Vegetation Structure, Tree Volume and Biomass Estimation using Terrestrial Laser Scanning Remote Sensing: A Case Study of the Mangrove Forests in the Everglades National Park

Mangrove forests are being threatened by accelerated climate change, sea level rise and coastal projects. Carbon/above ground biomass (AGB) losses due to natural or human intervention can affect global warming. Thus, it is important to monitor AGB fluctuations in mangrove forests similar to those inhabiting the Everglades National Park (ENP). Tree volume and tree wood specific density are two important measurements for the estimation of AGB (mass = volume * density). Wood specific density is acquired in the laboratory by analyzing stem cores acquired in the field. However, tree volume is a challenging task because trees resemble tapered surfaces. The majority of published studies estimate tree volume and biomass using allometric equations, which describe the size, shape, volume or AGB of a given population of trees. However, these equations can be extremely general and might not give a representative value of volume or AGB for a specific tree species. In order to have precise biomass estimations, other methodologies for tree volume estimation are needed. To overcome this problem, we use a state-of-the-art remote sensing tool known as ground-based LiDAR a.k.a Terrestrial Laser Scanner (TLS), which can be used to precisely measure vegetation structure and tree volume from its 3-D point cloud. We surveyed three mangrove communities: (Rhizophora mangle, Laguncuria racemosa and Avicennia germinans) in three different sites along Shark River Slough (SRS), which is the primary source of water to the ENP. Our sites included: small-, intermediate- and tall- size mangroves. Our ground measurements included both: traditional forestry surveys and TLS surveys for tree attributes (tree height and diameter at breast height (DBH)) comparison. These attributes are used as input to allometric equations for the estimation of tree volume and AGB. A total of 25 scans were collected in 2011 with a Leica ScanStation C10 TLS. The 3-D point cloud acquired from the TLS data revealed that trees could be modeled as a combination of tapered geometric surfaces called frustums. The volume of the various geometric frustums can be estimated using specific mathematical formulas. We compared our TLS-derived tree volume/AGB versus published allometric equation-derived tree volume/AGB for our mangrove sites. Preliminary results yielded a good agreement between the TLS-derived and the allometric-derived biomass measurements, with TLS slightly underestimating it. In addition, we compared hand-measured tree attributes vs. TLS-measured tree attributes. Preliminary results showed that DBH can be precisely estimated with TLS. In the other hand, tree height can be estimated in places where there is low vegetation density due to point cloud occlusion and shadowing. We suggest that TLS shows potential to fill the gap between traditional forestry measurements and airborne LiDAR for forestry applications.