5.3 Implications for meander morphodynamics
Since the generation and propagation of helical flow are hampered, questions arise regarding what are the chief morphodynamic processes driving meander evolution in unvegetated intertidal mudflats. Previous studies suggested that mudflat meanders can form and develop without significant secondary circulations. For example, the evolution of small mudflat meandering channels (about 1 m wide) in the Westerschelde estuary (Netherlands) was found to be primarily driven by late-ebb flows, which determined the erosion of channel bed due to backward-migrating steps generated by hydraulic jumps, which in turn promoted channel bank erosion due to bank undercutting and pronounced flow separation in sharp bends (Kleinhans et al., 2009). In our studied channel, sustained velocities at low water stages (Figures 5,6,8,9,10,11), together with direct visual inspections of sustained flow velocities near the end of ebb tides (see Figure 4), support the idea proposed by Kleinhans et al. (2009) that the morphodynamics of intertidal mudflat meanders is strongly controlled by late-ebb flows rather than by classic bar‐hugging helical flow produced by curvature-induced secondary flows at high-water stages. Reduced control of helical flows on channel morphodynamics is also testified by the symmetric, V-shaped form of the studied channel cross-sections (Figure 4b), which attests to the scarce development of secondary (i.e., cross-sectional) flows and contrasts with the asymmetrical U-shaped cross-sections displayed by meandering channels in vegetated tidal marshes (Finotello, Ghinassi, et al., 2020; Zhao et al., 2022).
In contrast to late-ebb flows, we speculate that tidal flows at early-flood stages are not likely to have significant effects in terms of bank undercutting and sediment transport because velocities increase more slowly than during late ebb, and rates of water depth change through time (\(\dot{Y}\)), though sustained, do not produce significant variations in DAVs (Figure 6). Our analyses indeed confirm that tidal flows tend to be ebb-dominated at low water depths (Figure 5e,f), and also highlight that at \(Y\)<\(Y_{B}\) ebb velocities attain values close to the maximum for much longer periods than during the flood (Figure 6), thus likely enhancing the morphodynamic control of late-ebb flows on channel evolution. Moreover, late-ebb flows are likely to occur even for tidal oscillations lower than those we monitored here, whereas pronounced overbank flows and related secondary circulations require significant tidal oscillations to be formed. Because intense late-ebb flows act at every tidal cycle and operate for extended periods, the total morphodynamic work they produce is in all likelihood much more significant than that produced during other tidal phases, further supporting the hypothesis that late-ebb tidal stages are the most morphodynamically relevant for mudflat meander evolution.
The above-described morphodynamic control of late-ebb stages is likely to be even more relevant compared to vegetated tidal landscapes due to the absence of vegetation not only on intertidal plains but also within tidal channels. In fact, previous studies focusing on salt-marsh channels demonstrated how in-channel aquatic vegetation can enhance bottom roughness and flow turbulence (e.g., Finotello, Ghinassi, et al., 2020; Folkard, 2005), further limiting tidal flow velocities at low stages, especially in relatively small channels with widths comparable to the characteristic size of vegetation patches. The presence of aquatic vegetation would clearly prevent significant morphodynamic work to be performed by late-ebb tidal stages, which is likely not the case in the unvegetated mudflat channel we investigated here.
In addition to the above, meander morphodynamics in unvegetated intertidal flats can also be driven by episodic and seasonal increases in discharges due to heavy rainfalls and melting snows (Choi et al., 2013; Choi & Jo, 2015). Choi (2011, 2014) observed that these episodic events are likely to cause abrupt morphologic changes, pronounced point bar migration, frequent meander-bend cutoff, as well as channel abandonments. Particularly, significant rainfall-induced runoff during low tides would mimick late-ebb flows, thus further increasing the morphodynamic relevance of seaward-directed, shallow, in-channel flows. New field measurements will however be required to support this hypothesis in the Yangkou tidal flat, since the data presented in this study were collected in October, which is outside the monsoon season.
Storm waves could also induce bank collapses in unvegetated tidal channels (Choi, 2011; Choi & Jo, 2015; Ghinassi et al., 2019), thus critically affecting meander morpho-sedimentary evolution. In spite of the absence of vegetation that can help stabilize banks and prevent erosion, no collapsed slump blocks were observed within our study channel (Gabet, 1998; Hackney et al., 2015), although such blocks could be easily disgregated and removed, once formed, by sustained in-channel velocities combined with the absence of additional cohesion given by vegetation roots.
Bank collapses can also form due to significant tidal oscillations and pore-excess pressure between channel and banks driven by rapid changes in water levels (Zhao et al., 2022; Zhao et al., 2019), which generate significant seepage flows (e.g., Gardner & Wilson, 2006; Wilson & Morris, 2012). Seepage flows during late-ebb tides, also favored by extensive bioturbation due to fiddler-crab and mudskipper burrowing (Harvey et al., 2019; Ishimatsu et al., 1998; Perillo et al., 2005; Xin et al., 2022), are likely responsible for the widespread bank slumps that we observed at the middle and lower portions of channel cross-sections in the studied channel (Figure 13). Notably, strong seepage flows can also help explain why sustained velocities are observed over nearly the entire duration of the ebb phase (Figure 6 a,b,c,d), and further support the idea that the ebb-late phases exert a strong control on the morphodynamics of intertidal mudflat meanders.