Silvilaser 2019 - Poster Presentations »
Adjustment of leaf area conversion functions in managed conifer stands in Sweden
The structure of contemporary managed forests are complex and deviate from experimental forests which are usually even-aged monocultures and single-storied. To apply growth functions developed from experimental forests on managed forests, adjustments are required especially for leaf area index (LAI) which is a key biophysical variable in many process-based growth models. To asses this, the effect of heterogeneity on the performance of canopy LAI in modelling the basal area (BA) of managed boreal forests dominated by Norway spruce (Picea abies (L). Karst) and Scots pine (Pinus sylvestris L.) in Sweden was investigated. The study was based on the assumption that crown leaf area and stem diameter are strongly related and are vital for estimating stand productivity. Managed forests were represented by field data from the Swedish National Forest Inventory (NFI). Species-specific LAI conversion parameters (developed from experimental stands) were applied on general plant area index (PAI) values from hemispheric fish-eye photos which were taken from the sample plots during the 2016 and 2017 NFI campaigns. The heterogeneity was studied in two parts by (a) ground-based stand structural heterogeneity (SSH) described by species composition, coefficient of tree diameter variation, tree social status and height-diameter ratio and (b) spectral heterogeneity (SPH) by vegetation and textural indices developed from Sentinel-2. Species-specific final (with heterogeneity metrics) and base (without heterogeneity metrics) models were fitted for BA-LAI and BA-PAI relationships by nonlinear least squares and generalised additive regression, respectively. For both species, a significant positive nonlinear relationship was found between BA-LAI whereas, the fit between BA-PAI demonstrated a positive linear trend. Generally, BA-LAI final models (FMs) accounting for heterogeneity resulted in larger explained variance (RMSE, m2 ha-1) than the base models (BMs). Relative to the BMs, FMs with SSH reduced the variance by 55% in Norway spruce (RMSE = 3.330, relative-RMSE = 15.398%) and 43% in Scots pine (RMSE = 3.701, relative-RMSE = 17.377%). The fit between BA-LAI with SPH was also significant and showed an improvement in the RMSEs over the BMs for Norway spruce (5.561) and Scots pine (5.659), suggesting the potential use of Sentinel-2 in future growth models. The conclusion from this study indicates that in growth models, when extrapolating functions developed from experimental forests to managed forests, there is a need to take into account the stand structural heterogeneity. This is important for effective growth and yield modelling of managed stands of Norway spruce and Scots pine.