An assessment of ground reference methods for estimating LAI of boreal forests

The amount of green leaf area, quantified by the leaf area index (LAI), is one of the key factors determining ecosystem net primary production and energy exchange between land surfaces and the atmosphere. LAI can be measured indirectly using optical methods that are based on the tight relationship between LAI and canopy light transmittance, or through allometric regression models. Until now there has been very little research to compare LAI estimated by the two different approaches. In this study we compare optically-based estimates of LAI to estimates produced using local foliage biomass models for three boreal tree species. Our study is based on 661 Scots pine, Norway spruce and Silver birch stands in the southern boreal zone in Finland. Routine stand inventory and LAI-2000 Plant Canopy Analyzer measurements were performed in all stands. We used three allometric foliage mass models to calculate foliage masses for each stand. The foliage mass estimates were then converted to LAI using specific leaf area (SLA) values. A theoretical stem diameter distribution was also used together with the only diameter-dependent foliage mass models. Optical LAI was corrected for shoot-level clumping (i.e., the clumping of needles into shoots). Finally, we estimated LAI from measured canopy gap fractions by inverting a canopy radiation model. Allometric foliage mass models produced significantly different foliage mass estimates (and the LAI estimates based on them). Among the allometric methods, however, the most sophisticated in terms of input tree variables agreed fairly well with each other and also with the clumping-corrected optical methods. The results indicate a need for caution when using foliage biomass or LAI estimates calculated using different models, especially if the estimates are to be used as an input for other models. The simple shoot-level clumping correction of optical LAI showed the best fit with allometric LAI. Our results support the use of a simple shoot-level clumping correction instead of more complicated correction methods in optical LAI estimation in coniferous forests.