas along the Oldman River in southern Alberta. Airborne images were spatially resampled to spatial resolutions between 18 cm and 500 m, and the Modified Simple Ratio (MSR) was calculated from the imagery. LAI regression models were created at each spatial resolution, and changes in the relationship between MSR and LAI were observed at each spatial resolution, as well as changes in the modelled LAI estimates. The relationship between MSR and LAI was scale invariant at spatial resolutions as low as 10 m, and only moderately changed until 30 m. MSR and predicted LAI gradually reduced as resolution coarsened further, with large changes occurred beyond 100 m. No relationship was evident between MSR and LAI at spatial resolutions coarser than 300 m.