Precipitation & Runoff modeling of Savinja catchment

36. Goljevščkov spominski danLjubljana, 23.03.2017Email: andrej.vidmar@fgg.uni-lj.si

1

Hidrološko modeliranje porečja Savinje

The drainage basin hydrological cycle

Source: http://www.alevelgeography.com/drainage-basin-hydrological-system

The drainage basin hydrological system

2

Drainage Basin Flow Chart

Source: http://www.alevelgeography.com

Lakes/ Reservoars

3

The basicprinciple

4

Which hydrological model?

• various ongoing researches are there on topics like which model will give more compatible results compared to P-R relations

• HSPF, HSPEXP+, TOPMODEL, HBV-96, WFLOW Py(distributed HBV), MIKE-SHE, SWAT, SWMM, RS Minerve…

HBV-light & PEST

HBV-light: http://www.geo.uzh.ch/en/units/h2k/services/hbv-model/PEST: http://www.pesthomepage.org/

5

HBV - Hydrologiska Byråns Vattenbalansavdelning(Hydrological Agency Water Balance Department)

• The HBV model (Bergström, 1976, 1992) is a rainfall-runoff model, which includes conceptual numerical descriptions of hydrological processes at the catchment scale. The general water balance can be described as

WhereP = precipitationE = evapotranspirationQ = runoffSP = snow packSM = soil moistureTZ = storage in soil top zone (introduced in HBV-light) UZ = upper groundwater zone storageLZ = lower groundwater zone storagelakes = lake volume

[ ]lakesLZUZTZSMSPdtdQEP +++++=−−

6Source: http://www.smhi.se/forskning/forskningsomraden/hydrologi/hbv-1.1566

Semi-distribution

7

Subdivides a large problem into smaller, simpler parts with unique characteristic

• Elevation zones• Vegetation zones

HBV overview

8

• The HBV model is a simple multi-tank-type model for simulating runoff.

• Rainfall and air temperature data as well as estimated potential evaporation data based on the American Society of Civil Engineers Penman–Monteith method are inputs to the model, which consists of four commonly used routines: snow; soil moisture; response; and routing. Picture: Help HBV-light – An Overview of the HBV Model

Model of Computed Runoff

9Source: Help HBV-light – Three GW Box (Distributed STZ and SUZ) Model

Equations Overview

Source: Help HBV-light – An Overview of the HBV Model

Catchment Parameters

Three GW Box (Distributed STZ and SUZ) Model

Vegetation Zone Parameters

12Source: Help HBV-light – Parameter Overview

Effect of TT

13Picture: Help HBV-light

Savinja Catchment

14

• Enclosed Area of 1852.3 sq km• min_Elev_m=190.1 m• max_Elev_m=2429.0 m• avg_Elev_m=604.5 m• avg_Aspect=SE (127°)• Older_Celje_elev=~238.0 m

Flood 1954 Savinja-Celje

15

Josephinische Landesaufnahme (1763-1787)

16

Flood 1990 Savinja-Laško

17

Relational Hydro and Meteo Data

18

Data source:MOP -ARSO, 2015

HIGRIS –Hydrologic Graphical IS

- basically designed with Global Mapper and has more than 160 layers and a lot of external links

• GIS (Global Mapper [LiDAR], Map Window, SAGA, ILWIS, Google Earth Pro)• CAD (AutoCAD MAP 3D, QuickSurf, Surfer)• Graphic design (PhotoLine, PaintShop)• DB (ASCII, MS Access, PostGIS)• Statistic (MS Excell, Origin, Scilab)• Programming (SQL, PowerBasic_CC, Python with NumPy)• P-R model (HBV-light_CLI, HBV-light-GUI)• Calibration Tools (PEST, GAP and Monte Carlo are included in HBV-light)• File navigation and data preview (Total Commander, IrfanView, Acrobat) 19

Savinja 21 sub-catchments; I. model

20

Savinja 77 sub-catchments; II. model

21

Geology [Alluvi-Karst]

22

Vegetation zones

23

• 3 vegetation zones

Elevation zones

24

• 16 elevation zones

Precipitation RRhour

25

• 33 precipitation stations

What is PEST?

26

Parameter Estimation andUncertainty Analysis

* parameter are estimated on the basis of past events * some of them cannot be estimated uniquely

BeoPEST, PEST++ are tools for model calibration and uncertainty analysis

27

Model Calibration

28Source: Help HBV-light – Model Calibration

[PTQ screen]

29

Flood Wave Celje18.-20.09.1007

30

Goodnes of Fit for calibration period - year 2007

31

Average model efficiency of Savinja River to Gračnica inflow for whole calibration period 2007

0.952

Average model efficiency for flood wave 18.-22.09.2007 0.988

Savinja - vp Nazarje

32

0.00

100.00

200.00

300.00

400.00

500.00

600.00

17.09.2007 12:00 18.09.2007 00:00 18.09.2007 12:00 19.09.2007 00:00 19.09.2007 12:00 20.09.2007 00:00 20.09.2007 12:00

3+4i vp Savinja - Nazarje

Series1 Series2

Savinja - vp Letuš 1

33

0

100

200

300

400

500

600

700

17.09.2007 12:00 18.09.2007 00:00 18.09.2007 12:00 19.09.2007 00:00 19.09.2007 12:00 20.09.2007 00:00 20.09.2007 12:00

5 vp Savinja -Letuš I

Series1 Series2

Savinja – vp Medlog

34

0

100

200

300

400

500

600

700

800

900

1000

17.09.2007 12:00 18.09.2007 00:00 18.09.2007 12:00 19.09.2007 00:00 19.09.2007 12:00 20.09.2007 00:00 20.09.2007 12:00

11 vp Savinja - Medlog

Series1 Series2

Savinja – vp Celje II _ brv

35

0

200

400

600

800

1000

1200

17.09.2007 12:00 18.09.2007 00:00 18.09.2007 12:00 19.09.2007 00:00 19.09.2007 12:00 20.09.2007 00:00 20.09.2007 12:00

13 vp Savinja - Celje II - brv

Series1 Series2

Savinja – vp Laško

36

0

200

400

600

800

1000

1200

1400

17.09.2007 12:00 18.09.2007 00:00 18.09.2007 12:00 19.09.2007 00:00 19.09.2007 12:00 20.09.2007 00:00 20.09.2007 12:00

19 vp Savinja - Laško

Series1 Series2

Dreta – vp Kraše

37

0.00

50.00

100.00

150.00

200.00

250.00

17.09.2007 12:00 18.09.2007 00:00 18.09.2007 12:00 19.09.2007 00:00 19.09.2007 12:00 20.09.2007 00:00 20.09.2007 12:00

4 vp Dreta - Kraše

Series1 Series2

Bolska – vp Dolenja vas

38

0

20

40

60

80

100

120

140

160

17.09.2007 12:00 18.09.2007 00:00 18.09.2007 12:00 19.09.2007 00:00 19.09.2007 12:00 20.09.2007 00:00 20.09.2007 12:00

10 vp Bolska - Dolenja vas

Series1 Series2

Gračnica – vp Vodiško I

39

0

2

4

6

8

10

12

14

17.09.2007 12:00 18.09.2007 00:00 18.09.2007 12:00 19.09.2007 00:00 19.09.2007 12:00 20.09.2007 00:00 20.09.2007 12:00

20 vp Gračnica - Vodiško I

Series1 Series2

Goodnes of Fit for validation period - year 1990

40

WS2 WS2_Name NS (1.10-14.11.1990)

1 Savinja do VP Solčava I 0.85

8 Dreta do VP Kraše 0.90

38 Ložnica do VP Levec I 0.94

45 Savinja do VP Celje II - brv 0.97

53 Hudinja do VP Škofja Vas 0.8

62 Voglajna do VP Celje II 0.8

67 Savinja do VP Laško 0.97

76 Savinja do VP Veliko Širje I 0.84

Flood Wave vp Laško26.10.-05.11.1990

41

Savinja – vp Laško

42

0

200

400

600

800

1000

1200

1400

1600

26.10.1990 00:00:00 28.10.1990 00:00:00 30.10.1990 00:00:00 01.11.1990 00:00:00 03.11.1990 00:00:00 05.11.1990 00:00:00 07.11.1990 00:00:00 09.11.1990 00:00:00 11.11.1990 00:00:00

Savinja-Laško I

Series1 Series2

Ložnica – vp Levec I

43

0

10

20

30

40

50

60

70

80

90

26.10.1990 00:00 28.10.1990 00:00 30.10.1990 00:00 1.11.1990 00:00 3.11.1990 00:00 5.11.1990 00:00 7.11.1990 00:00 9.11.1990 00:00 11.11.1990 00:00

Ložnica-Levec I

Series1 Series2

Dreta – vp Kraše

44

0

50

100

150

200

250

300

26.10.1990 00:00:00 28.10.1990 00:00:00 30.10.1990 00:00:00 01.11.1990 00:00:00 03.11.1990 00:00:00 05.11.1990 00:00:00 07.11.1990 00:00:00 09.11.1990 00:00:00 11.11.1990 00:00:00

Dreta-Kraše

Series1 Series2

Voglajna – vp Celje II

45

0

10

20

30

40

50

60

70

26.10.1990 00:00 28.10.1990 00:00 30.10.1990 00:00 1.11.1990 00:00 3.11.1990 00:00 5.11.1990 00:00 7.11.1990 00:00 9.11.1990 00:00 11.11.1990 00:00

Voglajna-Celje II

Series1 Series2

1. Why use a P-R modeling?

• for education• for decision support• for data quality control• for water balance studies• for drought runoff forecasting (irrigation)• for fire risk warning• for runoff forecasting/prediction (flood warning and reservoir

operation)• for what happens if’ questions

46

2. Why use a P-R modeling?

• to compute design floods for flood risk detection• to extend runoff data series (or filling gaps)• to compute design floods for dam safety• to compute energy production• to investigate the effects of land-use changes within the catchment• to simulate discharge from ungauged catchments• to simulate climate change effects

47

Designed Flood Predictions

• based on flood event 2007 (50 year return period)24 hour Precipitation Event for Q10, Q20, Q50, Q100, Q200 and Q500

• based on flood event 1990 and 1998 (100 year return period)48 hour Precipitation Event for Q10, Q20, Q50, Q100, Q200 and Q500

48

Sprememba rabe tal – zmanjšanje površinegozda za 20%

49

0

100

200

300

400

500

600

700

800

0 10 20 30 40 50 60

VP Letuš Sim&Obs-20% gozd

12 Qobs_12 12_20

10 % povečanje konice vala13 % povečanje volumna4.5 mio. m3 bi zadržal gozd1h visokovodni val nastopi 1 uro prej

Razporeditev padavin

50

0

200

400

600

800

1000

1200

1400

1600

1800

2000

1 3 5 7 9 11 13 15 17 19 21 23 25 27 29 31 33 35 37 39 41 43 45 47

VP Veliko Širje

2007 Scn01 Scn02 Scn03 Scn04 Scn05

Ground water

51

Vpliv zadrževalnikov na Dreti -VP Celje

52

0

200

400

600

800

1000

1200

1 3 5 7 9 11 13 15 17 19 21 23 25 27 29 31 33 35 37 39 41 43 45 47 49 51 53 55 57 59 61 63 65 67 69 71

45_Savinja do VP Celje II – brv_ 2007

Qsim_WS2 Q_zadrzano

-5.78 sprememba Qvk [%]-63.24 sprememba Qvk [m3]1.63 zadržani volumen [mio. m3]

River Analysis System (HEC-RAS)5.0.3 (1D,2D)

53

Vodotok Dolžina [km] Število profilovPovprečna

oddaljenost med profili [m]

Povprečni padec odseka [%]

Savinja 58,4 934 62,5 0,3Dreta 15,7 407 38,8 0,4

Voglajna 15 400 37,5 0,2Hudinja 17,2 466 36,9 0,6Ložnica 2,9 87 33,3 0,3

Koprivnica 2,4 61 39,3 0,2Sušnica 1,6 56 28,6 0,4

Vpliv zajezitve Voglajne zaradi povečanihpretokov Hudinje

54

Vpliv zadrževalnikov na Savinji v spodnjiSavinjski dolini - VP Celje

55