Assessing current patterns behind hybrid dams

Abstract

The Demak coastal area faces severe erosion problems. The main cause of the erosion of the coastline, which is at some places as high as 1.5 km of land, is the deforestation of the green belt of mangroves for the purpose of building fishponds. This has disturbed the delicate sediment balance in the area drastically. The BioManCO project aims to develop a bio-morphodynamic model for mangrove mudcoasts like the coast of Demak, which will eventually be used to identify the conditions under which restoration of the sediment balance may lead to autonomous reforestation of a sustainable mangrove green belt and thus the natural coastal protection. A measure to restore this sediment balance that is already being implemented in the area is the construction of semi-permeable dams. It is not clear yet however, why some of these dams work as intended and why some do not. In an earlier study a 2DH model has been set up to better understand and optimise the design of these structures using the large-scale morphodynamics of the area. A large obstacle that the development of this model and the BioManCO model faces however, is the scarcity of data in these areas. The flow patterns around these semi-permeable dams have been identified in the model, but cannot yet be verified.
The aim of this thesis is to visualise flow patterns around and the permeability of a selected dam in the field to be able to get a better understanding of the hydrodynamics of the area. As the coastal area of Demak is quite remote, low-tech drifters will be used for this purpose, made only of locally available, inexpensive materials. They will be released at different locations at different moments in the tidal cycle. Data acquisition will take place by visual observation and by a camera attached to a drone. An algorithm will be developed to process the data digitally. It will automatically detect the drifters in the images taken by the drone. Using several Ground Control Points, of which the GPS-coordinates are known, the images can be rectified, so that the relation between the pixels in every image and their location in the real world is known. Using this information and the pixel-coordinates of the detected drifters, the paths of the drifters can be extracted from the images in real-world coordinates. From these paths, flow velocites can be deduced. The permeability of the dam will be assessed by releasing milk into the flow in several configurations during which data will be acquired by means of a GoPro camera mounted above the semi-permeable dam.
The results show, that tide is the main driving force behind the flow patterns around the dam focused
on in this research. During falling tide, a circulation pattern can be seen, which is caused by a combination of the local bathymetry and a wind driven mass transport over the plot. The flow patterns however do not resemble the patterns found in earlier model runs of the area. The cause of this is
thought to be the combination of simplifications in the model and the permeability of the dams, which is not taken into account by the model, but is shown to be influential by the permeability tests. These
tests show that there is definitely mass transport through the dam considered, but also that this may not be universally applicable to all semi-permeable dams in the area, depended on their location. The drifter paths found by the developed algorithm resemble the visual observations convincingly. The found velocities are difficult to interpret, because of the large scatter on the found drifter paths. The scatter is believed to be a result of errors in the rectification process, which should be improved in the future if this technique is to be used more frequently. Also the design of the drifters and the flight plan of the drone should be revisited to increase the detected amount of drifters in the images. All in all, the technique does show great promise to acquire more data in this remote area, but needs further refinement before it can be used for scientific purposes.