Hydrodynamic modelling of a mangrove system in Singapore

Hydrodynamic modelling of a mangrove system in Singapore

Broekema, Y.

Verhagen, H.J. (mentor)
Bayen, S. (mentor)

Civil Engineering and Geosciences

Hydraulic Engineering

2013-10-25

Modelling tidal dynamics in mangrove systems is of great use in studying effects of vegetation on flow and may prove to be a useful tool to support the management of mangrove areas. Since there are lots of limitations on obtaining field data inside mangrove forests, numerical models are indispensable to analyse the effects of changes in mangrove forest parameters on hydrodynamics and vice versa.This research is performed to gain insight in the performance of different vegetation model representations and to apply them in a model of a real mangrove area, Sungei Buloh Wetland Reserve. Vegetation imposes an extra source term of friction in the momentum equation in the form of a drag force. Different model representations to include the extra friction induced by vegetation on the flow are tested in this research. The first approach is the ‘classical’ way of specifying vegetation: the influence of the increased friction is simply taken into account by increasing the bed friction. The second approach of specifying vegetation is by using the directional point model. This approach models vegetation as a set of rigid cylinders in the flow, where the density and diameter of these cylinders may vary over the depth. Both different model representations have first been tested in highly simplified test-cases to get a first feel on the performance of these model representations and to get familiar with using them. Before including vegetation in the model of Sungei Buloh, first some adaptations to the existing model were made. The existing model inhibited some instabilities in the velocity field, that where caused by some errors in the grid and in the bathymetry. Therefore the grid and bathymetry were adapted to overcome these instabilities, with good results. The vegetation is added in this adapted model in the two different ways described above, and both 2D and 3D modeling is tested for these vegetation inclusions. The added mangrove vegetation results in a significant decrease in horizontal depth averaged flow velocity in the northern part of the model domain, whereas the southern part is barely influenced by the vegetation. An explanation for this is found by looking at the order of magnitude of the different terms in the momentum equation. It turns out that in the northern part of Sungei Buloh Wetland Reserve the order of magnitude of the drag force imposed by vegetation is of the same order as the other dominant terms in the momentum equation, while in the southern part of Sungei Buloh Wetland Reserve the order of magnitude of the drag force is smaller than the other dominant terms. It is also observed that vegetation does not only alter the depth averaged velocity, but also changes the vertical structure of the horizontal velocity. When looking at the performance of different vegetation representations, the use of increased bed roughness and the use of the directional point model deliver completely different results. Part of that can be assigned to calibration of the Manning’s coefficient (which is done simultaneously while running the 3D model due to time limitations), but that does not cover the whole explanation. Bed roughness and drag force (induced by vegetation) are taken separately in the momentum equation, and it turns out that when the influence of drag is relatively low increasing the bed roughness delivers completely different predictions of the flow velocity than the directional point model. If the influence of the bed roughness is not negligible in the momentum equation, then increasing the bed roughness will influence the flow strongly, as opposed to the directional point model that only exerts a (relatively small) drag force on the water column due to vegetation. From the results also the difference between 2D depth averaged modelling and 3D modelling becomes clear. It is more accurate to run the model in 3D instead of 2D, because in 3D also the effect of vegetation on the turbulence is taken into account, as opposed to only the extra drag force that is taken into account in 2D depth averaged modelling. Furthermore, 2D depth averaged modelling introduces large errors when considering fully submerged vegetation. Therefore, to accurately model the influence of vegetation in a mangrove system, it is recommended to use a three dimensional model with vegetation specified by the directional point model.

mangrove
vegetation

http://resolver.tudelft.nl/uuid:c334b017-4c4c-4a70-b28c-dceca6e04131

Student theses

student report

(c) 2013 Broekema, Y.