The need for soft sand engines

To better cope with the rising sea levels and maintain fresh water security during periods of droughts, the committee Veerman suggested to gradually increase the water table of Lake IJsselmeer with up to 1.5 meter (Deltacommissie 2008). This will not only have severe consequences for important functions in the region, like nature areas and recreation, but puts additional pressure on existing engineering structures designed to maintain water safety. In 2009 the Building with Nature programme was asked to initiate a pilot study along the Frisian coast of the IJsselmeer to renew the landscape, improve existing – and develop new – nature areas, providing new opportunity for recreation whilst at the same time improving water safety. With these objectives in mind, the idea of creating soft sand engines was born. By creating sand banks at strategic locations in combination with pole screens sedimentation of foreshores can be stimulated. Of the three pilot projects planned, the sand engine in front of the Workumerwaard is the only one that has been constructed, is fully functional and is being monitored and evaluated. For an overview of the other pilot projects, see the Sand Engines - Friese IJsselmeer coast.

The Workumerwaard

The Workumerwaard is a nature reserve in front of the dike near the village of Workum. It consists of former tidal flats and salt marshes that have been vegetated after the closure of the Zuiderzee by the Afsluitdijk in 1932 and the subsequent desalination of the lake. On maps from 1850, the area is clearly recognizable as a tidal flat, with a salt marsh in the most landward area. Presently, the area consists of agricultural land that extends about 1.5 m from the dike. This land is bounded by a low levee, which separates it from a wetland nature reserve area of about 300 m wide. The shore consists of a pronounced sandy beach ridge of up to 1.2 meter height and up to 30 meters wide. Such a beach ridge is typical of an erosive coast (Reading and Collinson, 1996). At the lake side of this beach ridge there is a shallow submerged platform of about 800 m wide, on which a series of shore-parallel sandbars have developed. The depth of this area is up to NAP -1.2 m (NAP = chart datum). At the offshore edge of the platform, a steep slope leads to the bottom of the lake, at about NAP - 3.5 m. (Wiersma et al., 2016).

The important role of foreshores

Just a dike with the only function to be a barrier to high water might not be the optimal solution in terms of space usage and costs. By integrating natural processes and ecosystems in a flood defense scheme, a multitude of goals can be served (Tromp and Wesenbeeck, 2011). Shallow foreshores can play an important role in this approach:

• Gradually sloping foreshores can be a key factor in coastal defense. The shallow foreshores and the vegetation on top of it dissipate wave energy and enable the dikes behind them to withstand bigger storms (Tromp and Wesenbeeck, 2011).
• Foreshores form an important habitat for many (pioneering) species. Increasing the natural value will improve the quality of the landscape, may further reduce the hydrodynamic forcing on the dikes, add capacity for water purification and help stabilizing sandy foreshores (Tromp and Wesenbeeck, 2011).
• Foreshores can fulfill important recreational functions (Van Slobbe et al., 2012).
• As compared with the traditional approach of hard coastal defenses, an integrated hard/soft coastal defense system such as a dike with an elevated foreshore will increase the capacity to adapt to environmental changes like climate change and lake level rise. Thus, these alternatives enhance functional resilience (Gersonius, 2012).
  
  

Soft sand engines

Water safety, nature quality and recreational opportunities can all be served by creating or improving shallow foreshores. Constructing soft sand engines is a new and innovative Building with Nature approach to maintain or extend such foreshores. Rather than direct engineering of the foreshores in front of the dikes of lake IJsselmeer, the basic idea is to let nature take its course. If a ridge of sand is deposited on the edge of the shore face platform, waves and currents will slowly transport the sand onshore, where it can settle. In order to stimulate this settling at the right place, strategically placed pole screens can be used. Thus the coast is enabled to build out in a natural way and adapt itself to the prevailing hydrodynamic conditions, even if these are changing gradually (It Fryske Gea 2011, Van Slobbe et al. 2012).

Pre-feasibility

During the pre-feasibility study several workshops and meetings were held with different (governmental) parties involved, in order to affirm interests, to gain a clear picture of the experiments and to reach a statement of intent regarding the lake area. In this phase, negotiations were facilitated by a movie with decision makers expressing the importance of a building with nature approach for Fryslan. This movie showed support and provided informal legitimization from high-ranking authorities for the experiments. It was no longer questioned whether the project would be carried out, discussions rather shifted to how it would optimally succeed.
 

Feasibility

In the feasibility phase two studies were conducted: pre-conditions of eco dynamic design (Building with Nature, 2011) and an ecological assessment (Deodatus & de Jong, 2010).

A soft sand engine in the Workumerwaard

The Workumerwaard nourishment started in June 2011 and was completed in October of that year. The overarching goal of the project was to revitalize nature by turning coastal erosion into gradual accretion. At first sight, the process seems fairly simple: deposit an amount of sand on the shore face and let waves and currents bring it onshore. Yet, the actual creation of a sand engine in front the Workumerwaard turned out to be a complex process. It Fryske Gea (2011) has reported many of the steps taken in detail. The present page describes the most important aspects, from location selection to monitoring and evaluation.

Location selection

The pilot Workumerwaard consists of a concentrated shore face nourishment in the form of an elongated sand ridge combined with a pole screen perpendicular to the coast meant to trap the sediment at the right place. If the ridge is properly located, waves and currents will be able to gradually transport sediment onshore. Since Workumerwaard coastal system already includes a sandy shore and a wide floodplain with valuable wetland vegetation in front of a dike, this pilot aims at enhancing the robustness of nature, and thereby indirectly improving flood safety. Location is key in making the sand engine function as intended. Some important characteristics that need to be taken into account are:

• Sediment dynamics.
• Nature values and hydrological impact on water vegetation.
• Use of the area and municipal borders.
  
  

Sediment dynamics

Even after the Afsluitdijk dam was built, the lake bed has remained dynamic (see the knowledge page on sediment and ecology). Wind-induced waves and currents are strong enough to move sediment around. Due to differences in local characteristics like depth and average wind speeds, large variations in sediment transport capacity occur along the Frisian IJsselmeer coast. Several hydrodynamic and sediment transport models have been used in order to identify the best location for an effective sand engine.

In wave-induced sediment transport modelling a distinction is made between long shore and cross-shore current and transport components. The CROSMOR model has been used to describe them in a series of cross-shore transects. Below we describe how the model functions and what results have been accomplished. The following paragraph gives an overview of the sand dynamics model based on a comprehensive overview by Folmer et al. (2010).

The CROSMOR model is a "wave-by-wave" model where the wave equation is solved for each wave separate. In this model the following parameters have been varied:

• Cross shore profile
• Sediment size
• Water level
• Wave conditions
   

To gain a good overview of the pilot area, the model has been used in eleven cross-shore transects evenly distributed along the coast. With the mean grain size set at 200 μm and water levels between NAP -0,4 m en NAP +0,4 m, long shore and cross-shore sediment transport rates have been modeled. Transport rates for different water levels have been multiplied by the probability of occurrence of these water levels (Figure 1), in order to gain insight into where the sediment will go. Figure 1 shows the probability density function at measuring station Lemmer, based on 1988-2009 MWTL data (Monitoring Waterstaatkundige Toestand des Lands, a National Surface Water Monitoring Program) (Folmer et al. 2010).
 

Fig 1: Water levels station Lemmer

The yearly sediment transport in the area is the integral of this product over all water levels. Water levels are not exclusively correlated to wind and waves, because the water level regime differs between winter (relatively low, whereas wind speeds are generally higher) and summer (higher water levels, less wind). The computed sediment transport is shown in the picture in one of the eleven transects (top), for different water levels (middle) and integrated over the year (bottom). At higher water levels, sediment transport increases towards shallower water. At the same time, shore ward transport increases. As conditions favoring onshore transport are not very frequent, offshore transport at water depths of NAP -2 m to -1 m is the most frequent (Folmer et al. 2010).

Fig 2: Sediment transport in a transect

This has been repeated for every selected cross section in the pilot area. The findings are visualized in Figure 3, for the Frisian IJsselmeer coast. The arrows indicate the sediment transport direction, the thickness of the arrow the transport capacity. The biggest arrow represents a transport of about 1000 m3 pure sediment per year.

Fig. 3: Longshore sediment transport

The CROSMOR results have been used to select a location for the soft sand engine somewhere between the harbour of Workum and Gaast. For this stretch of coast a detailed design of the pilot has been made. Within this area waves and wind conditions, as well as water depths, vary greatly. On the southern part of the Workumerwaard coast a low breakwater and a number of shoals determines to a large extent the wave dynamics. Further north wave activity increases and reaches a maximum in the northernmost part of the area. Generally speaking: the greater the wave energy, the larger the sediment dynamics. On the other hand, water depths in the north are relatively small, also further offshore. These shallow parts reduce the wave energy, hence the sediment transport capacity. Yet, from a sediment transport perspective, the optimal location for the nourishment is as far north as possible in not too shallow water. Since the sediment transport alongshore is much larger than cross-shore, the sand needs to be deposited south of the location where it is meant to feed the coast and where the pole screen is placed to trap it.
Figure 4 shows that the breakwater and the shoals in the south are reducing local sediment dynamics. In the northernmost part of the area the lake is too shallow, which also limits sediment transport. The red line indicates the most suitable location for the pole screen (Folmer et al. 2010).

Fig. 4: Dam and islands effect

Apart from the location, the height of the sand ridge to be deposited and the length of the pole screen are very important. Again the CROSMOR model has been used, this time to investigate the relation between depth and sediment transport. Model runs have shown sediment transport to be biggest at depths between 1 and 2 m. Wave energy – even during storms – is too low to activate sediment transport in deeper parts of the lake. At depths less than 1 m the waves will break on the lake side of the sandbank, with offshore sediment transport as a possible consequence. Figure 5 shows the optimal location for a sand ridge creating onshore transport (Folmer et al. 2010). Based on CROSMOR model runs a location has been selected where – from a sediment perspective – a sand ridge of 25,000 m³ could best be built (Figure 5).

Fig. 5: Sand bank location

Nature values and hydrological impact on aquatic vegetation

The CROSMOR model only takes sediment dynamics into consideration. The sand engine, however, will also change the other dynamics of the region and can have far-reaching consequences for existing aquatic vegetation. Impacts on nature need to be investigated, especially because the area is part of the European ecological network Natura2000. Based on aquatic plants monitoring program of Rijkswaterstaat, the potential impacts on submerged vegetation have been examined. The next paragraphs give an overview of the impacts on two types of submerged aquatic vegetation: Charales species and Stuckenia pectinata (It Fryske Gea 2011 and Folmer et al. 2010).

Charales species and Stuckenia pectinata

Alongside the Frisian IJsselmeer different types of pondweeds – or more specifically Charales – can be found: Chara aspera, C. connivens, C. contraria, C. globularis, C. hispida, C. virgata and C. vulgaris. Data provided by the Ministry of Infrastructure and the Environment do not distinguish between the different species of Charales and do not cover the area south of the central part of the Workumerwaard coast. Figure 6 shows the dispersal of Charales species in 2008: they are exclusively found in the most shallow parts near the shoreline. Depositing sand between NAP -2 and -1 m will therefore not have an impact on the Charales population. The largest patches of Stuckenia pectinata can be found at the northern part of the study area (Figure 6). From a sediment transport perspective this is not the best location for a soft sand engine.

Fig. 6: Pondweed and S. pectinata

The effects of the sand engine on submerged aquatic vegetation

As shown, a sand engine at the location and with the dimension described above will have a limited impact on the two most common species of submerged aquatic vegetation. Impacts of induced sediment deposition further onshore are expected to be minor, as well, since this deposition process will be gradual. The sediment-trapping pole screen will be located in a zone where both types of aquatic vegetation are present. The direct impact is expected to be limited if the poles are placed carefully. Besides, as both species are annual, populations are expected to recover quickly from any temporary disturbance.

Land use in the Workumerwaard

Land use in and around the Workumerwaard is diverse (figure 7). The westernmost part of the area is only protected by a small natural sand ridge. The open and wet nature here is characterized by shell banks and sand sheets and forms a natural transition from the waters of the IJsselmeer towards the mainland. From a hydrodynamic perspective this is the most suitable location for the sediment-catching pole screen. Further landward the area is used for low-intensity agriculture and plays an important role for resting and nesting godwits. Figure 7 shows the land use in and around the Workumerwaard in the year 2003.

Recreational area It Soal, located south of the Workumerwaard, has sandy beaches with surfing opportunities. If the pole screen would be constructed south of this area, the sand engine should be placed even further south. This is not an attractive option, as the resulting sedimentation is likely to interfere with current land and water uses, in particular surfing activities.

Fig. 7: Land use

Conclusion location selection

In conclusion, the northern part of the Workumerwaard is the most suitable for the pole screen, with the nourishment to the southwest. Impact on aquatic vegetation is expected to be minor and temporary and there is no interference with other forms of land use. Figure 8 shows to best location for the nourishment (yellow lines) and the pole screen (the thick red line marks the most suitable location, the thin red lines are other options).

Fig. 8: Pole screen

Design of the soft sand engine

The soft sand engine consists of two parts, a concentrated shore face nourishment in the form of an elongated sand ridge and a sediment-trapping pole screen perpendicular to the coast.

Design of the sand bank

Once the location had been chosen, the final design of the soft sand engine was made. The budget available allowed for some 25.000 m³ of sand, which would be obtained from dredging projects near the recreational area of It Soal, south of the Workumerwaard. As this is a fairly small volume, the ambitious goals of the project of the three sand engines had to be adapted (It Fryske Gea 2011). Creating a sand engine in front of the Workumerwaard aims:

• To give insight into natural processes (wind and waves) as a mechanism to transport nourished sand onshore in a low-dynamic (as compared with an open coast) wind-driven environment.
• To gradually supply the coast of the Workumerwaard with additional sand, stimulated by a sediment-trapping pole screen. The resulting morphological changes should be such, that the existing vegetation is able to survive.
• To monitor the effects of this soft sand engine and learn from this experiment for other foreshore management projects in shallow delta lakes; as storms trigger sediment dynamics, monitoring should last long enough to have had enough storms to yield noticeable morphological changes (erosion and sedimentation).

An impression of the soft sand engine near the Workumerwaard is shown in the Figure 9 (It Fryske Gea, 2011). This figure shows the location of the soft sand engine pilot; arrows show the direction and magnitude of the sediment transport, the grey oval is the nourishment the black line represents the pole screen.
 

Fig. 9: Soft sand engine location

The characteristics of the nourishment are given in Table 1.

Table 1 – Characteristics of the soft sand engine in front of the Workumerwaard

 
A cross-section of the nourishment (figure 10) shows that its crest will be at about 0.3 m below the target mean water level during winter time. This keeps the nourishment from surfacing under average conditions. Offshore winds, however, can force the water level to drop by 0.5 m, which would result in emergence of the crest. The biggest morphological changes are expected during south-westerly and north-westerly storms. Under these conditions, water is set up against the Frisian coast and may rise with more than 0.5 m and wave action is expected to induce extensive sediment dynamics, mainly onshore directed.

Fig. 10: Sand engine dimensions

Design of the pole screen

The goal of the pole screen is to slow down long shore currents and stimulate sediment deposition in its vicinity. The degree of flow reduction must be determined carefully, as too much reduction may keep sediment away from the screen. In Bangladesh research on pole screens has shown that flow blockage should be between 30 and 70% (Government of Bangladesh, Ministry of Water Resources, 2001). In the Workumerwaard cage a blockage of about 50% is expected to give the best results.

Design criteria for the pole screen were therefore (It Fryske Gea, 2011):

• The pole screen should consist of untreated wooden poles (C14 quality) placed to reach a blockage of about 50%.
• The poles shall be vertically placed in a line perpendicular to the shoreline of about 500 m length.
• The mean winter water level is NAP -0.40 m.
• No erosion shall take place along the pole screen.
• The screen consists of two parts, each with a different pole dimension; in the first zone the lake bed is between NAP - 0.6 m and - 0.9 m, in the second zone between NAP -0.9 m and - 1.5 m.

Fig. 11: Location pole screen

Fig. 12 Cross section pole screen location

Figure 11 shows the exact location of the pole screen, Figure 12 a cross-section of the area where the pole screen is constructed. The light blue box indicates the zone between NAP -0.6 and 0.9 m, the black box the zone between NAP -0.9 m and -1.5 m. In Zone 1, between 150 and 550 m from the shoreline, the poles are submerged about 40 cm (below the winter water level). In Zone 2, between 550 and 650 m from the shoreline, the poles have a larger diameter (120 mm) and they are pushed deeper into the bed, in order to withstand the larger wave forces. The pole screen does not extend all the way to the shoreline, but only up to 0.5 m water depth. Trapping sediment at lower depths will have only a limited effect. A summary is provided in table 2.

Table 2 – Dimensions of the pole screen

First the pole row was constructed. This needed to be done quickly, as the breeding season was about to start. Furthermore, a 4 km long fibre-optical cable was placed in a zigzag pattern at the location where the sand would be deposited. This fibre-optical cable grid is an important innovative means to monitor the actual sand movement and dynamics in the pilot area (more on its functioning below). Subsequently 25.000 m³ of sand was put in place, which turned out to be a challenge. Close to the Workumerwaard the lake IJsselmeer is too shallow for big ships. The sand was transferred to smaller barges and subsequently pumped to the right location. Rainbowing was not allowed, as this would cause too much turbidity.

1.Monitoring and evaluation

The construction of a soft sand engine on the coast of the Workumerwaard is an important pilot project, aiming at several objectives (It Fryske Gea, 2011):

• To increase coastal safety by reducing wave attack on the dike and reducing or even stopping coastal erosion.
• Adaptation of the vegetation to morphological changes on the shore face and onshore and to the clearer and calmer waters in the lee of the nourishment.
• Insight into the wave-attenuating effect of the sand nourishment in relation to the coast during storms.
• Insight into the pathways of the nourished sand, into deeper water as well as towards the shore.
• Insight into and predictive capability of morphological developments in response to a sand nourishment offshore of a vegetated foreshore; one question is whether the nourished sand will segregate (according to grain size) when being transported, as this may influence vegetation development.

Several techniques have been used to monitor morphological changes and sediment dynamics in the area: LiDAR land surface elevation modelling, bathymetric surveys and fibre-optical measurements.

Fig. 1: Soft sand engine

2.LiDAR height model

LiDAR is short for Light Detection and Ranging, an optical remote sensing technology that can measure the distance to – and therefore the height of – an object by illuminating it with light and measuring the reflection, often using pulses from a laser. To monitor morphological changes to the coastal zone – in this case including the effects of the sand nourishment – every year a height map is generated using LiDAR. A helicopter flies over the area generating a map using the laser pulses reflected by objects below (figure 2). As this technique is unable to penetrate a water body (laser reflects on the water surface), morphological changes of the lake bed cannot be monitored.

Fig. 2: Helicopter LiDar use
 

In the context of the soft sand engine pilot, the first LiDAR maps have been generated in February of 2011 and 2012, and the water board Wetterskip Fryslân has provided similar images from February 2008. Figure 3 shows the LiDAR height images of three different years. Note that, because the sand engine is constructed below the water surface, it is not visible on these images. Legend; height in m.

Fig. 3: LiDar height images
 

More interesting than height images at different points in time are the differences between them. Figure 4 shows the LiDAR images of the height differences between the years 2008 and 2011 (left) and 2011 and 2012 (right). Legend; height differences in m.

Fig. 4: Heights 2008 and 2011
 

The height differences are most likely caused by the transport of sediment. The difference in land surface elevation between 2008 and 2011 is very limited. Some areas are slightly decreasing, some areas, particularly in the south are slightly rising. The limited sand dynamics in this time period suggest that the system has been relatively stable over the years. However, the dynamics are much more apparent in the period between 2011 and 2012. Results of the LiDAR map (figure 5; right) show coastal erosion, reflected by a retreat of the beach ridge. This retreat of 10 up to 25 meters was probably the result of one storm event in the first week of 2012. So far, no effects of the nourishment can be derived from these images.

Fig. 5: Coastal erosion 2011/2012

3.Bathymetric survey

To get a complete picture of the morphological changes in the area, it is important not only to look at changes above the water level. The bathymetry needs to be monitored as well. In the case of the soft sand engine Workumerwaard a jet ski with echo sounder has been used to measure the lake bed elevation. The interval from the emission of a pulse to reception of its echo is recorded, and the depth calculated from the known speed of sound through water. Where waters were too shallow for the usage of a jet sky a DGPS device has been used that could be used when walking. All bathymetric measurements have been conducted by a company – Shore Monitoring – specialized in shallow water surveys. Several surveys have been made. T₀ is the lake bed before construction of the soft sand engine, acquired in April 2011. The measurements following the creation of the sand bank – T­₁, T₂, … T_x – cover a larger area in order to cover all changes (figure 6).

Fig. 6: Echo sounding monitoring
 

The results of the measurements are shown in figure 7. In the T1 bathymetry grid, acquired immediately after the placement in October 2011, the sand nourishment is clearly visible as an elongated shallow structure. The top of the nourishment lies around NAP 0 m. The difference between the T1 and T0 measurements also clearly shows the location where the sand was placed. Around this elongated structure, scattered anomalies show in general shoaling of the shallowest parts, and deepening to the north of the nourishments. The anomalies are however minimal and may be the result of measuring artefacts.

Fig. 7: Lake bed visualization
 

After the first autumn and winter storm season, the T2 bathymetry was acquired in April 2012. The resulting grid (figure 8) shows that the nourishment has lost height and become less pronounced. The difference between the T2 and T1 data (figure 8) shows that immediately coastward of the nourishment an increase in elevation is present. This suggests a landward transport of sediment. Furthermore, the largest changes in elevation are along the zone with the most pronounced sandbars at the lake side of the shallow water area.

Fig. 8: Lake bed morphology
 

To further analyse the observed changes in bathymetry, elevation profiles were drawn across five sections of the nourishment (figure 9). The cross sections show that immediately north (Profile 1) and south (Profile 5) of the nourishment, the sandbars appear to be more pronounced in T1 and T2 in comparison with the T0 data. The profiles across the nourishment show the fresh nourishment in the T1 measurements with a sharp crest on the landward side. The T2 data show the erosion of the higher parts, and the formation of sandbars on top of the nourished sand. Also the landward transport of the sand as a result of the flattening of the nourishment is clearly visible (Wiersma, 2012 in prep).

Fig. 9: Lake bed cross section
 

4.Fibre optic distributed temperature sensing

In the area north of the sand bank a grid consisting of a 4 km long zigzagging fibre-optical cable has been constructed. Part of this grid is covered by the sand bank. Using laser technology it is possible to measure the temperature in each metre of this cable with a time interval of about two hours. As sediment has a lower heat conductivity than water, parts of the cable covered with sediment will yield a different temperature signal than parts that are not covered at all. Also, the thickness of the sediment layer on top of the cable will influence the temperature signal. Based on this difference in temperature signal the location and thickness of the sediment layer can be calculated. This technique and its results are described extensively in the tool page on fibre optic distributed temperature sensing.

1. Designing an operational sand engine in shallow lakes requires very careful consideration of local conditions. Elements like: dominant wave directions and energies, grain size of locally available sediments, coastal morphology vary and are key to a successful sand engine
2. Monitoring water temperatures as a proxy for sedimentation processes with optical glass fibre cable works. More research is needed to develop as a way to measure sediment thickness.
3. The sand engine slowly adapts its forms to the original lake bottom. There is sand movement in north-easterly direction, but it is slower than expected.
4. Select the right scale. The discussions in Hindeloopen partly were about the choice of scale. The selection of the beach as experiment location did not fit in the larger scale Masterplan process of the municipality. The intention to negotiate with local stakeholders did create friction with the municipalities ambition to plan for the whole coast. Also larger scale sand transport processes in front of the Frisian coast were not sufficiently understood.
    

Literature

• Deltacommissie (2008). ‘Samen werken met water - Bevindingen van de Deltacommissie 2008’, Deltacommissie Nederland, september 2008 (Dutch)
• Folmer E.O., Wilms T., Cleveringa J., Steijn R.C. (2010). ‘Pilot eco-dynamiek Fryske kust. Alkyon: Marknesse’. Accessed from: http://public.deltares.nl/display/BWN1/Theme+3+The+pilot (augustus, 2012).
• Gersonius, B., Ashley, R. and Zevenbergen, C. (2012). ‘The identity approach for assessing socio-technical resilience to climate change: example of flood risk management for the Island of Dordrecht’. Natural Hazards and Earth System Sciences, vol. 12, pp. 2139-2146
• Government of Bangladesh Ministry of Water resources (2001). ‘Bank Protection Pilot Project FAP-21, 2001’. Guidelines and Design Manual for Standardized Bank Protection Structures. Chapter 6 Design of permeable groynes.
• It Fryske Gea (2011). ‘Building with nature pilot Workumerwaard. MIJ 3.2 – Ecodynamic design deliverable report C04021.002448
• Reading, H.G., Collinson, J.D. (1996). Clastic coasts. In: Reading, H.G. (ed.) Sedimentary Environments: Processes, Facies and Stratigraphy. Blackwell Science, 154-231.
• Slobbe, E. van (2010) ‘Ecodynamic Design as a Boundary Object – A case study in Fryslan’. ERSCP-EMSU conference, Delft, The Netherlands, October 25-29, 2010
• Slobbe, E. van, Block, D. de, Lulofs, K. and Groot, A. (2012). ‘The role of experimentation in governance. Lessons from a Building with Nature experiment’ Water Governance, vol. 1, pp. 56- 62;
• Tromp, E. and Weesenbeeck, B. van (2011). ‘How ecological engineering can serve in coastal protection’. ICID, Groningen, May 2011
• Wiersma, A. (2011). ‘Memo about the lessons learned from the first of three Building with Nature pilots’. Memo (_1203678-000-BGS-0003) _to Erik van Slobbe
• Wiersma, A. (2012), presentation "Woelige Wateren hebben Ondiepe Gronden", De werking van zandmotoren langs de Friese IJsselmeerkust,
•  Wiersma, A., van Hattum, T., de lange, M., van Slobbe, E., (2016), 'Building with Nature pilot Zandmotor Friese IJsselmeerkust, Hoe effectief is de zandmotor als ecodynamische strategie voor het versterken van de Friese IJsselmeerkust?', Deltares report 1220414-000.