analysis of parameters for distributed modelling of...
TRANSCRIPT
![Page 1: Analysis of parameters for distributed modelling of ...erb2010.boku.ac.at/uploads_erb/08_50_tacheci.pdfAnalysis of parameters for distributed modelling of hydrologic regime in three](https://reader036.vdocuments.site/reader036/viewer/2022062505/5d5d520888c993c7058b5668/html5/thumbnails/1.jpg)
Analysis of parameters for Analysis of parameters for distributed modelling of hydrologic regime in three subcatchments of regime in three subcatchments of
Kopaninský tok catchment
Pavel Tachecí , DHI a.s., Prague, CZ, , g ,
Pavel Žlábek, University of South Bohemia, České Budějovice CZČeské Budějovice, CZ
Markéta Kaplická, Research Institute for Soil and Water Conservation, Prague, CZ
13th ERB conference, Seggau 2010
![Page 2: Analysis of parameters for distributed modelling of ...erb2010.boku.ac.at/uploads_erb/08_50_tacheci.pdfAnalysis of parameters for distributed modelling of hydrologic regime in three](https://reader036.vdocuments.site/reader036/viewer/2022062505/5d5d520888c993c7058b5668/html5/thumbnails/2.jpg)
Th F The Frame
t f C h Mi i t f A i lt R&D part of Czech Ministry of Agriculture R&D project (2008-2012)
The impact of land use in particular catchment on runoff and some substances loss
measurement + modelling
loss
sign NAZV QH 82095
coordinator: Research Institute for Soil and Water C ti P )Conservation, Prague)
continuation of projects on hydrology nad water quality
Analysis of parameters for distributed modelling
![Page 3: Analysis of parameters for distributed modelling of ...erb2010.boku.ac.at/uploads_erb/08_50_tacheci.pdfAnalysis of parameters for distributed modelling of hydrologic regime in three](https://reader036.vdocuments.site/reader036/viewer/2022062505/5d5d520888c993c7058b5668/html5/thumbnails/3.jpg)
Th AiThe Aim
current point:p
To use distributed deterministic mathemetical modellig for sensistivity mathemetical modellig for sensistivity analysis of rainfall/runoff response of small subcatchments
Analysis of parameters for distributed modelling
![Page 4: Analysis of parameters for distributed modelling of ...erb2010.boku.ac.at/uploads_erb/08_50_tacheci.pdfAnalysis of parameters for distributed modelling of hydrologic regime in three](https://reader036.vdocuments.site/reader036/viewer/2022062505/5d5d520888c993c7058b5668/html5/thumbnails/4.jpg)
K i ký t k t h tKopaninský tok exp. catchment
Operated by RISWC since 80‘s
Basic data:
467 - 578 m a. s. l. , hilly area
area 6.99 km2
arable land 45 %
grassland 13 % forest 36 %.
tile drainage 10 %
P: 665 mm
T: 7.0 °C
bedrock: paragneiss,
Analysis of parameters for distributed modelling
soil cover : Dystric Cambisol, 100 cm.
![Page 5: Analysis of parameters for distributed modelling of ...erb2010.boku.ac.at/uploads_erb/08_50_tacheci.pdfAnalysis of parameters for distributed modelling of hydrologic regime in three](https://reader036.vdocuments.site/reader036/viewer/2022062505/5d5d520888c993c7058b5668/html5/thumbnails/5.jpg)
Kopaninský tok Kopaninský tok exp. catchment
Analysis of parameters for distributed modelling
![Page 6: Analysis of parameters for distributed modelling of ...erb2010.boku.ac.at/uploads_erb/08_50_tacheci.pdfAnalysis of parameters for distributed modelling of hydrologic regime in three](https://reader036.vdocuments.site/reader036/viewer/2022062505/5d5d520888c993c7058b5668/html5/thumbnails/6.jpg)
Subcatchments
Area km2
Arable forest drainage
P6 0 157 96 0 100P6 0.157 96 0 100
P52 0.649 31 62 16
P53 0.049 98 100
measured:
discharge, temperature (10 min)
l d WQ i tsampled: WQ, isotopes
available: meteostation data
GW le els (1 pie ometer)
Analysis of parameters for distributed modelling
GW levels (1 piezometer)
![Page 7: Analysis of parameters for distributed modelling of ...erb2010.boku.ac.at/uploads_erb/08_50_tacheci.pdfAnalysis of parameters for distributed modelling of hydrologic regime in three](https://reader036.vdocuments.site/reader036/viewer/2022062505/5d5d520888c993c7058b5668/html5/thumbnails/7.jpg)
D t dData used
1. Maps (Arc GIS layers)Contour lines – DEM of surface
Land use map (1982 – 2002)
S il t ( t d)Soil types map (aggregated)
Soil substrates map (4 types) + depth map
Drainage mapDrainage map
Analysis of parameters for distributed modelling
![Page 8: Analysis of parameters for distributed modelling of ...erb2010.boku.ac.at/uploads_erb/08_50_tacheci.pdfAnalysis of parameters for distributed modelling of hydrologic regime in three](https://reader036.vdocuments.site/reader036/viewer/2022062505/5d5d520888c993c7058b5668/html5/thumbnails/8.jpg)
D t dData used
Topography - DEM
Analysis of parameters for distributed modelling
![Page 9: Analysis of parameters for distributed modelling of ...erb2010.boku.ac.at/uploads_erb/08_50_tacheci.pdfAnalysis of parameters for distributed modelling of hydrologic regime in three](https://reader036.vdocuments.site/reader036/viewer/2022062505/5d5d520888c993c7058b5668/html5/thumbnails/9.jpg)
D t dData used
2. Time series (2004 – 2009)
Precipitation – 10 min meteostation combined with daily data
Air temperature: hourly averages (AS1)Air temperature: hourly averages (AS1)
Ref. Evapotranspiration: computed (P-M eq.) daily from AS1 data
Discharge: 10 min data
Vegetation: LAI, root depth assesed di t t f according to type of crop
Analysis of parameters for distributed modelling
![Page 10: Analysis of parameters for distributed modelling of ...erb2010.boku.ac.at/uploads_erb/08_50_tacheci.pdfAnalysis of parameters for distributed modelling of hydrologic regime in three](https://reader036.vdocuments.site/reader036/viewer/2022062505/5d5d520888c993c7058b5668/html5/thumbnails/10.jpg)
D t dData used
Vegetation parameters time series
Analysis of parameters for distributed modelling
![Page 11: Analysis of parameters for distributed modelling of ...erb2010.boku.ac.at/uploads_erb/08_50_tacheci.pdfAnalysis of parameters for distributed modelling of hydrologic regime in three](https://reader036.vdocuments.site/reader036/viewer/2022062505/5d5d520888c993c7058b5668/html5/thumbnails/11.jpg)
K i ký t k t h tKopaninský tok exp. catchment
Data – 3. Parameters
Soil hydraulic characteristics:
SZ hydraulic characteristics: previous studies, calibrated
Drainage parameters: calibrated
Snow melt model: calibrated
Interception and vegeration parameters: measurement + literature
Analysis of parameters for distributed modelling
![Page 12: Analysis of parameters for distributed modelling of ...erb2010.boku.ac.at/uploads_erb/08_50_tacheci.pdfAnalysis of parameters for distributed modelling of hydrologic regime in three](https://reader036.vdocuments.site/reader036/viewer/2022062505/5d5d520888c993c7058b5668/html5/thumbnails/12.jpg)
D t dData used
Soil hydraulic parameters: (Ks + retention curve) based on 81+10 core samples and infiltration measurement
Analysis of parameters for distributed modelling
![Page 13: Analysis of parameters for distributed modelling of ...erb2010.boku.ac.at/uploads_erb/08_50_tacheci.pdfAnalysis of parameters for distributed modelling of hydrologic regime in three](https://reader036.vdocuments.site/reader036/viewer/2022062505/5d5d520888c993c7058b5668/html5/thumbnails/13.jpg)
M d lli t dModelling system used
MIKE SHE 2009 WM (DHI software)MIKE SHE 2009 WM (DHI software)
Analysis of parameters for distributed modelling
![Page 14: Analysis of parameters for distributed modelling of ...erb2010.boku.ac.at/uploads_erb/08_50_tacheci.pdfAnalysis of parameters for distributed modelling of hydrologic regime in three](https://reader036.vdocuments.site/reader036/viewer/2022062505/5d5d520888c993c7058b5668/html5/thumbnails/14.jpg)
Model setup overview
• Channel FlowChannel Flow1 -D kinematic wave
• Overland Flow2 D diff i 2-D diffusive wave
• EvapotranspirationKristensen & Jensen method for ETa calculation
• Unsaturated ZoneRichards Equation 1D vertical in all grid cells
• Saturated Zone Flow• Saturated Zone Flow2D finite difference approx. of Boussinesq eq.
Analysis of parameters for distributed modelling
![Page 15: Analysis of parameters for distributed modelling of ...erb2010.boku.ac.at/uploads_erb/08_50_tacheci.pdfAnalysis of parameters for distributed modelling of hydrologic regime in three](https://reader036.vdocuments.site/reader036/viewer/2022062505/5d5d520888c993c7058b5668/html5/thumbnails/15.jpg)
Model setup overview• Time step: 1h / 10 min
G id ll 6 10 • Grid cells: 6-10 m• Simplified river channel schematisation• 3 soil types and land use distribution• Soil profile: 3 horizons• 2 layers (constant) of saturated zone• Drainage schematised (distributed)g ( )• Initial conditions estimated according to
previous simulationsp
Analysis of parameters for distributed modelling
![Page 16: Analysis of parameters for distributed modelling of ...erb2010.boku.ac.at/uploads_erb/08_50_tacheci.pdfAnalysis of parameters for distributed modelling of hydrologic regime in three](https://reader036.vdocuments.site/reader036/viewer/2022062505/5d5d520888c993c7058b5668/html5/thumbnails/16.jpg)
Model setup overview
Analysis of parameters for distributed modelling
![Page 17: Analysis of parameters for distributed modelling of ...erb2010.boku.ac.at/uploads_erb/08_50_tacheci.pdfAnalysis of parameters for distributed modelling of hydrologic regime in three](https://reader036.vdocuments.site/reader036/viewer/2022062505/5d5d520888c993c7058b5668/html5/thumbnails/17.jpg)
Modelling steps• Calibration: on 1h (10min) discharge data
(summer periods of 2007-2009) (summer periods of 2007 2009) manual/automated (Shuffled Complex Evolution method, Weighted sum of squares + RMSE as Objective functions)
• Validation: against 1h (10min) discharge Validation: against 1h (10min) discharge data (summer periods of 2005-2006) correlation and effectivity coeff. usedy
• Sensitivity analysis (local for selected parameters backward pertubation fract ) parameters, backward, pertubation fract.)
Analysis of parameters for distributed modelling
![Page 18: Analysis of parameters for distributed modelling of ...erb2010.boku.ac.at/uploads_erb/08_50_tacheci.pdfAnalysis of parameters for distributed modelling of hydrologic regime in three](https://reader036.vdocuments.site/reader036/viewer/2022062505/5d5d520888c993c7058b5668/html5/thumbnails/18.jpg)
Results
• validation: 69mm event No.2
Analysis of parameters for distributed modelling
![Page 19: Analysis of parameters for distributed modelling of ...erb2010.boku.ac.at/uploads_erb/08_50_tacheci.pdfAnalysis of parameters for distributed modelling of hydrologic regime in three](https://reader036.vdocuments.site/reader036/viewer/2022062505/5d5d520888c993c7058b5668/html5/thumbnails/19.jpg)
Results
• validation: 28mm low intensity event
Analysis of parameters for distributed modelling
![Page 20: Analysis of parameters for distributed modelling of ...erb2010.boku.ac.at/uploads_erb/08_50_tacheci.pdfAnalysis of parameters for distributed modelling of hydrologic regime in three](https://reader036.vdocuments.site/reader036/viewer/2022062505/5d5d520888c993c7058b5668/html5/thumbnails/20.jpg)
Conclusions (P6)• the most important parameters were
ranked: ranked: 1. Bypass ratio of macropores in
unsaturated zone unsaturated zone, 2. Horizontal and vertical saturated
hydraulic conductivity of saturated zonehydraulic conductivity of saturated zone
Analysis of parameters for distributed modelling
![Page 21: Analysis of parameters for distributed modelling of ...erb2010.boku.ac.at/uploads_erb/08_50_tacheci.pdfAnalysis of parameters for distributed modelling of hydrologic regime in three](https://reader036.vdocuments.site/reader036/viewer/2022062505/5d5d520888c993c7058b5668/html5/thumbnails/21.jpg)
Conclusions (P6)• Optimum values of parameters differ
according to different grid cell size of according to different grid cell size of model
• The influence of initial conditions is • The influence of initial conditions is crucial at least for first simulated year
Analysis of parameters for distributed modelling
![Page 22: Analysis of parameters for distributed modelling of ...erb2010.boku.ac.at/uploads_erb/08_50_tacheci.pdfAnalysis of parameters for distributed modelling of hydrologic regime in three](https://reader036.vdocuments.site/reader036/viewer/2022062505/5d5d520888c993c7058b5668/html5/thumbnails/22.jpg)
Thank you for your attentionThank you for your attention
p tacheci@dhi [email protected]
Analysis of parameters for distributed modelling