You are here
Harvard Forest REU Symposium Abstract 2011
- Title: Is Digital Cover Photography a Viable Method for Calculating Leaf Area Index For Phenological Research in Closed Forest Canopies?
- Author: Elizabeth S Felts (Harvard College (Harvard University))
- Abstract:
The use of digital hemispherical (fisheye) photography (DHP) as a method for calculating Leaf Area Index (LAI) based on gap-fraction has been established through past studies; however this method has relied on very specific light conditions and the 180° viewing angle degrades image resolution. A promising alternative is digital cover photography (DCP). The method has been developed and tested in various open ecosystems such as Eucalyptus forests and oak savannas where enough light can penetrate the canopy. This research seeks to explore the viability of DCP as a method of obtaining LAI and thus tracking phenological changes in closed forest canopies where light limitation might pose a methodological constraint. To test DCP, weekly imaging of 33 long-term incremental biomass plots located at the Harvard Forest in Petersham, Massachusetts, a deciduous-dominated temperate forest, using a digital single-lens reflex camera (Pentax K100D), was undertaken. To examine the role of scene illumination, the images were acquired in RAW format to allow maximum control over image exposure in the post-processing. LAI values were calculated after image binarization using the recently introduced two-corner method. These estimates were then compared to LAI estimates obtained from gap-fraction measurements made with the LAI-2000 instrument at the same plots, recomputed using only the first 7° ring, to better align this method with the field-of-view given by DCP. The data was further compared against canopy greenness obtained from a fixed webcam at Harvard Forest. Our results demonstrate that DCP is a viable method to track spring phenological changes at Harvard Forest, especially in comparison to canopy greenness. Furthermore, our results demonstrate the need to account for the changing contributions of woody canopy elements to light interception. We show that a normalized color index might be useful for separating green image elements (i.e., foliage) from non-green image elements (i.e., wood, sky).
- Research Category: Biodiversity Studies
