Fullyintegrated Flow and Transport Modelling

1
Canadian Water Network Researcher
Retreat Victoria, BC, June 24-26, 2008
Integrated Surface-Subsurface Flow Transport
Modelling from the Watershed-Scale and Beyond A
Framework for Quantitative Analysis
E. A. Sudicky Department of Earth Environmental
Sciences University of Waterloo Waterloo,
Ontario, Canada N2L 3G1 (email
sudicky_at_sciborg.uwaterloo.ca)
2
Integrated Surface/Subsurface Hydrologic
ModellingA Grand Challenge in Water Resources
Simulation

• Attempt to account for all interactions between
  the surface and subsurface flow regimes
• Simultaneously solve surface and subsurface flow
  transport equations
• Simulate the complete hydrologic cycle

3
Some Issues

• Quantify spatial/temporal distribution of GW
  recharge, surface water/groundwater interactions
  in streams and wetlands
• Impacts of groundwater extraction on surface
  water
• Effects of urbanization/land-use/climate change
  on water quantity quality, health of aquatic
  ecosystems
• Restoration of adversely-impacted streams,
  wetlands, etc.
• Subsurface versus overland migration pathways of
  contaminants pathogens

4
Challenges

• Disparate time frames between surface/subsurface
  flow and transport regimes
• Very large unstructured grids, irregular
  topography, complex boundary conditions, surface
  properties geological features
• Strong nonlinearities in governing equations
• Data availability and upscaling issues

5
Waterloo/Laval HydroGeoSphere Model

• First Generation Code Integrated Hydrologic
  Model (InHM, VanderKwaak, U of Waterloo, PhD
  Thesis, 1999), Evolved from FRAC3DVS (Therrien
  Sudicky, JCH, 1996)
• Consortium of HydroGeoSphere developers/research
  ers
• Sudicky et al. (Earth Sciences, Univ. Waterloo,
  Ontario)
• Therrien et al. (Geology, Univ. Laval, Quebec)
• Forsyth (Computer Science, Univ. Waterloo,
  Ontario)
• Panday, Guvanasen Huyakorn (HydroGeoLogic,
  Herndon, VA)
• Matanga et al. (US BOR, Sacramento, CA)

6
Overview of Current HydroGeoSphere Model
Features

• 2D overland/stream flow (Diffusion-wave
  equation), including stream/surface drainage
  network genesis
• 3D variably-saturated flow (Richards equation
  ET) in porous medium
• 3D variably-saturated flow in macropores,
  fractures and karst conduits (dual-porosity,
  dual-permeability or discrete fractures)
• Advective-dispersive, reactive solute/thermal
  transport in all continua, snow
  accumulation/melting, soil freeze/thaw
• Groundwater age, life expectancy (Park et al.,
  WRR, 2008 Cornaton et al., WRR, 2008)
• Allows for complex topography, irregular surface
  subsurface properties, density-dependent flow,
  subgridding subtiming (Park et al., AWR, 2008)
• Fully-coupled, simultaneous solution of
  surface/subsurface flow and transport via
  Control-Volume Finite Element Method.

7
Integrated Model Parameterization

• Geology and Stratigraphy Aquifer
  characterization
• DEM Surface elevation, Channel definition
• Radar data Precipitation
• Met. stations Temp., Humidity, Wind, Reference
  Evapotranspiration
• Hydrology databases Lakes, ponds and hydraulic
  structures location, connectivities and
  characteristics structure operations channel
  geometries
• Well permitting databases pumping estimates,
  depths

8
Integrated Model Parameterization Continued

• GIS coverages surface parameterization
• Land-Use / Land Cover (at various times) LAI
  Canopy Interception Mannings friction Root
  Distribution Evaporation Distribution Surface
  Leakance Crop Coefficient Rill and Obstruction
  heights
• Soils Data Moisture retention, porosity
  hydraulic conductivity field-capacity and
  wilting-point saturations
• Other GIS coverages Hydrography Septic systems
  Agricultural returns Water Supply System Service
  Areas pumping locations Geographic landmarks

9
Catchment-Scale Study Laurel Creek Watershed,
Waterloo, Ontario, Canada
10
Laurel Creek
Rural
Urban (Waterloo)
Discharge to Grand River
11
Physical System Geometry
12
Land Use
Manning Coefficients

• Water
• Wetland
• Forest
• Urban
• Agricultural

0.04 0.05 0.6 0.012 0.2
13
Surficial Soils
14
Subsurface Hydrostratigraphy
15
Simulated vs. Observed Drainage
16
Simulated vs. Observed Hydrograph
Simulated
Observed
17
Water Exchange Fluxes
Note negative values denote discharge from
subsurface to surface
18
Analysis of the Impact of Surficial Contaminant
Releases on Stream Water Quality
19
Subcatchment of Laurel Creek Watershed,Waterloo,
Ontario, Canada
20
Surficial Contaminant Source
Rainfall to the surface is the only water into to
the system
First-type solute boundary condition Co 1.0
Critical depth boundary
21
Plume Migration on Land Surface
t 0
Under mean annual rainfall of 0.36 m/year
22
Plume Migration on Land Surface
t 1 year
Under mean annual rainfall of 0.36 m/year
23
Plume Migration on Land Surface
t 2 years
Under mean annual rainfall of 0.36 m/year
24
Plume Migration on Land Surface
t 3 years
Under mean annual rainfall of 0.36 m/year
25
Plume Migration on Land Surface
t 5 years
Under mean annual rainfall of 0.36 m/year
26
Plume Migration on Land Surface
t 10 years
Under mean annual rainfall of 0.36 m/year
27
Plume Migration on Land Surface
t 30 years
Under mean annual rainfall of 0.36 m/year
28
Breakthrough Curve at Stream Outlet
29
Subsurface Plume Migration
Under mean annual rainfall of 0.36 m/year
Cross-Section
30
Subsurface Plume Migration
t 0
Under mean annual rainfall of 0.36 m/year
31
Subsurface Plume Migration
t 1 year
Under mean annual rainfall of 0.36 m/year
32
Subsurface Plume Migration
t 2 years
Under mean annual rainfall of 0.36 m/year
33
Subsurface Plume Migration
t 3 years
Under mean annual rainfall of 0.36 m/year
34
Subsurface Plume Migration
t 5 years
Under mean annual rainfall of 0.36 m/year
35
Subsurface Plume Migration
t 10 years
Under mean annual rainfall of 0.36 m/year
36
Subsurface Plume Migration
t 30 years
Under mean annual rainfall of 0.36 m/year
37
Solute Breakthrough at Channel Outlet
38
Solute Exchange Fluxes at Stream Nodes A, B and C
Stream Node A
Stream Node B
Stream Node C
39
Application of HydroGeoSphere to a Large-scale
Watershed in 3D Duffins Creek, Ontario
40
Duffins Creek Watershed

• 286 km2 in area
• Hydro-eco concerns due to urban development

41
Geology and Hydrogeology

• Bedrock shale Whitby Formation (Late Ordovician)
• Quaternary sediments 0 m (absent) to 200 m thick
• Eight hydrostratigraphic units including three
  aquifers

42
Discretization
43
Calibration Results Steady-State Subsurface Heads
RMS error 14 m
44
Calibration Results Stream Baseflow
Mean error 0.0024 m3/s
45

Steady-state Surface Drainage Network
Surface/Subsurface Exchange Fluxes
Surface Water Depths
Surface/Subsurface Exchange Fluxes
46
Transient Model Hydrographs
47
Impact of The Wisconsinian Glaciation on Canadian
Continental Groundwater Flow
48
Last Glacial Maximum (-21kyr)
49
Computational Domain
50
Wisconsinian Glaciation Progression
GM3
GM2
GM1
51
Topography Ice Sheet Thickness
Tarasov and Peltier, 2005
52
Isostasy
Surface elevation and ice thickness variation in
Waterloo, Ontario
53
Ice Sheet Margins Surface Water Depths
Tarasov and Peltier, 2005
54
Permafrost Evolution
Last Interglacial
Last Glacial Maximum
Tarasov and Peltier, 2005
55
Subglacial Meltwater Production at LGM
Tarasov and Peltier, 2005
56
Hydraulic Properties Distribution
57
Boundary Conditions

• Model Forcing University of Toronto Glacial
  Systems Model
• Ground surface elevation
• Relative sea Level
• Ice thickness

• Permafrost thickness
• Meltwater rate
• Surface water depth

58
Cross Section Location
59
Geological Cross Section
Vertical exaggeration 125
60
Hydraulic Head and Permafrost Distribution
Hydraulic Head
Vertical exaggeration 100
Permafrost
61
Relative Concentration at LGM (1 265 g/l)
Groundwater Age Distribution at LGM
Vertical exaggeration 100
62
Surface/Subsurface Water Interaction

• Groundwater Recharge and Meltwater Ratio

63
Impact of Climate Change on Canadian Water
Resources

• Full coupling of HydroGeoSphere with NCARs
  Community Climate System Model (CCSM)
• Replace CCSM land surface module with
  HydroGeoSphere to include a groundwater component
• Examine impact of global warming on both surface
  and groundwater resources (quantity and quality)
• Basin to Canadian-landmass scale simulations

64
Community Climate System Model
Atmosphere
Coupler
Land
HydroGeoSphere
Ice
Ocean
65
Model Domain
66
3D Grid Resolution for Future Climate Computations

• covers entire Canadian landscape
• timescales of 1-2 centuries
• includes both unsaturated
• saturated zones
• framework could be applied to
• other regions

67
HydroGeoSphere Computed (Hydrodynamic) Surface
Water Depths For Present-Day Climate
68
Lessons from Experience

• Integrated models are successful in
  characterizing hydrologic cycle processes at
  multiple scales in watersheds
• Parameterization There is always missing data
• Computational challenges remain, but with modern
  numerical solution methods HPCs, there is
  optimism for handling very large complex systems
• Fully coupled solution is robust and provides a
  holistic view of water, contaminant heat
  transport
• Model uncertainty is reduced due to simultaneous
  examination of coupled processes