Characteristics of ships' depression waves and associated sediment resuspension in Venice Lagoon, Italy

Abstract

The character of ships' wakes and the consequent sediment resuspension induced by the passage of commercial vessels in Venice Lagoon, Italy was investigated during July, 2009. Suspended sediment concentration (SSC), water depth, and water velocity were measured on the shoals alongside the shipping channel after the passage of forty vessels. The ships' passage produced large waves of depression with troughs oriented at an average angle of 44° to the channel. High water velocities are recorded opposite the direction of wave propagation and may exceed 2 m s^− 1 at the rear edge of the trough. Ten vessel-induced wakes led to SSC concentrations above 400 mg L^− 1, ~ 30 times higher than the average background concentration. When large wakes passed over the shoals, maximum concentrations persisted for several minutes and elevated concentrations persisted for nearly an hour. The height of a ship's wake can be related to the product of the depth-based Froude number, Fr, and a blocking coefficient, S, according to Fr^3.5S^1.6. Because of the sensitivity of the height of the wake to the ship's speed, we suggest a method for calculation of a ship-by-ship velocity threshold below which shear stress does not exceed the critical erosive stress level for sediments.

Introduction

As the shipping industry continues its rapid expansion it is important to gain a better understanding about the development and propagation of ships' wakes from channels toward shoals and to understand their role in erosion processes. Ships' wakes have been found to be an important factor in shoreline erosion and transport in rivers (Nanson et al., 1994), bays (Parnell et al., 2008, Soomere et al., 2009), lakes (Hofmann et al., 2008), tidal creeks (Ravens and Thomas, 2008), and on salt marshes in estuaries (Houser, 2010). Several of these studies investigated the relative importance of wind waves compared to ship waves and found that, due to the frequency of wind wave events, wind waves were the most important contributor to erosion. Houser (2010), however found that the energy from ships' wakes was much higher than that from wind waves and, therefore, ships' wakes remain an important factor in salt marsh erosion in the Savannah River estuary. Meanwhile a limited number of studies analyzed benthic sediment resuspension and sediment transport associated with ships' wakes (Erm et al., 2009, Hofmann et al., 2008, Schoellhamer, 1996, Wiberg and Sherwood, 2008). These studies show that increased bottom water velocities from ships' wakes lead to significant sediment resuspension along shoals and transport of the sediment with the water current. Such resuspension can redistribute contaminants, contribute to the shoaling of channels and reduce light levels in the water column.

As ships pass through a narrow channel they produce both Airy or oscillating waves, and a wake, which is marked by a significant drawdown and surge of water (Houser, 2010). The latter has been called the Bernoulli wake (PIANC, 2003), depression wake (Soomere, 2007) drawdown and surge wave (Houser, 2010, Nanson et al., 1994) or ship-induced bore (Ravens and Thomas, 2008). Soomere (2007) in a review of non-linear components of ship wakes describes this type of wake as a depression-area wake or consequently, ship squat (Gourlay, 2006, Gourlay, 2008). (For simplicity we will use the term “Bernoulli wake” in this document). This depression wave is a nonlinear effect strongly connected to the vicinity of the sailing regime to critical velocities, rather than an effect of the local water velocity fields. In coastal waters with narrow fairways, the Bernoulli wake propagates across the neighboring shoals. The exact vessel parameters which determine the characteristics of the Bernoulli wake, the propagation of the wake onto the shoals, and how it interacts with bottom sediments remain poorly understood (PIANC, 2003). While studies have shown that ships' wakes can be an important factor controlling shoreline morphology (Soomere et al., 2009), it is often unclear as to the type of wake (Bernoulli or Airy waves) responsible for erosion, sediment remobilization, and transport (Houser, 2010). In large bays or open water areas it may be difficult even to distinguish ships' wakes from wind waves using conventional instruments, thereby making it complicated to isolate the effect of vessel-generated waves (Nanson et al., 1994, PIANC, 2003). However fairways, enclosed lagoons and protected harbors make this separation possible and also provide environments where the different ship wakes can be seen (PIANC, 2003). In shallow coastal waters the proximity of shoals to the shipping channel allows the Bernoulli wake to frequently propagate onto the shoals where it can then be separated by shape and frequency. Venice Lagoon, therefore, is an ideal location for understanding vessel impacts on the environment.

While there has been some discussion of the possible negative impacts shipping has on lagoons, such as pollution from ballast water and emissions and the introduction of invasive species (e.g. Goldberg, 1995, Hayes and Sliwa, 2003, Ricciardi and Rasmussen, 1998) less attention has been paid to the impact of sediment resuspension from commercial vessels in lagoons. Lagoons, such as Venice Lagoon, represent highly dynamic and complex ecosystems. Although lagoons are naturally turbid due to their relatively shallow depths, terrestrial runoff, and anthropogenic development (Neil et al., 2002), an increase in suspended sediment can be detrimental to the health of a lagoon by reducing sunlight penetration and releasing particle-bound contaminants into the water column (Streets and Holden, 2003). In this context, there is an urgent need to understand the effect ships' wakes have on sediment dynamics and morphology.

Sediment resuspension and erosion of shoals from ships' wakes are of concern in Venice Lagoon; there is a long history of ships' wakes which propagate unimpeded across the lagoon shoals. Some of these wakes are generated by very large commercial vessels, cruise ships or ferries navigating a deep, narrow channel dredged across shallow shoals. The vessels may be slow moving but they may also occupy a significant fraction, say more than 10%, of the cross-sectional area of the channel and the wakes they generate propagate immediately into very shallow water. Over the shoals, these wakes produce high, near-bottom current velocities (Bauer et al., 2002), leading to substantial sediment resuspension (Erm et al., 2009, Schoellhamer, 1996, Wiberg and Sherwood, 2008). Some of the sediment resuspended by wakes on the shoals may find its way into the channel, where preferential deposition and shoaling will increase dredging demand (Umgiesser et al., 2004). Indeed, there was a marked increase of water depth (adjusted for local subsidence rates) in the vicinity of the Malamocco–Marghera shipping channel during the period from 1970 to 2000 (Molinaroli et al., 2009) possibly caused by ships' wakes propagating across the shoals. Industrial, sediment-bound contaminants will also be spread by periodic resuspension due to ships' wakes.

We have documented how ships influence suspended sediments in a shallow-water area of Venice Lagoon, Italy. We aim to ascertain which parameters of the vessel most influence the formation of the Bernoulli wake. We describe the characteristics of ships' wakes caused by large, slowly moving vessels which transverse a narrow, deep channel across a shallow lagoon that are especially important in the resuspension of sediment. We will also discuss how changes in the ships' speed and the timing of transit in relation to the tide might minimize adverse impacts due to sediment resuspension.

A substantial amount of literature exists concerning the development and propagation of waves from passing vessels (e.g. Velegrakis et al., 2007), including investigations concerned with the properties of the wake once it leaves the channel (e.g. Didenkulova et al., 2009, Houser, 2010, Parnell and Kofoed-Hansen, 2001, Schoellhamer, 1996) and consequent resuspension (Erm et al., 2009, Hofmann et al., 2008, Nanson et al., 1994, Osborne and Boak, 1999, Schoellhamer, 1996, Soomere and Kask, 2003, Soomere et al., 2007). Of these investigations, most are concerned with the study of a far-field wake from fast ferries and its effect on sediment erosion and resuspension on the shoreline (e.g. Parnell et al., 2008). However, in Venice Lagoon the shoals are immediately adjacent to the shipping channel and are thus affected by the Bernoulli wake which develops on the side of the hull (Soomere, 2007). Gourlay (2008) has shown that it is possible to predict the squat of a ship based on a Fourier transform method for different channel geometries and that this method holds best for slender ships. Gourlay, 2001, Soomere, 2006 discuss the development of a supercritical bore or a soliton which advances in front of the ship; however in these cases channel walls limit the propagation of the bore. Ravens and Thomas (2008) however demonstrate that laterally propagating waves can be converted into bores. Therefore, while the bore normally follows the ships motion, solitons and, as pointed out by a reviewer, bore-like events may occur at some distance from the channel. The height of waves is expected to increase greatly as the ship-induced waves move onto the shoals and optionally interact with one another. Moreover, the particular configuration of the cross-section of the channel may lead to effects similar to those occurring at high Froude numbers even for ships sailing at moderate Froude numbers (Torsvik et al., 2009). While theoretically no bore can occur in front of a ship moving at sub-critical Froude velocities, the width-averaged Froude number of some of the ships is high enough to cause a large displacement of water in the channel (Torsvik et al., 2009). Water velocity then must increase around the sides of the ship to replace the water displaced by the large ship (PIANC, 2003). The increase in velocity must be balanced by a decrease in pressure creating a noticeable depression around the sides of the ship (Oebius, 2000). It is important to note that as the velocity of the ships increase, their sailing speed moves closer to the critical regime, which leads to greater height of the depression wake. When recorded on an instrument within the deep channel near the ship, the Bernoulli wake produced by this pressure difference may take the shape of a symmetrical solitary wave (Oebius, 2000) however the shape of the wave will be modified as it propagates over the shoals. It is likely that the shape of the wave at critical speed is determined by the interaction of two undulating bores (El et al., 2007, Grimshaw and Smyth, 1986).

The study was completed along the shoals outside the Malamocco–Marghera industrial channel of Venice Lagoon, Italy. With a surface area of more than 500 km² Venice Lagoon, located in north-eastern Italy, is the largest Italian lagoon (Fig. 1). It is shallow with an average depth of 0.8 m; however it is crossed by numerous natural and man-made channels, which are all maintained by an active dredging program and reach depths of 10 m or more. The lagoon is connected to the sea by three inlets with a mean total tidal discharge of 6500 m³ s^− 1 (Cucco and Umgiesser, 2006, Gaĉić et al., 2002). The tidal excursion in the lagoon is about 0.3 m to 1.1 m during neap and spring tide respectively.

Concurrent with the development of the Porto Marghera Industrial Zone (PMIZ) between 1920 and 1970, a new channel known as the Malamocco–Marghera channel (or locally known as Canale dei Petroli) was dredged from the Malamocco Inlet to the PMIZ. This channel has a total length of 20 km, mean width of 200 m, and a depth of 12 m to accommodate medium-sized container ships, bulk carriers, and tankers. After entering the lagoon, the channel heads directly west towards the mainland cutting through natural tidal channels. It then makes a sharp right-hand turn north towards the PMIZ, where it stretches alongside the mainland for 14 km (Fig. 1) in an almost straight path.

Grain size at the channel bottom ranges from sandy-silt to silty-sand sediments, the finer fraction more easily resuspended (Saretta et al., 2010, Zonta et al., 2007). Our measurements indicate that the sides of the channel are exceptionally steep rising towards the shoal with a slope of about 25% on the east side and 30% on the west side. The slopes at the edge of the channel are very steep and must be maintained by a dredging program. East of the channel the seafloor quickly shoals to depths ranging between 1 and 2 m (Fig. 1). This shallow-water area extends 5 to 8 km lagoon-ward of the channel for its entire length. In most locations a wave can travel unimpeded along the shoal for several hundred meters to kilometers until it is dissipated naturally via bottom friction, white capping, depth-induced breaking and, possibly, deformation of the sea bed (Chan and Liu, 2009).

Our study site is located in these eastern shoals, 2000 m south of the PMIZ and directly seaward of Fusina (Fig. 1). Tidal currents in this part of the lagoon are generally low (< 0.05 m s^− 1) due to its distance from the inlet and relatively open water location, however winds can increase currents to over 0.15 m s^− 1 (Coraci et al., 2007). The lagoon is naturally sheltered from high wind waves, due to limited fetch length. Although we do not have any data of annual wave climate in this area, our measurements show that southeasterly winds caused significant wave heights in the study period of up to 0.2 m (the fetch is 8 km in this direction). An analysis of annual hourly wind data measured at a nearby meteorological station from the year 2009 provided by the Venice municipal authority (Istituzione Centro Previsione e Segnalazione Maree, Comune di Venezia) shows that our study period was representative of the annual wind climate with an average wind speed of 3.23 m s^− 1, which is only slightly higher than the annual mean (3.11 m s^− 1). Sustained winds above 7.5 m s^− 1 occurred 2.1% of the time during our study period and 1.96% of the time during the year. Six storms were recorded in 2009 in which the sustained wind speed reached 10 m s^− 1, though none occurred during our study period. Empirical relationships for the forecasting of wind-waves (Goda, 2003) predict that the maximum significant wave height reached in 2009 could be no more than 0.5 m, and that this height was likely only attained during these 6 storms. The fetch length and the presence of Scirocco conditions, however, can lead to waves creating bottom stress values above 0.7 Pa, and may be a cause of the extensive erosion seen in the area (Umgiesser et al., 2004).

Ships are near maximum allowed speed (5.56 m s^− 1) when passing our study location. An analysis of shipping statistics from the Port Authority of Venice website implies that between 2000 and 3000 large (length > 100 m) ships pass through this shipping channel every year (Port of Venice website, http://www.port.venice.it/pdv/Home.do?metodo=carica_home).

During a preliminary sampling in March 2009, an acoustic Doppler current profiler (600 kHz Teledyne-RDI Workhorse Rio Grande ADCP) was used to collect current data prior to and directly after a ship passage across the shipping channel. Acoustic backscatter data of the instrument was additionally used to estimate the concentration of suspended particles by calibration with direct measurements of SSC in water samples (Defendi et al., 2010).

An across channel ADCP transect taken after the passage of a vessel shows resuspension directly in the path of boats with drafts greater than 8 m (e.g. Fig. 2). The ADCP data show that significant resuspension (greater than 100 mg L^− 1) occurs throughout the water column directly below the ship; these sediments remain in suspension within the channel for 5 to 20 min after the passage of the ship. There is a distinct dark cloud near the surface of the water column but this cloud quickly disappears suggesting that it is most likely bubbles created by the propeller motion. The figure shows some resuspension on the shoals outside the channel as well; however no connection between resuspension in the channel and on the shoals is found (Fig. 2).