Accepted Manuscript

Tidal impact on suspended sediments in the Macuse estuary in

Mozambique

Lucas Lavo António Jimo Miguel, João Wagner Alencar Castro, Fialho
Paloge Juma Nehama

PII: S2352-4855(17)30004-X
DOI: http://dx.doi.org/10.1016/j.rsma.2017.07.002
Reference: RSMA 261

To appear in: Regional Studies in Marine Science

Received date : 17 December 2016

Revised date : 5 July 2017
Accepted date : 8 July 2017

Please cite this article as: Miguel, L.L.A.J., Castro, J.W.A., Nehama, F.P.J., Tidal impact on
suspended sediments in the Macuse estuary in Mozambique. Regional Studies in Marine Science
(2017), http://dx.doi.org/10.1016/j.rsma.2017.07.002

This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to
our customers we are providing this early version of the manuscript. The manuscript will undergo
copyediting, typesetting, and review of the resulting proof before it is published in its final form.
Please note that during the production process errors may be discovered which could affect the
content, and all legal disclaimers that apply to the journal pertain.
 1 TIDAL IMPACT ON SUSPENDED SEDIMENTS IN THE MACUSE ESTUARY IN
2 MOZAMBIQUE
3
4 1, 2, 3
Lucas Lavo António Jimo Miguel; 1, 3 João Wagner Alencar Castro; 2 Fialho Paloge Juma Nehama;
5 1
Programa de Pós-graduação em Geologia PPGl–UFRJ, Brasil.
6 2
Escola Superior de Ciências Marinhas e Costeiras – ESCMC/ Universidade Eduardo Mondlane, Moçambique.
7 3
Laboratório de Geologia Costeira, Sedimentologia e Meio Ambiente - LAGECOST / Museu Nacional–UFRJ
8 Corresponding author e-mail: lucaslavomiguel@yahoo.com.br
9
10 Abstract
11 Many studies around the world have found that tidal dynamics control the transport of suspended
12 sediments in estuaries, as well as their relationship with runoff discharges and bathymetry
13 morphology. This manuscript examines tidal impact on suspended sediments in the Macuse estuary
14 in Mozambique, southern Africa. Data was collected at 42 stations, including tidal current
15 measurements with an Acoustic Doppler Current Profiler (ADCP), tide elevation with a tidal gauge
16 and suspended sediments with a Niskin bottle. A 14-month model simulation was implemented
17 beginning in November 2013, forced by tides, wind regime and runoff discharges. The results
18 indicated a tidal elevation of 4 m that generated tidal currents of 120 cm s-1. These currents
19 combined with the runoff of 500 m3 s-1 and bathymetry configuration, drove the suspended
20 sediments during the ebb and flood tides. Suspended sediment concentrations suggested the
21 occurrence of sedimentation where water flow was less than 0.10 cm s-1 for silt and less than 10 cm
22 s-1 for fine sand. Results of silt concentrations was about ~ 300 mg l-1 while fine sand was found to
23 be 0.2 mg l-1 in both tidal periods in ebb and flood tides during the rainy and dry seasons. The
24 model accuracy of about R2 = 0.88 between the observed and simulated results demonstrated the
25 ability to predict the impact of tides and runoff discharges in the Macuse estuary and may be
26 extended to other estuaries. The model matches in reproducing the sediment transport mechanism
27 will help making political decisions about the estuary sustainability and local coastal management.
28
29 Keywords: Tidal currents; Suspended sediments; Sedimentation; Macuse Estuary.
30
31
32
33
34
35
1
36 Highlights
37 1- The tidal regime of 4 m in height generates tidal currents of 120 cm s-1 between the ebb and
38 flood tides;
39 2- Silt sedimentation occurred with 0.10 cm s-1 and fine sand was found to be 10 cm s-1;
40 3- Besides tidal currents, runoff discharge and wind energy has controlled the mixing in both
41 rainy and dry seasons;
-1
42 4‐ The observed and predicted sediment indicated silt concentration of 300 mg l in rainy
-1
43 season and 200 mg l in dry season. 
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79

2
 80 1. Introduction
81 Estuaries at land-sea interfaces are critical to the functioning of coastal systems. Their dynamics
82 are caused primarily by a combination of tidal elevation and storm surges, changes to the wave
83 climate, and high runoff discharges which intensify during rainy seasons (Quinn et al., 2014).
84 Estuarine flooding or ebbing caused by tidal elevation combined to storm surge and to extreme
85 river flow can increase flood hazards in many estuaries (Maskell et al., 2013). Tidal elevation and
86 runoff discharges may change daily and seasonally, increasing flood risk, intensifying tidal current
87 dynamics, tidal asymmetry and the supply of suspended sediments. The sediment supply can
88 produce highly productive environments that may serve as important routes through which
89 terrestrial material enters into the ocean (Simpson et al., 2001). These materials are highly
90 influenced by environmental dynamics that change their sedimentological, chemical and biological
91 structure in different spatial and temporal scales (Valle-Levinson, 2010). Many studies confirm that
92 these changes may be caused by the combination of tidal dynamics and runoff discharges that
93 control the transport of suspended sediments and their morphology.
94 Several studies have indicated that potentially negative impacts of estuarine changes caused by
95 tidal changes and river flooding might include, sediment supply that may damage infrastructure and
96 population displacement (Hanson et al., 2011; Nicholls et al., 2011), increased erosion and loss of
97 land to the sea (Zhang et al., 2004), pollution and increased human health risk, and significant loss
98 of wetland habitats and ecosystems (Day et al., 1995; Nicholls et al., 1999). In this case, any
99 modelling process involving environmental factors might bring output results into questions.
100 Different researches about the impact of tides on the dynamic of suspended sediments in estuaries
101 are found around the world (Biggs, 1990; Gelfenbaum, 1983; Fain et al., 2001; Patchineelam and
102 Kjerfve, 2004; Grabemann and Krause, 2001) in which the principles were extended to the East
103 coast of Africa (Liu, 2014; Kitheka et al., 2005); Nehama, 2012; Chevane et al., 2016; Maskell et
104 al., 2013, Van der Lubbe, et al., 2014). These studies analyzed the importance of developing
105 modelling approaches to predict suspended sediments, water discharges and tidal flooding risk,
106 which will help assess impacts and support mitigation decisions for port construction.
107 Decadal increase of suspended sediment supply from both the river and sea may cause depth
108 changes and flood risk due to the lower depths caused by sediment deposition over long time scales
109 (Miguel, 2013). Specifically in East Africa, Liu (2014), Kitheka et al. (2005), Nehama (2012),
110 Chevane et al. (2016) provided various results of coastal processes in the region and have helped

3
111 provide better undeerstanding oof tidal dynaamics and ruunoff dischaarges. The effect
e of theese factors oon
112 suspendded sedimennts modifiess not only estuary
e morrphology buut also lighht penetratioon as well aas
113 affects navigability
n y and harborr logistics. Meanwhile, there is noo studies inn Mozambiqque that havve
114 focusedd on the imppact of tidall elevation and river diischarges onn suspendedd sedimentss in estuariees
115 stronglyy dominatedd by tides. T
This study exxamines thee impact of ttides on susspended seddiments at thhe
116 Macuse estuary inn Central M
Mozambiquue, where iis located a harbour with regionnal strategiic
117 importaance for centtral Mozambbique and laandlocked ccountries in the interior.
118
119 1.1. Stu
udy site
120 Locaated in Zam
mbezia Provvince in cenntral Mozam
mbique, the Macuse river estuary sserves as onne
121 of the m
major outletss for discharrge of Nam mated to aveerage 500 m3
macurra Riveer runoff, whhich is estim
122 s-1 durinng the rainyy season (Fiig. 1). The Macuse flooodplain hass been usedd for rural ddevelopmennt,
123 particulaarly rice cultivation and a reecent port project byy the Moozambican governmennt.
124 Intensifi
fication of aagriculture and
a the portt project maay increase sediment suupply to thee Namacurrra
125 River ass well as allter its tidall regimes annd thus cauuse immensee changes inn the bathyymetry of thhe
126 river channel.

127
128 Fig 1. Macuuse estuary wiithin the conteext of southeasst Africa in M
Mozambique. M M1 is the entraance boundaryy, while the
129 Namacurra River
R and righht tributary are the upstream boundaries.
4
130 Previous studies find that the combination of transgressions and regressions and the sediment
131 supply have formed several tidal plains characterized by beaches with very fine and fine sand and
132 several mangrove fields that extend up to about 40 km inland to the Namacurra district. These fields
133 are composed of several species of typical mangrove vegetation and an abundance of very fine
134 material (mud, clay and silt), and are colonized by populations of benthos and molluscs.
135 Moore et al. (2008) state that the regional climate is tropical humid according to the Köppen
136 scale and is affected by the warm Mozambique current. This current has a strong effect on the
137 annual seasons according to studies by Ramsay (1995), Armitage et al. (2006), Halo et al. (2014),
138 Lutjeharms et al. (1988a). Two annual seasons are quite evident in the region - summer from
139 September to March and winter from April to August. Annual rainfall averages roughly 1000 mm
140 year-1 with annual average temperatures of 24 oC (Moore et al., 2008). Maximum rainfall and
141 Macuse riverine runoff occur during the austral summer between January and March when the
142 Intertropical Convergence Zone (ITCZ) and associated rainfall belts reach their southernmost
143 position (Moore et al., 2008). The rainfall have fed runoff discharges that reach maximum
144 discharges of 800 m3 s-1 between February and March.
145 The region has weak nebulosity and prevailing winds from the east and the southeast. The
146 monthly 10 m winds vary between 3.5 m s-1 and 5.5 m s-1 with the predominating frequencies being
147 from the NE, E, SE, S and SW (Langa, 2007, Miguel, 2013). Although the wind intensities are
148 relatively weak, the local winds are the main factor responsible for generating waves averaging 1.5
149 m in height. The wave regime is predominantly from the southeast and is influenced by SE winds
150 that generate the south-north longshore water current (Lutjeharms et al., 1988a). This longshore
151 water current is responsible for the sediment transportation and coastal hydrodynamic circulation
152 along the Mozambique coastline. According to the National Institute of Hydrography and
153 Navigation of Mozambique - INAHINA (2013), the region has a tidal range of about 4 m during
154 spring tides, which generates tidal currents of about 120 cm s-1, that may cause erosion and
155 transport of particulate suspended sediments.
156
157
158
159
160
161
5
162 2. Materials and methods
163 2.1.1. Synoptic data collection
164 We set up locations of 42 stations along the estuary that were identified in the field with GPS-
165 Garmin, to get correct coordinates on the base map. The stations were selected to ensure greater
166 coverage of the study area, determining longitudinal and transversal vertical sections. In November
167 2013 and November 2014, was conducted a field survey with CTD profilers (conductivity,
168 temperature and depth sensors) and a tidal gauge to collect synoptic data, including water
169 temperature, tidal currents, tidal elevation, water salinity, suspended sediments and water turbidity.
170 Salinity units are in PSU or g Kg-1, however in this research have been considered unitless.
171 Suspended sediments were measured by sampling 3.5 liters of water in a Niskin bottle. The
172 collected sediments were filtered using 0.47 μm porosity fiberglass, 0.45 mm in diameter.
173 The meteorological data (from 2000 to 2013) were acquired from the National Meteorological
174 Institute of Mozambique (INAM), which routinely collects precipitation, air temperature and their
175 derivatives at a dense network of stations. Hydrological data were acquired from the National Water
176 Directorate of Mozambique for the period between 2000 and 2013 at the Licungu River at station
177 91 in Mocuba district, which continuously records water heights.
178
179 2.1.2. ELCOM module
180 The Estuary Lake and Coastal Ocean Model (ELCOM) was developed at the University of
181 Western Australia. ELCOM is a numeric model that uses thermodynamic and hydrodynamic
182 principles to simulate the spatial and temporal variations of physical, chemical, biological, and
183 geological parameters in natural water bodies. This particular model was chosen because studies by
184 Hipsey and Hamilton (2008), Miguel (2013), Hodges and Dallimore (2013), Hodges and Dallimore
185 (2007), Hodges et al. (2000) have found that ELCOM is a strong and reliable tool for predicting
186 physical, biological and chemical parameters.
187 Indeed, the ELCOM model is a tool used to predict oceanographic parameters in natural water
188 bodies subjected to external environmental forces. These environmental forces are usually wind
189 stress, tides, precipitation, surface heating or cooling as well as inflows and outflows of water. The
190 model uses hydrostatic pressure assumptions based on the Navier-Stokes equations for
191 incompressible flow (Hodges and Dallimore, 2007). Besides the hydrostatic assumption for
192 pressure, the Boussinesq approximation and Reynolds averaged transport equations (1, 2 and 3) are

6
193 implemented (Hodges and Dallimore, 2013). The modeled and simulated processes include
194 baroclinic and barotropic responses, rotational effects, tidal forcing, wind stresses, surface thermal
195 forcing, inflows, outflows and transport of salt, heat and passive scalars.
u u u u p  
196  u  v  w    2v sin   2w cos    x   b 1
t x y z x z x

v v v v p  y 
197  u  v  w    2u sin     b 2
t x y z x z y
w w w w p
198 u v w   g 3
t x y z z
1  n
199 Where :   .  o is the water density. are the winds at 10m height, and n stands for the east
o z
200 and north wind components. P is water pressure.  is a local latitude. To simulate a Macuse
201 scenario, the Coriolis force was neglected in the simulation. Meanwhile, the water pressure, tidal
202 currents water density, wind regime, water temperature and salinity were used in the model.
203
204 2.1.3. CAEDYM module for suspended sediments
205 The ELCOM model is the main physical driver that controls the coupled CAEDYM model. It
206 provides spatially explicit, three-dimensional simulations of current and water temperature patterns
207 based on processes including baroclinic and barotropic responses, rotational effects, tidal forcing,
208 wind stress, surface thermal forcing, water inflows and outflows, passive scalars and sediment
209 modules in all classes based on the Wentworth-1922a classification. The physical predictions
210 generated, such as those for temperature, tidal elevation, salinity and tidal currents, establish a
211 physical context for the use of the CAEDYM model to generate spatially explicit prediction
212 processes of biological, chemical and geological parameters in different time scales (Zhao et al,
213 2009). The CAEDYM structure allows incorporating a submodule for suspended sediments, which
214 generates spatial and explicit simulations of temporal variations. In this research submodule of the
215 suspended sediment simulation is based on the calculation of suspended particles (inorganic
216 particles) (equation 4) as proposed by Hipsey and Hamilton’s (2008).
SS s v    cs SS s  sed 1
217  s SS s  S s
t Z  ref K ss s  SS s  sed Z bot 4
218 In which, SSs is the concentration of inorganic suspended solids for group s (g m−3); vs is the settling

7
219 velocity (m s−1 ); ∆z is the grid or layer thickness,  S s is the resuspension rate parameter in g m−2 s
−1 ;
220 τ is the shear stress (N m−2 ), τcs is the critical shear stress for group s; τref is a reference shear
221 stress (1 N m−2 ); K ss controls the effect of sediment limitation on resuspension (g); and SSs−sed is
s

222 the sediment SS mass of group s (g). According to Hipsey and Hamilton (2008), a critical shear
223 stress parameter τcs can be estimated from the dimensionless critical shield’s parameter, which in
224 turn is estimated from the dimensionless particle diameter D. Parameter D may be estimated as:

G 1
225 D  Ds s 3 g 5

226 Then G is the specific gravity of the re-suspended particles, g is gravitational acceleration (9.8 m s-2
227 ),  is the kinematic viscosity (10−6 m2 s−1 at 20 oC) and DS is the particle size. The critical shield
s

228 parameter was calculated according to value of D:

0.5 tan   D  0.3


0.25 D tan   0.3  D  19
 0 .6

D
0.013 D tan   19  D  50
229 0 .4 6
0.06 tan   D  50

230 In which α is the angle of repose of the sediment from 30o to 42o. The critical shear stress is then

231 estimated from the equation,  cs   ( s   m ) Ds;  s   s g is the specific sediment weight; and m =
232  m g = 9800 N m−3 is the specific weight of the water-sediment mixture. Based on Stoke’s Law the
233 settling velocity is calculated by using the particle density,  ss and diameter Dss :

S s   w (T , S )
234 vs  g ( Ds ) 2 7
18 (T ) s

235 Where  w is water density calculated dynamically as a function of temperature and salinity; and μ is
236 the dynamic viscosity of water calculated internally as a function of temperature.
237
238 2.1.4. Model implementation
239 2.1.4.1. Pre-processing
240 The model configuration for the Macuse estuary was implemented according to the Hodges and
241 Dallimore (2003) scheme (Fig 2). The numerical code of the ELCOM was written in the Fortran-90
242 language. The diagram summarizes the key variables for the module execution, including pre-
243 processing, model execution and visualization. The configuration of suspended sediment variables
8
244 was bassed on the C
CAEDYM m
module undeer initial connditions andd their preparation thatt consider thhe
245 geochem
mical and bbiological variables. Thhe data defi
fined in ELC
COM - CA
AEDYM weere processeed
246 simultanneously for the producttion of Netw
work Comm
mon Data Foorm (NetCD
DF) files in three types -
247 vertical,, horizontal curtain andd mesh proffiles. The poost-processiing phase coorresponds tto the resultts
248 in NetC
CDF format that were converted aand visualizzed throughh Fortran-strructured com
mmands annd
249 Matlab.

250
251 Fig 2. Scchematic diagrram for ELCO M module (Hoodges and Dalllimore, 2013)).
OM - CAEDYM

252 m totaling 600 km2 at a 1:10000 scale

A map s (1998--99 at GCS--WGS-19844, Tete 36) w
was digitizeed
253 with MIIKE21 from
m a bathymeetric chart obbtained by IInstituto Nacional de H
Hidrografia e Navegaçãão
254 de Mozaambique - IN
NAHINA. T
The bilinearr interpolation techniquue was implemented wiith the modeel
255 algorithhm providedd automaticaally by the M
Mike21. Altthough this ttechnique iss not more aaccurate thaan
256 the trianngular interppolation tecchnique, thee errors did not affect ssignificantlyy the specifiic purpose oof
257 this studdy. Therefoore, for the bathymetryy accuracy we updatedd bathymetrry points w
with the fielld
258 measureements condducted with an echo-soounder Garm
min 150. The final interrpolated batthymetry waas

9
259 configured with a mesh grid of 20 m x 20 m (Fig 4), with 230 x-rows and 190 y-columns and 10
260 vertical layers. The ELCOM geographic position was identified as -0.707107 north-x and +0.70711
261 north-y, which corresponds to 17.82oS and 37.07oE. The mesh grid was converted into an ELCOM
262 - CAEDYM configuration within four open boundary conditions, including: 1-ocean boundary in x-
263 row 230 and between y-column 80 and y-column 129 and at all layers; 2-runoff flow to the estuary
264 in row 1 and between columns 112 and 120; 3-lateral runoff flow between rows 42 and 55 in the
265 first column; and 4-lateral river flow between rows X = 171 and Y = 176 in y-column 61. In the two
266 upstream boundaries (2 and 3), we applied the water discharge forces and sediment concentrations.
267 In the downstream boundary 1 (ocean boundary), the tidal elevation from the Model-tpxo7 with a
o
268 global resolution of 1/4 x 1/4o between January 2000 and December 2014 was implemented
269 (www.coas.oregonstate.edu/research/po/research/tide/index.html). These included eleven tidal constituents:

270 M2, S2, N2, K2, K1, O1, P1, Q1, MF, MM and M4, which began at 00:00, 01/01/2014 (17.76o S
271 and 37.22 oE). Their contributions are presented in table 1.
272 Table 1. Tidal constituents at water level boundary 1.

Constituent Amplitude (m) Phase (o)

M2 1.1001 41.05
S2 0.6328 82.59
N2 0.1862 25.80
K2 0.1784 79.10
K1 0.0315 354.23
O1 0.0499 345.74
P1 0.0092 10.20
Q1 0.0135 321.14
MF 0.0121 2.57
MM 0.0058 355.49
M4 0.0085 334.63
273
274 2.1.4.2. Initial conditions
275 Initial conditions were implemented at the three boundaries (BC-1, BC-2, and BC-3) to execute
276 the model, including water level, water temperature, water salinity, runoff and fine sand and silt
277 concentrations (Table 2). At all of the boundaries, constant values of water salinity and water
278 temperature were defined, and every volume of water in the computational domain was considered
279 to be quiescent and vertically well mixed. A time interval of 120 seconds was defined for the
280 simulation, for which 210000 iterations were performed, and the output data were recorded hourly.

10
281
282 Table 2. The initial conditions at the boundary interfaces. Where BC-1 represents the Macuse estuary entrance, BC-2
283 the right tributary and BC-3 the Namacurra River.
Model boundary
Parameter of Initial condition BC-1 BC-2 BC-3
Initial uniform water level (m) 2.6 1.0 1.0
Initial uniform temperature (o) 29 29 29.2
Initial uniform salinity (PSU) 29 29 27.5
-1
Initial uniform sand concentration (mg l ) 0 0 0
-1
Initial uniform silt concentration (mg l ) 0 0 0
3 -1
Initial uniform water runoff (m s ) 0 300 500
284
285 2.1.4.3. Model execution
286 The model was configured to run a simulation of 14 months, which included the summer and
287 winter seasons. This was due to significant variations in seasonal conditions that may have relevant
288 impact on the runoff discharges, sediment supply, wind force and solar radiation. These factors may
289 affect the water temperature, water salinity, biological activity and estuarine and ocean circulation.
290 Regarding the simulation applied to the Macuse estuary, a spin-up period of 5 days was
291 implemented according to the model stability to guarantee the initial hydrodynamic conditions
292 (Table 2). A time step of 60 seconds was adopted to ensure numerical stability with 555,000 time
293 steps, with hourly saving of output data. The model was run for one month for spin-up verification
294 and then performed a one-year run, which was completed after two weeks of simulation. The
295 stabilization time was calculated according to the equation:
t
296 g 'D  2 8
x
297 In which g´ is equal to 100 (maximum parameter estimated for shallow water), D = 15 m (average
298 depth of the estuary) and x = 20 m (corresponding to the dx size of the bathymetry). The model
299 stability constant was estimated using the equation:
 tU
300 CFL  9
x
301
302
303

11
304
305 3. Ressults
306 3.1. Evaluatingg model ressults
307 To eensure the aaccuracy of tthe simulateed tidal elevvation withinn the modell domain, thhe model waas
308 calibrateed and validdated using data obtainned from a tidal gauge and CTD profilers
p fixxed at statioon
309 M3 (thee port zone). In this casse, the waterr levels werre reproduceed excellenttly, indicatinng the abilitty
310 of the nuumerical model to desccribe the feaatures of thee tidal signaal such as phhase, amplittude, and ebbb
311 to floodd tides, andd the springg to neap variation
v off the tidal rrange. The results of this analysiis
312 model reprodduced well the tidal eleevation of 4 m, tidal cuurrents of 1220 m s-1, annd
indicateed that the m
313 salinity variation between
b 34 and 35 (Figg. 3). Thesee excellent m
matches bettween the oobserved annd
314 modelleed data, connfirm the acccuracy of thhe model, caapturing botth the tidal amplitudes and the tidaal
315 i a significcant correlaation of R2 = 0.86, wheereas the coorrelation off observed aand modeleed
phases in
316 salinity was estimatted at R2 = 00.79. These significant correlationss indicate thhe accuracy of the modeel
317 perform
mance in thee Macuse R
River estuaary for bothh neap andd spring tiddes for all conservativve
318 parametters.

319
320 Fig. 3. A - Comparisonn between the observed andd modelled tidees. B - Compaarison betweenn the observedd and simulateed
321 salinity.
322 For sediment verification,
v the comparrison betweeen the obseerved and simulated seediments waas
323 conductted at statioons M1, M33 and M5 (T
Table 3, Figg. 4). Both observed aand the moddelled resultts
324 matchedd well with R2 = 0.88. Moreover,, it is imporrtant to notte that the 888 % accuraacy obtaineed
112
325 indicatees the robusstness of thee model peerformance for satisfacttorily reprooducing dataa outputs foor
326 suspendded sedimennts.
327 Tabble 3. Observeed and simulatted suspendedd sediments annd their correlaation on 16 annd 17/11/2013 in Fig. 4.

328 Stations Depths (m) S (mg l-1)

Obs. SS Siimulated SS ( mg l-1)
329 M5 1 111 120
330 M5 5 159 170
331 M5 10 2240 250
332 M3 1 100 110
333 M3 5 168 160
334
M3 10 661 130
M1 3 3301 250
335
336 M1 7 3307 260
337 M1 12 2212 170
338
339

340
341 Fiig. 4. Compariison between tthe observed aand simulated suspended sediments
342
343 Speccific evaluaations indicaated that thee tidal regim
me with a m
maximum heeight of 4 m had the rolle
344 m velocity oof 120 cm s-1
of generrating the tidal currentss (Fig. 5). Thhese tidal cuurrents withh a maximum
345 reachedd their peak after the flood tides inn both the spring
s and nneap tides. IIt is importtant to noticce
346 that thee estuary geeometry deffined the U-velocity
U a the mainn factor thaat controls the
as t estuarinne
347 o 70 cm s-1 at station M
circulatiion with maaximum tidaal currents of M3 (Fig. 6). The ebb annd flood tidaal
348 currentss had a stronng correlatioon with the suspended sediments, which weree estimated at R2 = 0.911.
113
349 This coorrelation is highly significant aand reflecteed in the relative low
w suspendded sedimennt
350 concenttrations of 1110 mg l-1 during ebb tides, while flood tidees were chaaracterized bby relativelly
351 o 160 mg l-1.
high-susspended seddiment conccentrations of

352
353 Fig. 5. A – Tidal elevaation and U - velocity.
v B-T
Tidal elevationn and suspendeed sediments. C – Correlatioon between U-
354 velocity and simulated susspended sedim
ments.
355 3.2. Synoptiic data
356 The 1:10000 sscale bathym
metry interppolated witth Mike21 demonstrated a configguration thaat
357 indicateed shallow ddepths in thhe entrance region
r and relatively deep
d depths along the river (Fig. 6).
358 The EW
W river orienntation com
mbined with the geomorrphology, peermitted thee U tidal currrents flux tto
359 be stronnger than thee V-velocityy componennt. Accordinng to the deppths presentted, site A ddemonstrateed
360 that it iss narrower and deeper (14 m deptth) than B which
w explaains why it is favorablee for the port
361 construcction. Despiite the miniimum depthh of 2 m at the
t river enntrance, the 4 m tidal reegime allow
ws
362 water deepths of 6 m at flood tiddes, which hhas set miniimum vessel draft for navigation.
n
363 Partticular analyyses are focuused on bedd level variaations beyonnd the vicinnity of the Macuse
M Riveer
364 and the proposed aareas (A andd B) for the port constrruction. Speecial attentioon was giveen to the areea
114
365 of the tiidal entrance, which is very imporrtant due to its effects oon sedimentt transport bby U-velocitty
366 (Fig. 100). Indeed, tthe U-velocity variationn between --20 cm s-1 aand 80 cm s-1 is responnsible for beed
367 level vaariations caaused by hyydraulic eroosion. The erosion proocess is thee primary factor
f in thhe
368 channel deepening of the Nam
macurra Riveer and sedim
mentation att its entrancce. This sediimentation iis
369 very stroong at the riiver entrancce due to thee weak U-veelocity intennsity, thereffore, it beginns to weakeen
370 along thhe channel to the Macuse River. Thhe lower depths found at the entrannce to the M
Macuse Riveer
371 give waay to the prim
mary tidal fflow, eventuually becom
ming a depossitional areaa due to sediiment supplly
372 from thee rivers.

373
374 Fig. 6. Innterpolated batthymetry usingg the bilinear m
method in the Mike21 modeel. A and B inndicate the prooposed sites foor
375 port construcction.
376 The bathymetryy defined haas different water fluxees accordinng to the sloope and geoometry of thhe
377 riverbedd. Thereforee, the wateer dischargees are the ssecond facttor that conntributes too the water’s
378 current flux and sediment
s traansport com
mbined withh the sedim
mentation oor erosion pprocess. Thhe
379 maximuum runoff ddischarges of 850 m3 s-11 in March ffed by averaage rains off 1000 mm yyear-1 have a
380 strong rrole in the ssupply and transport off sediment and the seddimentation that accum
mulates at thhe
381 entrancee of the Maacuse Riverr. This runooff is dilutedd by ocean water durinng the rainyy season annd
382 leaves the
t Macusee River witth two watter types annd in a quiite turbulennt state. Inddeed, in sittu
383 observaations of tiddal free-surfface elevatioon data indicated that high tidal aand low tiddal elevationns
384 drove thhe tidal currents and sallinity distribbution alongg the estuaryy.
115
385 A cclose associiation betweeen the tiddal elevationn and tidal current was observedd during thhe
386 transitioon between the ebb annd flood tiddes and vice-versa. T
There was a strong corrrespondencce
387 betweenn salinity vaariation and daily tidal variation measured
m bettween 07:000:00 hours aand 09:17:000
388 hours frrom 16 to 17 Novembeer/2014, witth salinity ffound to be 35.9 at floood tides andd 35.5 at low
w
389 tides. The observedd in situ tem
mperature averaged
a 299 oC with noo expressivve changes in
i the periood
390 from 077:00:00 houurs to 09:17:00 hours on 16 andd 17 Noveember 20144. Field obsservations oof
391 turbidityy concentrattion indicateed relativelyy high valuees at the enttrance of rouughly 160 N
NTU and 1220
392 NTU att the catchhment. In thhe entrancee zone, theese concenttrations inddicated the strong tidees
393 influencce, which arre associatedd with bottoom roughnesss, the waterr velocity annd sedimentt size.
394
395 3.3. Moodel output data
396 The model outpputs confirm model for repproducing laarge-scale tidal patterns,
m the accuraacy of the m
397 tidal currrents, salinnity variationn, temperatuure changes and suspennded sedimeents. The tiddal range of 4
398 m provides good documentedd informatiion to explaain the imppact of thee tides on tthe estuarinne
399 hydrodyynamics studied. For exxample, at stations
s M1,, M3 and M
M5, the wateer temperatuure and wateer
400 salinity correlated well to expplain the tw
wo possible water typees in the T--S diagramss. These tw
wo
401 water tyypes includde the oceaan water wiith salinity changes beetween 34 and 36 - A
A, while thhe
402 freshwaater and bracckish water had salinityy of less thann 34 - B.

403
404 Fig. 7-a. T
The annual T--S diagrams (aa, b and c) andd the relation between
b saliniity and suspennded sedimentts (d, e, f) at thhe

116
405 defined sttations M1, M
M2 and M3. B - Indicates the brackish waater and runoff
ff discharges. A - Indicates tthe ocean wateer
406 type.
407 The average annnual salinitty at all stattions (Fig. 77-b), exhibitted increasinng salinity profiles witth
408 depth vaarying betw
ween 34.6 annd 34.9 at M
M1, 32.4 andd 34 at M3,, and between 32.5 andd 33.5 at M55.
409 The chaanges presented make it quite eviident that M
M1 had salttier water thhan M3 andd M5, whicch
410 presenteed fresh watter dischargges that dilutted the oceaan water to bbrackish waater. Meanw
while, verticaal
411 temperaature profilees with deptth decreasedd from the ssurface to thhe bottom cchanging at M1 betweeen
412 23.20 oC and 23.4 oC, at M3 beetween 20.88 oC and 21.3 oC and at M5 betweeen 21.6 oC annd 22.3 o C..

413
414 Fig. 7--b. Average annnual salinity and temperatuure profiles at stations M1, M2
M and M3.
415 Baseed on the pproposed eqquation 7, we c s-1, and a
w identifiedd a silt-settlling velocitty of 0.10 cm
416 fine-sannd settling velocity
v of about
a 10 cm s-1 or less. T
These veloccities explaiin why the ffine sand waas
417 usually transportedd by dragginng mechanics on the riiverbed whiile the silt rremained inn suspensionn.
418 Indeed, the Macusse geomorphhology indiicated that U-velocity is the mainn factor thaat drives thhe
117
419 hydrodynamics of the Macuse estuary. The velocities presented in Fig. 8 indicate a maximum of 20
420 cm s-1 in the rainy season, which is higher than the settling velocity for both fine sand and silt. The
421 maximum velocity was reached during the spring tides in the ebb period with a maximum of 120
422 cm s-1. Minimum velocity of 0 cm s-1 was detected during the transition tides between ebb and flood
423 periods. Figure 8-a) and b), makes it quite evident that zero velocities are found at the bottom layers
424 suggesting proper conditions for fine sand and silt settling.
425 In both rainy and dry seasons, it is evident that the mixing energy was driven by water current
426 patterns. In response, strong mixing occurred when the water speed was over 10 cm s-1, and
427 possible stratification when water speed was less than 10 cm s-1 for fine sand and 0.10 cm s-1for silt.
428 Indeed, for example, the fine sand was usually transported on the bottom layers by dragging, while
429 the silt fluxes were distributed at all water layers with concentrations ranging between 180 mg l-1
430 and 260 mg l-1. During the flood tide, due to the lower river intake and increased tidal elevation, the
431 suspended sediment (SS) concentrations reached between 180 mg l-1 and 300 mg l-1. Smaller SS
432 concentrations were observed in the dry season, influenced by the modulation of the local tide,
433 which ranged between 180 mg l-1 and 200 mg l-1. During ebb tides, the highest SS concentrations
434 occurred near the entrance of the estuary and was influenced by the resuspension process.
435 The rising tide increased the SS concentration from the inner estuary to the harbour zone. In the
436 transition between stations M1 and M5, the U-velocities drop expressively because of the
437 converging movements between the river discharge and tidal propagation (Fig. 8-a, Fig. -b). It is
438 evident that the main source of silt and fine sand is the Namacurra River. The water discharge from
439 the Namacurra combined with tidal currents drives silt and fine sand downstream in the estuary. In
440 addition, this combination is responsible for the advective transport of suspended matter, water
441 salinity, and water temperature in the Macuse estuary (Fig. 8-a, Fig. 8-b).
442 It was observed that in the rainy season, the transport of salt was predominantly associated with
443 runoff from the Namacurra River and the right tributary, forcing the salt water to the entrance and
444 demonstrating the strong fluvial influence during the rainy season. In the dry season, the runoff was
445 low and the salt water was transported upstream in the estuary to the catchment zone. The
446 downstream transport during the ebb tides the water salinity had values close to 35. Salt
447 accumulation at the catchment may result between May and December, when was observed a
448 negative water balance and salinization oriented upstream (from the ocean to the Namacurra River).
449 The transport of salt associated with the river discharge was more strongly influenced by tides at all
450 the stations in the dry season due to the reduced storage of freshwater during the tidal change. The
18
451 simulateed temperatture data inddicates thatt the warmeer water cam
me from thee river, whiile the coldeer
452 and salttier water m
moved from the ocean tto the river. The combbination of ttemperaturee and salinitty
453 ma-t) betweeen 20 kg m-3 and 24 kkg m-3, whiich suggest a mixture oof
generateed water deensities (sigm
454 more occean water tthan runoff. The mixturre was expressive due tto the oceann water retenntion-time oof
455 5 days.
456 w about 300x10-6 J annd
In bboth the rainy and dryy seasons, thhe wind ennergy at thee surface was
457 reachedd bottom layyers at abouut 3 m, whiich contribuuted to wateer mixing and
a suspendded sedimennt
458 m was aboutt 10-4 J in thhe rainy andd dry seasonns, while thhe
transporrt. The sheaar energy att the bottom
459 shear ennergy along the water ccolumn was about 200xx10-6 J in booth the rainyy and dry seeasons. Thesse
460 parametters had a sttrong role inn the water mixing
m in both seasons due to the ebb and floood tides. Thhe
461 dissipatiion energy was
w about 11J with highh values obseerved at thee entrance.

462
463 Fig. 8-a. Typiical results of simulated dataa during the raainy season in 2014.

119
464
465 Fig. 8-b. Typpical results off simulated datta during the ddry season in 2014.
2
466 The water salinnity distribution reflectss the combiination of alll the transpport processes, includinng
467 density changes, sttratification and mixingg. These proocesses in tturn controll density cirrculation annd
468 modify the water--column miixing. Turbbulent mixiing plays a critical roole in deteermining thhe
469 stratificaation and residual circculation of Macuse esttuary. Meannwhile, the suspended silt and finne
470 sand conncentrationss are resultss of the mixxing processs and hydroodynamics, iin which thhe spatial annd
471 time disstribution is controlled by persistennt tidal channges. In this case, the maximum
m c
concentratio
on
472 predicteed was abouut 300 mg l-1 with higgh concentrrations durinng spring tides
t and cooncentrationns
473 lower thhan 100 mg l-1 during ebb
e tides in neap tides ((Fig. 9-a annd 9-b). Thee results inddicated stronng
474 currentss where the bathymetryy presented a steep topoographic sloope that usuually projects suspendeed
475 sedimennts downwarrd or leftwaard of site A (see Fig. 4)).
476 m s-1 and 1220
The 4 m high ttides were rresponsible for the absoolute speed changes beetween 0 cm
477 cm s-1 (F
Fig. 9-a andd 9-b). The periodic speeed changess were stronng during sppring tides aand relativelly
478 weak during
d neapp tides. Thee role of thhese speedds ware refl
flected in thhe suspendded sedimennt
479 concenttration in thee water coluumn during the rainy annd dry seasoons.

220
480
481 Fig. 9-a. Short timee serious of finne sand (a, b, cc), silt (d, e, f) and U-velocitty (g, h, i ), in the rainy seasson

482
483 Fig. 9-b. Short tim
me serious of fiine sand (a, b, c), silt (d, e, ff) and U-veloccity (g, h, i), inn the dry seasoon
221
484 In the transition period from the neap to spring tides, the concentration of suspended sediments
485 changed from a 100 mg l-1 to 270 mg l-1. The separated investigation of the rainy and dry seasons
486 facilitated a more thorough analysis of suspended particles, which allowed us to identify and
487 quantify the variation of several physical parameters correlated to sediment transport and their
488 characteristics. In this case, for the sake of simplicity, results from the rainy season indicated a
489 strong correlation between the suspended sediments and tidal oscillation. This was observed in the
490 high concentration of both fine sand and silt during the spring and floodtides (Fig. 9-a and Fig. 9-b).
491 The fine sand flux was not observed expressively in the simulation because of the high weight and
492 relatively high density. Despite the relatively unexpressive concentration of the fine sand in the
493 rainy and dry seasons, low concentrations were observed with a maximum of 0.15 mg l-1. During
494 the dry season, there were strong changes in the suspended sediments with relatively lower
495 concentrations than that of the rainy season. The changes were caused with no significant
496 discharges from the main boundary rivers and were immediately reflected in the low concentrations
497 of suspended sediments.
498 Pronounced changes in concentration between the rainy (in February) and dry season (in June)
499 were observed in our analysis (Fig. 9-a, Fig. 9-b). At station M3, silt concentrations were relatively
500 low and unexpressed fine sand concentrations were observed in the two seasons. The results reflect
501 a strong relationship between the variation in the tidal currents and direct changes in the suspended
502 sediment concentrations. While the runoff discharges are observed seasonally with relatively large
503 concentrations in the rainy season between January and May and relatively low concentrations
504 during the dry season between July and December (Fig. 9-a, Fig. 9-b). More generally, these results
505 illustrate the intimate relationship between the tidal flow and sediment transport patterns of the two
506 groups (silt and fine sand). In response, may have a role in the morphodynamics that may be
507 expected from any alteration of the Macuse estuary bathymetry.
508 Specific results in February illustrated that the estuary geomorphology and topography
509 configuration may control the U-velocity, which governs the suspended sediment in the Macuse
510 estuary. This process was confirmed at station M3 where the maximum concentration of sand was
511 6x10-6 mg l-1, while 8x10-3 mg l-1 was observed at station M1 and 0.10 mg l-1 at station M1. These
512 concentrations indicate each station’s dependence on the main sources of sediment and tidal energy.
513 Silt concentration values between 220 mg l-1 and 261 mg l-1 were observed at station M5 with the
514 U-velocity ranging between -21 cm s-1 and 20 cm s-1. The concentration was about 200 mg l-1 and
515 240 mg l-1 when the U-velocity was between -50 cm s-1 and 50 cm s-1 at M3. Moreover, at the
22
516 U-velocity bbetween -500 cm s-1 annd 50 cm s--1 was regisstered silt concentratio
station M1 with U c on
517 betweenn 170 mg l-1 and -230 mg
m l-1. Both results
r for silt and fine sand presennted differennces betweeen
518 Februarry and June and an intim
mate relationn with the driving
d U-veelocity and rrunoff dischharges.
519 Verttical profilees of suspennded fine sannd at stationns M1, M3 and M5 inccreased withh depth from
m
520 the surfface to the bbottom withh a maximuum of 0.0155 mg l-1 (Fiig. 10). Thiis suggests the dragginng
521 transporrt process oon the riverbbed by tidal currents coombined with the river slope that increases thhe
522 gravity water flux. Contradictoory to the fiine sand, the silt decreaased with depth from thhe surface tto
523 the botttom at the M
M1, M3 annd M5 statioons. At stattion M1, thee annual avverage silt concentratio
c on
524 changedd between 179 mg l-1 annd 118 mg l-1. At statioon M3 the silt
s changedd between 2004 mg l-1 annd
525 214 mg l-1. At M5 the
t annual aaverage silt concentratioon changedd between 2008 mg l-1 annd 212 mg l--1.
526 These aannual profiiles suggestt that silt annd fine sannd sources aare the Nam
macurra Rivver and righht
527 tributaryy that decreaase expressiively their concentratio
c ons towards the river enntrance.

528
529 Fig. 10. Annuual vertical prrofiles of fine ssand and silt aat M1, M3 andd M5.
223
530 4. Results discussion
531 The field and modeled data outputs agreed that there is a strong relationship between tidal
532 oscillation and tidal currents, as well as with suspended sediments. Tidal currents, gravity flow, and
533 river discharge were previous identified by Dias et al. (2016) as the three forces responsible for the
534 advective transport of salt in estuarine systems. The results obtained in this research indicated that
535 might cause additionally the advective transport of suspended sediments (silt and fine sand).
536 The tidal range of 4 m in height proved that the Macuse estuary is dominated by tidal currents
537 with sufficient tidal prism that permits the river entrance maintain against longshore and cross-
538 shore wave-driven littoral sediment transport. Despite the mesotidal regime found in Macuse
539 estuary, the results agree with those of Cooper (2001) in microtidal estuaries of Republic of South
540 Africa. Furthermore, the magnitude of the tide and the runoff discharges determined the
541 concentration of suspended sediments closer to the catchment and entrance. Thus, the higher the
542 magnitude of the tides or runoff discharges, the higher the suspended sediments concentration
543 during the dry or rainy seasons. These results are in good agreement with those of Dias et al.
544 (2016), Kitheka et al. (2005) and D’Aquino et al. (2010) in similar systems.
545 The unidirectional flows of U-velocity indicated that in the dry season, marine waters have a
546 greater influence on the estuary system, which results in flood tides that are higher than ebb tides at
547 stations M1, M3 and M5. Dias et al. (2016), Stanev et al. (2007) and Dias et al. (2013b) recently
548 experienced same related results in similar environments. The U-velocity results suggest that the
549 Macuse estuary is well mixed and the ocean water is transported upstream predominantly via tidal
550 propagation. The stationary U-velocity pattern in the estuary revealed the fluvial influence on the
551 estuarine system, which diminished the water salinity to 30 during the rainy season between
552 January and March. Dias et al. (2013b) state that positive values for U-velocity during the rainy
553 season may transport freshwater downstream. This process was found at the Macuse estuary where
554 the freshwater was exported by a unidirectional U-velocity with a maximum of 80 cm s-1. The
555 transport of runoff from the Namacurra River in the rainy season generated positive current
556 velocities (downstream from the estuary) during both the flood and ebb tides. However, in the dry
557 season, the flood velocities indicated the entrance of ocean water with salinity of about 35 (Fig. 8).
558 Field results revealed expressive suspended sediment concentrations with a maximum of 307
559 mg l-1, while the predicted suspended sediment concentrations indicated a maximum of about 300
560 mgl-1. In fact, the concentration of suspended sediments changed cyclically as a function of tidal

24
561 variation and seasonal runoff discharge volume. These concentrations revealed a vertical light
562 attenuation caused by the concentration of suspended sediments of 300 mg l-1 and runoff discharges
563 from the main boundaries. On the other hand, the turbidity between 120 NTU and 160 NTU
564 revealed the attenuated light penetration in the water column, which corroborates the explanation
565 given by (Bowers et al., 2004); Mills et al. (2002); Painting et al. (2007); Devlin et al. (2008) and
566 Corbett (2010), who found that opaque particles in the water block the light penetration.
567 Seasonal suspended sediments results indicated relatively higher suspended sediment
568 concentrations in the rainy season than in the dry season. The tidal modulation combined with river
569 discharge influenced the highest suspended concentrations during the rainy season. While the low
570 suspended concentrations during the dry season were influenced by the tidal modulation. These
571 results are similar to those reported for other similar estuaries (Kitheka et al., 2005; Schettini, 2003;
572 Cooper, 2001). The deposition and resuspension was related to cyclic changes in tidal currents (Fig.
573 8), which corroborate conclusions of Brockmann and Dippner (1987), Boon and Duinevel (1996),
574 Stanev et al. (2007), Schettini (2003), Shi et al. (2003) and Chen et al. (2005) in similar
575 environmental conditions. The suspended particulate matter and resuspension matter from the
576 bottom was usually caused by tidal changes, which agrees with the findings of Tian et al. (2009) in
577 a similar environment. Tidal currents are the main controlling factor that governs the water
578 circulation and re-suspended sediments in the Macuse estuary. Because of the tidal currents, the
579 fine sand settled quickly at the main boundaries because of the weight, viscosity, gravity, density
580 and bottom roughness. In contrast, silt was transported significant distances in the water column.
581 These results agreed with those of Hurzeler et al. (1995) in a similar study, who concluded that the
582 suspended sediment deposition changes the buoyancy of the inlet flow density and changes the
583 background flow to the surface over a relatively short distance. In this case, the tidal dynamics
584 provide strong evidence of their role in the transport of suspended sediments during the spring and
585 neap tides of about 300 mg l-1.
586 Various studies around the world (Dias et al., 2016; Cooper, 2001; Brockmann and Dippner,
587 1987; Kitheka et al., 2005; Tian et al., 2009; Boon and Duinevel , 1996; Stanev et al., 2007; Shi et
588 al., 2003; Chen et.al., 2005) agree that sediment transport in estuaries is governed by tidal dynamics
589 and tidal asymmetry, plus their relationship with the runoff discharges and geomorphology.
590 However, it is worth noting that other factors may control sediment transport such as bed
591 roughness, margin friction, sediment density and fluid viscosity. This may raise questions about the
592 scale of uncertainty from the modeling of complex natural environments such as the Macuse
25
593 estuary. Indeed, there are many factors to be considered, including environmental patterns, relative
594 sea-level rise/fall, sediment supply and river discharges in different scales. These factors may be
595 responsible for sediment transport and any modeling process should question the quantities of
596 output data. Despite these issues, general numerical modeling approaches in estuarine systems tend
597 to support analysis, contributing to the understanding of existing processes (e.g. Macuse estuary
598 system).
599 Recent analyses have allowed researchers to increase the accuracy of models used in the
600 planning of estuary development and of changes caused by different development phases. For
601 example, the data results obtained in this study show that the two sites proposed for harbor
602 construction by the Mozambique government have stronger potentialities at site A than B. The
603 suitable oceanographic potential found at site A is supported by the relative insignificant suspended
604 sediment deposition and the 25 m depth of the site. These conclusions sustain the hypothesis that no
605 dredging would be needed at site A over a long time scale, while site B would be strongly affected
606 by the weak current speed of 20 cm s-1 and suspended sediment deposition index. These factors are
607 aggravated by the 4 m depth, which is not suitable for ship navigation. The benefits of
608 understanding the sediment transport and estuarine morphodynamics may extend beyond their
609 importance to the hydrodynamics and bathymetry at the Macuse estuary to policy and political
610 decisions about the estuary’s sustainability and local coastal management.
611
612 5. Conclusions
613 1 - The 4 m high tidal regime generated tidal currents of 120 cm s-1 during the transition period
614 between the ebb and flood tides and vice-versa. The critical silt settling velocity was 0.10 cm s-1,
615 while the fine sand velocity settling speed was found to be 10 cm s-1, suggesting that the suspended
616 sediments were maintained in the water column when the tidal currents were greater than critical
617 velocities.
618 2 - Two water types were detected in both seasons, including water type A (ocean water) with
619 salinity greater than 34 during the dry season and type B (brackish water) with salinity less than 30.
620 Moreover, the wind mixing energy combined with the bottom and shear stress energy were
621 controlling factors in the mixing of the estuary waters in both the rainy and dry seasons.
622 3 - The field results indicated maximum suspended sediment concentrations of 310 mg l-1, which
623 confirmed the model’s prediction of 300 mg l-1, in which the comparison indicated a correction of

26
624 R2 = 0.88. During the rainy season, the suspended sediments had a greater concentration with a
625 maximum of 300 mg l-1, while in the dry season it was about 200 mg l-1. The variation in tidal
626 currents had a lead role in the transportation of suspended sediments during the rainy and dry
627 seasons.
628 4 - The comparison between the observed and simulated data demonstrated the accuracy of the
629 ELCOM-CAEDYM model for predicting the role of the tides and river discharges in the dynamics
630 of suspended sediments in the Macuse estuary, and that it may be applied to studies of other
631 estuaries. The efficacy of the model in understanding the sediment transport mechanism will help
632 making political decisions about the estuary’s sustainability and local coastal management.
633
634 Acknowledgement
635 We are grateful to the MCT – Mozambique for financing this project and to the Eduardo
636 Mondlane University for the lab analyses and the availability of the multi-sensor instruments for
637 the field-data collection. We are grateful to the Lwandle Company of South Africa for the second
638 phase of fieldwork in November 2014. Finally, we would like to thank Dra Kate Munnik for her
639 special attention to the data and her contributions. Moreover, we would like to thank Stella Vaz
640 (Student at PPGl – Federal University of Rio de Janeiro in Brazil) for English grammar corrections.
641
642 6. References
643 Armitage, S. J., Botha, G. A., Duller, G. A. T., Wintle, A. G., Rebêlo, L. P., Momad, F. J., 2006. The formation and
644 evolution of the barrier islands of Inhaca and Bazaruto, Mozambique. Geomorphology 82 (2006) 295 - 308.
645 Biggs, R. B., 1970. Sources and distributions of suspended sediments in Northern Chesapeake Bay. Marine Geology 9,
646 87 - 201.
647 Boon, A., Duineveld, G., 1996. Phytopigments and fatty acids as molecular markers for the quality of near-bottom
648 particulate organic matter in the North Sea. J. Sea Res.35 (4), 279 - 291.
649 Bowers, D., Evans, D., Thomas, D., Ellis, K., Williams, P., 2004. Interpreting the colour of an estuary . Estuar. Coast.
650 Shelf Sci. 59, 13 - 20.
651 Brockmann, C., Dippner, J., 1987. Tidal correction of hydrographic measurements. Ocean Dyn. 40 (6), 241- 260.
652 Chen, S. L.; Zhang, G.; Yang, S. L., Shi, J. Z., 2005. Temporal variations of fine suspended sediment concentration in
653 the Changjiang River estuary and adjacent coastal waters, China. Journal of Hydrology. 331,137-145.
654 doi:10.1016/j.jhydrol.2006.05.013
655 Chen, W., Liu, W., Hsu, M., Hwang, C., 2015. Modeling investigation of suspended sediment transport in a tidal
656 estuary using a three-dimensional model. Taiwan. Applied Mathematical Modelling, 39, 2570 - 2586.
657 doi.org/10.1016/j.apm.2014.11.006

27
658 Chevane, C. M., Penven, P., Nehama, F. P. J., Reason, C. J. C., 2016. Modeling the tides and their impacts on the
659 vertical stratification over the Sofala Bank, Mozambique. African Journal of Marine Science. ISSN: 1814-232X
660 (Print) 1814-2338. DOI: 10.2989/1814232X.2016.1236039
661 Cooper, J. A. G., 2001.Geomorphological variability among microtidal estuaries from the wave-dominated South
662 African coast. Geomorphology 40 (2001). 99-122. PII: S0169-555XŽ01.00039-3
663 Corbett, D. R., 2010. Resuspension and estuarine nutrient cycling: insights from the Neuse River Estuary. Department
664 of Geological Sciences, Institute for Coastal Science and Policy, East Carolina University, Greenville, North
665 Carolina, USA. Published in Biogeosciences. 7, 3289–3300. doi:10.5194/bg-7-3289-2010
666 D’Aquino, C. A., 2010. Processos de transporte e retenção de sedimentos finos em estuários dominados por rios. Tese
667 de Doutorado. UFRGS. Porto Alegre. Brasil.
668 Day, J. W., Pont, D., Hensel, P. F., Ibanez, C., 1995. Impacts of sea-level rise on deltas in the Gulf of Mexico and the
669 Mediterranean: the importance of pulsing events to sustainability. Estuaries 18, 636 - 647.
670 Devlin, M. J., Barry, J., Mills, D. K., Gowenc, R. J., J. Foden, J., Sivyerb, D., Tett, P., 2008. Relationships between
671 suspended particulate material, light attenuation and Secchi depth in UK marine waters. Estuarine, Coastal and
672 Shelf Science. 79 (2008) 429 - 439. doi.org/10.1016/j.ecss.2008.04.024
673 Diagnóstico Preliminar e Acções de Reconstrução Pós-Calamidades - DPARPC, 2013. República De Moçambique
674 Ministério Da Planificação E Desenvolvimento, Maputo, Março de 2013.
675 Dias, F. J. S., Castro, B. M., Lacerda, L. D., 2013a. Continental shelf water masses off the Jaguaribe River, North-
676 eastern Brazil. Cont. Shelf Res. 66, 123 - 135.
677 Dias, F. J. S., Castro, B. M., Lacerda, L. D., Miranda, L. B., Marins, R. V., 2016. Physical characteristics and
678 discharges of suspended particulate matter at the continent-ocean interface in an estuary located in a semiarid
679 region in northeastern Brazil. Estuarine, Coastal and Shelf Science 180 (2016) 258-274. http://dx.doi.org/10.1016/j.ecss.2016.08.006
680 Dias, F. J. S., Marins, R.V., Maia, L.P., 2013b. Impact of drainage basin changes on suspended matter and particulate
681 copper and zinc discharges to the ocean from the Jaguaribe River in the semiarid NE Brazilian coast. J. Coast.
682 Res. 290, 1137-1145.
683 Dippner, J. W. A., 1993. Frontal - resolving model for the German Bight. Continental Shelf Res. 13(1), 49 - 66.
684 Direcção Nacional de Águas Delegação da Província da Zambézia em referência ao DPARH., 2012. Water National
685 Direction at Zambezi Province - Mozambique.
686 Fain, A. M. V., Jay, D. A., Wilson, D. J., Orton, P. M., Baptista, A. M., 2001. Seasonal and tidal monthly patterns of
687 particulate matter dynamics in the Columbia River estuary. Estuaries-24 (5), 770 - 786.
688 Gelfenbaum, G., 1983. Suspended-sediment response to semidiurnal and fortnightly tidal variations in a mesotidal
689 estuary: Columbia River, USA. Marine Geology 52, 39 - 57.
690 Grabemann, I., Krause, G., 2001. On different time scales of suspended matter dynamics in the Weser Estuary.
691 Estuaries 24, 688 - 698.
692 Halo, I., Backeberg,B, Penven, P., Ansorge, I., Reason, C., Ullgren, J., 2014. Eddy properties in the Mozambique
693 Channel: A comparison between observations and two numerical ocean circulation models. Deep Sea Res.,Part
694 II, doi: 10.1016/j.dsr2.2013.10.015, in press.
695 Hanson, S., Nicholls, R., Ranger, N., Hallegatte, S., Corfee-Morlot, J., Herweijer, C., Chateau, J., 2011. A global

28
696 rannking of port ccities with highh exposure to climate extrem
mes. Clim. Chhange 104, 89 -111.
697 Hipsey, M.
M R., Hamiltoon, D. P., 20008. Computational Aquatic Ecosystem Dyynamics Moddel: CAEDYM
M v3. Centre fo
for
698 Waater Research University off Western Austtralia.
699 Hodges, B.,
B Dallimore, C., 2007. Esstuary, Lake aand Coastal O
Ocean Model: ELCOM v2.22. Science mannual Centre fo
for
700 Waater Research,, University off Western Ausstralia. May 233, 62p.
701 Hodges, B
B., Dallimore,, C., 2013. Esttuary, Lake annd Coastal Oceean Model: EL
LCOM v2.2. 2013.
2 Sciencee manual Centrre
702 forr Water Reseaarch, Universitty of Western Australia. Maay 23, 62p.
703 Hodges, B.R., Imbergger, J., Angello Saggio, A.., Winters, K Modeling basiin-scale internnal waves in a
K. B., 2000. M
704 stratified lake. L
Limnol. Oceannogr., 45 (7), 22000, 1603 -16620.
705 Hurzeler, B. E., Ivey, G
G.N., Imbergeer, J., 1995. Sppreading modeel for a turbidiity current witth reversing buuoyancy from
ma
706 connstant-volumee release. Mar.. Freshw. Res.. 46, 393 - 4088.
707 Imberger,, J., Hamblin, B., 1982. Dynnamics of lakees, reservoirs, and cooling ponds. Ann. Reev. Fluid Mechh.14, 153-1877.
708 Imberger,, J., Patterson,, J. C., 1990. Physical limnnology. In: Wuu, T. (Ed.). Addvances in Appplied Mechannics, vol. 27, ppp.
709 3022 - 475.
710 Kitheka, J. U., Obiero M., Nthenge,, P., 2005. Rivver discharge,, sediment traansport and exxchange in thee Tana Estuary
ry,
711 Keenya. Estuarinee, Coastal andd Shelf Sciencee 63 (2005) 4555 - 468. httpss://doi.org/10.1016/j.ecss.20004.11.011
712 Langa, J. V. Q., 2007. Problemas naa zona costeirra de Moçambbique com ênffase para a cossta de Maputoo. Universidadde
713 Edduardo Mondlaane, Moçambiique. Revista dde Gestão Cossteira Integradda. 7 (1) : 33 - 44.
714 Lima, G. M. P., Lessa, G. C., 2001. The Fresh-Water Dischargee in Santos Baay (Ba) and its Significancee to the Generral
715 Waater Circulatioon. Revista de Pesquisas em
m Geociencias, 28 (2): 85 - 996.
716 Liu, X., 22014. Sedimennt dynamics ooff the east A
African contineental margin dduring the lastt deglaciation and Holocenne:
717 Coonstrained by cchanges in clim
mate and sea llevel. Bremen University. D
Dissertation.
718 Lutjeharm
ms, J. R. E., van Ballegoooyen, R. C., 1988a. Anom
malous upstreaam retroflectiion in the Aggulhas Currennt.
719 Science; Washinngton 240.48600 (Jun 24, 19888): 1770
720 Maskell, J., Horsburghh, K., Lewis, M., Bates, P., 2013. Investtigating river--surge interacttion in idealissed estuaries. J.
721 Cooast. Res. 30, 2248 - 259.
722 L. L. A. J., 22013. Estudo da circulaçãoo hidrodinâmicca e transportte de sedimenntos no estuárrio de Macuse:
Miguel, L
723 Coontribuição paara a navegaçção e construção de um pporto, Provínccia da Zambézia - Universsidade Eduarddo
724 Moondlane, Moçaambique. 74ppp.
725 Mills, D. K., Rutgers, M., Laane, R. W. P. M., Rees,
R J. M., 20002. Continuoous measuremeent of suspendded matter. Seea
726 Teechnology 43 ((10), 5.
727 Moore, A Main, M. P. L.,, Williams, H. B., 2008. Thhe Zambezi R
A. E., Cotterilll, F. P. D., M River. In: Guptta, (Ed.), Largge
728 Rivvers: Geomorpphology and Management.
M
729 National IInstitute of Hyydrography annd Navigation of Mozambiqque –INAHINA
A, 2013. Tabeela de marés dde 2013.
730 Nehama, F. P. J., 20122. Modelling the Zambezi River plume using the Reggional Oceaniic Modelling System. Thessis
731 preesented for thee Degree of Doctor
D of Philoosophy. Oceannography Deppartment, Facuulty of Sciencee, University oof
732 Caape Town, Souuth Africa. 1777pp.
733 Nicholls, R. J., Hoozem
mans, F. M. J., Marchand, M.,
M 1999. Incrreasing flood risk and wetlaand losses duee to global seaa-

229
734 level rise: regional and global analyses. Glob. Environ. Change Human Policy Dimens. 9, 69 - 87.
735 Nicholls, R. J., Marinova, N., Lowe, J. A., Brown, S., Vellinga, P., de Gusmão, D., Hinkel, J., 2011. Sea-level rise and
736 its possible impacts given a beyond 4 degrees world in the twenty-first century. Philos. Trans. R. Soc. A 369,
737 161-181.
738 Painting, S. J., Devlin, M. J., Malcolm, S. J., Mills, C., Mills, D. K., Parker, E. R., Tett, P., Wither, A., Burt, J., Jones,
739 R., Winpenny, K., 2007. Assessing the impact of nutrient enrichment in estuaries: susceptibility to
740 eutrophication. Marine Pollution Bulletin 55 (2007) 74 - 90. doi:10.1016/j.marpolbul.2006.08.020
741 Patchineelam, S. M., Kjerfve, B., 2004. Suspended sediment variability on seasonal and tidal time scales in the Winyah
742 Bayestuary, South Carolina, USA. Estuarine. Coastal and Shelf Science 59, 307-318.
743 Quinn, N. D., Lewis, M., Wadey, M., Haigh, I., 2014. Assessing the variability in extreme storm-tide time-series for
744 coastal flood risk assessment. J. Geophys. Res. 119, 4983 - 4998.
745 Ramsay, P. J., 1995. 9000 years of sea-level change along the southern African coastline. Quaternary International,
746 vol.31,71-75.
747 Schettini, A. C. A., 2003. Estudo do regime de correntes e material particulado em suspensão ao largo do estuário do rio
748 Itajaí-açu. Brasil. II Congresso sobre Planejamento e Gestão das Zonas Costeiras dos Países de Expressão
749 Portuguesa.
750 Shi, Z., Kirby, R., 2003. Observations of fine suspended sediment processes in the turbidity maximum at the North
751 Passage of the Changjiang Estuary, China. Journal of Coastal Research.19 (3), 529 - 540.
752 Simpson, J. H., Vennell, R., and Souza, A. J., 2001. The salt fluxes in a tidally - energetic estuary. Estuarine, Coastal
753 and Shelf Science, 52 (1), 131-142.
754 Stanev, E., Brink-Spalink, G., Wolff, J., 2007. Sediment dynamics in tidally dominated environments controlled by
755 transport and turbulence: a case study for the east Frisian Wadden Sea. J. Geophys. Res. 112, 40 -18.
756 Tian, T., Merico, A., Jian-Su, J., Staneva, J., Wiltshire, K., Wirtz, K., 2009. Importance of resuspended sediment
757 dynamics for the phytoplankton spring bloom in a coastal marine ecosystem. Journal of Sea Research 62 (2009)
758 214 - 228. doi:10.1016/j.seares.2009.04.001
759 Valle-Levinson, A., 2010. Contemporary issues in estuarine physics. New York: Cambriges University Press, 315p.
760 Van der Lubbe, J. J. L., Tjallingii, R., Prins, M. A., Brummer, G. A., Jung, S. J. A., Kroon, D., Schneider, R. R.,
761 Sedimentation patterns off the Zambezi River over the last 20,000 years. Marine Geology 355 (2014) 189 - 201.
762 http://dx.doi.org/10.1016/j.margeo.2014.05.012
763 Zhang, K. Q., Douglas, B. C., Leatherman, S. P., 2004. Global warming and coastal erosion. Clim. Change 64, 41-58.
764 Zhao, Y., Jones, M. L., Shuter, B. J., Roseman, E. F., 2009. A biophysical model of a Lake Erie Walleye (Sander
765 vitreus) explains interannual variation in recruitment. Can.J. Fish Aquat. Sci. 66 (2009): 114 -
766 125.doi:10.1139/F08-188

30