On improving the accuracy of digital hemispherical photography measurements of seasonal leaf area index variation in deciduous broadleaf forests

Optical methods have been widely used to estimate seasonal changes of the leaf area index (LAI) in forest stands because they are convenient and effective; however, their accuracy in deciduous broadleaf forests has rarely been evaluated. We estimate the seasonal changes in the LAI by combining periodic observations of leaf area variation with litter collection (LAIdir) in deciduous broadleaf forests and use these estimates to evaluate the accuracy of optical LAI measurements made using digital hemispherical photography (DHP). We also propose a method to correct DHP-derived LAI (LAIDHP) values for seasonal changes in major factors that influence the determination of LAI, including the woody to total area ratio (α), the element clumping index (ΩE, using three different methods), and the photographic exposure setting (E). Before these corrections were made, LAIDHP underestimates LAIdir by 14%–55% from 21 May to 1 October but overestimates it by 78% on 12 May and by 226% on 11 October. Although pronounced differences are observed between LAIdir and LAIDHP, they are significantly correlated (R2 = 0.85, RMSE = 0.32, P < 0.001). After considering seasonal changes in α, ΩE, and E, the accuracy of LAIDHP improves markedly, with a mean difference between the corrected LAIDHP and LAIdir of <17% in all periods. The results suggest that the proposed scheme for correcting LAIDHP is useful and effective for estimating seasonal LAI variation in deciduous broadleaf forests.