Climate change impacts on hydrology The chain of

Climate Change Impacts on Groundwater
Warsaw 12-13 October 2011
Uncertainties in climate change
impact modeling
(groundwater)
Torben O. Sonnenborg
Jens Christian Refsgaard
Geological Survey of Denmark and Greenland (GEUS)
Ministry of Climate and Energy
Acknowledgements
•
•
Lieke van Roosmalen, Flinders University Adelaid (previously
PhD student, GEUS/University of Copenhagen)
Lauren P. Seaby, PhD student, GEUS
Dorthe Seifert, Post Doc, GEUS/ALECTIA A/S
•
Other colleagues at GEUS
•
HYACINTS team – Research project: “Hydrological Modelling for
Assessing Climate Change Impacts at different Scales”
http://www.hyacints.dk
•
Outline
1. Why is uncertainty particularly important in climate
change studies?
2. Climate change impact predictions for groundwater –
methodology illustrated by example
3. Uncertainty in climate change impact predictions cascade of uncertainties
4. Example – uncertainty on climate projections versus
uncertainty on geology
5. Conclusions
Why is uncertainty particularly
important in relation to climate change?
• Hydrological models used for climate impact
predictions can not be calibrated against data from
future climate conditions larger prediction
uncertainties
Calculations of climate change
effects on hydrology
Present
climate
Downscaling
Bias
correction
Future
climate
Global
Regional
Hydrological
Models
100-250 km
10-25 km
50-500 m
Scale
DK-model
- The hydrological model
Seven sub-models
– Horisontal discretization:
500 m
– Vertical discretization –
layers; varying numbers
and geometry
Submodel
Total area
(km2)
Active grids
pr. layer
Omr. 1 – Sjælland
7163
37569
Omr. 2 – Sydhavsøerne
2042
13885
Omr. 3 – Fyn
3473
24009
Omr. 4 – Sønderjylland
7897
35869
Omr. 5 – Midtjylland
11578
49993
Omr. 6 – Nordjylland
9934
47649
Omr. 7 – Bornholm
2358
10106
Model code
MIKE SHE/MIKE11
3D groundwater flow
2D overland flow
Drain flow (pipes, ditches)
1D river routing
1D unsaturated zone,
(evapotranspiration)
• Degree-day snow
melt/accumulation
•
•
•
•
•
Geology/
hydrostratigraphy
• Borehole data
• Geological interpretation
• Incorporation of
knowledge from more than
50 local geological models
established by regional
authorities (incl.
geophysical data)
DK-model data basis beyond geology
Model setup
• Rivers
•
•
River cross-sections (MIKE11) for all major streams
Discharge from urban sewage treatment plants
• Water supply – all as groundwater abstraction
•
23,500 plants (40,000 intakes) for water supply, including
irrigation
• Precipitation: DMI’s 10 km grid daily values
• Temperature, potential evapotranspiration: DMI’s 20 km
grid
Model calibration/validation (1990-2006)
• 183 stream discharge stations, daily values
• > 10,000 wells with groundwater head observations
Example of predictions of climate
change impacts on hydrology
(van Roosmalen et al., 2007, 2010)
Climate model projections:
• A2 and B2 scenarios
• HIRHAM
• 2071-2100
Hydrological model
• DK-model
Impacts on
• Groundwater heads
• River discharges
Western
Jutland
5460 km2
Zealand
7230 km2
Change in groundwater head and
discharge
Discharge,
Skjern
river
Skjern
Å
Relativ ændring
Relative
change(%)
(%)
40
Change in head (m)
A2
30
B2
20
10
0
-10
J F M A M J J A S O N D
-20
Måned
Month
Van Roosmalen et al. (2007)
Climate change impacts on hydrology
The cascade of uncertainties
• Emission scenarios (CO2)
Present
climate
Downscaling
Bias
correction
Future
climate
Global
Regional
Hydrological
Models
100-250 km
10-25 km
50-500 m
Scale
IPCC Greenhouse Gas Emission Scenarios
Climate change impacts on hydrology
The chain of uncertainties
• Emission scenarios
• Climate models (GCM + RCM)
Present
climate
Down
scaling
Bias
correction
Future
climate
Global
Regional
Hydrological
Models
100-250 km
10-25 km
50-500 m
Scale
Uncertainties on climate models’ projections
- Delta change factors on precipitation 2071-2100
1,6
1,6
1,6
Factor
Change
Delta
Factor
Change
Delta
Factor
Change
Delta
1,4
1,4
1,4
ARPEGE-CNRM
ARPEGE-DMI
BCM-DMI
BCM-SMHI
ARPEGE-DMI
ECHAM-DMI
ECHAM-ICTP
ECHAM-KNMI
ARPEGE-DMI
ECHAM-MPI
ECHAM-SMHI
HADQ0-ETHZ
BCM-DMI
HADQ0-HC
Mean
1,2
1,2
1,2
1
11
0,8
0,8
0,8
0,6
0,6
0,4
0,4
0,4
Jan
Jan
Jan
Feb
Feb
Feb
Mar
Mar
Mar
Apr
Apr
Apr
Lauren P Seaby
PhD project www.hyacints.dk
Preliminary results
May
May
May
Jun
Jun
Jun
Jul
Jul
Jul
Aug
Aug
Aug
Sep
Sep
Sep
Oct
Oct
Oct
Nov
Nov
Nov
Dec
Dec
Dec
Data from 11 climate models
in the ENSEMBLES project
(A1B)
Climate change impacts on hydrology
The chain of uncertainties
• Emission scenarios
• Climate models (GCM + RCM)
• Downscaling / bias correction
Present
climate
Down
scaling
Bias
correction
Future
climate
Global
Regional
Hydrological
Models
100-250 km
10-25 km
50-500 m
Scale
HIRHAM model results
- A2 scenario 2071-2100, different resolution and sea
surface temperature forcings over Baltic Sea
Present climate/control period (1961-90)
Future climate (2071-2100)
Van Roosmalen et al. (2010)
Bias correction (statistical downscaling) of precipitation
Delta Change Method
(correction of observed precipitation)
Skalering
af nedbør
nedbør
Observeret
Observed
precipitation
Scaled precipitation
Nedbør (mm/dag)
Transformation
of precipitation
Critical
assumption:
40
30
M fut = present dynamics
Future
dynamics
35
25
Pfut
change
Pobs in number of rainfall days
No
30
M cont
20
No
change in distribution of rainfall intensity
25
15
20
Etc.
M:15Mean monthly precipitation (30 years period)
10
: Future climate
10Mfut
5
5M
cont : Present climate (control period)
0
1-12-99
11-12-99
21-12-99
Dato
Observeret
Delta Change
Observeret
31-12-99
Statistical downscaling/bias
correction
• Many different methods for making statistical
downscaling  different results
• We cannot know beforehand which downscaling
method will turn out to be the best one
• Example – comparison of two methods for future
precipitation
• Delta change (monthly correction factors to observed
precipitation)
• Direct method – Histogram Equalisation Method
(Gamma function correction of RCM simulated
precipitation)
Statistical downscaling of precipitation
- Delta change versus Direct method
6000
50
Antal hændelser
40
4000
3000
30
20
10
2000
75-80
70-75
65-70
60-65
55-60
50-55
45-50
35-40
30-35
40-45
30-35
25-30
0
1000
Nedbørsintensitet
(mm/dag)
Intensity
(mm/day)
RCM
Delta Change
Direkte
75-80
70-75
65-70
60-65
55-60
50-55
45-50
40-45
35-40
20-25
15-20
10-15
7-10
5-7
3-5
1-3
0.1-1
0
0
Antal hændelser
5000
Climate change impacts on hydrology
The chain of uncertainties
•
•
•
•
Emission scenarios
Climate models (GCM + RCM)
Downscaling / bias correction
Hydrological model (geology, process equations, parameter
values, input data)
Present
climate
Down
scaling
Bias
correction
Future
climate
Global
Regional
Hydrological
Models
100-250 km
10-25 km
50-500 m
Scale
Uncertainty on climate models
versus geological uncertainty
Case study
Dorthe Seifert, ALECTIA
Postdoc www.hyacints.dk
Preliminary results
Uncertainty on climate models
versus geological uncertainty
Step 1
Six different
geological models
Step 2
Six groundwater models.
Calibrated against head
and discharge data
Step 3
Case study
Dorthe Seifert, ALECTIA
Postdoc www.hyacints.dk
Preliminary results
Apply climate 2071-2100
from 11 different climate
models
Predicted change in groundwater head
- 11 climate model projections 2071-2100
- 6 geologies/groundwater models
ECHAM-MPI
ECHAM-SMHI
HADQ0-ETHZ
HADQ0-HC
-1.08
0.14
0.40
0.60
0.14
0.34
0.15
0.40
0.28
-0.47
-0.02
0.60
N2
-0.94
-0.88
0.12
0.33
0.49
0.11
0.27
0.12
0.32
0.23
-0.41
-0.02
0.49
R1
-0.73
-0.69
0.11
0.31
0.48
0.13
0.28
0.13
0.33
0.24
-0.32
0.02
0.42
R2
-0.61
-0.59
0.08
0.25
0.39
0.09
0.21
0.10
0.25
0.18
-0.28
0.01
0.34
L1
-1.01
-1.00
0.18
0.49
0.71
0.23
0.45
0.21
0.50
0.39
-0.39
0.07
0.60
L2
-1.12
-1.07
0.19
0.52
0.82
0.23
0.48
0.21
0.55
0.40
-0.47
0.07
0.66
Mean models
-0.93
-0.89
0.14
0.38
0.58
0.16
0.34
0.15
0.39
0.29
-0.39
0.02
Std. models
0.22
0.21
0.04
0.10
0.16
0.06
0.11
0.05
0.11
0.09
0.08
Decrease in Dhmean
Increase in Dhmean
Std. clim-.mod
ECHAM-KNMI
Mean clim. mod.
ECHAM-ICTP
-1.16
(m)
ECHAM-DMI
BCM-DMI
N1
Dhmean
BCM-SMHI
ARPEGE-DMI
ARPEGE-CNRM
Mean change over model area: Dh = hfuture - hpresent
24
Conclusions
• Climate change predictions involves large
uncertainties  cascade of uncertainties
• Uncertainty on climate model projections often
dominating
• Different geological interpretations result in different
predictions on climate change impacts on
groundwater
Further information
Journal papers
•
van Roosmalen L, Christensen BSB, Sonnenborg TO (2007) Regional differences
in climate change impacts on groundwater and stream discharge in Denmark.
Vadose Zone Journal, 6, 554–571.
•
van Roosmalen L, Christensen JH , Butts MB, Jensen KH, Refsgaard JC (2010)
An intercomparison of regional climate model data for hydrological impact
studies in Denmark. Journal of Hydrology, 380, 406-419.
•
van Roosmalen L, Sonnenborg TO, Jensen KH, Christensen JH (2011)
Comparison of hydrological simulations of climate change using pertubation of
observation and distribution-based scaling. Vadose Zone Journal, 10, 136-150.
•
Henriksen HJ, Troldborg L, Nyegaard P, Sonnenborg TO, Refsgaard JC, Madsen
B (2003) Methodology for construction, calibration and validation of a national
hydrological model for Denmark. Journal of Hydrology, 280(1–4), 52–71.
•
Henriksen HJ, Troldborg L, Højberg AL, Refsgaard JC (2008) Assessment of
exploitable groundwater resources of Denmark by use of ensemble ressource
indicators and numerical groundwater-surface water model. Journal of
Hydrology, 348(1-2), 224-240.
Hydrological Modelling for Assessing Climate
Change Impacts at different Scales
http://www.hyacints.dk