Monsoon wave transmission at bamboo fences protecting mangroves in the lower mekong delta

Highlights

• Porosity of bamboo fences drastically affects attenuation of both high- and low-frequency waves and therefore should be specified for the fence functionality.

• Fence height is relatively more influential than fence width in damping waves.

• High-frequency waves are generally more effectively dissipated by bamboo fences than low-frequency waves.

• Empirical formulation of wave transmission at bamboo fences is derived based on a large field dataset, in which the fence porosity effect is explicitly included.

Abstracts

This paper is concerned with a study on monsoon wave transmission at bamboo fences protecting newly-planted mangroves in the Lower Mekong Delta (LMD) of Vietnam. For this, new field experiments were carried out in combination with the use of an existing field dataset by Albers et al. (2013).

In support of the analysis of the field data, influences of governing parameters on wave transmission are systematically examined using a numerical wave tank (i.e. a RANS-VOF model). It follows that the fence height is relatively more influential than the fence width. Also, high-frequency waves are generally more effectively dissipated by the fences than long-frequency waves. Importantly, the fence porosity is found to affect attenuation of both high- and low-frequency waves drastically, indicating its important role in the quantitative understanding of wave transmission as well as in the fence functional design.

An empirical formulation of wave transmission at bamboo fences, needed for the fence functional design, was eventually derived from all the field data. As hinted at by the numerical analysis, the fence freeboard, the fence width and especially the fence porosity are found the major governing parameters.

Introduction

Re-plantation of mangroves at erosion-prone areas requires supporting structures such as wooden fences to mitigate hydro-morphological disturbances and thus accommodate seedlings of pioneer vegetation species [1]. Porous fences made of local timbers like bamboo and melaleuca (hereinafter generally referred to as bamboo fences) are being used as a sustainable and technically-feasible measure against beach erosion in mangrove-mud coasts, particularly in Thai Land, Indonesia and lower Mekong delta (LMD) - Vietnam. While hard structures (breakwaters, concrete piles, etc.) bring about negative effects, the bamboo fences are considered softer as they are sufficiently porous to allow partial wave passage (less reflective), across-fence mud exchange and thus result in minimal adverse effects [2]. In the sense of wave damping and promoting sediment entrapment, the bamboo fences function in a similar manner to mangrove plants.

Nevertheless, due to high porosity structure the fences are moderate in damping waves. The fences could not sustain high waves and require frequent maintenance and repair. Their use is therefore most suitable in areas of low wave energy such as monsoon waves, particularly aiming at supporting newly-planted mangroves in the early stage of reforestation (see Fig. 1).

Quantitative understanding of wave attenuation by bamboo fences is required for their functional design. However, despite the popular use, there exist only a few studies on this matter. Mai et al. [3] examined wave damping efficiency of brushwood fences on summer-dike forelands along the German North Sea coast. Fences of various heights of timber poles with brushwood bundles in between were therefore tested at prototype scale. Wave transmission was found to mainly relate to the relative crest freeboard R_c/H_s (the ratio of the fence freeboard above the still water level R_c to the incident wave height H_s) in the same manner as by Angremond et al. [4] for permeable structures. The fence porosity and wave period were also reported to affect wave attenuation, however without further quantitative description.

Under the framework of a GIZ-funded (German Agency for International Cooperation - GIZ) coastal zone management project in Vietnam, Albers et al. [5] showed the high effectiveness of bamboo T-groins and bamboo fences in sediment entrapment at several eroding sites in Soc Trang province. The fence (henceforth referred to as GIZ fence) was basically constituted of two parallel rows (at 0.5 m apart) of bamboo poles and fill-in unstructured bamboo/brushwood branches. Wave damping by the fence was monitored with a six-month field survey campaign. The authors reported that wave transmission is mainly characterized by the fence relative freeboard. The effect of fence porosity was not explicitly addressed but it was claimed that fences with flexible bundles were better in damping waves than those with stiff bundles. The effect of bundle stiffness on wave energy dissipation was not evidenced but a different brushwood type apparently results in a different fence porosity and thus affects wave attenuation. The transmission coefficient K_t (ratio of transmitted wave height to incoming wave height, see also Section 2.1) generally reduces as the relative freeboard R_c/H_s increases, from approximately 0.75 when R_c/H_s ≤−1.0 (fence crest is below water) to about 0.20 (for flexible bundles or bundles of small-size branches, hence small fence porosity) and 0.60 (for stiff bundles or bundles of large-size branches, hence large fence porosity) when R_c/H_s ≥ 1.0 (fence crest is above water). At laboratory scales, Albers and Von Lieberman [6] also asserted the important dependency of the bundle porosity.

In another GIZ's attempt to assist mangrove restoration in LWD, Cuong et al. [7] introduced two variants of melaleuca fences at a low wave energy coast in Kien Giang province. These were modified versions of the bamboo fences with melaleuca poles and branches in combination with fish nets and bamboo mats. The melaleuca fences were relatively narrower and more porous than one by Albers et al. [5]. Field monitoring showed that the fences were effective in promoting sedimentation behind the fence line. Unfortunately, no quantitative wave measurements were reported, though a general wave height reduction rate of 63% was claimed.

There exist several numerical studies on wave attenuation over fences in the literature. These all follow the modelling approach of wave transmission through a vegetation field (i.e. mangroves or salt marshes), in which plants are schematized as tiers of vertical cylinders of various densities and diameters so that wave energy dissipation can be described through the work done by drag forces on cylinders [8,9]. Halide et al. [10] used a numerical model to assist the design of wave fences of vertical bamboo poles with several alternatives of pole densities and diameters. The model outputs showed that fences of this type generally demonstrate a weak wave damping capacity and thus very large fence width (dozens of metres) is needed to yield a significant wave reduction level. In preparation of their physical experiments, Dao et al. [11] numerically examined wave damping efficiency of GIZ-type fences at a reduced model scale using SWASH [12]. Albeit the bamboo fences are largely composed out of horizontal elements (bundles) they were represented in the model as a rigid vegetation field, whose density and diameter are derived from the cross-sectional fence porosity. The simulation results showed that transmitted waves reduce with the increase of wave nonlinearity represented by the Ursell number. Scattering in the GIZ wave transmission data by Schmitt et al. [13] was explained through the effect of wave nonlinearity. However, the variation magnitude of the transmission coefficient with wave nonlinearity was rather insignificant (± 0.10) and moreover no model validation against the measured data was possible. Therefore, the dependency of wave non-linearity appears to be inconclusive.

The paper is organized as follows. New field experiments and the use of the new dataset in combination with the existing one are described in Section 2. Section 3 addresses a numerical analysis using a numerical wave tank to increase understanding of influences of the fence parameters on wave attenuation. In Section 4 regression analysis of all the field data, with support from the numerical analysis, is performed to formulate wave transmission at bamboo fences. Discussion on the effect of wave reflection and hints for optimizing the fence design are given in Section 5. Summary and concluding remarks are finally drawn out in Section 6.