2016 lienard lynn strigul norris gatziolis mullarney bryan henderson Estuarine, Coastal and Shelf Science accepted manuscript.pdf
Accepted version, 6.630Mb
Publicly accessible from 2018-09-06
Request a copy
Request a copy
Liénard, J., Lynn, K., Strigul, N., Norris, B. K., Gatziolis, D., Mullarney, J. C., … Henderson, S. M. (2016). Efficient three-dimensional reconstruction of aquatic vegetation geometry: Estimating morphological parameters influencing hydrodynamic drag. Estuarine, Coastal and Shelf Science, 178, 77–85. http://doi.org/10.1016/j.ecss.2016.05.011
Permanent Research Commons link: https://hdl.handle.net/10289/10410
Aquatic vegetation can shelter coastlines from energetic waves and tidal currents, sometimes enabling accretion of fine sediments. Simulation of flow and sediment transport within submerged canopies requires quantification of vegetation geometry. However, field surveys used to determine vegetation geometry can be limited by the time required to obtain conventional caliper and ruler measurements. Building on recent progress in photogrammetry and computer vision, we present a method for reconstructing three-dimensional canopy geometry. The method was used to survey a dense canopy of aerial mangrove roots, called pneumatophores, in Vietnam’s Mekong River Delta. Photogrammetric estimation of geometry required 1) taking numerous photographs at low tide from multiple viewpoints around 1 m2 quadrats, 2) computing relative camera locations and orientations by triangulation of key features present in multiple images and reconstructing a dense 3D point cloud, and 3) extracting pneumatophore locations and diameters from the point cloud data. Step 3) was accomplished by a new ‘sector-slice’ algorithm, yielding geometric parameters every 5 mm along a vertical profile. Photogrammetric analysis was compared with manual caliper measurements. In all 5 quadrats considered, agreement was found between manual and photogrammetric estimates of stem number, and of number × mean diameter, which is a key parameter appearing in hydrodynamic models. In two quadrats, pneumatophores were encrusted with numerous barnacles, generating a complex geometry not resolved by hand measurements. In remaining cases, moderate agreement between manual and photogrammetric estimates of stem diameter and solid volume fraction was found. By substantially reducing measurement time in the field while capturing in greater detail the 3D structure, photogrammetry has potential to improve input to hydrodynamic models, particularly for simulations of flow through large-scale, heterogenous canopies.
This is an author’s accepted version of an article published in the journal: Estuarine, Coastal and Shelf Science. © 2016 Elsevier.