Graduation of Kelvin Jerez Nova

Geometrical Mangrove Models: quantifying frontal surface area distribution for Avicennia marina vegetation: an important parameter for estimating wave attenuation

• Professor of graduation: Dr. ir. B.K. van Wesenbeeck

• Supervisor(s) of graduation: Dr. ir. B. Hofland (TU Delft), Ir. A.G.M. Gijón Mancheño (TU Delft), Ir. S.A. Kalloe (TU Delft) 

For the purpose of estimating wave attenuation, mangrove trees are often schematized as uniform cylinders that in almost all cases only represent the stem of the tree. However, trees have much more complex geometrical features that influence the behaviour of the interaction between waves and tree structures. In some cases, mangrove trees are scanned with terrestrial laser scanners in the field. This is a time-consuming endeavour to obtain the frontal surface area of trees. Numerical models often take into account these frontal surface areas for estimations of wave dissipation trough drag-type forces. This thesis aims to model the complex geometry of the Avicennia marina mangrove species that are currently being grown in the greenhouse of Deltares. The research question becomes: How do we obtain the complex, projected frontal surface area distribution over the vertical coordinate of Avicennia marina vegetation?. The wave attenuation properties of Avicennia m. vegetation are, at a later stage, tested in the delta flume, a wave generating facility at Deltares, Delft. Firstly, measurements were performed to obtain valuable tree structure parameters. At the same time, many observations were made to map the branching patterns of the investigated trees. Both canopy measurements and root measurements were conducted for Avicennia marina trees with varying characteristics: two saplings (1.5 years old) with varying canopy density and two young trees (5 years old) with varying canopy density. Canopy measurements were stored in a so-called tree data structure consisting of nodes (representing branches) that contain information of the individual branches and edges (links that establish direct relations from one branch to another). Thereafter, a search tree algorithm was developed that could loop through all tree data, compute parameters of interest and again store the results in assigned data arrays. Parameters of interest were branch dimensions and diameter ratios of branches (between different branch classes). This information is necessary to create a blueprint for the construction of geometrical tree models. The tree modelling started with the construction a root (pneumatophore) model for Avicennia m. vegetation. Additionally, the projected frontal surface area distribution over the vertical, Av(z), of these roots was computed and displayed in clear graphs as function of the vertical height. two canopy models were constructed. The first model is based on relations between individual branches and is less probabilistic compared to the second canopy model. The second canopy model is based on the normal distribution of branch diameters and, is therefore, more probabilistic by nature. However, the second model displays large variability within branch orders and shows great similarities with the real trees that were observed in the greenhouse. The tree heights of the measured trees were best approached by canopy model 2 (probabilistic method). An algorithm (canopy model 1 and 2) has been scripted to construct the individual branches based on information from branches that were constructed at an earlier stage. The model accounts for proper placement of branches within the three-dimensional coordinate system by rotation- and translation operations. The algorithm outputs the projected frontal surface area of a system of branches for a given direction. This information can then be used to implement in numerical simulations in combination with, for example, bulk drag coefficients to estimate wave dissipation by Avicennia m. vegetation.