Fruit of Swedish fieldwork: canopy-height relationships
Way back in March 2012 I went to Abisko, Sweden on behalf of Tim Reid to collect data to help link forest canopy density with remotely measured tree height. As I’m learning, the birthing process for academic work is often a long one! That work has now found its place in the larger picture and been published in the Agricultural and Forest Meteorology journal under the title
You can read about my fieldwork here and here, which include some great snowy photos! The paper covers much more than those data I collected in 2012, it also focuses on canopy data collected during a field campaign led by my supervisor, Richard Essery, in 2011. The paper abstract is as follows:
Leafless deciduous canopies in boreal regions affect the energy available for snowmelt and reduce overall surface albedo during winter, thereby exerting a strong influence on weather and climate. In this work, ground-based measurements of leafless canopy structure, including hemispherical photography, terrestrial laser scanning (TLS) and manual tree surveys were collected at 38 sites in an area of mountain birch forest in northern Sweden in March 2011 and 2012. Photo-derived sky view fraction was strongly inversely correlated (r < −0.9) to the total tree basal area in a 5 m radius around the photo site. To expand findings to wider areas, maps of canopy height for a 5 km × 3 km area were obtained from airborne lidar (ALS) data collected during summer 2005. Canopy heights derived from TLS were used to validate the ALS estimates, and simple models were developed to establish relationships between hemispherical sky view and ALS canopy height (RMSE < 5%). The models and ALS data provide useful methods for estimating canopy radiative transfer and biomass over wide areas of birch forest, despite the relatively low ALS resolution (∼1 return m−2).
If you’re not used to scientific writing this might seem a little impenetrable. The bottom line is that a good relationship was found between forest canopy density and tree trunk area, and between canopy height (as measured by an aeroplane mounted LiDAR) and forest density. The latter, in particular, is useful as it could allow some forest parameters to be categorised remotely. This is really common sense – bigger trees have denser canopies. What’s useful is quantifying that relationship.
The paper is available from ScienceDirect at this DOI: http://dx.doi.org/10.1016/j.agrformet.2013.12.005. Unfortunately this is behind a pay-wall, but as I understand it I am allowed to send copies of the paper as personal communication (correct me if I’m wrong!), so please get in touch if you need to.