Precipitation & Runoff modeling of catchment Hidrološko modeliranje porečja Savinje

36. Goljevščkov spominski dan Ljubljana, 23.03.2017 Email: [email protected]

1 The drainage basin hydrological cycle

The drainage basin hydrological system

Source: http://www.alevelgeography.com/drainage-basin-hydrological-system 2 Drainage Basin Flow Chart

Lakes/ Reservoars

Source: http://www.alevelgeography.com 3 The basic principle

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 d = + + + + + Where P − E −Q [SP SM TZ UZ LZ lakes] P = precipitation dt E = evapotranspiration Q = runoff SP = snow pack SM = soil moisture TZ = storage in soil top zone (introduced in HBV-light) UZ = upper groundwater zone storage LZ = lower groundwater zone storage lakes = lake volume Source: http://www.smhi.se/forskning/forskningsomraden/hydrologi/hbv-1.1566 6 Semi-distribution

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

• Elevation zones • Vegetation zones

7 HBV overview

• 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 8 Model of Computed Runoff

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

Source: Help HBV-light – Parameter Overview 12 Effect of TT

Picture: Help HBV-light 13 Savinja Catchment

• 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

14 Flood 1954 Savinja-

15 Josephinische Landesaufnahme (1763-1787)

16 Flood 1990 Savinja-Laško

17 Relational Hydro and Meteo Data

Data source: MOP -ARSO, 2015 18 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 • 3 vegetation zones

23 Elevation zones • 16 elevation zones

24 Precipitation RRhour • 33 precipitation stations

25 What is PEST? Parameter Estimation and Uncertainty Analysis

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

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

27 Model Calibration

Source: Help HBV-light – Model Calibration 28 [PTQ screen]

29 Flood Wave Celje 18.-20.09.1007

30 Goodnes of Fit for calibration period - year 2007

Average model efficiency of Savinja River to Gračnica 0.952 inflow for whole calibration period 2007

Average model efficiency for flood wave 18.-22.09.2007 0.988

31 Savinja - vp Nazarje

3+4i vp Savinja - Nazarje 600.00

500.00

400.00

300.00

200.00

100.00

0.00 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

Series1 Series2

32 Savinja - vp Letuš 1

5 vp Savinja -Letuš I 700

600

500

400

300

200

100

0 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

Series1 Series2

33 Savinja – vp Medlog

11 vp Savinja - Medlog 1000

900

800

700

600

500

400

300

200

100

0 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

Series1 Series2

34 Savinja – vp Celje II _ brv

13 vp Savinja - Celje II - brv 1200

1000

800

600

400

200

0 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

Series1 Series2

35 Savinja – vp Laško

19 vp Savinja - Laško 1400

1200

1000

800

600

400

200

0 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

Series1 Series2

36 – vp Kraše

4 vp Dreta - Kraše 250.00

200.00

150.00

100.00

50.00

0.00 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

Series1 Series2

37 – vp Dolenja vas

10 vp Bolska - Dolenja vas 160

140

120

100

80

60

40

20

0 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

Series1 Series2

38 Gračnica – vp Vodiško I

20 vp Gračnica - Vodiško I 14

12

10

8

6

4

2

0 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

Series1 Series2

39 Goodnes of Fit for validation period - year 1990

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 do VP Škofja Vas 0.8 62 do VP Celje II 0.8 67 Savinja do VP Laško 0.97

76 Savinja do VP Veliko Širje I 0.84 40 Flood Wave vp Laško 26.10.-05.11.1990

41 Savinja – vp Laško

Savinja-Laško I 1600

1400

1200

1000

800

600

400

200

0 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

Series1 Series2

42 Ložnica – vp Levec I

Ložnica-Levec I 90

80

70

60

50

40

30

20

10

0 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

Series1 Series2 43 Dreta – vp Kraše

Dreta-Kraše 300

250

200

150

100

50

0 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 Series1 Series2 44 Voglajna – vp Celje II

Voglajna-Celje II 70

60

50

40

30

20

10

0 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 Series1 Series2 45 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 površine gozda za 20%

VP Letuš Sim&Obs-20% gozd 800

700

600

500

400

300

200 10 % povečanje konice vala

100 13 % povečanje volumna 4.5 mio. m3 bi zadržal gozd 0 0 10 20 30 40 50 60 1h visokovodni val nastopi 1 uro prej 12 Qobs_12 12_20

49 Razporeditev padavin 2000 VP Veliko Širje 1800 1600 1400 1200 1000 800 600 400 200 2007 Scn01 Scn02 Scn03 Scn04 Scn05 0 1 3 5 7 9 11 13 15 17 19 21 23 25 27 29 31 33 35 37 39 41 43 45 47

50 Ground water

51 Vpliv zadrževalnikov na Dreti -VP Celje

45_Savinja do VP Celje II – brv_ 2007 1200

1000

800

600

400

200

0 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

Qsim_WS2 Q_zadrzano

-5.78 sprememba Qvk [%] -63.24 sprememba Qvk [m3] 1.63 zadržani volumen [mio. m3] 52 River Analysis System (HEC-RAS)5.0.3 (1D,2D)

Povprečna Povprečni padec Vodotok Dolžina [km] Število profilov oddaljenost med odseka [%] profili [m] Savinja 58,4 934 62,5 0,3 Dreta 15,7 407 38,8 0,4 Voglajna 15 400 37,5 0,2 Hudinja 17,2 466 36,9 0,6 Ložnica 2,9 87 33,3 0,3 Koprivnica 2,4 61 39,3 0,2 Sušnica 1,6 56 28,6 0,4 53 Vpliv zajezitve Voglajne zaradi povečanih pretokov Hudinje

54 Vpliv zadrževalnikov na Savinji v spodnji Savinjski dolini - VP Celje

55