Epsilon Archive for Student Projects

Abstract

Leaf or plant area index (LAI/PAI) is a useful biophysical indicator to characterize the interrelationships between forests and the atmosphere and offers greater potential to estimate productivity of forested landscapes. Recently, hemispherical photography has been used in a pilot study implemented in the Swedish National Forest Inventory (NFI) to estimate LAI. However, using this indirect approach to estimate stand basal area has been less explored in boreal forests of Sweden. This study sought to evaluate the use of LAI in estimating stand basal area for different forest structures (species composition, age, density) and site characteristics using data from the 2016 and 2017 NFI. A 10-year average of absorbed radiations and precipitation for summer months obtained from the Japanese Reanalysis-55 were used to augment a stepwise regression modeling of measured basal area for monocultures of Norway spruce, Scots pine, mixed coniferous and broad-leaved forests. Models with indirect estimates of leaf area were significant (p < 0.001) for all species. The explained variation was higher for models with LAI functions in Norway spruce (77 %) and Scots pine (71 %) compared to mixed coniferous (60 %) and broad-leaved forests (60 %) with general PAI estimates. Other predictors such as absorbed radiation, stand age and density contributed to the explained variations. It is evident that leaf area index could enhance current predictions of stand basal area and increase the sensitivity of these models to climate change. It is also acknowledged that spectral and textural variables from higher resolution satellite imagery and digital elevation models would substantially improve the model estimates of basal area in boreal forest systems.