Transcript PowerPoint Presentation - Impact of Climate Change on Flow

Climate Change Impacts on the
Hydrology of the Upper
Mississippi River Basin
Eugene S. Takle
with significant assistance from
Manoj Jha, Chris Anderson, Phil
Gassman, and Mahesh Sahu
Atmospheric Science Seminar, ISU, 13 September 2005
If we had perfect predictability of
low-resolution global climate
fields, how well can we
downscale this predictability to
stream flow at one point?
Outline



Upper Mississippi River Basin
Soil and Water Assessment Tool (SWAT)
Climate information






Observations
Contemporary climate
Future climate
Flow simulations
Water quality
Summary
Sub-Basins of the
Upper Mississippi
River Basin
119 sub-basins
Outflow measured
at Grafton, IL
Approximately one
observing station
per sub-basin
Approximately one
model grid point
per sub-basin
Soil Water Assessment Tool
(SWAT)
Long-term, continuous watershed
simulation model (Arnold et al,1998)
 Daily time steps
 Assesses impacts of climate and land
management on yields of water, sediment,
and agricultural chemicals
 Physically based, including hydrology, soil
temperature, plant growth, nutrients,
pesticides and land management

Calibration of SWAT:
Annual Stream Flow at Grafton, IL
Calibration of SWAT:
Monthly Stream Flow at Grafton, IL
Validation of SWAT:
Annual Stream Flow at Grafton, IL
Validation of SWAT:
Monthly Stream Flow at Grafton, IL
Downscaling Methods




Dynamical downscaling (use GCM to
provide b.c. for RCM)
Statistical or empirical transfer functions to
relate local climate to GCM output
Climate analog procedures
Combinations of statistical and dynamical
methods
Downscaling
For applications of global
climate model results to
hydrology, there is a significant
mismatch between the spatial
scales of the model resolution
and features of drainage basins.
Approximate locations
of points for a 2.5o x
2.5o global model grid
RegCM2 Simulation Domain
Red = global model grid point
Green/blue = regional model grid points
SWAT Output with Various Sources
of Climate Input
Annual Stream Flow Simulated by SWAT
Driven by the RegCM2 Regional Climate
Model with NNR Lateral Boundary Conditions
Seasonal Stream Flow Simulated by SWAT
Driven by the RegCM2 Regional Climate
Model with NNR Lateral Boundary Conditions
Mean Monthly Precipitation Simulated by
the RegCM2 Regional Climate Model
with NNR Lateral Boundary Conditions
120
(a)
RegCM2
Snowfall (mm)
100
SWAT
80
60
40
20
0
Jan
Feb
Mar
Apr
May
Jun
Jul
Aug
Sep
Oct
Nov
Dec
60
(b)
RegCM2
SWAT
50
Runoff (mm)
40
30
20
10
0
Jan
Feb
Mar
Apr
May
Jun
Jul
Aug
Sep
Oct
Nov
Dec Aver.
120
(c)
RegCM2
Evapotranspiration (mm)
100
SWAT
80
60
40
20
0
Jan
Feb
Mar
Apr
May
Jun
Jul
Aug
Sep
Oct
Nov
Dec
120
(d)
RegCM2
100
Snow melt (mm)
SWAT
80
60
40
20
0
Jan Feb Mar
Apr May
Jun
Jul Aug Sep Oct
Nov Dec
Hydrological component comparison
between RegCM2 and SWAT
RegCM2
SWAT
Evapotranspiration
588
528
Surface runoff
151
166
Snowmelt
256
240
Note: All values are in mm per year averaged for 1980-1988 in NNR run.
Ten-Year Mean Precipitation Generated by the RegCM2
Regional Climate Model Driven with HadCM2
Global Model Results for the Contemporary and
Future Scenario (2040s) Climate
Ten-Year Mean Monthly Stream Flow Generated by the
RegCM2 Regional Climate Model Driven
with HadCM2 Global Model Results for the
Contemporary and Future Scenario (2040s) Climate
Hydrologic Budget Components
Simulated by SWAT under Different Climates
Hydrologic budget
components
Calibration
(19891997)
Validation
(19801988)
NNR
(19801988)
CTL
(around
1990s)
SNR
(around
2040s)
% Change (SNR-CTL)
Precipitation
856
846
831
898
1082
21
Snowfall
169
103
237
249
294
18
Snowmelt
168
99
230
245
291
19
Surface runoff
151
128
151
178
268
51
GW recharge
154
160
134
179
255
43
Total water yield
273
257
253
321
481
50
Potential ET
947
977
799
787
778
-1
Actual ET
547
541
528
539
566
5
All units are mm
Yield is sum of surface runoff, lateral flow, and groundwater flow
Relation of Runoff to Precipitation
for Various Climates
Summary of RCM Studies



RCM provides meteorological detail needed by
SWAT to resolve sub-basin variability of
importance to streamflow
There is strong suggestion that climate change
introduces changes of magnitudes larger than
variation introduced by the modeling process
Relationship of streamflow to precipitation might
change in future scenario climates
Alternative to Dynamical
Downscaling Global Model Results
“…a de facto minimum standard of any useful downscaling
method for hydrologic applications: the historic (observed)
conditions must be reproducible.” Wood, et al., 2004: Climatic
Change 62:189



Linear interpolation of GCM results
Spatial disaggregation
Bias corrected spatial disaggregation
Note: These methods could be applied to downscaled (RCM)
results as well.
Hypothesis:
Simple linear interpolation of global climate model
results as input to SWAT is incapable of reproducing
historical (observed) hydrological conditions in the
Upper Mississippi River Basin
Global models used in the SWAT-UMRB simulations (20C)
Institution
Model Name
Lon x Lat
Resolution
W/m2
Cl. Sens
NOAA Geophysical Fluid Dynamics Laboratory
(USA)
GFDL-CM 2.0
2.5 o x 2.0 o
2.9
NOAA Geophysical Fluid Dynamics Laboratory
(USA)
GFDL-CM 2.1
2.5 o x 2.0 o
2.0
Center for Climate System Research (Japan)
MIROC3.2(medres)
2.8 o x 2.8 o
1.3
Center for Climate System Research (Japan)
MIROC3.2(hires)
1.125 o x 1.125o
1.4
Meteorological Research Institute (Japan)
MRI
2.8 o x 2.8 o
0.86
NASA Goddard Institute for Space Studies
(USA)
GISS-AOM
4 o x 3o
2.6
NASA Goddard Institute for Space Studies
(USA)
GISS-ER
5 o x 4o
2.7
Institut Pierre Simon Laplace (France)
IPSL-CM4.0
3.75 o x 2.5 o
1.25
Canadian Centre for Climate Modeling &
Analysis Canada)
CGCM3.1(T47)
3.8 o x 3.8 o
n/a
O
C
M
R
I
L
s
_m
ed
er
s
ns
em
bl
e
ul
tim
od
el
E
M
IR
_h
ire
IP
S
IS
S_
ER
O
C
G
M
2.
1
2.
0
3.
1
_A
O
L
L
M
FD
IS
S
M
IR
G
G
C
FD
G
G
C
BS
ag
e
O
G
Streamflow (mm)
Simulation of Streamflow by 9 Global Models and Model Ensemble
800
700
600
500
400
300
200
100
0
P-values of T-test of individual GCM/SWAT streamflow
and pooled GCM/SWAT streamflow (labeled as GCM
POOL) compared to OBS/SWAT
GCMs
P-value
GFDL-CM 2.0
4.8303E-17
GFDL-CM 2.1
3.3774E-5
MIROC3.2(medres)
4.1050E-5
MIROC3.2(hires)
MRI
0.8312
0.3963E-8
GISS-AOM
0.0098
GISS-ER
0.0124
IPSL-CM4.0
0.0050
CGCM3.1(T47)
0.0229
GCM POOL
0.5979
Hypothesis:
Simple linear interpolation of global climate model
results as input to SWAT is incapable of reproducing
historical (observed) hydrological conditions in the
Upper Mississippi River Basin
Results:
Hypothesis is true for individual models (except
MIROC-hires)
Hypothesis is false for MIROC-hires
Hypothesis is false for the ensemble of GCMs
Hydrological components simulated by SWAT.
Hydrological
components
OBS/
SWAT
(19681997)
HadCM2/
Measured
RegCM2
Data
~1990
GFDLCM 2.0
GFDLCM 2.1
MIROC
MIROC
3.2
3.2 (hires)
(medres)
MRI
GISS_AO GISS-ER
M
IPSL-CM
4.0
CGCM
3.1
Precipitation
846
846
900
1032
910
736
821
707
746
746
793
859
Snowfall
118
-
244
213
196
110
104
134
125
95
202
140
Snowme lt
116
-
241
211
193
107
100
130
120
94
200
138
Surface runoff
100
-
148
215
140
55
75
58
63
51
147
80
Baseflow
181
-
213
330
223
145
213
109
170
182
196
161
Potential ET
967
-
788
759
854
1054
984
1011
744
729
692
970
Evapotranspiration (ET)
557
-
533
484
540
531
527
532
505
506
445
611
Total water yield
275
253
350
531
353
194
279
162
227
227
336
232
113
-
101
110
78
78
88
66
63
56
87
86
84
81
95
108
55
58
69
50
52
53
90
57
Standard
Precipitation
Deviation
of annual
Streamf low
values
Notes: 1. Measured streamflow data is at Grafton, IL (USGS gage # 05587450).
2. All values are average annual values (in mm) averaged over 1963-2000 (unless otherwise specified); Years 1961 and 1962
are simulated as initialization period.
3. HadCM2/RegCM2 SWAT simulations (Jha et al., 2004) are average over 10-year period.
Table 4.for
Results
for the multi-model
ensemble
mean
SWAT driv
Results
the multi-model
ensemble
mean
of of
SWAT
driven by GCMs and observed meteorological conditions
Table 4. Result s for the mult i-model ensemble mean of SWAT driven by GCMs and
for sub-periods
ofthethe
observed meteorological conditions
for sub-periods of
20C.20C
Hydrological
components
OBS/
SWAT
Measured
data
Precipitation
846
Snowfall
GCM/SWAT
MIROC 3.2 (hires)
HadCM2/RegCM2
/SWAT
Amount % Diff .
Mean
% diff .
Amount
% diff .
846
817
-03
821
-03
900
+ 06
118
-
147
+25
104
-12
244
+206
Snowme lt
116
-
144
+24
100
-13
241
+208
Surface runoff
100
-
98
-02
75
-25
148
+ 48
Baseflow
181
-
192
+06
213
+18
213
+ 18
Potential ET
967
-
866
-10
984
+02
788
- 15
ET
557
-
520
-07
527
-05
533
- 04
Total water yield
275
253
282
+11
279
+10
350
+ 38
Note: Percent differences are calculated from measured data when available and
otherwise from results of SWAT driven by observed meteorology. Different averaging
periods were used as follows: OBS/SWAT : 1968-1997; GCM/SWAT : 1963-2000;
MIROC3.2 (hires)/SWAT : 1963-2000; and HadCM2/RegCM2/SWAT : 1990-1999.
Results in the last two columns are from Jha et al. (2004).
Global models used in the SWAT-UMRB simulations
(2082-2099)
Institution
Model Name
Lon x Lat
Resolution
W/m2
Cl.
Sens
NOAA Geophysical Fluid Dynamics
Laboratory (USA)
GFDL-CM 2.0
2.5 o x 2.0 o
2.9
Center for Climate System Research
(Japan)
MIROC3.2(medr
es)
2.8 o x 2.8 o
1.3
Center for Climate System Research
(Japan)
MIROC3.2(hires)
1.125 o x
1.125o
1.4
Meteorological Research Institute
(Japan)
MRI
2.8 o x 2.8 o
0.86
NASA Goddard Institute for Space
Studies (USA)
GISS-AOM
4o x 3o
2.6
NASA Goddard Institute for Space
Studies (USA)
GISS-ER
5o x 4o
2.7
Institut Pierre Simon Laplace (France)
IPSL-CM4.0
3.75 o x 2.5 o
1.25
Model biases and climate change for each hydrological
cycle component (2082-2099)
Hydrologic
Component/
Model
Bias(%)
Change (%)
Precipitation
GFDL 2.0
Hydrologic
Component/
Model
Change (%)
Snowfall
22
1
GISS AOM
-12
17
GISS AOM
GISS ER
-12
25
GISS ER
IPSL
-6
0
MIROC-hi
-3
-4
MIROC-med
-13
-12
MRI
-16
16
-6
6
Mean
Bias(%)
GFDL 2.0
81
-32
6
-22
-19
3
71
-43
-12
-80
MIROC-med
-7
-65
MRI
13
-18
Mean
19
-37
IPSL
MIROC-hi
Model biases and climate change for each hydrological
cycle component (2082-2099)
Hydrologic
Component/
Model
Bias(%)
Change
(%)
Hydrologic
Component/
Model
Snowmelt
GFDL 2.0
Bias(%)
Change (%)
Runoff
83
-32
GFDL 2.0
155
-30
5
-20
GISS AOM
-24
-2
-19
5
GISS ER
-39
32
73
-43
IPSL
73
-31
-12
-79
MIROC-hi
-9
-38
MIROC-med
-6
-65
MIROC-med
-30
-63
MRI
13
-17
MRI
-21
-7
Mean
20
-36
Mean
15
-20
GISS AOM
GISS ER
IPSL
MIROC-hi
Model biases and climate change for each hydrological
cycle component (2082-2099)
Hydrologic
Component/
Model
Bias(%)
Change (%)
Baseflow
GFDL 2.0
Hydrologic
Component/
Model
Bias(%)
Change (%)
Potential ET
176
4
GFDL 2.0
-54
45
GISS AOM
50
43
GISS AOM
-42
5
GISS ER
76
45
GISS ER
-49
5
IPSL
22
-5
IPSL
-34
46
MIROC-hi
63
-12
MIROC-hi
-24
37
MIROC-med
27
-32
MIROC-med
-29
32
MRI
11
38
MRI
-34
14
Mean
61
12
Mean
-38
27
Model biases and climate change for each hydrological
cycle component (2082-2099)
Hydrologic
Component/
Model
Bias(%)
Change (%)
ET
Hydrologic
Component/
Model
Bias(%)
Change (%)
Total Water
GFDL 2.0
-37
16
GISS AOM
-26
7
GISS ER
-30
IPSL
GFDL 2.0
154
-8
GISS AOM
16
33
12
GISS ER
27
43
-25
12
IPSL
33
-17
MIROC-hi
-18
6
MIROC-hi
29
-18
MIROC-med
-20
3
MIROC-med
0
-40
MRI
-22
12
MRI
-7
25
Mean
-25
10
Mean
36
3
Preliminary Interpretation




Models consistently under-estimate ET
and PET (likely due to coarse resolution)
Low ET forces more water to baseflow
High baseflow increases total water yield
Hence I assert that low-resolution models
over-predict streamflow because they are
incapable of resolving high daily max
temps that have a disproportionate
influence on ET
Current Work




Look at more global models
Look at ensembles of individual models
Look at the low, medium, and highresolution results for MIROC
Extend SWAT to better simulate subsurface effects of riparian buffer strips
(Mahesh Sahu)
Application of SWAT model
to simulate riparian buffer
zone
Mahesh Sahu, Graduate Research Assistant CCEE
Present Scheme in Swat for
riparian buffer simulation
Crop
Crop
Buffer Strip
Buffer Strip
River
River
Hill Slope scheme
Conventional SWAT: Present
Mahesh Sahu, Graduate Research Assistant CCEE
Flows simulated in SWAT
Existing Hill slope option & its limitations
Q surface
Q lateral
Q GW
The existing hill slope option has the capability to incorporate the surface
flow from the crop area through the buffer zone area. Lateral and
groundwater flow links are NOT present in the existing hill slope scheme.
Mahesh Sahu, Graduate Research Assistant CCEE
Future Directions


Couple GCM, RCM, SWAT, Crop Model
and Economic Model
Evaluate policy alternatives:
Impact of introducing conservation practices
 Impact of introducing incentives


Hypothesis:
It is possible to balance profitability with
sustainability in an intensively managed
agricultural area under changing climate
through development of robust policy
GCM
OBS
NNR
RCM
Climate
Over UMRB
Crop
Model
Crop Yield
Soil
Drainage
Land-use
SWAT
Management
Choices
Economic
Model
OBS
Stream
flow
Soil
Carbon
Crop
Production
Water
Quality
Evaluate Sustainability
and Profitability
Incentives
Public
Policy
Summary


Changes to the hydrological cycle
associated with climate change are of high
societal importance
Dynamical downscaling of global model
results by a regional model gives 20%
increase in precipitation in the basin and
50% increase in streamflow
Summary



Linear interpolations of individual low-resolution
GCMs are incapable of simulating historical
streamflows in the UMRB
Linear interpolation of a high-resolution global
model is capable of simulating historical
streamflows in the UMRB
An ensemble of linear interpolations of individual
low-resolution GCMs is capable of simulating
historical streamflows in the UMRB