5.1 Overbank flows and stage-velocity relationships
Horizontal velocity distributions and stage-velocity diagrams of our
studied channel display critical differences compared to those observed
in channels wandering through vegetated salt marshes and mangrove
forests (D’Alpaos et al., 2021; Fagherazzi et al., 2008; Hughes, 2012;
Kearney et al., 2017; McLachlan et al., 2020; van Maanen et al., 2015;
see Figures 5 and 6). Owing to the characteristic geomorphic structure
of vegetated intertidal plains, peaks of ebb and flood velocities in
tidal channels typically occur just below or above the bankfull stage
(i.e., for \(Y\)>\(Y_{B}\)), with velocities being
significantly reduced at \(Y\)<\(Y_{B}\) and approaching null
values when \(Y\) is minimum (see Bayliss-Smith et al., 1979; Boon,
1975; Fagherazzi et al., 2008; Hughes, 2012; Kearney et al., 2017). In
contrast, our monitored mudflat channel is characterized by sustained
velocities at \(Y\)<\(Y_{B}\), with both horizontal
(\(\overrightarrow{v}\)) and depth-averaged velocities (DAVs)
peaks occurring when tidal flows are confined within the channel banks
(Figures 5 and 6). Notably, in all the monitored sites velocities are
relevant (DAV≈0.8 m/s) even for reduced water depth
(\(Y\)<0.5 m), especially during the ebb (see Figure
6a,b,c,d). Overbank stages are instead characterized by reduced
velocities, both in terms of \(\overrightarrow{v}\) and DAVvalues (Figures 5 and 6).
These discrepancies in velocities fields between channels found in
vegetated and unvegetated intertidal settings are likely due to the
relative speed at which tides can propagate within and outside tidal
channel networks. Specifically, frictionally-dominated tidal flows
across vegetated intertidal plains make channels preferential pathways
for tide propagation even when water levels exceed the bankfull (i.e.,
for \(Y\)>\(Y_{B}\); D’Alpaos et al., 2007; Rinaldo et al.,
1999a, 1999b). In contrast, flow resistance in unvegetated mudflats is
comparable between tidal channels and intertidal plains, such that tide
propagation through unvegetated intertidal mudflats is dominated by
sheet flow. This hypothesis is supported by field data from the
meso-macrotidal Scheldt Estuary (Vandenbruwaene et al., 2015)
highlighting similar velocities within tidal channels
(0.3~1 m/s) and across bare intertidal mudflats
(0.1~0.4 m/s), in contrast to salt marshes wherein tidal
flow velocities are typically lower than 0.1 m/s. In our studied
channel, flow velocities for above- and below-bankfull stages are found
to be in the range 0.2~0.4 m/s and 0.2~1
m/s, respectively (Figure 5,6), which roughly correspond with the
results of Vandenbruwaene et al. (2015). The latter data also suggest
that instantaneous water levels are not significantly different within
channels and across mudflats, in contrast to frictionally-dominated
vegetated intertidal plains where significant differences in
instantaneous water levels occur moving away from tidal channels
(D’Alpaos et al., 2021; Rinaldo et al., 1999a, 1999b; Sullivan et al.,
2015).
Besides differences in bottom friction at overbank stages, one should
also appreciate that hydrodynamic dissimilarities are to be expected in
mudflat vs. salt-marsh channels as a consequence of distinct
characteristic elevations of both their banks and the adjoining
intertidal platforms. Mudflat channels typically occupy the lower
portions of the intertidal frame, their bank elevation typically ranging
between the mean sea level (MSL) and the mean low water springs (MLWS).
This allows for significant water depths at above-bankfull stages, which
reduce flow confinement within the channel and limit channel flow
velocities (Brooks et al., 2021). In contrast, channel banks in salt
marshes are typically located in the highest portions of the intertidal
frame, which ensures in-channel flow confinement and sustained flow
velocities even for large tidal oscillations, effectively limiting
above-bankfull water depths. In our study case, high\(\overrightarrow{v}\) during overbank stages are only observed when
peak tidal levels exceed \(Y_{\max}\)>3.2 m, that is, for
spring tidal cycles (Figure 4a,b,c,d) or, more generally, for
high-amplitude (HAT ) tidal cycles (Figure 7b). Such high\(\overrightarrow{v}\) values are however likely related to overbank
circulations occurring at the scale of the entire mudflat systems, which
are not necessarily related to flow dynamics within the channel. This is
confirmed by the analysis of \(\overrightarrow{v}\) directions along the
water column (Figure 7c), which testifies clear deviations of tidal
flows at \(Y\)>\(Y_{B}\) relative to the orientation of the
channel axis both for HAT and LAT tidal cycles. Such a
deviation produces consistent flow directions in all the monitored
sites, with tidal flows being directed to the South-East and South-West
during ebb and flood tides, respectively (Figure 7c).
These observations altogether support the idea that differences in the
character of overbank flows result in marked hydrodynamic
dissimilarities between tidal channels dissecting vegetated and
unvegetated intertidal plains. Such differences are also likely to
affect curvature-induced secondary circulations and the related meander
morphodynamic evolution, as we discuss in detail in the next sections.