Evaluating the impact of dredging strategies at tidal inlets: Performance assessment

Evaluating the impact of dredging strategies at tidal inlets: performance assessment Carmen Zarzueloa, Alejandro L´opez-Ruiza,∗, Miguel Ortega-S´anchezb aDepartamento de Ingenier´ıa Aeroespacial y Mec´anica de Fluidos, Universidad de Sevilla, Camino de los Descubrimientos s/n, 41092, Seville, Spain bAndalusian Institute for Earth System Research, University of Granada, Avda. del Mediterr´aneo, s/n, 18006, Granada, Spain Abstract Despite relevant advances achieved in recent years, sediment transport and sedimentation problems at tidal inlets are still worldwide issues to be addressed. Furthermore, dredging strategies are carried out following traditional layouts, such as channel deepening, lasting short periods of time despite the high economic expenditures and the potential environmental impacts. This work proposes a new dredging strategy for tidal inlets and analyzes its morphodynamic evolution by means of numerical modeling. This numerical model, used to perform hydro–morphodynamic simulations, is applied to a highly altered tidal inlet (Punta Umbr´ıa inlet, Southern Spain) with a navigational capacity being continuously compromised. After calibrated and tested, the model is applied to different dredging strategies, including channel deepening, littoral drift barrier and shoal removal. Among these strategies, the shoal removal, which is a new soft–engineering strategy, is found to be the most efficient to improve the navigational channel oper- ∗Corresponding author alop[email protected] Preprint submitted to Science of the Total Environment December 13, 2018 ativity, defined as the percentage of navigable hours per year for different vessel drafts; this operativity improves up to 60% compared to the other strategies. This solution, which reduce the frequency of maintenance interventions and hence the environmental impacts, may be suitable for other inlets with compromised navigational capacities due to the presence of ebb shoals. The relation between the main maritime drivers and the morphodynamic changes is analyzed, concluding that the morphodynamic evolution of the navigational channels is closely related not only to the wave energetic content, but largely to the wave directionality. Finally, the shoal removal also increases the flow velocities at the inlet modifying the stability of the mouth and hampering its long–term closure. The potential environmental impacts derived from the shoal removal are also discussed. Keywords: navigational channel, morphodynamics, dredging strategy, numerical model, operativity, environmental impact 1. Introduction1 Tidal inlets constitute the hydraulic connection between the open ocean2 and coastal environments such as estuaries, rivers or bays, among others3 (Militello and Kraus, 2001; Ray, 2001). They are frequently used for human4 activities including commercial routes or recreational areas, which drive im-5 pacts on their hydro–morphodynamics and environmental qualities, such as6 reductions on sediment supply, water discharge or tidal prism, what in turn7 modify the morphodynamic equilibrium and the water quality of the inlets8 (Davis and Fitzgerald, 2004; Seminack and McBride, 2018). Moreover, tidal9 2 inlets are essential for shore and dune processes and the exchange of both10 water and sediments (Fortunato et al., 2014). Their dynamics are mainly11 controlled by complex interactions between different drivers (wind, tides, or12 waves) and the interactions with their own topographic features (Vikas et al.,13 2015).14 All these interactions are the main reason of the tidal inlet stability and15 morphodynamics, determining also the evolution of human interventions.16 Among these interventions, the commercial and/or recreational exploitation17 of ports located inside tidal inlets usually requires periodic dredging works18 to maintain minimum water depths along the navigational channels assuring19 their operational capacity. These dredging works not only impact the econ-20 omy of public administrations, driving also environmental impacts derived21 from the sand removal and the associated increase in suspended sediment22 (Varriale et al., 1985; Brown et al., 1990; Wilber and Clarke, 2001; Harff23 et al., 2009; Oberle et al., 2014).24 Given the significant sediment transport rates and the rapid morpho-25 dynamic variations at tidal inlets, the life–time of the dredging strategies26 around ebb tidal shoals is generally reduced to some years (in some cases27 only months) thus increasing the predicted impacts of these activities (Je28 et al., 2007; Dabees and Kraus, 2008; Wang et al., 2014b,a). Hence, im-29 proving the understanding of the tidal inlet dynamics, the complex processes30 herein, and the morphodynamic evolution of dredging strategies is essential31 for the sustainable management of these coastal areas in view of reducing32 the dredging maintenance costs (Knowles and Cayan, 2004; Hinwood and33 McLean, 2018).34 3 While a number of studies have evaluated the stability of tidal inlets35 through the analysis of the sediment dynamics (Bales and Holley, 1989; Ro-36 man et al., 1997; Duong et al., 2016; Hinwood and McLean, 2018), the effi-37 ciency and impact of dredging strategies have received much less attention38 (Van Maren et al., 2015). These dredging activities, mainly designed to39 maintain the navigation requirements, frequently alter the natural environ-40 ment (Montero et al., 2013) and imply extracting millions of cubic meters41 of sand and gravel (annually) in developed countries (Meng et al., 2018).42 Furthermore, the excavation, transportation and disposal of these sediments43 may lead to various adverse impacts on the marine environment (Erftemeijer44 and Robin Lewis, 2006), that can be especially relevant when dredging or45 disposal are performed in the vicinity of sensitive marine environments, such46 as coral reefs (Erftemeijer et al., 2012) and seagrass beds (Erftemeijer and47 Robin Lewis, 2006).48 One of the few analyses on the efficiency of dredging activities was per-49 formed by Garel (2017), who studied the relation between dredged volumes50 and the frequency of maintenance operations at the Guadiana ebb tidal delta51 using a simplified version of the Inlet Reservoir Model (Kraus, 2000; ´ Alvarez52 et al., 2017). More recently, Reyes-Merlo et al. (2017) proposed a new dredg-53 ing strategy based on the reduction of the flow energy fluxes and its di-54 vergence, although neither its performance nor the sediment transport were55 analyzed in detail. There are two main reasons for this small number of stud-56 ies: (1) the performance assessment of dredging strategies using numerical57 models is challenging due to the complexity of the simultaneous simulation of58 different drivers, and (2) these simulations require robust and reliable mor-59 4 phodynamic predictions, demanding an accurate calibration and testing of60 the model.61 The main objective of this work is to define a new efficient dredging strat-62 egy, analyzing its morphodynamic evolution by means of numerical modeling.63 The efficiency of the strategy is measured in terms of navigational capacity64 and operativity of the main navigation channels for different vessel drafts.65 This operativity is defined as the percentage of navigable hours per year for66 different vessel drafts. The model is calibrated and tested both for hydro– and67 morphodynamics using mid–term simulations and multi–beam bathymetries.68 Different alternatives of dredging strategies are simulated with the model, in-69 cluding channel deepening, littoral drift barrier and shoal removal, which is70 a new soft–engineering strategy previously proposed by Reyes-Merlo et al.71 (2017) but which efficiency was not analyzed. This strategy may be suitable72 for any other tidal inlet where the presence of an ebb shoal compromise the73 navigational capacity. The performance of the strategies is assessed obtaining74 the operativity on the main navigational channels. Finally, the relation be-75 tween the main maritime drivers and the morphodynamic evolution, as well76 as the potential hydrodynamic and environmental impacts, are discussed for77 the most efficient strategy. The methodology is applied to Punta Umbr´ıa78 inlet (Southern Spain), a highly human–altered environment where the nav-79 igation capacity is continuously compromised despite the frequent dredging80 strategies. This area constitutes a prototypical navigable inlet affected by81 the presence of an ebb tidal shoal where periodic dredging works are carried82 out.83 5 2. Study site84 The R´ıa de Huelva is a shallow mesotidal estuary, located on the south-85 western Spanish coast facing the Gulf of C´adiz (37◦11’N, 6◦57’W; Fig. 1). It86 occupies an area of 250 km2including intertidal zones, the intersection of two87 main rivers (Tinto and Odiel) and the Punta Umbr´ıa inlet (PUI hereinafter).88 The Tinto River (Fig. 1) has a length of 100 km with a drainage basin of89 720 km2, whereas the Odiel River (Fig. 1) flows 140 km until the mouth of90 the estuary (Sainz et al., 2004).91 PUI is an 8 km long (NW–SE) and 0.5 km wide (SW–NE) channel with92 a maximum depth of 12 m below mean sea level (MSL hereinafter). It is an93 ebb–tidal system with minor ebb channels, shoals and frontal lobes (Reyes-94 Merlo et al., 2017), which is characterized by large salt marshes with a high95 density of shallow meandering tidal creeks, sand flats and a complex network96 of natural and partially dredged channels. Its mouth is bound at the Atlantic97 Ocean side by a littoral barrier and a mixture of sandbanks and highly mobile98 shallow channels with depths of 2–3 m respect to MSL (Barba-Brioso et al.,99 2010). The PUI system is characterized by the presence of an ebb shoal with100 a averaged water depth of 3 m respect to MSL (Fig. 1d). The shoal slightly101 migrated to the East during last decades due to the dredging works performed102 in the area (Reyes-Merlo et al., 2017). The presence of this shoal have caused103 navigational problems at PUI for decades, promoting the construction of a104 jetty in the mid 80’s of the 20th century at the western side of the inlet105 mouth, extending to approximately 4 m depth. Two navigational channels,106 NC–W and NC–E in Fig. 1d, are used by the local vessels to cross the shoal107 towards the two main port areas along PUI (Fig. 1d).108 6 Morales et al. (2014) analyzed the sedimentary evolution of the PUI en-109 vironment, pointing out that it has been intensified by recent human inter-110 ventions with extensive and intense dredging works performed during the111 last years with the aim of avoiding the closure of the inlet and the silting of112 the channel. Different designs of navigational channel deepening have been113 carried out, although they had a limited life–time (up to 4 years) and have114 been unable to resolve the navigational issues at PUI (Reyes-Merlo et al.,115 2017).116 Tidal data obtained from a tidal gauge (REDMAR 3329, Puertos del117 Estado, Spanish Ministry of Public Works), located at Huelva Port (Fig. 1),118 indicate that tides are semi–diurnal with a tidal range varying between 1119 and 4 m. According to Rodr´ıguez-Ram´ırez (2008) this tidal range controls120 the hydrodynamics and sediment transport in the area. Furthermore, Reyes-121 Merlo et al. (2017) found that tide–induced currents can reach 1.2 m/s and122 0.8 m/s at its mouth during ebb and flood, respectively. Different works123 (Mu˜noz et al., 1997; Sainz and Ruiz, 2006; Ruiz et al., 2014) concluded that124 the morphological framework of the PUI is controlled by the tides in the125 deeper channels where silt and clay are deposited, whereas in the middle126 part of the PUI other factors such as wind waves, the presence of vegetation127 or the weak fluvial action affect the sediment dynamics. At the South of the128 PUI, the morphology is largely influenced by wind and swell waves. This129 variability has an effect on the sediment distribution, which is coarser (D50 130 ≈2 mm) upstream and finer (D50 ≈0.53 mm) downstream.131 The local wave climate, obtained from the hindcast data of SIMAR132 1052048 (Puertos del Estado, Spanish Ministry of Public Works, Fig. 1), is133 7 characterized by moderate sea states (wave heights and periods in the range134 0.5–1 m and 4–6 s, respectively), predominantly approaching from West to135 the Southwest. Storms typically approach PUI from Southwest with wave136 heights between 3 and 6 m. This wave climate results in a local net long-137 shore sediment transport to the west of approximately 0.5–3 ×105m3/year138 (CEDEX, 2013; Reyes-Merlo et al., 2015). Furthermore, according to I.E.O.139 (1992) and Mu˜noz et al. (1997) the highest runoffs occur from December to140 February for both rivers, with averaged discharges of 38.6 m3/s, whereas the141 lowest are concentrated during the summer months (≤0.4 m3/s).142 FIGURE 1143 3. Materials and methods144 3.1. Bathymetry and topography data145 Between 2002 and 2015 regular bathymetric surveys were carried out146 at the mouth and dredged channels of PUI to control the maintenance147 (dredging) works of the navigational channels. A total of 18 multi–beam148 bathymetry surveys were provided by Agencia P´ublica de Puertos de An-149 daluc´ıa (APPA, Andalusian Regional Government), although the majority150 of the dataset only covered some cross–sections of the dredged area along151 the main navigation channel (NC–E). The spatial resolution of the bathyme-152 tries is 2×2 m, and data were corrected and referenced to the MSL at the153 study site using water levels from the tidal gauge REDMAR 3329 (Fig. 1).154 APPA bathymetries covering the complete PUI were used in the anal-155 ysis, being completed with offshore bathymetry data provided by the Hy-156 drographic Marine Institute (IHM, Spain) with a spatial resolution of 5×5157 8 m. The topography close to the inlet was defined with data from a Digital158 Elevation Model (DEM, provided by National Geographic Institute, Spain)159 with a resolution of 5×5 m at tidal flats and 25×25 m elsewhere.160 3.2. Field survey161 Field measurements of water levels, currents and wave heights were used162 to calibrate and test the numerical model. The survey was carried out during163 the spring 2014 collecting data at 5 locations along the main channel of PUI164 and the inner continental shelf with 4 current profiles (ADCP hereinafter)165 and 1 tidal gauge (Fig. 1). Instruments were deployed on May 2014 and166 retrieved on June 2014. For further details on the survey the reader is referred167 to Reyes-Merlo et al. (2017).168 3.3. Numerical model169 The Delft3D model, which is a finite–difference numerical model devel-170 oped by WL/Delft Hydraulics and Delft University of Technology (Lesser171 et al., 2004), was used to study the PUI hydro–morphodynamics. It is a172 widely used computational model to simulate the main physical processes173 that are relevant in coastal environments, such as embayments and estuar-174 ies (Van Leeuwen et al., 2003; Van Rijn, 2007; Ruggiero et al., 2009; Iglesias175 et al., 2012; Hansen et al., 2013). We used two modules of the model: FLOW176 and WAVE. The former is based on the two dimensional (depth–integrated)177 Navier–Stokes equations for an incompressible fluid under the shallow water178 and the Boussinesq assumptions. It also includes morphodynamic evolution179 equations, for which the total transport is obtained as the sum of bed and180 suspended load transports based on the depth–integrated advection–diffusion181 9 observed bed evolution, with values of 0.66 (0.54) for the calibration (testing)304 periods.305 FIGURE 3306 3.4. Dredging strategies307 According to Reyes-Merlo et al. (2017), the strong siltation caused by308 the litoral drift resulted in the growth of ebb shoal near the mouth of PUI309 after each channel deepening performed in the area, reducing its navigational310 capacity and requiring recurrent interventions. Thus, the present geometry311 of the shoal is influenced by past human interventions. In this work four312 different strategies were defined to improve the navigational capacity of the313 inlet and reduce the number and impact of the interventions. Among these314 strategies, the most efficient, i.e. the one that maintains the highest naviga-315 tional capacity during a 3-month period, was chosen based on the results of316 the morphodynamic simulations. After that, we used the selected strategy317 to perform a deeper analysis of its morphodynamic evolution and its effects318 on the PUI hydrodynamics. The four strategies defined are characterized as319 follows.320 3.4.1. S1: PUI in 2014321 This first strategy does not imply any intervention in PUI. The bathymetry322 corresponds to the most unaltered configuration of the study site (Fig. 4a),323 since it was obtained after the longest period without dredging works in the324 area (4 years). According to the bathymetric dataset, it is the most natural325 configuration of PUI.326 16 3.4.2. S2: Channel deepening327 This strategy deepen the main navigation channel in S1(NC–E, Fig. 4b),328 similar to the past dredging works performed in PUI. It increases the water329 depth in the shallowest zone of the submerged sandbar of PUI to reduce its330 recurring siltation. The average dredged volume is 1.3 ×105m3and the331 dredged volume in the shoaling area represents ≈30% of all the mobilized332 material. The depth of the dredged channel is 4 m due to the maximum333 draft of the main vessels that navigate in the area.334 3.4.3. S3: Littoral drift barrier335 This strategy enlarges the jetty of the river mouth (Fig. 4c). The current336 jetty extends to approximately 4 m depth (respect to MSL), allowing the337 sediment bypass of the eastern–oriented littoral drift generated up–drift PUI,338 being the main reason for sedimentation in the channel. With this strategy,339 the jetty is extended to 8 m depth (300 m in length), which is approximately340 the closure depth of the adjacent beach, contributing to significantly reduce341 the sand bypass. This is the most expensive and long–term strategy, and can342 be used in combination with any other strategy. However, to simplify the343 analysis these combinations were not considered.344 3.4.4. S4: Shoal removal345 This strategy, originally proposed by Reyes-Merlo et al. (2017), consists346 of a leveling and partial removal of the shoal at the lee–side of the jetty (Fig.347 4d), to reduce the energy divergence and hence the sediment transport. The348 main goal of this solution is to extend the life–time of the intervention imitat-349 ing how nature works, reducing the external gradients and hence minimizing350 17 the sedimentation at PUI. It is a soft alternative (no structures required)351 that promotes ecosystem services and the resilience of the system. The wa-352 ter depths are consistent with the requirements for navigation at PUI (≃4 m353 below MSL), with maximum differences between the initial and final profiles354 of ≃2 m. The total affected region covers 7.4 ×105m2and the volume of355 removed material is 2 ×105m3(Reyes-Merlo et al., 2017). This strategy is356 presented as a more natural–adapted and sustainable strategy in the mid to357 long–term maintenance of the inlet, being applicable to any other worldwide358 inlet with navigational issues derived from the presence of ebb tidal shoals.359 FIGURE 4360 4. Results361 4.1. Efficiency of the strategies362 Morphodynamic simulations of the 4 strategies defined (Fig. 4) were363 performed by means of morphological evolution and operativity. The sim-364 ulations span three months of a winter period with extreme wave climate365 conditions (December 22th, 2009 to March 20th, 2010).366 Fig. 5 shows the wave, wind and water level climate during the 3–month367 period using the data from SIMAR 1052048 and REDMAR 3329 (Fig. 1).368 Three different phases in terms of wave climate can be defined: (1) a se-369 quence of 4 important storms until the middle of January, (2) a period of370 approximately 20 days of milder wave energy conditions, and (3) another371 20 days of consistent storm conditions with wave heights above 2 m, storm372 surges up to 0.5 m and important wind velocities (red dashed boxes-Fig. 5).373 During the entire period, waves predominantly reached PUI from the374 18 WSW with heights typically up to approximately 4 m at the peak of the375 storms. Wind directions were predominantly NW and NE with speeds vary-376 ing between 2 and 15 m/s and a monthly average close to 10 m/s. Given the377 energetic content of waves and winds, important bathymetry changes were378 expected.379 FIGURE 5380 4.1.1. Morphodynamic evolution of the inlet381 Fig. 6 shows the bathymetric differences between the beginning and the382 end of the 3–month simulations for each strategy. Results were obtained not383 only over PUI (left panels), but also along the two navigational channels384 NC–W and NC–E.385 For S1, S2and S3the erosion/accretion patterns are similar: erosion is386 concentrated at the surroundings of the main channel, specially at the eastern387 shoreline of PUI, whereas accretion is observed at large areas over the central388 part of the PUI where the NC transect. This is clearly observed in Fig.389 6 (panels e and f) where the differences along NC are plotted: significant390 reductions of water depths are identified, specially for NC–W where this391 reduction is above 10% of the initial water depth. Along these NC, erosion392 is only slightly present (≃3-5%) at the southern end of NC–W. These results393 highlight that S2and S3scarcely reduce the accretion at the southern end394 of NC–W, where the loss of water depth decreased 1% compared to the395 unaltered strategy (S1).396 Results are clearly different for S4: although there is also a generalized397 accretion along the NC, water depths only reduced up to 3.5%. The morpho-398 logical changes are weaker compared to S1and S3even for the areas located399 19 outside the shoal removal area; furthermore, the water depth reduction ob-400 served at NC–W for the other strategies is vanished. Hence, this strategy401 clearly reduces the morphodynamic variations and its gradients along the402 NC. Results are summarized in Table 4.403 FIGURE 6404 4.1.2. Operativity405 Although S4was found to be the most efficient strategy in terms of re-406 ducing the morphodynamic changes (i.e. less accretion), we quantified the407 effects of these changes on the navigational capacity of PUI assessing the op-408 erativity along the NC. We define the operativity as the percentage of time409 during the simulation period for which the minimum water depth along the410 complete NC is greater than a threshold value in correspondence to differ-411 ent vessel drafts. Hence, the operativity depends on the clearance at the412 minimum water depth point along the complete NC.413 The results for each strategy are shown in Fig. 7, where the water depth414 of 3 m is highlighted with a vertical black line. According to local adminis-415 trations, this value corresponds to the draft (including the safety clearance)416 of the design vessel expected to navigate along the NC of PUI. The opera-417 tivities for S1and S3are closely related with the morphodynamic evolution418 obtained in Fig. (6), with values of approximately 60% and 40% for NC–E419 and NC–W, respectively, for 3 m water depth. The improvements on the420 operational capacity for S2are very narrow and only perceptible for NC–W,421 where a slight increase of ≃2% is observed. The operativity for vessels with422 drafts above 5 m is 0 using neither NC–W or NC–E.423 On the other hand, results for S4highlight the improvement of the op-424 20 Max. erosion (%) Max. accretion (%) Operativity, 3m (%) Strategy NC–W NC–E NC–W NC–E NC–W NC–E S12 0 12 4.4 60 40 S22 0 10 4.6 60 42 S31 0 10 4.4 60 40 S40 0 4 3.5 80 100 Table 4: Summary of the results after the evaluation of strategies. erativity for both NC. In the case of NC–E, the operativity for 3–m water425 depth increases from 60% to 80%, whereas it rises from 40% to 100% for426 NC–W. Hence, with the shoal removal the design vessel is able to navigate427 during the complete period of the simulation. Moreover, vessels with drafts428 over 6 m are able to navigate during certain time windows (high water lev-429 els at spring tides) of the simulation period. According to these results, S4 430 was chosen as the most efficient strategy, providing not only a significant431 improvement in terms of navigational capacity, but also a reduction of mor-432 phodynamic changes, thus increasing its life–time. The results of the analysis433 of the different dredging strategies is summarized in Table 4.434 FIGURE 7435 4.2. Analysis of the shoal removal436 We performed longer morphodynamic simulations to analyze in detail how437 S4evolves in terms of operativity. The simulations were also performed for438 S1to compare with the unaltered strategy, and spanned a complete year to439 analyze the influence of the entire range of climate conditions. With the aim440 21 of performing realistic simulations, the selected period starts on October 1st,441 2014, approximately the date for which the bathymetry of S1was obtained.442 Fig. 8 shows the sea levels and wind and wave climates during the sim-443 ulated period. Between November and March, storms were frequent with444 maximum wave heights typically over 3 m (both from the East and West)445 and wind velocities up to 20 m/s. However, during spring and summer milder446 conditions were recorded, with wave heights barely reaching 1.5 m and pre-447 dominantly approaching PUI from the Southwest, and wind velocities usually448 below 10 m/s.449 FIGURE 8450 4.2.1. Morphodynamic evolution451 Fig. 9 a and b show the non–dimensional bed level differences between452 the initial and final bathymetries for each strategy. A general sedimentation453 over PUI (approximately 8% of the initial water depth) is observed, except454 for the shallow water area located at the East of the river mouth, where455 erosion is observed. This general sedimentation, which was negligible for the456 results on section 4.1, agree with the tendency of the inlet to be filled up457 with sediments described by Reyes-Merlo et al. (2017).458 The results in the area where the NC diverge are clearly different for S1 459 and S4: whereas consecutive areas of important sedimentation (up to 40%)460 and erosion (20%) are obtained for S1, the removal of the shoal reduces sig-461 nificantly the morphodynamic activity of the area, with only sedimentation462 in the area close to the NC–W bend. These differences are also observed463 in Fig. 9c, where the bathymetry changes along the NC are depicted: the464 maximum water depth reduction is decreased from 40% (S1) to 12% (S4).465 22 Furthermore, for the latter the bed level variations are smoothed not only466 over the shoal, but also upstream the area usually dredged.467 FIGURE 9468 4.2.2. Operativity469 The consequences of these differences on bed level evolution between S1 470 and S4were quantified in terms of operativity. Fig. 10 shows the operativity471 during the complete year for both strategies. A significant increase for both472 NC is observed for S4, specially in the case of NC–W, where the results for473 3 m draft improves from 65% to 100%. Although these increases were also474 observed for the 3–month simulations (section 4.1), some differences arise:475 whereas the operativity for 3 m draft along NC–E is approximately 80% af-476 ter the 3–month simulations, it increases up to 90% for the complete year.477 These differences demonstrate that the operativity is not entirely dependent478 on the initial bathymetry (and hence on the dredging strategy) but also on479 the local wave climate and the subsequent morphodynamic evolution. This480 motivates the analysis of the relation between wave climate and morphody-481 namic evolution performed in section 5.1.482 FIGURE 10483 To complete the analysis, the operativity over the entire PUI was also484 obtained for the complete year considering a draft of 3 m. Fig. 11a shows485 how the shoal at the PUI mouth significantly reduces the operativity for the486 unaltered conditions (S1), with values slightly over 50% in some areas. The487 effects of the shoal removal are evident for the area between the NC Fig.488 11b, where the operativity is almost 100%. Fig. 11c depicts the differences489 between both strategies showing that the operativity improvement is gener-490 23 alized along the area between the NC, being only reduced in small and very491 shallow areas at the eastern margin of PUI. In these areas, small variations492 in water depths results in large reductions in operativity due to the limited493 initial water depth.494 FIGURE 11495 5. Discussion496 In this section, we use the 1–year simulations for S1and S4to discuss497 (1) the relation between the drivers and the morphological evolution of the498 area; (2) the effects of the shoal removal on PUI hydrodynamics and (3) the499 potential environmental impacts derived from the shoal removal.500 5.1. Relation of tides and wave climate with morphodynamic evolution501 To deepen the understanding of the morphodynamic behavior of PUI, the502 shoal removal, and the loss of operativity, we analyze the relation between503 tides and wave climate with the erosion/deposition patterns. Because this504 kind of analysis is difficult due to the complex interactions between waves,505 tides, wind and sediment, we tried to isolate the role of waves and tides on506 the morphodynamic evolution of PUI.507 Fig. 12 shows the monthly variations of the water depth along the NC,508 including their net values and the monthly–averaged wave power vector ob-509 tained at PUI mouth. The deposition/erosion values (panels a and c) were510 obtained as the accumulated mean deposition/erosion rates along the NC for511 each month, whereas their net variation is showed in panels b and d. The512 24 wave power vector was obtained as:513 −→ P=1 8ρg−→ H2cg(2) where the over–arrow indicates vector, ρis the water density, gis gravity,514 −→ His the significant wave height vector averaged in a monthly scale and cgis515 the group celerity.516 Results show that the bed elevation along the NC is closely related with517 both the intensity and orientation of wave power. Generally, the larger the518 intensity and the westerly the direction of the wave power, the larger mor-519 phodynamics changes along the NC are found. The wave directionality plays520 an important role: although the norm of the wave power vector is higher521 in February, larger morphodynamic variations are found for April and May,522 which are more westerly oriented. Considering the orientation of the coast,523 results show that the expected bathymetry changes are more intense for wave524 power vectors with higher obliquity. The importance of wave directionality525 on coastal morphodynamics was described in previous studies, such as L´opez-526 Ruiz et al. (2015), although its importance on navigational channel capacity527 was not previously addressed. These type of analyses are increasingly rel-528 evant, since one of the main impacts of future climate change is expected529 to be the variation on wave climate directionality (Fernandino et al., 2018),530 although no reliable information about the local future tendencies are avail-531 able.532 FIGURE 12533 25 The averaged tidal prism is reduced, which is a proxy of the inlet ten-672 dency of reaching its morphodynamic equilibrium, in accordance with673 the reduction of the bed level changes over the inlet observed for this674 strategy.675 •Results show that shoal removal is an effective solution for navigable676 inlets in which periodic dredging works are carried out due to the pres-677 ence of ebb shoals. The similarity between PUI and many others inlets678 located United States (Dabees and Kraus, 2008; Buonaiuto and Kraus,679 2003) or Europe (Garel et al., 2014, 2015; Garel, 2017) enhance the680 applicability of the findings described in this work, specifically for sci-681 entists and coastal managers dealing with operational, financial and/or682 environmental issues derived from channel deepening.683 Acknowledgments684 This work was partially funded by the European Union through FEDER685 founding (G-GI3002/IDII), project DRAGAPORT of the Programa Oper-686 ativo FEDER 2007-2013. We also thank the staff and researchers of the687 Agency of Public Works of the Andalusia Regional Government for their688 dedication and professionalism. Miguel A. Reyes-Merlo, Pedro Ot´ı˜nar and689 Rafael J. Bergillos are acknowledged for their help during the field campaign690 and the necessary data. Two anonymous reviewers and Filip Tack (Associate691 Editor) are also acknowledged for their comments and suggestions which im-692 proved significantly the manuscript.693 32 References694 ´ Alvarez, M., Carballo, R., Ramos, V., Iglesias, G., 2017. An integrated695 approach for the planning of dredging operations in estuaries. Ocean En-696 gineering 140, 73–83.697 Anthony, E.J., 2004. Sediment dynamics and morphological stability of estu-698 arine mangrove swamps in sherbro bay, west africa. Marine Geology 208,699 207–224.700 Bales, J.D., Holley, E.R., 1989. Sand transport in Texas tidal inlet. Journal701 of Waterway, Port, Coastal, and Ocean Engineering 115(4), 427–443.702 Barba-Brioso, C., Fern´andez-Caliani, J., Miras, A., Cornejo, J., Gal´an, E.,703 2010. Multi-source water pollution in a highly anthropized wetland sys-704 tem associated with the estuary of Huelva (SW Spain). Marine Pollution705 Bulletin 60, 1259–1269.706 Blott, S.J., Pye, K., Van der Wal, D., Neal, A., 2006. Long-term mor-707 phological change and its causes in the Mersey Estuary, NW England.708 Geomorphology 81, 185–206.709 Bolla Pittaluga, M., Tambroni, N., Canestrelli, A., Slingerland, R., Lanzoni,710 S., Seminara, G., 2015. Where river and tide meet: The morphodynamic711 equilibrium of alluvial estuaries. Journal of Geophysical Research: Earth712 Surface 120 (1), 75–94.713 Booij, N., Ris, R.C., Holthuijsen, L.H., 1999. A third-generation wave model714 for coastal regions: 1. Model description and validation. Journal of Geo-715 physical Research: Oceans 104 (C4), 7649.716 33 Bosboom, J., Reniers, A.J.H.M., Luijendijk, A.P., 2014. On the perception717 of morphodynamic model skill. Coastal Engineering 94, 112–125.718 Boyd, S.E., Limpenny, D.S., Rees, H.L., Cooper, K.M., 2005. The effects of719 marine sand and gravel extraction on the macrobenthos at a commercial720 dredging site (results 6 years post-dredging). ICES Journal of Marine721 Science 62, 145–162.722 Brown, B., Le Tissier, M., Scoffin, T., Tudhope, A., 1990. Evaluation of the723 environmental impact of dredging on intertidal coral reefs at Ko Phuket,724 Thailand, using ecological and physiological parameters. Marine Ecology725 Progress Series , 273–281.726 Buonaiuto, F.S., Kraus, N.C., 2003. Limiting slopes and depths at ebb-tidal727 shoals. Coastal Engineering 48, 51–65.728 CEDEX, 2013. Estudio de la din´amica litoral, defensa y propuesta de mejora729 en las playas con problemas: Estudio de actuaci´on del tramo de costa com-730 prendido entre las desembocaduras de los r´ıos Guadiana y Guadalquivir.731 Technical Report (in Spanish). Centro de Estudio y Experimentaci´on de732 Obras P´ublicas , Gobierno de Espa˜na.733 Cheng, R.T., Casulli, V., Gartner, J.W., 1993. Tidal, residual, intertidal734 mudflat (TRIM) model and its applications to San Francisco Bay, Califor-735 nia. Estuarine, Coastal and Shelf Science 36, 235–280.736 Dabees, M., Kraus, N., 2008. Cumulative effects of channel and ebb shoal737 dredging on inlet evolution in southwest Florida, USA. In: Proceedings738 34 of the 31st International Conference on Coastal Engineering, Hamburg,739 Germany, , 2303–2315.740 D’Alpaos, A., Lanzoni, S., Marani, M., Rinaldo, A., 2009. On the O’Brien–741 Jarrett–Marchi law. Rendiconti Lincei 20, 225–236.742 D’Alpaos, A., Lanzoni, S., Marani, M., Rinaldo, A., 2010. On the tidal743 prism–channel area relations. Journal of Geophysical Research: Earth744 Surface 115(F1).745 Davis, R., Fitzgerald, D., 2004. Beaches and coasts. Oxford: Blackwell. 419746 pp 300.747 Dias, J., Lopes, J., 2006a. Calibration and validation of hydrodynamic, salt748 and heat transport models for R´ıa de Aveiro lagoon (Portugal). Journal749 of Coastal Research , 1680–1684.750 Dias, J.M., Lopes, J., 2006b. Implementation and assessment of hydrody-751 namic, salt and heat transport models: the case of R´ıa de Aveiro Lagoon752 (Portugal). Environmental Modelling & Software 21, 1–15.753 Dias, J.M., Sousa, M., Bertin, X., Fortunato, A., Oliveira, A., 2009. Numer-754 ical modeling of the impact of the Anc˜ao Inlet relocation (R´ıa Formosa,755 Portugal). Environmental Modelling & Software 24, 711–725.756 Dissanayake, P., Wurpts, A., 2013. Modelling an anthropogenic of a tidal757 basin evolution applying tidal and wave boundary forcings: Ley Bay, East758 Frisian Wadde Sea. Coastal Engineering 82, 9–24.759 35 Duong, T.M., Ranasinghe, R., Walstra, D., Roelvink, D., 2016. Assessing760 climate change impacts on the stability of small tidal inlet systems: Why761 and how? Earth-science reviews 154, 369–380.762 Dyer, K.R., 1995. Sediment transport processes in estuaries. volume 53, pp.763 423–449. Elsevier.764 Eelkema, M., Wang, Z., Stive, M., 2012. Impact of Back-Barrier Dams on765 the Development of the Ebb-Tidal Delta of the Eastern Scheldt. Journal766 of Coastal Research 285, 1591–1605.767 Egbert, G., Erofeeca, S., 2002. Efficient inverse modeling og barotroptic768 ocean tides. Journal of Atmosphere Ocean Tech. 19, 183–204.769 Elias, E., Hansen, J., 2012. Understanding processes controlling sediment770 transports at the mouth of a highly energetic inlet system (San Francisco771 Bay, CA). Marine Geology 345, 207–221.772 Erftemeijer, P., Riegl, B., Hoeksema, B., Todd, P., 2012. Environmental773 impacts of dredging and other sediment disturbances on corals: a review.774 Marine Pollution Bulletin 64, 1737–1765.775 Erftemeijer, P.L., Robin Lewis, R.R., 2006. Environmental impacts of dredg-776 ing on seagrasses: A review. Marine Pollution Bulletin 52, 1553–1572.777 Fernandino, G., Elliff, C.I., Silva, I.R., 2018. Ecosystem-based management778 of coastal zones in face of climate change impacts: Challenges and inequal-779 ities. Journal of Environmental Management 215, 32–39.780 36 Fortunato, A.B., Nahon, A., Dodet, G., Pires, A.R., Freitas, M.C., Bruneau,781 N., Azevedo, A., Bertin, X., Benevides, P., Andrade, C., Oliveira, A.,782 2014. Morphlogical evolution of an ephermeral tidal inlet from opening783 to closure: The Albufeira inlet, Portugal. Continental Shelf Research 73,784 49–63.785 Garel, E., 2017. Efficient dredging strategy for channel maintenance of the786 Guadiana ebb–delta, in: Proc. of Coastal Dynamics 2017, Helsingor, Den-787 mark.788 Garel, E., Sousa, C., Ferreira, 2015. Sand bypass and updrift beach evolution789 after jetty construction at an ebb-tidal delta. Estuarine, Coastal and Shelf790 Science 167, 4–13.791 Garel, E., Sousa, C., Ferreira, ´ O., Morales, J.A., 2014. Decadal morphological792 response of an ebb-tidal delta and down-drift beach to artificial breaching793 and inlet stabilisation. Geomorphology 216, 13–25.794 Geyer, W., Trowbridge, H., Bowen, M., 2000. The Dynamics of a Partially795 Mixed Estuary. Physical Oceanography 30, 2035–2048.796 Goodwin, C.R., 1987. Tidal-flow, circulation, and flushing changes caused797 by dredge and fill in Tampa Bay, Florida. Technical Report. Dept. of the798 Interior, U.S. Geological Survey,.799 Hansen, J., Elias, E., List, J., Erikson, L., Barnard, P., 2013. Tidally in-800 fluenced alongshore circulation at an inlet-adjacent shoreline. Continental801 Shelf Research 56, 26–38.802 37 Harff, J., Graf, G., Bobertz, B., 2009. Dynamics of natural and anthropogenic803 sedimentation (dynas). Journal of Marine Systems 75, 315–316.804 Hinwood, J., McLean, E., 2018. Tidal inlets and estuaries: Comparison of805 Bruun, Escoffier, O’Brien and attractors. Coastal Engineering 133, 92–105.806 I.E.O., 1992. (Instituto Espa˜nol de Ocenaografr´ıa) Variaci´on espacio–807 temporal de par´ametros f´ısico-qu´ımicos y biol´ogicos en la R´ıa de Huelva y808 ´area de influencia, en el per´ıodo 1987–1991 (in Spanish). Internal report809 no 138:103.810 Iglesias, G., Carballo, R., 2009. Seasonality of the circulation in the R´ıa de811 Muros (NW Spain). Journal of Marine Systems 78, 94–108.812 Iglesias, G., S´anchez, M., Carballo, R., Fern´andez, H., 2012. The TSE in-813 dex e A new tool for selecting tidal stream sites in depth-limited regions.814 Renewable Energy 48, 350–357.815 Je, C.H., Hayes, D.F., Kim, K.S., 2007. Simulation of resuspended sedi-816 ments resulting from dredging operations by a numerical flocculent trans-817 port model. Chemosphere 70(2), 187–195.818 Jonge, V.N.D., 1992. Tidal flow and residual flow in the Ems Estuary. Es-819 tuarine, Coastal and Shelf Science 34, 1–22.820 Knowles, N., Cayan, D.R., 2004. Elevational dependence of projected hy-821 drologic changes in the San Francisco estuary and watershed. Climatic822 Change 62, 319–336.823 38 Kraus, N.C., 2000. Reservoir model of ebb-tidal shoal evolution and sand824 bypassing. Journal of Waterway, Port, Coastal, and Ocean Engineering825 126, 305–313.826 Lesser, G., Roelvink, J., Van Kester, J., Stelling, G., 2004. Development and827 validation of a three-dimensional morphological model. Coastal Engineer-828 ing 51, 883–915.829 Lewis, M.A., Weber, D.E., Stanley, R.S., Moore, J.C., 2001. Dredging impact830 on an urbanized Florida bayou: Effects on benthos and algal-periphyton.831 Environmental Pollution 115, 161–171.832 L´opez-Ruiz, A., Solari, S., Ortega-S´anchez, M., Losada, M., 2015. A simple833 approximation for wave refraction–application to the assessment of the834 nearshore wave directionality. Ocean Modelling 96, 324–333.835 Luijendijk, A.P., Ranasinghe, R., de Schipper, M.A., Huisman, B.A.,836 Swinkels, C.M., Walstra, D.J., Stive, M.J., 2017. he initial morphological837 response of the Sand Engine: A process-based modelling study. Coastal838 Engineering 119, 1–14.839 Meng, X., Jiang, X., Li, Z., Wang, J., Cooper, K.M., Xie, Z., 2018. Re-840 sponses of macroinvertebrates and local environment to short-term com-841 mercial sand dredging practices in a flood-plain lake. Science of the Total842 Environment 631-632, 1350–1359.843 Militello, A., Kraus, N.C., 2001. Re-alignment of an inlet entrance channel844 by ebb-tidal eddies, in: Coastal Dynamics’01, pp. 423–432.845 39 Montero, N., Belzunce-Segarra, M.J., Del Campo, A., Garmendia, J.M., Fer-846 rer, L., Larreta, J., Gonz´alez, M., Maidana, M.A., Espino, M., 2013. Inte-847 grative environmental assessment of the impact of Pasaia harbour activities848 on the Oiartzun estuary (southeastern Bay of Biscay). Journal of Marine849 Systems 109-110, S252–S260.850 Morales, J.A., Borrego, J., Davis, R.A., 2014. A new mechanism for chenier851 development and a facies model of the Salts Island chenier plain (SW852 Spain). Geomorphology 204, 265–276.853 Mu˜noz, F.R., Gonz´alez-Regalado, M., Flores, J.B., Morales, J., 1997. The854 response of ostracod assemblages to recent pollution and sedimentary pro-855 cesses in the Huelva Estuary, SW Spain. Science of the Total Environment856 207, 91–103.857 Murphy, A., 1988. Skill scores based on the mean square error and their858 relationships to the correlation coefficient. Monthly Weather Review 116,859 2417–2424.860 Nienhuis, J.H., Ashton, A.D., Nardin, W., Fagherazzi, S., Giosan, L., 2016.861 Alongshore sediment bypassing as a control on river mouth morphodynam-862 ics. Journal of Geophysical Research: Earth Surface 121, 664–683.863 Oberle, F.K., Storlazzi, C.D., Hanebuth, T.J., 2014. Wave-driven sediment864 mobilization on a storm-controlled continental shelf (northwest iberia).865 Journal of Marine Systems 139, 362–372.866 O’Brien, M.P., 1967. Equilibrium flow areas of tidal inlets on sandy coasts.867 Coastal Engineering 1966 , 676–686.868 40 Ray, R.D., 2001. Inversion of oceanic tidal currents from measured elevations.869 Journal of Marine Systems 28, 1–18.870 Reniers, A.J., Roelvink, J., Thornton, E., 2004. Morphodynamic modeling871 of an embayed beach under wave group forcing. Journal of Geophysical872 Research: Oceans 109(C1).873 Reyes-Merlo, M. ´ A. Tintor´e-Parra, A., D´ıez-Minguito, M. andOrtega-874 S´anchez, M., Losada, M.A., 2015. Morphodynamic evolution and influ-875 ence of dredging activities in small–scale mesotidal estuaries: the case of876 Punta Umbr´ıa (Southwestern Spain). In Proceedings of the 36th IAHR–877 International Association for Hydro–Environment Engineering and Re-878 search World Congress 100.879 Reyes-Merlo, M.A., Ortega-S´anchez, M., D´ıez-Minguito, M., Losada, M.A.,880 2017. Efficient dredging strategy in a tidal inlet based on an energetic881 approach. Ocean and Coastal Management 146, 157–169.882 Ris, R.C., Holthuijsen, L.H., Booij, N., 1999. A third-generation wave model883 for coastal regions: 2. Verification. Journal of Geophysical Research:884 Oceans 104(C4), 7667.885 Rodr´ıguez-Ram´ırez, A., 2008. The impact of man on the morphodynamics of886 the Huelva coast (SW Spain)/Efectos antr´opicos en la morfodin´amica de887 la costa de Huelva (SO Espa˜na). Journal of Iberian Geology 34, 313–327.888 Roman, C.T., Peck, J.A., Allen, J., King, J.W., Appleby, P.G., 1997. Accre-889 tion of a New England (USA) salt marsh in response to inlet migration,890 41 Figure 3: Bottom elevation observed (a) and computed (b) and differences of the bottom elevation between the initial and final steps of the simulation for the measured (c) and modeled (d) data. Red and blue colors indicate erosion and sedimentation, respectively. 48 Figure 4: Strategies tested: a) S1, initial configuration of PUI, b S2, channel deepening, c) S3, littoral drift barrier and d) S4, shoal removal. 49 0 2 4 6 Wave Height [m] 0 5 10 15 20 Peak Period [s] 0 100 200 300 400 Wave Direction [º] Jan10 Feb10 Mar10 -2 -1 0 1 2 Water Level [m] -0.2 0 0.3 0.6 Storm Surge [m] -15 0 15 -15 0 Wind Velocity [m/s] (3) (2) (1) Figure 5: From top to bottom: Panel 1 shows the peak period and wave direction at SIMAR 1052048 in blue and red lines, respectively (Fig. 1). Panels 2 and 3 represent the wave height and wind velocities measured off the coast at SIMAR 1052048 (Fig. 1). Panel 4 represents the water level and the storm surge measured at REDMAR 3329 (Fig. 1) in blue and red line, respectively. The red dashed boxes indicate three different phases in terms of wave climate identified: (1) sequence of storms, (2) milder wave energy conditions, and (3) consistent storm conditions. 50 Figure 6: Non–dimensional bed level differences between the initial and the end of the simulations (blue = sedimentation): (a) S1, (b) S2, (c) S3and (d) S4. The third column represents the differences of water depth between the end and the beginning of the simulation for NC–W (e) and NC–E (f). Negative values represent sedimentation. The NC are marked with solid lines in panels a-d. 51 0 0.2 0.4 0.6 0.8 1 Operativity S1 NC-W NC-E S2 0 2 4 6 Depth [m] 0 0.2 0.4 0.6 0.8 1 Operativity S3 0 2 4 6 Depth [m] S4 a) b) c) d) Figure 7: Operativity (non–dimensional ratio of navigable hours per year) for each NC for S1(a) to S4(d). The NC are plotted in panel e. 52 0 2 4 6 Wave Height [m] 0 10 20 Peak Period [s] 0 100 200 300 400 Wave Direction [º] Sep14 Oct14 Nov14 Dec14 Jan15 Feb15 Mar15 Apr15 May15 Jun15 Jul15 Aug15 Sep15 Oct15 -2 0 2 Water Level [m] 0 1 Storm Surge [m] -18 -9 0 9 18 Wind Velocity [m/s] Figure 8: From top to bottom: Panel 1: Peak period and wave direction (SIMAR 1052048, Fig. 1) are represented in blue and red line, respectively. Panels 2 and 3 show, respectively, the wave height and wind velocities (SIMAR 1052048, Fig. 1). Panel 4 represents the water level and the storm surge measured at REDMAR 3329 (Fig. 1) in blue and red line, respectively. 53 Figure 9: Non–dimensional bed level differences between initial and final bathymetries of PUI (blue colors indicate sedimentation): a) S1, b) S2and c) results along NC–W (solid line) and NC–E (dashed line). The NC are marked with solid lines in panels a and b. 54 01234567 Depth [m] 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 Operativity Figure 10: Operativity along NC–W (green) and NC–E (red) for S1and S4(solid and dashed lines, respectively). The inset represents the location of NC and initial bathymetry of S1. 55 Figure 11: Operativity over PUI for a 3 m draft vessel: a) S1, b) S4and c) non–dimensional differences between S1and S4. 56 Figure 12: Monthly variations in cumulative sedimentation/erosion along the NC and monthly–averaged wave power vector at the PUI mouth. Panels a and c (b-d) depicts the cumulative mean deposition/erosion and the net water depth changes along NC–W (NC–E) on a monthly basis. Panels e and f show the wave power vector for each month. 57