HYDRAULIC STUDY OF THE NERETVA RIVER (FROM MOSTAR TO THE BORDER WITH THE REPUBLIC OF CROATIA)

B.Sc. Draženka Kvesić, civ. eng. PRONING DHI d.o.o., Račkog 3, Zagreb – Croatia. Email: drazenka@proning-dhi.hr. Contact author. M.Sc. Božidar Deduš, civ. eng. PRONING DHI d.o.o., Račkog 3, Zagreb – Croatia. Email: deda@proning-dhi.hr

Key words: MIKE 11, steady mode, unsteady mode, calibration, MIKE GIS, flooding mapping ABSTRACT The main purpose of the study about the Neretva river project is the creation of a mathematical model, which will be used for analyzing possibilities of establishing a flood forecast system as well as simulation and presentation of flood waves by using mapping technology.

The basis of this system represents the mathematical model MIKE 11, which is used for the creation of the Neretva river model from the hydroelectric power plant HE Mostar until the border with the Republic of Croatia (i.e. up to the water reservoir (VS) Metković), an area of app. 50 km. The model includes 5 tributaries.

For the purpose of this project, both geodetic and hydrometric surveys have been conducted as well as detailed hydrological analysis, which are used for conducting hydraulic simulations.

Available time series data came from 14 water monitoring stations, for the period from 1995 to 2006.

Hydraulic simulations were made with MIKE 11, which has a possibility of simulating 1-dimensional flow. Flow simulations of steady and unsteady mode were made. A mathematical model calibration was done for both types of flows, followed by verification, and afterwards simulation of selected water waves.

A flow simulation was done in steady mode for the case of the bank protection which is located in the area from VS Počitelj until VS Gabela. It was also done in the case of being without a bank protection on the river Neretva.

An analysis of the flow regime transformation was made in order to evaluate the anthropogenic influence from gravel exploitation.

A flooding map was created by using MIKE 11 GIS based on the obtained steady mode flow simulation results, topographic data and digital maps. Flood zone mapping creates an important step towards better understanding of flood problems in urban areas, as well as in areas of special interest.

1. INTRODUCTION

The river Neretva and its tributaries is the largest and most significant river in this water area that lies in the area of responsibility of JP for “Vodno područje slivova Jadranskog mora“ (“Catchments areas of the Adriatic Sea”), which ordered this project. It is also one of the larger rivers in Bosnia and Herzegovina.

The river Neretva is 225 km long and runs with its longest part through BH (203 km), and only 22 km through Croatia before inflowing into the Adriatic Sea. It rises underneath the Jabuka mountain in BH. Together with its tributaries, it represents an individual environmental unit and a unique ecological system in this part of world. It rises in the mountainous part of Herzegovina and has features of a mountain river throughout its major part. Because of these features, several hydroelectric power plants have been built: Jablanica, Grabovica, Salakovac and Mostar. Tributaries to the Neretva river are as follows: Ljuta, Rama, Drežanka, Radobolja, Jasenica, Buna, Bregava, Trebižat and Krupa.

On its way to the Adriatic Sea, the Neretva river runs through several cities which are considered as BH’s most beautiful cities, such as: Konjic, Jablanica, Mostar, Čapljina, Počitelj and the Croatian cities Metković and Opuzen. It is famous for its emerald- green colour, clean and entirely drinkable water in its upper flow. The Neretva river flows into the Adriatic Sea in the vicinity of Opuzen within the Neretva delta.

Figure 1 – Orthophoto of the Neretva river at the examined section

The respective section of this hydraulic study is 50 km long and starts at its upstream part at the hydroelectric power plant HE Mostar and runs to the border with the Republic of Croatia, i.e. up to Metković, where the downstream water monitoring station is located. This project consists of three parts: hydrological analyses, hydraulic calculation and mapping of flooding zones. According to the project assignment, the hydraulic study should be done in the 1-D model program, i.e. in this case with MIKE 11. 2. HYDROLOGICAL ANALYSES The basic task of the hydrological analyses is creating historical and current flow curves for the existing automatic hydrological stations at the catchments area of the Neretva river- from Mostar to Gabela, defining water balances (spatial and time distribution of the flow in the catchments) for the selected years based on created flow curves; and calculation of probabilities on high water occurrences (flows and water level) of various recurrent periods for the respective hydrological stations.

Table 1 – Synoptic survey of calculated hydrological parameters which are presented in this elaborate

No.

VS

Flow Flow curve

Probability of

occurrence HW

Water balance 1980. ;

2005/2006.

1 Mostar Noretva Yes Yes Yes

2 Baćevići Noretva Yes Yes Yes

3 Žitomislići Noretva Yes Yes Yes

4 Gabela Noretva Yes Yes Yes

5 Dom Jasenica No Yes No

6 Buna Buna Yes Yes Yes

7 Stolac-Do Bregava No Yes Yes

8 Studenci Trebižat No Yes Yes

9 Stubica Trebižat Yes No No

10 Humac Trebižat Yes Yes No

11 Studenci Studenčica Yes Yes No

Figure 2 – Cartographic survey of the respective catchments of the Neretva river with appurtenant hydrological stations

The possibility of having large waters of various recurrent periods, which is later on used as input data for stationary hydraulic simulations, is displayed in the following diagram:

VS Baćevići, Neretva

Qmax.1/20 = 1909 m3/s

Qmax.1/100= 2318 m3/s

Qmax.1/500= 2720 m3/s

HE Mostar

Border of RH

Radobolja

Jasenica

Krupa

Bregava

Buna

Trebižat

VS Mostar, Neretva

Qmax.1/20 = 1814 m3/s

Qmax.1/100= 2216 m3/s

Qmax.1/500= 2610 m3/s

VS Žitomislići, Neretva

Qmax.1/20 = 2046 m3/s

Qmax.1/100= 2433 m3/s

Qmax.1/500= 2900 m3/s

Profile 2, Neretva

Qmax.1/20 = 2096 m3/s

Qmax.1/100= 2490 m3/s

Qmax.1/500= 2980 m3/s

VS Dračevo

?

VS Gabela, Neretva

Qmax.1/20 = 2208 m3/s

Qmax.1/100= 2600 m3/s

Qmax.1/500= 3150 m3/s

Profile 1, Neretva

Qmax.1/20 = 1929 m3/s

Qmax.1/100= 2330 m3/s

Qmax.1/500= 2740 m3/s

Figure 3

3. HYDRAULIC CALCULATIONS and ANALYSIS OF RESULTS 3.1. MATHEMATICAL MODEL Measurements from different water monitoring stations have been available for this project in such a way that boundary conditions of the mathematical model have been determined at their locations. As the time measure data is not the same for each station, several mathematical models have been created for which hydraulic simulations have been conducted later on. Therefore, we gained geometric models for the following sections:

HE Mostar – VS Gabela,

HE Mostar – VS Dračevo,

HE Mostar – VS Metković,

VS Žitomislići - VS Dračevo. The mathematical model also included tributaries of the river Neretva as follows: Radobolja, Jasenica, Buna, Bregava, Trebižat and Krupa. For the area of Počitelj up to the border with the Republic of Croatia where the river bed changes due to anthropogenic influence, it is necessary to repeat the recording of the Neretva river bed quiet often because of gravel exploitation. Hence, we have cross section profiles from 1984, 1997, 2002-2005. At this section, the cross section profiles are recorded with high density (in ca. 200 locations, about every 75m), in comparison with the section upstream of Počitelj towards Mostar (ca. 80 locations, about every 450 m). It is necessary to define three models which will differ in their year of recoding at the flow analysis due to anthropogenic influence. The first model is a model that consists of cross section from 1984, the second one is of cross section from 1997 and the third one of cross section from 2003-2005, including cross section from 1997. This all refers to the part from Počitelj to the border with Croatia. Upstream of Počitelj, the cross sections of all three models are the same, which were recorded after 1998. In this way, the three models differ from each other:

Model 1984

Model 1997

Model 2003/05

The following figure displays the changes of one cross section profile throughout the year:

Figure 4 – cross sections at chainage 32+487

3.2. BOUNDARY CONDITIONS The typical upstream boundary condition of the mathematical model of river flows is the hydrogram Q(t) on the specific water wave at non-chainage flow, i.e. constant inflow for chainage flow. The upstream boundary conditions are defined at all entries into the mathematical model, which means at the upper model boundary of the Neretva river, and at the upper boundaries of all Neretva river tributaries at the examined section. The downstream boundary condition as the water level diagram h(t), where the water level in the function of time, is usually used as a suitable boundary condition. On the other hand, the unambiguous value Q(h) is the downstream boundary condition which is required for chainage flow. Furthermore, during the flow simulation, the condition h(t) will generally speaking not always be available, except when it is about reservoirs or tides’ conditions. The tides of the Adriatic Sea have a rather small amplitude. This difference rarely rises over 40 cm in the southern part. However, the tide may rise to a significant height in narrow channels and bays during strong north winds. This phenomenon is characteristic for large and deep bays of the southern Adriatic Sea. The sea phases consist of a mixed type and their amplitudes are quiet irregular. The Q(h) condition is therefore used as the downstream boundary condition for chainage flow, while h(t) is used as the downstream border condition for non-chainage flow.

1984. 1997.

2005.

3.3. CALIBRATION AND VERIFICATION OF THE MATHEMATICAL MODEL OF THE NERETVA RIVER The calibration of the unsteady flow consists of adjusting model features in such a way to gain coincidences between measured and simulated time dependant hydrograms. The water level diagram is in this case particularly emphasized due to the fact that flows are never precise or accurate and are rarely recorded for the entire flooding period. In the same time, the water level in the function of time is mostly accurate at the majority of locations along the river flow. The measured unsteady flows might be used for additional verification at the calibration. By having available the measurement data from the water monitoring stations at the Neretva river and its tributaries from HE Mostar to VS Metković, seven historical events from 1995-2005 have been selected as well as one extracted historical event for which a model verification has been conducted. They are as follows: Table 2 – events for which the calibration of the model has been conducted

Simulation start

Simulation end

Simulation duration (days)

Section Upstream boundary condition

Downstream boundary condition

12.12.1995. 22.4.1996. 130 Žitomislići-Dračevo

Q(t) VS Žito.

H(t) VS Dračevo

15.12.1999. 20.12.1999. 5 HE Mostar–

Gabela Q(t)

VS Mostar H(t)

VS Gabela

9.10.2002. 16.10.2002 7 Žitomislići-Dračevo

Q(t) VS Žito.

H(t) VS Dračevo

21.3.2004. 6.4.2004. 15 HE Mostar–

Gabela Q(t)

VS Mostar H(t)

VS Gabela

10.4.2004. 11.5.2004. 31 HE Mostar–

Gabela Q(t)

VS Mostar H(t)

VS Gabela

27.3.2005. 10.5.2005 44 HE Mostar–

Dračevo Q(t)

VS Mostar H(t)

VS Dračevo

25.12.2005. 11.1.2006. 17 HE Mostar–

Dračevo Q(t)

VS Mostar H(t)

VS Dračevo

Table 3 – event for which the verification of the model has been conducted

Simulation start

Simulation end

Simulation duration (days)

Section Upstream boundary condition

Downstream boundary condition

30.11.2004. 7.12.2004. 7 HE Mostar-

Metković Q(t)

VS Mostar H(t)

VS Metković

23-3-2004 25-3-2004 27-3-2004 29-3-2004 31-3-2004 2-4-2004 4-4-2004 6-4-2004

100.0

200.0

300.0

400.0

500.0

600.0

700.0

800.0

900.0

1000.0

1100.0

1200.0

1300.0

[m^3/s] HE Mostar - Gabela 21.3.2004. - 6.4.2004. Discharge

NERETVA-MO-HR 59275.15

External TS 1

protok-bacevici

23-3-2004 25-3-2004 27-3-2004 29-3-2004 31-3-2004 2-4-2004 4-4-2004 6-4-2004

31.8

32.0

32.2

32.4

32.6

32.8

33.0

33.2

33.4

33.6

33.8

34.0

34.2

34.4

34.6

34.8

35.0

35.2

35.4

35.6

35.8

36.0

36.2

36.4

36.6

36.8

[meter] HE Mostar - Gabela 21.3.2004. - 6.4.2004. Water Level

NERETVA-MO-HR 58790.00

External TS 1

nivo-bacevici

The following figure displays only one result of the calibration of unsteady flow, for one location, i.e. VS Bačevići: Figure 5 measured simulated

Manning’s numbers M have been used in the mathematical model of the Neretva river, which values are marked at the schematic display as shown in figure 6. Figure 6

Discharge

Water level

0.0 5000.0 10000.0 15000.0 20000.0 25000.0 30000.0 35000.0 40000.0 45000.0

[m]

-15.0

-10.0

-5.0

0.0

5.0

10.0

15.0

20.0

25.0

30.0

35.0

40.0

45.0

50.0

55.0

60.0

65.0

70.0

75.0

80.0

85.0

90.0

[meter] 14-7-2004 00:00:00

NERETVA-MO-HR

65027

65057

65067

65322

65663

66032

661316622866526

66680

66722

66896

670466

7521

67851

682906

8660

69245

69593

70377

71073

71240

71852

72271

72554

NERETVA-MO-HR 65027 - 55561

55561

56554

57309

5814858790

59760

60141

61252

61783

6252763457

63775

64058

64404

64691

5340954020

54688

NERETVA-MO-HR 53409 - 30145

30145

30242

30325

30402

30485

30571

30651

30732

30819

30906

30974

31084

31117

31191

31278

31363

31470

31543

31585

31695

31763

31863

32003

32070

32190

32273

32370

32487

32557

32616

32688

32764

32832

32922

33009

33100

33184

33286

33333

33434

33495

33564

33657

337423

3806

33864

33947

34007

34096

34149

3421834286

34397

34481

34564

34632

34712

34789

34887

34964

35048

35195

35306

35427

35510

35573

35660

35677

35735

35788

35840

35906

35983

36074

36116

36184

3626736349

36439

36511

36571

37267

38118

389463

9425

40172

41677

42977

43517

44398

45137

4563846174

46801

47333

48185

48985

4996450869

51608

5174552355

27841

27940

28041

28154

28279

28379

28482

28587

28684

28784

28883

29028

29133

29367

29475

29566

29681

29759

29830

29931

30057

25323

25354

25423

25540

25620

25715

25812

25926

25983

26101

26183

26297

26382

26476

26578

26664

26766

26863

26957

27061

27151

27249

27338

27433

27544

27637

27734

vs b

acevic

i

vs m

osta

r

jasenic

a

buna

vs c

apljina

vs z

itom

islic

i

vs g

abela

bre

gava

trebiz

at

radobolja

vs s

utin

a

he m

osta

r

3.4. HYDRAULIC CALCULATION FOR STEADY FLOW, FOR FLOW OF RANK 1/20, 1/100 AND 1/500 YEAR Steady flow has been conducted for two different scenarios: in cases with embankments (which is the real case) and without embankments. Simulations have been conducted for water waves of the recurrent period 20, 100 and 500 years, which are defined by the Hydrologic analyses of the Neretva river. Figure 7 – Schematic display of the determined boundary conditions for RP=100 g. Figure 8 –Longitudinal section of the Neretva river with marked maximum water level for RP=100 years

3.5. ANALYSIS OF THE CHANGES IN THE FLOW REGIME IN ORDER TO EVALUATE THE ANTHROPOGENIC INFLUENCE DUE TO GRAVEL EXPLOITATION ALONG THE SECTION FROM THE BORDER TO ČAPLJINA Cross section profiles from three different time periods have been available for the analysis of the anthropogenic influence due to gravel exploitation. Based on gained results from the hydraulic simulations, it is visible that the border is at chainage 33 km where the tendency of changes in the water surface level shifts. With time, the river bed is becoming deeper downstream of that chainage, so that the maximum level of the water is at the lowest for model 2003/05, regardless of the water wave rank. Upstream of the 33 km at the Neretva river is the highest water level for model 2003/05, which chronologically speaking means that it has a tendency of rising. All this is only valid under the assumption that the riverbed is as measured in 1997, i.e. that the flow curve at VS Gabela and for the other models is as it was defined in the hydrological analysis.

Razine vode za stacionarno tečenje bez nasipa, povratnih perioda 20, 100, 500 godina

0

10

20

30

40

50

60

70

25300 30300 35300 40300 45300 50300 55300 60300 65300 70300

stacionaža

razin

a v

od

e (

m.n

.m)

PP=20g. Hmax

PP=100g. Hmax

PP=500g. Hmax

Figure 9 –Water levels for steady flow without enbankments, recurrent periods 20, 100, 500

Antropogeni utjecaj na modelu HE Mostar - VS Gabela PP=100 god.

5

7

9

11

13

15

17

19

25000.00 27000.00 29000.00 31000.00 33000.00 35000.00 37000.00 39000.00

stacionaža

niv

o v

od

e (

m.n

.m)

PP100 1984

PP100 1997

ITP100 2003/05

Figure 10 – Antropogenic influence

4. MAPPING FLOODS Within the project “Hydraulic study of the Neretva river”, one of the assignments was the cartographic layout of hydraulic simulation results which, combined with orthophoto maps and 3D Digital Elevation Model (DEM) in shapes as lines, contour lines and individual dots, give a survey on the events during floodings in different versions that represent the flows of the recurrent periods 20, 100 and 500 years. The digital map of the area which was used to determine the flooding area was created by using all available geodetic and GIS documents as follows:

1. Cross section profile of the Neretva river and its tributaries, gained by geodetic recording, from chainage km. 20+705 – st. 73+500 in ACAD format, from where they were translated into table format which is suitable for using it in the MIKE11 Cross Section Editor.

2. Different geodetic records of the Neretva river bed where heavy gravel exploitation occurs (control records) in ACAD format

3. 3D Digital Elevation Model (DEM) that is defined by lines, contour lines and individual dots at the examined section up to above the level of the 500 year water by developing the digital orthophoto

4. colored DOF in the scale 1:5000 of the entire area 5. colored DOF in the scale 1:1000 of part of the area (from the RH border to

beyond the city Čapljina). Geodetic profiles are used in cases when they exist (riverbed and surrounding, mostly up to the embankment, but not always. For parts where the geodetic profile is missing, data from the DEM follows up. In this way, a model has been created which is used in mapping. 70 maps (dimensions 90/60) have been created as the result of mapping flooding zones of certain recurrent periods and models with existing bank protections and without them, similar to this one:

Figure 11