A Survey Near Tambara Along the Lower Zambezi River

environments

Article A Survey near Tambara along the Lower Zambezi River

Francesco Pasanisi 1,*, Carlo Tebano 1 and Francesco Zarlenga 2

1 ENEA, Italian National Agency for New Technologies, Energy and Sustainable Economic Development, Sustainable Territorial and Production Systems Department, Models and Technologies for Mitigation of Human Impacts and Natural Hazards Division, Seismic Engineering and Prevention of Natural Hazards Laboratory; Portici Research Centre, Piazzale Enrico Fermi 1, Portici (NA) 80055, Italy; [email protected] 2 ENEA, Italian National Agency for New Technologies, Energy and Sustainable Economic Development, Energy Technologies Department, Technical-Strategical Support Unit; Casaccia Research Centre, Via Anguillarese 301, Santa Maria di Galeria (Roma) 00123, Italy; [email protected] * Correspondence: [email protected]; Tel.: +39-081-7723306; Fax: +39-081-7723344

Academic Editor: Yu-Pin Lin Received: 27 October 2015; Accepted: 16 February 2016; Published: 25 February 2016

Abstract: The paper reports a ﬁeld investigation on a reach of the lower Zambezi River about 230–240 km downstream of the Cahora Bassa dam in the Republic of Mozambique. In the framework of a wider research program, bathymetric measures of the riverbed were performed on a 10 km stretch of the river using an echo sounder, a GPS receiver, and an integrated navigation-acquisition system. Field observations and measures revealed a general agreement with macro-features of river morphology reported in early literature, dealing with the morphological response of the river to the construction of large dams, in the second half of the 20th century. Results hereby reported are some of the few examples of direct ﬁeld measures in the lower Zambezi reported in literature, and could be used by researchers and practitioners either as a knowledge base for further surveys or as input data for validation and calibration of mathematical models and remote sensing observations.

Keywords: ﬁeld survey; bathymetry; morphology; Zambezi; Mozambique

1. Introduction The Zambezi River is one of the main rivers in southern Africa and the longest African river flowing into the Indian Ocean. The river basin, lying between latitudes 9–20˝ South and longitudes 18–36˝ East, covers an area about 1.4 million km2 wide, and is the fourth-largest catchment area in Africa after the Congo, Nile, and Niger watersheds. It is shared by eight countries (Angola, Botswana, Malawi, Mozambique, Namibia, Tanzania, Zambia, and Zimbabwe), and has an estimated population of 30 million people. Figure1 illustrates the major African river basins and the Zambezi river basin, divided into its main sub-basins, representative of major tributaries and drainage areas [1]. A large body of literature exists on the Zambezi River system, dealing with its geological, hydrological, morphological, ecological, and socio-economic aspects [2–8]. The main stem of the river has an estimated length of 2524 km [2], subdivided into three major segments (upper, middle and lower Zambezi), each having distinct geomorphological and hydrological features. The upper Zambezi, about 1078 km long [2], flows from the river source near Kalene Hill (in Zambia, at an altitude of 1450 m) to Victoria Falls. In this reach, the river flows across a vast plateau with low gradients and is, to a large extent, confined in a well-defined channel; exceptions are the Barotse and Chobe swamps, two extended flood plains in which the river cross-section locally widens up to

Environments 2016, 3, 6; doi:10.3390/environments3010006 www.mdpi.com/journal/environments Environments 2016, 3, 6 2 of 14 Environments 2016, 3, 6 2 of 14 tens(Mosi-oa-Tunia, of kilometres. “the The smoke upper Zambezithat thunders” abruptly in the ends local at theidiom), well-known with a drop Victoria of 108 Falls m and (Mosi-oa-Tunia, a width of “the1700 smoke m. that thunders” in the local idiom), with a drop of 108 m and a width of 1700 m.

FigureFigure 1. 1.(a )(a Map) Map of of the the African African major major riverriver basins.basins. (b) Overview of of Zambezi Zambezi catchment catchment area. area. Maps Maps werewere adapted adapted from from Grid-Arendal Grid-Arendal website website [1[1].].

The middle Zambezi is comprised of the stretch between Victoria Falls and the Cahora Bassa Lake,The with middle an approximate Zambezi is comprised length of 853 of the km stretch [2]. Th betweenis is the Victoriareach with Falls the and steepest the Cahora gradient Bassa of the Lake, withentire an approximatecourse of the lengthZambezi, of 853characterized km [2]. This by narrow is the reach gorges with and the rapids. steepest Along gradient this reach of the in entirethe coursesecond of half the Zambezi,of the last characterizedcentury, the Kariba by narrow and the gorges Cahora and Bassa rapids. dams Along were thisbuilt reach for hydroelectric in the second halfenergy of the production. last century, The the Kariba Kariba dam, and thealong Cahora the border Bassa damsbetween were Zambia built for and hydroelectric Zimbabwe, energywas production.completed Thein 1959 Kariba and dam, created along an artificial the border lake between with a storage Zambia capacity and Zimbabwe, of 180.6 billion was completed m3 and a in 1959surface and createdarea of 5364 an artificial km2 [9]. lakeThe Cahora with a storageBassa dam, capacity in Mozambique, of 180.6 billion was completed m3 and a surfaceat the end area of of 53641974; km it2 [gives9]. The rise Cahora to an artificial Bassa dam, lake inthat, Mozambique, according to wasdata completed provided by at the electrical end of 1974; company it gives riseHidroeléctrica to an artificial de Cahora lake that, Bassa according [10], has ato useful data capacity provided of by52 thebillion electrical m3, a length company of 270Hidroeléctrica km, a width de Cahoraof 30 Bassakm, and[10 ],a hassurface a useful area of capacity 2900 km of2 at 52 maximum billion m 3water, a length level. of Other 270 km,minor a width dams are of 30 present km, and in a surfacethe basin area of of the 2900 middle km2 Zambezi;at maximum among water them, level. the most Other important minor dams are the are Kafue present Gorge in the dam basin and ofthe the middleItezhi-Tezhi Zambezi; dam, among on the them, Kafue the ri mostver, which important flows are into the the Kafue main Gorge stem of dam the andZambezi, the Itezhi-Tezhi about 80 km dam, ondownstream the Kafue river, of the which Kariba flows dam. into the main stem of the Zambezi, about 80 km downstream of the The lower Zambezi, about 593 km long [2], is the segment of the river downstream of the Cahora Kariba dam. Bassa dam, descending from the Central African Plateau to the Mozambique coastal plain and the The lower Zambezi, about 593 km long [2], is the segment of the river downstream of the Cahora Indian Ocean. The first reach, upstream of Tete, is characterized by boulder and bedrock outcrops Bassa dam, descending from the Central African Plateau to the Mozambique coastal plain and the with gorges and rapids [9]. The alluvial plain begins downstream of Tete, where the river flows on Indian Ocean. The first reach, upstream of Tete, is characterized by boulder and bedrock outcrops with Karoo sediments and is interrupted at Lupata gorge, near Mount Nhamalongo, where the river cuts gorgesvolcanic and rock rapids outcrops [9]. The (basalt alluvial and plainrhyolite) begins [11–1 downstream3] and the cross-section of Tete, where narrows the river to a flowsfew hundred on Karoo sedimentsmeters. Downstream and is interrupted of the Lupata at Lupata gorge, gorge, a large near low-gradient Mount Nhamalongo, alluvial plain where characterizes the river cutsthe course volcanic rockof the outcrops river with (basalt multiple and rhyolite) channels [and11– 13islands] and covered the cross-section by vegetation. narrows The mouth to a few of hundredthe Zambezi meters. is Downstreama wide delta of with the very Lupata low gorge, gradients a large and low-gradient altitudes, frequently alluvial subjected plain characterizes to flooding,the depending courseof on the riverriver with flow multiple and tide, channels which can and extend islands up covered to 40–50 by km vegetation. inland [7]. TheThe mouthdelta is ofcharacterized the Zambezi by is a arich wide deltaalluvial with soil very with low flooded gradients grasslands and altitudes, near the frequently banks and subjected mangroves to flooding,in the coastal depending zone. on river flow and tide,The which banks can and extend riverbed up toof 40–50the lower km inlandcourse [of7]. the The Zambezi delta is are characterized incised on the by aCretaceous rich alluvial and soil withTertiary flooded sedimentary grasslands cover near of the the banks alluvial and plain, mangroves and are in mostly the coastal composed zone. of unconsolidated gravel andThe sand banks [7,9]. and Unpublished riverbed of previous the lower results course of grain of the size Zambezi analyses are on incised sediment on thesamples Cretaceous taken on and TertiaryZambezi sedimentary river banks cover near Tete of the confirm alluvial the plain, predominance and are mostly of coarse composed sand, with of unconsolidatedan average of 76% gravel of andsediment sand [7 in,9]. the Unpublished range 0.5–1.0 previous mm. results of grain size analyses on sediment samples taken on ZambeziFigure river 2a banks reports near a Tetemap confirm of the thelower predominance Zambezi region, of coarse created sand, using with QGIS an average software of 76%and of sedimentOpenLayers in the plugin range 0.5–1.0with MapQuest-Open mm. Aerial and OpenStreetMap data. In Figure 2b, a longitudinalFigure2a reportsprofile of a the map main of stem the of lower the river Zambezi is plotted, region, derived created from usingRonco QGISet al. data software [14]. and OpenLayers plugin with MapQuest-Open Aerial and OpenStreetMap data. In Figure2b, a longitudinal profile of the main stem of the river is plotted, derived from Ronco et al. data [14].

Environments 2016, 3, 6 3 of 14 Environments 2016, 3, 6 3 of 14

FigureFigure 2. (a) MapMap ofof the the lower lower Zambezi Zambezi and and main main tributaries. tributaries. (b) ( Longitudinalb) Longitudinal proﬁle profile of the of mainthe main stem stemof the of lower the lower Zambezi; Zambezi; data fromdata from [14]. [14].

Regarding hydrology [6], the rainfall is not uniform over the river basin, due to its wide Regarding hydrology [6], the rainfall is not uniform over the river basin, due to its wide extension. extension. Estimated mean annual precipitation is about 950 mm/year; the highest values of rainfall Estimated mean annual precipitation is about 950 mm/year; the highest values of rainfall (up to (up to 1400 mm/year) occur in the northern parts of the basin and around Niassa Lake, while the 1400 mm/year) occur in the northern parts of the basin and around Niassa Lake, while the lowest lowest values are observed in Zimbabwe and the southern areas of the basin (about 500 mm/year). values are observed in Zimbabwe and the southern areas of the basin (about 500 mm/year). The climate The climate of the basin is influenced by the movement of air masses in the Intertropical Convergence of the basin is inﬂuenced by the movement of air masses in the Intertropical Convergence Zone (ITCZ), Zone (ITCZ), with major precipitations occurring during the summer months (November to March), with major precipitations occurring during the summer months (November to March), while winter while winter months (April to October) are generally dry. The mean annual runoff is estimated by months (April to October) are generally dry. The mean annual runoff is estimated by different authors different authors as either 103 billion·m3 [6], or 97 billion·m3 [7], or 130 billion·m3 [8]; the above values as either 103 billion¨ m3 [6], or 97 billion¨ m3 [7], or 130 billion¨ m3 [8]; the above values correspond to correspond to mean annual flows at the river mouth of 3264 m3/s, 3074 m3/s, and 4134 m3/s, respectively. mean annual ﬂows at the river mouth of 3264 m3/s, 3074 m3/s, and 4134 m3/s, respectively. Besides the main stem, the Zambezi River system comprises many important tributaries which can Besides the main stem, the Zambezi River system comprises many important tributaries which be regarded as major rivers on their own. The main tributaries are: the Kabompo, Lungue-Bunguo, can be regarded as major rivers on their own. The main tributaries are: the Kabompo, Lungue-Bunguo, Luanginga, and Cuando in the upper Zambezi; the Sanyati, Kafue, and Luangwa in the middle Luanginga, and Cuando in the upper Zambezi; the Sanyati, Kafue, and Luangwa in the middle Zambezi, and the Luia, Revuboe, Luenha-Mazoe, and Shire in the lower Zambezi. Zambezi, and the Luia, Revuboe, Luenha-Mazoe, and Shire in the lower Zambezi.

2.2. Impacts Impacts of of Large Large Dams Dams on on the the Lower Lower Zambezi Zambezi River River TheThe construction of of large large dams dams in in the middle Za Zambezimbezi valley valley has has profoundly profoundly altered altered the the flow flow pattern of the river downstream [5]. [5]. At At present, present, wi withth the the only only exception exception of of contributions contributions from from Shire Shire andand minor minor tributaries, tributaries, the the course course of of the the lower lower Zambezi Zambezi is substantiall is substantiallyy regulated regulated by the by theoutflow outflow at the at Cahorathe Cahora Bassa Bassa dam, having dam, having lost the lostnatural the variability natural variability due to seasonal due to distribution seasonal distribution of rainfall and of runoff. rainfall and runoff.The change in the hydrological regime is clearly evident in Figure 3. On the left (Figure 3a), hydrographsThe change of mean in the monthly hydrological inflows regime and outflows is clearly at the evident Kariba in (1958–1998) Figure3. On and the the left Cahora (Figure Bassa3a), (1975–1999)hydrographs reservoirs of mean monthlyare plotted inflows [5]. On and the outflows right, Figure at the 3b Kariba illustrates (1958–1998) the mean and monthly the Cahora discharges Bassa at(1975–1999) the Tete gauging reservoirs station are plotted [15], about [5]. On 130 the km right, downstream Figure3b illustrates of the Cahora the mean Bassa monthly dam, referring discharges to threeat the different Tete gauging periods, station namely: [15], before about the 130 Kariba km downstream dam construction of the (1951–1958), Cahora Bassa after dam, the referringKariba and to beforethree different the Cahora periods, Bassa namely: dam construction before the (1959–197 Kariba dam4), constructionand after the (1951–1958),Cahora Bassa after dam the construction Kariba and (1975–2003).before the Cahora A strong Bassa reduction dam construction of maximum (1959–1974), flows in the and rainy after months the Cahora and an Bassa increase dam of construction minimum flows(1975–2003). in the dry A strong season reduction are observed. of maximum As a result, flows discharges in the rainy tend months to be rather and an uniformly increase of distributed minimum during the year. From Figure 3b data, considering the period after the Cahora Bassa dam flows in the dry season are observed. As a result, discharges tend to be rather uniformly distributed construction, a mean annual runoff at the Tete gauging station of 69.4 billion·m3 can be estimated, during the year. From Figure3b data, considering the period after the Cahora Bassa dam construction, corresponding to a mean annual flow of 2198 m3/s. a mean annual runoff at the Tete gauging station of 69.4 billion¨ m3 can be estimated, corresponding to a mean annual flow of 2198 m3/s.

Environments 2016, 3, 6 4 of 14 Environments 2016, 3, 6 4 of 14

Figure 3. (a) Mean monthly inflows and outflows at the Kariba and Cahora Bassa dams; data from [5]. Figure 3. (a) Mean monthly inﬂows and outﬂows at the Kariba and Cahora Bassa dams; data from [5]. (b) Mean monthly discharges at the Tete gauging station; data from [15]. (b) Mean monthly discharges at the Tete gauging station; data from [15].

The effects of evaporation from the reservoir lakes are also to be considered. Regarding the KaribaThe reservoir, effects of evaporationit can be estimated from the that reservoir about lakes 11% areof alsoinflows to be are considered. lost during Regarding the peak the flooding Kariba reservoir,months, with it can a mean be estimated annual net that evaporation about 11% rate of inflowsof 955 mm/year, are lost duringwhereas the an peakannual flooding net evaporation months, withrate of amean 1523 annualmm/year net can evaporation be assumed rate at of the 955 Cahora mm/year, Bassa whereas reservoir an annual[5]. Overall, net evaporation the mean annual rate of 1523volume mm/year of water can belost assumed at the atCahora the Cahora Bassa Bassa reservoir reservoir can [5be]. Overall,estimated the as mean about annual 4.7 volumebillion·m of3, 3 waterapproximately lost at the equal Cahora to Bassa7% of reservoirthe total inflows can be estimated [5]. as about 4.7 billion¨ m , approximately equal to 7%Another of the total important inflows [consequence5]. of the impoundments is the trapping of sediments in the reservoirs.Another As important reported by consequence Bolton [16,17], of the the impoundments rate of sediment is theaccumulation trapping ofin sedimentsKariba Lake in was the reservoirs.briefly considered As reported in the bypre-impoundment Bolton [16,17], the studies rate ofbased sediment on a limited accumulation number inof Karibasamples, Lake whereas was brieflyno quantitative considered estimations in the pre-impoundment of sediment transpor studiest were based reported on a limited for the number Cahora of samples,Bassa dam whereas at the nodesign quantitative stage. The estimations author performed of sediment an independent transport were study reported and estimated for the Cahora a mean Bassa annual dam sediment at the designinput in stage. the range The author7–70 million performed·m3/year an independentat Kariba Lake, study and and20–200 estimated million a·m mean3/year annual at Cahora sediment Bassa 3 3 inputLake. inMore the recently, range 7–70 Ronco million et al.¨ m[15],/year combining at Kariba previous Lake, turbidity and 20–200 measures million and¨ m hydrological/year at Cahora data, Bassaestimated Lake. the More total recently, sediment Ronco inputet al. [15at ],Cahora combining Bassa previous Lake turbidityas 28.6 measuresmillion·m and3/year, hydrological of which 3 data,14.6 million estimated·m3/year the totalflows sediment in from the input main at Cahorastem of Bassa the middle Lake asZambezi 28.6 million and ¨14.0m /year,million of·m which3/year 3 3 14.6from million the Luangwa¨ m /year tributary. flows in fromUnfortunately, the main stemthe above of the middleestimations Zambezi cannot and be 14.0 confirmed million¨ bym /yeardirect frommeasures, the Luangwa since no tributary.data from Unfortunately,bathymetric surveys the above or other estimations monitoring cannot activities be confirmed aimed to by quantify direct measures,the sediment since deposition no data from at the bathymetric Kariba and surveys Cahora or Bassa other reservoirs monitoring are activities reported aimed in literature. to quantify the sedimentThe depositionconstruction at of the large Kariba dams and had Cahora important Bassa reservoirs consequences are reported on the morphology in literature. of the river downstream.The construction According of largeto the dams classical had Lane’s important formulation consequences [15,18,19], on the the morphology flow regulation of the and river the downstream.reduction of sediment According load to alter theclassical the natural Lane’s equi formulationlibrium of the [15 ,18river,19 ],with the two flow opposite regulation effects. and theNamely, reduction considering of sediment the sediment load alter grain the natural size and equilibrium stream slope of the constant, river with the twopeak opposite flow reduction effects. Namely,lowers the considering transportthe capacity sediment and grain would size tend and to stream cause slope sediment constant, deposition the peak and flow aggradation reduction lowers of the theriverbed, transport whereas capacity the andreduction would in tend sediment to cause load sediment would induce deposition riverbed and aggradation erosion downstream. of the riverbed, Thus, whereasthe actual the reductionmorphological in sediment changes load are would the induceresult riverbedof complex erosion interactions downstream. between Thus, thedifferent actual morphologicalcomponents affecting changes the are equilibr the resultium of conditions complex interactions of the river. between different components affecting the equilibriumBased on aerial conditions photographs of the river. and field measures, Guy [20] reported severe erosion of the southernBased banks on aerial of the photographs middle Zambezi and field valley measures, downstream Guy [20 of] reportedthe Kariba severe dam. erosion The author of the concluded southern banksthat, besides of the the middle nature Zambezi of soil, erosion valley downstream can be due to of out-of-season the Kariba dam. floods, The induced author by concluded dam operation that, besidesand the therelease nature of sediment-free of soil, erosion water can be downstream. due to out-of-season floods, induced by dam operation and the releaseSuschka of sediment-freeand Napica [21] water discussed downstream. the modifications induced on the lower Zambezi by the CahoraSuschka Bassa anddam, Napica reporting [21] an discussed almost complete the modifications lack of investigation induced on during the lower the 10 Zambezi years after by thethe Cahoraconstruction Bassa of dam, the reporting dam. The an authors almost completeconcluded lack that, of investigationdespite some during consequences the 10 years which after were the constructionobviously expectable, of the dam. no The evident authors changes concluded in the that, riverbed despite somemorphology consequences could whichhave been were estimated, obviously expectable,and a much no longer evident period changes of observation in the riverbed should morphology have been necessary could have to been find estimated,noticeable effects. and a much longerDavies period et of al. observation [9] distinguished should different have been macro-scale necessary morphological to find noticeable zones effects. in the lower Zambezi River,Davies comprisinget al. [9 ]narrow distinguished gorges, different transitional macro-scale zones morphologicalreferred to as zones“mobile in thesands lower zones”, Zambezi an River,anabranching comprising zone, narrow and a coastal gorges, distributary transitional zone. zones Based referred on observations to as “mobile from sands an aerial zones”, survey an conducted in 1996, dramatic changes in the river system were reported due to flow regulation, with a different response for each of the above morphological zones. Namely, erosion of the bars in the

Environments 2016, 3, 6 5 of 14 anabranching zone, and a coastal distributary zone. Based on observations from an aerial survey conducted in 1996, dramatic changes in the river system were reported due to flow regulation, with a different response for each of the above morphological zones. Namely, erosion of the bars in the gorges was noticed, with removal of larger sediment stores due to the increase of low flows during the dry season. Sediments eroded in the gorges were conveyed in the mobile sands zones, causing stabilization of the braids and bar and increasing areas of stable floodplain. In the anabranching and distributary zone, even with slight differences, the reduction in the magnitude and frequency of inundation in the rainy period seemed to cause the dominance of one main channel, with isolation and progressive obliteration of small secondary channels. Finally, the authors reported a huge erosion in the delta and coastal zone, with significant coastline retreat due to reduction of sediment supply. It is to be noted that the studies summarized above are mainly based either on macro-scale observations or on first-approximation measures. As for siltation at water impoundments, no bathymetric surveys of riverbed or systematic in situ monitoring studies, aimed at a quantitative estimation of geomorphological changes after the construction of dams, are available in literature. Ronco et al. [14,15] applied a simplified one-dimensional mathematical model based on the “local uniform flow” hypothesis to simulate the sediment transport and long-term morphological evolution of the Zambezi River between the Cahora Bassa dam and the Indian Ocean. Simulations were performed, both in the present condition (with dams), and by assuming that the dams were not built. Simulations without dams showed a general tendency toward bottom aggradation for the largest part of the lower Zambezi, especially in the flat areas downstream of the gorges. As regards the effects of the dams, for the first time after the start of the simulations, the flow reduction seemed to prevail, inducing an increase in the sediment deposition rate in the flat areas, whereas a reduction in sediment load due to siltation in the reservoirs began to show its influence at a longer simulation time, with a smaller aggradation or a higher erosion rate of the riverbed compared to the natural conditions. The computed reduction of sediment supply to the deltaic area yielded results satisfactorily consistent with the rates of coastal erosion estimated from satellite images. The one-dimensional model described above was recently implemented by adding a quasi-two-dimensional module which simulates the evolution of the cross-section shape, also taking into account the dynamics of riparian vegetation [22]. Simulations in the presence of the dams showed that the river would tend to simplify its path, with the deepening of a main channel, and the increase of vegetation density and progressive obliteration of secondary channels, in agreement with Davies et al. [9]. The simplified formulation of the model allows us to significantly reduce the amount and resolution of input data for simulations, making it possible to use remote sensing images and large-scale digital elevation models derived by satellite observations [23]. More generally, the overall impact of water impoundments on the Zambezi River system and community is an interesting matter of debate among scientists, practitioners, politicians, and both governmental and non-governmental organizations—sometimes with very different opinions. The debate deals with different topics such as ecology, agriculture, fishery, and other economic activities, as well as ethical, social and cultural issues, which are not discussed in the present paper. It is just noteworthy to mention that the first concerns about potential impacts of the Cahora Bassa dam were expressed even before the filling of the reservoir [24,25], and that the discussion is still open as documented, besides already cited works, by a number of papers, dealing with both the evaluation of impacts and the proposal of mitigation measures, considering expected climate change scenarios as well [26–33].

3. The Field Activities During March 2013, a bathymetric survey was performed along a reach of the lower Zambezi River near Tambara, about 230–240 km downstream of the Cahora Bassa dam. The survey was conducted in the framework of a wider research program, in cooperation with Minas Rio Bravo Lda., a mining company based in Mozambique. A detailed description of operations and results can be found Environments 2016, 3, 6 6 of 14 in an ENEA technical report, in Italian [34]. The month of March, immediately after the end of the

Environmentsrainy reason, 2016 was, 3, 6 chosen in order to work with both dry weather, and because the expected high water6 of 14 level would enhance navigability. The Rio Bravo dredging platform, about 30 m long and 10 m wide (Figure 44a),a), waswas usedused toto transport personnel and instrumentation along the Zambezi and served as a logistic base during the fieldﬁeld survey.

Figure 4. (a) The Rio Bravo dredging platform. (b) Operations on board the Alfa1 vessel. Figure 4. (a) The Rio Bravo dredging platform. (b) Operations on board the Alfa1 vessel.

The main resources used for hydrographic surveys are listed as follows: The main resources used for hydrographic surveys are listed as follows: . Single-beam echo sounder Odom Echotrac DF3200 MKII, dual frequencies 24–200 kHz, for ‚ bathymetricSingle-beam measures. echo sounder Odom Echotrac DF3200 MKII, dual frequencies 24–200 kHz, for . Trimblebathymetric Modular measures. GPS receiver SPS461 (72-channel L1 C/A Code, L1/L2 Full Cycle Carrier GPS ‚ receiver)Trimble Modularfor positioning. GPS receiver SPS461 (72-channel L1 C/A Code, L1/L2 Full Cycle Carrier GPS . Trimblereceiver) HydroPro for positioning. software for survey planning, integrated real-time navigation-acquisition ‚ andTrimble first HydroProdata processing. software for survey planning, integrated real-time navigation-acquisition and . Goldenﬁrst data Software processing. Surfer for data elaboration, construction of riverbed digital models through ‚ interpolationGolden Software and graphic Surfer for presentation data elaboration, of results. construction of riverbed digital models through interpolation and graphic presentation of results. The survey vessels and personnel for navigation were provided by Minas Rio Bravo Lda., whereas the authorsThe survey attended vessels to the and navigation personnel forplan navigation and equipment were provided setup and by operation,Minas Rio providing Bravo Lda. ,real-time whereas instructionsthe authors attendedto the helmsman to the navigation during the plan surveys, and equipment based on GPS setup data. and operation, providing real-time instructionsAccording to the to indications helmsman duringprovided the by surveys, the authors, based ontwo GPS steel data. work boats (Alfa1 and Alfa2) were adaptedAccording to place to equipment indications and provided surveyors by the(Figure authors, 4b). twoThe echo steel worksounder boats transducer (Alfa1 and wasAlfa2 mounted) were atadapted the port to side place of equipment the survey and vessels surveyors using (Figurea pole and4b). an The ad echo hoc designed sounder transducersteel framework. was mounted at the portAccording side of theto the survey state vesselsof the art using [35], a the pole sound and anvelocityad hoc indesigned the echo steel sounder framework. was calibrated prior to eachAccording survey. Both to the the state bar ofcheck the arttechnique [35], the and sound indirect velocity estimations in the echo based sounder on water was temperature calibrated prior and salinityto each survey.obtained Both by multi-parameter the bar check technique probes andwere indirect used for estimations calibration. based Bar oncheck water calibration temperature was performedand salinity by obtained lowering by an multi-parameter iron plate at known probes depths were below used forthe calibration.transducer, Barand checkregulating calibration the sound was performedspeed in the by echo lowering sounder an ironas an plate adjustment at known pa depthsrameter. below Eutech the PH600 transducer, and Eutech and regulating COND600 the probes sound werespeed used in the to echomeasure sounder the water as an temperature adjustment parameter.and conductivity, Eutech respectively. PH600 and Eutech COND600 probes wereMore used todetails measure on the the instruments’ water temperature calibration and an conductivity,d measurement respectively. techniques are given in the next sectionMore which details is dedicated on the instruments’ to discussion calibration of the results. and measurement techniques are given in the next sectionDepths which below is dedicated the river to surface discussion were of simultaneously the results. acquired using both frequencies of the echo sounderDepths referred below to the as high river frequency surface were (200 simultaneously kHz) and low acquiredfrequency using (24 kHz), both frequenciesrespectively. of the echo sounderThe referredsurveyed to asarea high frequencylies between (200 kHz)latitudes and low16°38 frequency′43″–16°43 (24′03 kHz),″ South respectively. and longitudes 34°08The′35″–34°13 surveyed′45″ East. area Bathymetry lies between was measured latitudes along 16˝38 143river2–16 cross-sections˝431032 South with and spacing longitudes in the range34˝08 1800–1500352–34˝13 m;145 in2 East.the areas Bathymetry of major wasinterest, measured the distance along between river cross-sections survey lines withwas reduced spacing to in aboutthe range 100 800–1500m. Additional m; in theposition areas measures of major interest, were also the taken distance along between the river survey banks lines and was the reduced island shoresto about where 100 m. the Additional depth was positioncompatible measures with the were navigation also taken and along the operation the river of banks the echo and thesounder. island shoresIn whereFigure the 5, a depth geographical was compatible setting withof the the investigated navigation area and and the operationa plan-view of theof the echo survey sounder. data are plotted; maps in Figure 5 were created using QGIS software and OpenLayers plugin with Google Maps and OpenStreetMap data, respectively.

Environments 2016, 3, 6 7 of 14

In Figure5, a geographical setting of the investigated area and a plan-view of the survey data are plotted; maps in Figure5 were created using QGIS software and OpenLayers plugin with Google

MapsEnvironments and OpenStreetMap 2016, 3, 6 data, respectively. 7 of 14

FigureFigure 5. Overview5. Overview of of the theinvestigated investigated area area and and map map of ofsurvey survey data. data.

Submerged shoals and a general morphology of the river banks and islands, quite different from Submergedgeographical shoalsinformation and aavailable general at morphology the planning of thestage, river were banks noticed and during islands, the quite surveys. different fromEnvironmental geographical and information operating conditions available (in at theparticular planning the presence stage, were of sparse noticed crocodiles during and the bloats surveys. Environmentalof hippopotamuses) and operating also played conditions a role during (in particular the survey the operations, presence ofsometimes sparse crocodiles leading to andchanges bloats of hippopotamuses)in the execution also of survey played lines a role compared during theto the survey navigation operations, project. sometimes leading to changes in the executionAs of regards survey data lines processing, compared raw to survey the navigation data were project.first filtered to remove spikes and gross errors. Then, bathymetric grids were derived using the Surfer software. The Kriging algorithm, specifically As regards data processing, raw survey data were first filtered to remove spikes and gross errors. suited for irregularly spaced data [36], was chosen for interpolations. Contour maps and cross-sections Then, bathymetric grids were derived using the Surfer software. The Kriging algorithm, specifically were finally derived from the bathymetric grids. The UTM 36 South-WGS 84 coordinate system was suitedadopted for irregularly for cartographic spaced projection. data [36], was chosen for interpolations. Contour maps and cross-sections were finally derived from the bathymetric grids. The UTM 36 South-WGS 84 coordinate system was adopted4. Results for cartographic and Discussion projection. In Figures 6 and 7, the contour maps of the surveyed area derived from both high and low 4. Results and Discussion frequency data are plotted. As shown in the Figures, based on spatial distribution of data, bathymetric Ingrids Figures were derived6 and7 forthe eight contour different maps zones; of theamong surveyed them, zones area 1, derived 3, 4, 5, and from 7 are both comprised high and of a low frequencysingle datacross-section, are plotted. whereas As shownzones 2, in 6, theand Figures,8 also extend based along on spatialthe river distribution flow direction. of data, bathymetric grids wereThe derived highest forvalues eight of differentdepth, say zones; 8.0 m, amongwere me them,asured zonesclose to 1, the 3, 4,eastern 5, and limit 7 are of comprisedthe survey of a area, where the river flow is confined to a relatively narrow section delimited by an island (zone 8). single cross-section, whereas zones 2, 6, and 8 also extend along the river flow direction. As shown in the graphs, it was not always possible to investigate the entire river cross-section, The highest values of depth, say 8.0 m, were measured close to the eastern limit of the survey due to the presence of islands, shoals, and shallow water zones, where navigation and sounding were area,severely where the restricted. river flow is confined to a relatively narrow section delimited by an island (zone 8). As shownA map in of the the graphs, bathymetric it was differences not always between possible high to investigateand low frequency the entire data river is plotted cross-section, in due toFigure the presence 8. To a very of large islands, extent, shoals, a shift and of tens shallow of centimeters water zones, between where the two navigation signals can and be sounding observed, were severelywith restricted. low frequency depths generally greater than high frequency soundings. A map of the bathymetric differences between high and low frequency data is plotted in Figure8. To a very large extent, a shift of tens of centimeters between the two signals can be observed, with low frequency depths generally greater than high frequency soundings.

Environments 2016, 3, 6 8 of 14 Environments 2016, 3, 6 8 of 14 Environments 2016, 3, 6 8 of 14

Figure 6. Bathymetric map of surveyed area, derived from high frequency data (200 kHz). Figure 6. Bathymetric map of surveyed area, derived from high frequency data (200 kHz). Figure 6. Bathymetric map of surveyed area, derived from high frequency data (200 kHz).

Figure 7. Bathymetric map of surveyed area, derived from low frequency data (24 kHz). Figure 7. Bathymetric map of surveyed area, derived from low frequency data (24 kHz). Figure 7. Bathymetric map of surveyed area, derived from low frequency data (24 kHz). In Figure 9, some transects derived from bathymetric grids are plotted, which, despite some differences,In Figure can 9, be some considered transects representative derived from of baththe mainymetric morphological grids are plotted, types ofwhich, river cross-sectionsdespite some In Figure9, some transects derived from bathymetric grids are plotted, which, despite some differences,observed from can the be consideredsurvey data. representative The plotted cross-sectionsof the main morphological show the presence types ofof riverone orcross-sections a few main differences, can be considered representative of the main morphological types of river cross-sections channels,observed withfrom higherthe survey depth data. located The either plotted in cross-sectionsthe middle or nearshow the the bank presence of islands. of one As or expected, a few main the observed from the survey data. The plotted cross-sections show the presence of one or a few main highestchannels, values with ofhigher depth depth in the located main channeleither in were the middle measured or near in the the narrowest bank of islands. cross-sections. As expected, the highestchannels, values with of higher depth depth in the located main channel either in were the middle measured or near in the the narrowest bank of islands. cross-sections. As expected, the highest values of depth in the main channel were measured in the narrowest cross-sections.

Environments 2016, 3, 6 9 of 14 Environments 2016, 3, 6 9 of 14

Environments 2016, 3, 6 9 of 14

Figure 8. Map of bathymetric differences between high and low frequency data. Positive values Figure 8. Map of bathymetric differences between high and low frequency data. Positive values indicate low frequency depths greater than high frequency ones. indicateFigure low 8. Map frequency of bathymetric depths greater differences than highbetween frequency high and ones. low frequency data. Positive values indicate low frequency depths greater than high frequency ones.

Figure 9. A selection of cross-sections of investigated area (high and low frequency data).

Figure 9. A selection of cross-sections of investigated area (high and low frequency data). TheFigure instruments’ 9. A selection accuracy of cross-sections and resolution, of investigated and the areamain (high possible and low sources frequency of measurement data). uncertainties [35] are discussed below: The instruments’ accuracy and resolution, and the main possible sources of measurement . uncertaintiesTheDepth instruments’ sounding [35] are. According discussed accuracy below:to and Odom resolution, Echotrac and DF3200 the mainMKII technical possible documentation sources of measurement [37], the uncertaintiesdigitizer [35 accuracy] are discussed can be assume below: d as 0.0425% of the total measured depth, while resolution is . 1Depth cm in sounding the depth. According range 0–600 to Odomm. The Echotrac transducer DF3200 resolution MKII cantechnical be assumed documentation to be of the [37], same the ‚ Depthorderdigitizer sounding of magnitude accuracy. According can of thebe assumeacoustic to Odomd wavelengthas Echotrac 0.0425% ofand, DF3200 the basically, total MKII measured depends technical depth, on documentation operating while resolution frequency; [37], is the digitizer1 cm in accuracythe depth can range be 0–600 assumed m. The as 0.0425% transducer of theresolution total measured can be assumed depth, to while be of resolution the same is

order of magnitude of the acoustic wavelength and, basically, depends on operating frequency;

Environments 2016, 3, 6 10 of 14

1 cm in the depth range 0–600 m. The transducer resolution can be assumed to be of the same Environmentsorder 2016 of magnitude, 3, 6 of the acoustic wavelength and, basically, depends on operating frequency;10 of 14 for a reference sound speed of 1500 m/s, wavelengths can be estimated as 0.75 cm for 200 kHz for a reference sound speed of 1500 m/s, wavelengths can be estimated as 0.75 cm for 200 kHz frequency, and 6.25 cm for 24 kHz. The effect of acoustic beam angles was considered negligible frequency, and 6.25 cm for 24 kHz. The effect of acoustic beam angles was considered negligible due the relatively low values of depth and bottom slope. due the relatively low values of depth and bottom slope. ‚ Sound speed. Uncertainties in the bar check procedure basically depend on environmental and . Sound speed. Uncertainties in the bar check procedure basically depend on environmental and operating conditions, and are difficult to quantify. Regarding water parameters, according to the operating conditions, and are difficult to quantify. Regarding water parameters, according to the probes’ technical specifications [38], resolution and accuracy can be estimated as 0.01 µS/cm ˘ 1% probes’ technical specifications [38], resolution and accuracy can be estimated as 0.01 µS/cm ± 1% full scale +1 LSD for the range of measured conductivity and 0.1 ˝C for temperature measures. full scale +1 LSD for the range of measured conductivity and 0.1 °C for temperature measures. Salinity and sound speed were computed using the formulation proposed by UNESCO [39]. Salinity and sound speed were computed using the formulation proposed by UNESCO [39]. Overall, little variations in water parameters were observed during the survey, and calculated Overall, little variations in water parameters were observed during the survey, and calculated values of sound speed were substantially consistent with results of the bar check calibration. values of sound speed were substantially consistent with results of the bar check calibration. Observed differences of sound speed resulted in possible differences in depth estimation, in the Observed differences of sound speed resulted in possible differences in depth estimation, in the range 1–4 cm. range 1–4 cm. ‚ Horizontal Positioning. According to Trimble SPS461 technical documentation [40], a sub-metric . Horizontal Positioning. According to Trimble SPS461 technical documentation [40], a sub-metric horizontal accuracy can be achieved when the GPS receiver operates with satellite differential horizontal accuracy can be achieved when the GPS receiver operates with satellite differential correction. In cases where the receiver is operated in autonomous mode, the actual positioning correction. In cases where the receiver is operated in autonomous mode, the actual positioning accuracy is difficult to quantify, depending on a number of factors including the satellites’ number accuracy is difficult to quantify, depending on a number of factors including the satellites’ and position, meteorological conditions, and receiver features. Overall accuracy was also verified, number and position, meteorological conditions, and receiver features. Overall accuracy was comparing test positioning data and ground control points measurements. Plots of survey points also verified, comparing test positioning data and ground control points measurements. Plots of on recent satellite images, as in Figure5, showed a sufficient agreement. survey points on recent satellite images, as in Figure 5, showed a sufficient agreement. ‚ Vessel heave, pitch and roll. Vessel heave, pitch, and roll were minimized during the survey, limiting . Vessel heave, pitch and roll. Vessel heave, pitch, and roll were minimized during the survey, the survey speed to a maximum of 3 kn. Overall, the effects on bathymetric measures were limiting the survey speed to a maximum of 3 kn. Overall, the effects on bathymetric measures considered negligible compared to other possible error sources. were considered negligible compared to other possible error sources. In the absenceabsence of data from previousprevious inin situsitu measures, the results of the present survey were qualitatively compared with observations reported in literature on the river morphologymorphology after the construction of the CahoraCahora BassaBassa damdam (Figure(Figure 1010).).

Figure 10. (a) Morphological zonation of the lower Zambezi valley. (b) Geomorphological changes Figure 10. (a) Morphological zonation of the lower Zambezi valley. (b) Geomorphological changes due to Cahora Bassa dam. Adapted from Davies et al. [9] with permission from Prof. Richard Beilfuss due to Cahora Bassa dam. Adapted from Davies et al. [9] with permission from Prof. Richard Beilfuss and Schweizerbart Science Publishers [41]. and Schweizerbart Science Publishers [41].

Environments 2016, 3, 6 11 of 14

According to the morphological zonation proposed by Davies et al. [9], the present survey area lies in the “mobile sands zone”, close to the boundary with the “anabranching zone” (Figure 10a). As already discussed, the mobile sands zones can be regarded as transitional zones between the narrow gorge zones (dominated by boulder and bedrock outcrops and high stream energies) and the large Environments 2016, 3, 6 11 of 14 floodplains composed of a series of anabranch channels, characterized by low values of gradients streamand stream energy. energy. In the In themobile mobile sands sands zones, zones, the the gradients gradients and and stream stream energy energy are are lower lower than than in in the gorge zones, but still relatively high compared to the anabranching zone; the river is braided and the riverbed is mostly made of finefine gravel and sand, with high sediment fluxesfluxes [[9].9]. The Figure 11 illustrates two pictures taken du duringring the survey, showing evident analogies with schemes depicted in Figure 10b.10b. Pictures Pictures in in Figure Figure 11a,b11a,b were were taken, respectively, respectively, near the upstream and downstream boundary of the surveyed area.

Figure 11. (a) Picture taken near the upstreamupstream boundaryboundary of the investigatedinvestigated area. ( b) Picture taken near the downstream boundary of the investigated area. Typical morphological features of the lower Zambezi can be recognised.

Measured cross-sections’ shapes plotted in Figure 9 qualitatively show a reasonable agreement with Measured cross-sections’ shapes plotted in Figure9 qualitatively show a reasonable agreement morphological macro-features of expected post-impoundment cross-sections depicted in Figure 9b, with with morphological macro-features of expected post-impoundment cross-sections depicted in Figure9b, one or few main flow deeper channels incised in a relatively flat shallow water cross-section. In with one or few main flow deeper channels incised in a relatively flat shallow water cross-section. particular, cross-sections A and B have a single main channel, while sections C and D show a more In particular, cross-sections A and B have a single main channel, while sections C and D show a more complex bathymetry with multiple channels. In order to correctly compare the measured cross-sections complex bathymetry with multiple channels. In order to correctly compare the measured cross-sections with the Davies et al. schemes [9], it should be considered that the present survey was performed just with the Davies et al. schemes [9], it should be considered that the present survey was performed just at the end of the rainy season when the water level was near the highest annual values. at the end of the rainy season when the water level was near the highest annual values. Finally, the difference between the high and the low frequency depths can be explained, because Finally, the difference between the high and the low frequency depths can be explained, because the high frequency signal, besides having a higher measurement resolution, is more sensitive to low the high frequency signal, besides having a higher measurement resolution, is more sensitive to low density gradients than the low frequency one, and records as “bottom” the first interface at which a density gradients than the low frequency one, and records as “bottom” the first interface at which a certain density value is exceeded. On the other hand, a lower frequency signal is less sensitive and is certain density value is exceeded. On the other hand, a lower frequency signal is less sensitive and is also able to penetrate for a limited depth into the first bottom layer when it is made of soft materials. also able to penetrate for a limited depth into the first bottom layer when it is made of soft materials. A nearly uniform shift between the high and low frequency soundings would suggest the absence of A nearly uniform shift between the high and low frequency soundings would suggest the absence macroscopic anomalies in the riverbed coverage materials, whereas zones with higher differences could of macroscopic anomalies in the riverbed coverage materials, whereas zones with higher differences be hypothesized as the result of a more recent lower density sediment deposit or a fluid mud [42]. The could be hypothesized as the result of a more recent lower density sediment deposit or a fluid mud [42]. above qualitative considerations, however, should be taken with caution, considering the different The above qualitative considerations, however, should be taken with caution, considering the different resolutions and accuracies of high and low frequency signals. Moreover, in the present work, it was resolutions and accuracies of high and low frequency signals. Moreover, in the present work, it not possible to adequately calibrate the instrument’s response on different riverbed materials was not possible to adequately calibrate the instrument’s response on different riverbed materials and coverages. and coverages. 5. Conclusions 5. Conclusions The field investigation hereby reported provided quantitative information on a limited but The field investigation hereby reported provided quantitative information on a limited but significant reach of the lower Zambezi River. During the survey, it was also possible to recognize the significant reach of the lower Zambezi River. During the survey, it was also possible to recognize the main morphological features of the river downstream of Tete, described in early literature. The main main morphological features of the river downstream of Tete, described in early literature. The main limitations to the extension of the surveyed area were due to logistic and environmental conditions. In particular, it was not possible to extend the survey to the entire river cross-section in some areas, where navigation and sounding were restricted, due to obstacles or shallow water. Despite the fact that it was not possible to find the results of previous bathymetric measures for comparison in the literature, a qualitative agreement was found between the present results and Davies et al. [9] considerations on river morphology after the Cahora Bassa dam construction. A quantitative analysis could be attempted, comparing the present survey data with the results of mathematical models reported in the literature [14,15,22,23].

Environments 2016, 3, 6 12 of 14 limitations to the extension of the surveyed area were due to logistic and environmental conditions. In particular, it was not possible to extend the survey to the entire river cross-section in some areas, where navigation and sounding were restricted, due to obstacles or shallow water. Despite the fact that it was not possible to ﬁnd the results of previous bathymetric measures for comparison in the literature, a qualitative agreement was found between the present results and Davies et al. [9] considerations on river morphology after the Cahora Bassa dam construction. A quantitative analysis could be attempted, comparing the present survey data with the results of mathematical models reported in the literature [14,15,22,23]. The limited extension of the investigated area does not allow us to derive general information for navigation purposes. However, the survey data show an overall quite complex bathymetry, suggesting that it could be preferable to base potential studies for navigation assessment on extensive bathymetric investigations, rather than on isolated cross-section measures. Results hereby reported can be used by researchers and practitioners as a knowledge base for further ﬁeld surveys aimed at monitoring the morphological evolution of the lower Zambezi River, as well as for calibration and validation of mathematical models and remote sensing observations.

Acknowledgments: The present research was funded by an agreement between ENEA and mining company Minas Rio Bravo Lda., that provided technical and logistic assistance during transfers, and all the operations in Mozambique. The authors gratefully appreciated the Rio Bravo Lda. management and staff. Author Contributions: Francesco Pasanisi, Carlo Tebano, and Francesco Zarlenga took part in the survey in Mozambique, and equally contributed to the work hereby reported. Conﬂicts of Interest: The authors declare no conﬂict of interest.

References

1. GRID-Arendal. Environmental Knowledge for Change. Available online: http://www.grida.no (accessed on 8 October 2015). 2. Davies, B.R.; Jackson, P.B.N. The Zambezi River system. In The Ecology of River Systems; Davies, B.R., Walker, K.F., Eds.; Springer: Berlin, Germany, 1986; Volume 60, pp. 225–288. 3. Nugent, C. The Zambezi River: Tectonism, climatic change and drainage evolution. Palaeogeogr. Palaeoclimatol. Palaeoecol. 1990, 78, 55–69. [CrossRef] 4. Hughes, R.H.; Hughes, J.S. A Directory of African Wetlands; IUCN: Cambridge, UK, 1992. 5. Beilfuss, R.; dos Santos, D. Patterns of Hydrological Change in the Zambezi Delta, Mozambique. Available online: http://www.xitizap.com/zambeze-hydrochanges.pdf (accessed on 18 February 2016). 6. Euroconsult Mott MacDonald. Integrated Water Resources Management Strategy and Implementation Plan for the Zambezi River Basin. Final Report, Rapid Assessment. Available online: http://www.elmed- rostov.ru/Projects/Zamwis_php/Reports/RA%20Water%20Resources%20Zambezi%2002Jan2008.pdf (accessed on 18 February 2016). 7. Moore, A.E.; Cotterill, F.P.D.; Main, M.P.L.; Williams, H.B. The Zambezi River. In Large Rivers: Geomorphology and Management; Gupta, A., Ed.; John Wiley & Sons: Chichester, UK, 2007; pp. 311–332. 8. World Bank. The Zambezi River Basin. A Multi-Sector Investment Opportunities Analysis. Volume 3. State of the Basin; The International Bank for Reconstruction and Development/The World Bank: Washington, DC, USA, 2010. 9. Davies, B.R.; Beilfuss, R.D.; Thoms, M.C. Cahora Bassa retrospective, 1974–1997: Effects of ﬂow regulation on the Lower Zambezi River. Verh. Int. Ver. Limnol. 2001, 27, 2149–2157. 10. Hidroeléctrica de Cahora Bassa. Technical Characteristics of the Reservoir and the Dam. Available online: http://www.hcb.co.mz/eng/Engineering/The-Infrastructures/Technical-Characteristics-of-the-Reservoir- and-the-Dam (accessed on 8 October 2015). 11. Lächelt, S. The Geology and Mineral Resources of Mozambique; Direcção Nacional de Geologia: Maputo, Mozambique, 2004. Environments 2016, 3, 6 13 of 14

12. Westerhof, A.B.P.; Lehtonen, M.I.; Mäkitie, H.; Manninen, T.; Pekkala, Y.; Gustafsson, B.; Tahon, A. The Tete-Chipata Belt: A new multiple terrane element from western Mozambique and southern Zambia. In GTK Consortium Geological Surveys in Mozambique 2002–2007; Pekkala, Y., Lehto, T., Mäkitie, H., Eds.; Geological Survey of Finland: Espoo, Finland, 2008; pp. 145–166. 13. Westerhof, A.B.P.; Tahon, A.; Koistinen, T.; Lehto, T.; Åkerman, C. Igneous and tectonic setting of the allochthonous Tete Gabbro-Anorthosite Suite, Mozambique. In GTK Consortium Geological Surveys in Mozambique 2002–2007; Pekkala, Y., Lehto, T., Mäkitie, H., Eds.; Geological Survey of Finland: Espoo, Finland, 2008; pp. 191–210. 14. Ronco, P.; Fasolato, G.; Di Silvio, G. The case of the Zambezi River in Mozambique: Some investigations on solid transport phenomena downstream Cahora Bassa Dam. In River Flow 2006; Taylor & Francis Group: London, UK; pp. 1345–1354. 15. Ronco, P.; Fasolato, G.; Nones, M.; di Silvio, G. Morphological effects of damming on lower Zambezi River. Geomorphology 2010, 115, 43–55. [CrossRef] 16. Bolton, P. The Regulation of the Zambezi in Mozambique: A Study of the Origins and Impact of the Cabora Bassa Project. Ph.D. Thesis, University of Edinburgh, Edinburgh, UK, 1983. 17. Bolton, P. Sediment deposition in major reservoirs in the Zambezi Basin. In Challenges in African Hydrology and Water Resources; International Association of Hydrological Sciences (IAHS): Wallingford, UK, 1984; pp. 559–567. 18. Lane, E.W. The Importance of Fluvial Morphology in Hydraulic Engineering; U.S. Department of Interior, Bureau of Reclamation: Denver, CO, USA, 1954. 19. Brandt, S.A. Classification of geomorphological effects downstream of dams. Catena 2000, 40, 375–401. [CrossRef] 20. Guy, P.R. River bank erosion in the Mid-Zambezi valley, downstream of Lake Kariba. Biol. Conserv. 1981, 19, 199–212. [CrossRef] 21. Suschka, J.; Napica, P. Ten Years after Completion of Cabora Bassa Dam. In The Impact of Large Water Projects on the Environment; United Nations Educational, Scientific and Cultural Organization (UNESCO): Paris, France, 1986; pp. 171–203. 22. Nones, M.; Ronco, P.; Di Silvio, G. Modelling the impact of large impoundments on the lower Zambezi River. Int. J. River Basin Manag. 2013, 11, 221–236. [CrossRef] 23. Nones, M.; Guerrero, M.; Ronco, P. Opportunities from low-resolution modelling of river morphology in remote parts of the world. Earth Surf. Dyn. 2014, 2, 9–19. [CrossRef] 24. Davies, B.R. Cabora Bassa hazards. Nature 1975, 254, 477–478. [CrossRef] 25. Davies, B.R.; Hall, A.; Jackson, P.B.N. Some ecological aspects of the Cabora Bassa Dam. Biol. Conserv. 1975, 8, 189–201. [CrossRef] 26. Chenje, M.; Kullberg, M.; Kureya, T.; Laisi, E.; Mafuta, C.; Mutambirwa, C.; Saruchera, M.; Sola, L. O Estado do Ambiente na Bacia do Zambeze 2000. Available online: http://www.sardc.net/books/ ZamBasin2000Port/ZamBasin2000Port.pdf (accessed on 18 February 2016). 27. Isaacman, A.; Sneddon, C. Toward a social and environmental history of the building of Cahora Bassa dam. J. South. Afr. Stud. 2000, 26, 597–632. [CrossRef] 28. Basson, G. Hydropower dams and fluvial morphological impacts—An African perspective. In Proceedings of the United Nations Symposium on Hydropower and Sustainable Development, Beijing, China, 27–29 October 2004. 29. Scodanibbio, L.; Mañez, G. The World Commission on dams: A fundamental step towards integrated water resources management and poverty reduction? A pilot case in the Lower Zambezi, Mozambique. Phys. Chem. Earth 2005, 30, 976–983. [CrossRef] 30. The WorldFish Center. Proceedings of the International Workshop on the Fisheries of the Zambezi Basin; Davies, S., Ed.; The WorldFish Center: Penang, Malaysia, 2007. 31. Beilfuss, R.; Brown, C. Assessing environmental flow requirements and trade-offs for the Lower Zambezi River and Delta, Mozambique. Int. J. River Basin Manag. 2010, 8, 127–138. [CrossRef] 32. Mc Cartney, M.; Cai, X.; Smakhtin, V. Evaluating the Flow Regulating Functions of Natural Ecosystems in the Zambezi River Basin; International Water Management Institute (IWMI): Colombo, Sri Lanka, 2013. 33. Kling, H.; Stanzel, P.; Preishuber, M. Impact modelling of water resources development and climate scenarios on Zambezi river discharge. J. Hydrol. Reg. Stud. 2014, 1, 17–43. [CrossRef] Environments 2016, 3, 6 14 of 14

34. Pasanisi, F.; Tebano, C.; Zarlenga, F. Indagine Morfologica su un Tratto del Fiume Zambesi, nella Repubblica del Mozambico. Misure Batimetriche ed Analisi dei Dati; Technical Report RT/2015/14/ENEA; Italian National Agency for New Technologies, Energy and Sustainable Economic Development (ENEA): Rome, Italy, 2015. 35. USACE. Engineering and Design. Hydrographic Surveying; U.S. Army Corps of Engineers (USACE): Washington, DC, USA, 2002. 36. Golden Software Inc. Surfer User’s Guide; Golden Software Inc.: Golden, CO, USA, 1999. 37. Odom Hydrographic Systems Inc. Echotrac MKII Technical Manual; Odom Hydrographic Inc.: Baton Rouge, LA, USA, 2001. 38. Thermo Fisher Scientific. Eutech Instruments Water Analysis Solutions. Available online: http://www.eutechinst.com (accessed on 1 December 2015). 39. Fofonoff, N.P.; Millard, R.C., Jr. Algorithms for Computation of Fundamental Properties of Seawater; United Nations Educational, Scientific and Cultural Organization (UNESCO): Paris, France, 1983. 40. Trimble Navigation Ltd. SPSx61 Modular GPS Heading Receivers User Guide; Trimble Navigation Ltd.: Sunnyvale, CA, USA, 2009. 41. Schweizerbart Science Publisher. Available online: http://www.schweizerbart.de (accessed on 5 November 2015). 42. Schettini, C.A.F.; de Almeida, D.C.; Siegle, E.; de Alencar, A.C.B. A snapshot of suspended sediment and fluid mud occurrence in a mixed-energy embayment, Tijucas Bay, Brazil. Geo Mar. Lett. 2010, 30, 47–62. [CrossRef]

© 2016 by the authors; licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons by Attribution (CC-BY) license (http://creativecommons.org/licenses/by/4.0/).