Figure1. Examples of meandering channels in tidal mudflats along the World’s coast. (a) Baegmihang Port, South Korea (37°09′N, 126°40′E; ©Google, TerraMetrics; imagery date: March 14, 2019). (b) Boseong Bay, South Korea (34°52′N, 127°30′E; ©Google, Maxar Technologies; imagery date: August 30, 2020). (c) Cardiff Flats, England (51°28′N, 3°08′W; ©Google, Maxar Technologies; imagery date: July 11, 2013). (d) Fundy Bay, Canada (45°45′N, 64°38′E; ©Google, Maxar Technologies; imagery date: May 21, 2017). (e) Mühlenberger Loch, Germany (53°32′N, 9°48′E; ©Google, CNES/Airbus; imagery date: April 22, 2020). (f) The Wadden Sea, Germany (53°41′N, 8°02′E; ©Google, Maxar Technologies; imagery date: September 25, 2016). (g) I’Épinay Estuary, France (47°31′N, 2°36′W; ©Google, Maxar Technologies; imagery date: March 19, 2011). (h) Lanveur Bay, France (48°21′N, 4°17′W; ©Google, Landsat/Copernicus; imagery date: January 01, 2005). (i) Morlaix Bay, France (48°38′N, 3°51’W; ©Google, TerraMetrics; imagery date: January 01, 2005).

In spite of their prominence and ubiquity, however, meandering channels in tidal mudflats are still poorly studied especially from a hydrodynamic standpoint. Previous field measurements of flow fields in tidal meanders focused primarily on either tidally-influenced fluvial reaches, where flow dynamics are largely influenced by river discharges and density-stratification effects (Chant, 2002; Keevil et al., 2015; Kranenburg et al., 2019; Somsook et al., 2020), or on intertidal channels dissecting vegetated salt marshes and mangrove swamps (Finotello, Ghinassi, et al., 2020; Horstman et al., 2021). In contrast, field studies on tidal meanders wandering through unvegetated intertidal mudflats are still scarce (Choi et al., 2013; Kleinhans et al., 2009), and flow velocity measurements are virtually nonexistent to date. This is a critical knowledge gap because significant differences might exist in terms of flow fields between tidal channels wandering through vegetated and unvegetated intertidal plains, especially concerning overbank stages (i.e., water levels that exceed the channel bankfull capacity). Overbank velocities in vegetated settings dominated by turbulence and friction are typically a magnitude lower than those observed on unvegetated mudflats (Bouma et al., 2005; Christiansen et al., 2000; D’Alpaos et al., 2021; Friedrichs, 2011; Hughes, 2012; Rinaldo et al., 1999a; Sullivan et al., 2015). Besides, overbank stages are more frequent in mudflats than in salt marshes, owing to the relatively lower position occupied by mudflat channel banks within the intertidal frame. As such, stage-velocity relations in mudflat tidal channels can differ greatly from those observed in vegetated marshes and mangrove forests, and overbank stages might have stronger control on tidal channel morphodynamics (D’Alpaos et al., 2021; Hughes, 2012; Kearney et al., 2017; McLachlan et al., 2020; Sgarabotto et al., 2021), potentially justifying the observed morphological differences of tidal channel networks in distinct vegetational settings (Geng et al., 2021; Kearney & Fagherazzi, 2016; Schwarz et al., 2022; Wang et al., 1999a, 1999b). These differences in landforming hydrodynamic processes are also likely to affect the development of curvature-induced helical flow that is typically related to the development and growth of meander bends in both rivers and salt-marsh tidal channels (Azpiroz-Zabala et al., 2017; Finotello, Ghinassi, et al., 2020; Keevil et al., 2015; Kranenburg et al., 2019; Nidzieko et al., 2009; Thorne et al., 1985). Such helical flow forms as a consequence of secondary (i.e., cross-sectional) circulations, oriented toward the inner and outer bank in the near‐bed and near‐surface zone, respectively, which result from the imbalance between the upward-increasing centrifugal forces and the lateral pressure gradients created by the curvature‐induced superelevation of the water surface at the outer bank (Engelund, 1974; Prandtl, 1926; Rozovskiĭ, 1957; Solari et al., 2002). The downstream advection of secondary circulations operated by the main streamwise flow produces a helical flow, as extensively documented in a variety of field (Dietrich & Smith, 1983; Dinehart & Burau, 2005; Frothingham & Rhoads, 2003), laboratory (Blanckaert, 2011; Liaghat et al., 2014), and numerical studies (Blanckaert & de Vriend, 2003; Bridge & Jarvis, 1982; Ferguson et al., 2003).

Although secondary currents akin to those found in river meanders have been observed and modelled in meandering salt-marsh creeks and large estuarine tidal channels (Finotello, Canestrelli, et al., 2019; Finotello et al., 2022; Finotello, Ghinassi, et al., 2020; Kranenburg et al., 2019; Nidzieko et al., 2009; Pein et al., 2018; Somsook et al., 2020; Somsook et al., 2022), their presence in sinuous mudflat channels has yet to be demonstrated. In fact, previous studies (e.g., Choi, 2011; Choi & Jo, 2015; Ghinassi et al., 2019; Kranenburg et al., 2019) suggested that the morphodynamic processes governing meander evolution in intertidal mudflat settings can differ greatly from the classic secondary-current-driven lateral channel migration mechanism acting in vegetated fluvial and intertidal plains. For instance, Kleinhans et al. (2009) argued that owing to the higher thresholds for erosion that characterize mudflat deposits, bank erosion is primarily due to bank undercutting caused by backward-migrating steps along the channel bed driven by hydraulic jumps that form during ebb tides. They also demonstrated that bank migration occurs preferentially in very sharp bends, where flow separates from the meander inner (convex) bank and impinges directly against the outer (concave) bank. Choi (2011) noted enhanced tidal channel migration in association with episodic and seasonal increase of discharge due to, for example, heavy precipitations, pointing to a strong control of these non-tidal processes on the morphodynamic and sedimentology of tidal mudflat meanders. Accordingly, Choi and Jo (2015) measured pronounced meander migration in the Yeochari macrotidal flat (South Korea) during the summer rainy season, when point bars were observed to migrate as fast as 40 m per month due to increased runoff discharge caused by heavy rainfalls in the order of tens to hundreds of millimeters per hour, possibly compounded by monsoon precipitations. Finally, Ghinassi et al. (2019) suggested that wave winnowing of mudflats during high-tides modulates meander morphosedimentary evolution, leading to widespread bank collapses within the channel.

In view of the above, the structure of tidal flow fields in mudflat tidal meanders appears to be worth investigating. Here we present novel hydroacoustic data from a meandering tidal channel dissecting a macrotidal mudflat located along the Jiangsu coast (China). The aim of the study is threefold, as we intend to: (i) highlight the characteristics of tidal flows within a meander bend developed in an unvegetated tidal mudflat; (ii) unravel possible differences in meander hydrodynamics among below-bankfull and above-bankfull (i.e., overbank) water stages; and (iii) disclose the characteristics of secondary circulations and their relations with the overbank flows. To the best of our knowledge, this study represents the very first attempt to directly measure tidal flows in meandering mudflat channels.

2 Geomorphological setting and study-case

Our study case is found in the Yangkou tidal flat (YTF), an extensive mudflat system located on the southern Jiangsu coast, northward of the Yangtze River Delta, which is bordered by the Yellow Sea to the East and North and by the East China Sea to the South (Figure 2a). The YTF was formed by abundant sediment supply input from both the Yangtze River and the Yellow River, which historically allowed for seaward expansion of the whole Jiangsu province coastline (Shi et al., 2016; Wang & Zhu, 1990). Sediments consist mainly of silty-muddy material, with average grain sizes ranging between 10 and 45 μm (i.e., 4.5 ~ 6.6 φ) (Shi et al., 2016; Wang & Ke, 1997). In the last 2 centuries, however, the seaward extent of the YTF has decreased from 5 ~ 11 km to about 5 ~ 8 km as a consequence of changes in sediment transport regime driven by anthropogenic interventions, the latter including the diversion of the Yellow River to the Bohai Sea in 1855 (Ren & Shi, 1986), and the construction of the Three Gorges Dam in 2003, which significantly decreased sediment supply from the Yangtze River (Yang et al., 2014). In addition to this, land reclamation projects, the building of oceanic outfalls, aquaculture, and the construction of wind farms have further contributed to increasing anthropogenic pressures in the YTF area (Xu et al., 2019; Zhao et al., 2020; Zhao & Gao, 2015). Nowadays, the whole intertidal area in the YTF covers approximately 100 km², extending seaward from the shoreline with gentle slopes ranging between 0.5‰ and 1.2‰ on average (Wang & Ke, 1997; Zhu et al., 1986).

Intertidal mudflats in the YTF are dissected by extensive networks of tidal channels. These channels serve as the main conduits for the propagation of both the East China Sea progressive tidal wave and the southern Yellow Sea rotary tidal wave, which converge nearby the town of Yangkou giving rise to complex coastal circulations (Liu et al., 1989). The tidal regime in the study area is semidiurnal macro-tidal, with average and spring tidal ranges equal to 4.6 m and 8 m, respectively. Morphodynamic processes are also affected by the East Asian Monsoon, which blows with a mean winter wind speed of 4.2 m/s toward the southeast and a mean summer wind speed of 2.8 m/s toward the northwest, respectively (maximum measured wind speed is 34 m/s; (Li et al., 2011; Xing et al., 2012). As wave conditions in this region are mainly related to wind speeds, wave heights are smaller in the summer and larger in the winter, with annual average values ranging between 0.5 and 1.5 m (Chen, 2016). The annual precipitation is about 900 ~ 1000 mm on average, with the summer season accounting for more than 40% of the whole yearly rainfall (Wang & Ke, 1997; Xing et al., 2012).

Our study site is a blind tidal channel found within a natural reserve facing the Xiaoyaokou Scenic and the Xinchuan port, both located nearby the city of Yangkou (Figure 2b). The studied channel is 1.9 km long and is characterized by average width of 8 m. With an overall channel sinuosity equal to 1.5, it represents a well-developed meandering reach. The channel originates from a fringing salt marsh, which borders the Xiaoyangkou Scenic and is covered by Spartina alterniflora Loisel(Figure 2c,e), and extends seaward wandering through an unvegetated intertidal mudflat. Freshwater fluxes from the Beiling river to the North and the Bencha canal to the South do not interfere with the hydrodynamic regime of the studied channel, which is always submerged at high tide and drains out almost completely at low tide.

In this study, we focused specifically on a meander bend located in the central portion of the channel and surrounded by unvegetated tidal flats (Figure 2c,d,f). The studied bend is characterized by a cartesian wavelength (i.e., the linear distance between bend inflections)\(L_{xy}\)=37 m, whereas the along-channel bend length (\(L_{s}\)) is equal to 56 m. Hence, the bend attains a sinuosity\(\chi\)=\(L_{s}/L_{\text{xy}}\)=1.5. The average meander radius of curvature is \(R\)=19 m, and the amplitude, measured as the maximum distance from the line passing through both bend inflections, is equal to \(A\)=18 m. The cross-sectional width (\(W\)) decreases from 8.8 m to 8.3 m in the landward direction (average width\(\overset{\overline{}}{W}\)=8.5 m). Being the bankfull depth (\(Y_{B}\)) equal 1.20 m on average, the studied bend is characterized by an average width-to-depth ratio (\(\beta=\overset{\overline{}}{W}/Y_{B}\)) of about 7.1. All these morphometric parameters are in line with typical values observed for tidal channels worldwide (D’Alpaos et al., 2005; Finotello, D’Alpaos, et al., 2020; Hughes, 2012). While many regularly-spaced small erosional gullies cut through the channel banks (Figure 2f), a 4 m wide and 0.5 m deep side tributary, meandering in planform, is found landward of the apex of the studied bend (Figure 2d).