Alan F. HamletAnthony L. WesterlingTim P. BarnettDennis P. Lettenmaier
•JISAO/CSES Climate Impacts Group•Dept. of Civil and Environmental Engineering
University of Washington•Scripps Institute of Oceanography•School of Engineering, University of California, Merced
Late 20th Century Precipitation Variability in the Western U.S. in the Context of Long-Term Climate Variability and Global Change
DJF Temp (°C) NDJFM Precip (mm)
PNW
CA CRB
GB
Cool Season Climate of the Western U.S.
Snow Model
Schematic of VIC Hydrologic Model and Energy Balance Snow Model
PNW
CACRB
GB
Evaluation of Streamflow Simulations of the Colorado River at Lee’s Ferry, AZ
2000
3000
4000
5000
6000
7000
8000
9000
1916
1921
1926
1931
1936
1941
1946
1951
1956
1961
1966
1971
1976
1981
1986
1991
1996
2001
Str
eam
flo
w (
cms)
OBS
REG
0
50
100
150
200
250
300
350
400
450
1916
1921
1926
1931
1936
1941
1946
1951
1956
1961
1966
1971
1976
1981
1986
1991
1996
2001
Str
eam
flo
w (
cms)
OBS
REG
R2 = 0.83
R2 = 0.91
Columbia River
Sacramento River
Cool Season Precipitation Explains Most of the Variability in Annual Flow in the PNW and CA
Relationship Between Annual Flow and
Cool Season Precip.
Relationship Between Annual Flow and
Cool Season Precip.
200
300
400
500
600
700
800
900
1000
1100
1916
1921
1926
1931
1936
1941
1946
1951
1956
1961
1966
1971
1976
1981
1986
1991
1996
2001
Str
eam
flo
w (
cms)
OBS
REG
200
300
400
500
600
700
800
900
1000
1100
1916
1921
1926
1931
1936
1941
1946
1951
1956
1961
1966
1971
1976
1981
1986
1991
1996
2001
Str
eam
flo
w (
cms)
OBS
REG
R2 = 0.56
Colorado River
R2 = 0.18
Colorado River
Cool Season Precip Explains Most of the Variability in Annual Flow in the CRB, but the Summer Monsoon Also Plays a Role
Relationship Between Annual Flow and
Cool Season Precip.
Relationship Between Annual Flow and
Warm Season Precip.
Consensus Forecasts of Temperature and Precipitation Changes from IPCC AR4 GCMs
Pacific Northwest
°C
0.4-1.0°C0.9-2.4°C 1.2-5.5°C
Obse
rved 2
0th
centu
ry v
ari
abili
ty
+1.7°C+0.7°C
+3.2°C
Pacific Northwest
%
-1 to +3%-1 to +9% -2 to +21%
Obse
rved 2
0th
centu
ry v
ari
abili
ty
+1% +2%
+6%
-3
-2
-1
0
1
2
3
419
16
1920
1924
1928
1932
1936
1940
1944
1948
1952
1956
1960
1964
1968
1972
1976
1980
1984
1988
1992
1996
2000
Std
An
om
alie
s R
elat
ive
to 1
961-
1990
PNW
CA
CRB
Regionally Averaged Cool Season Precipitation Anomalies
PRECIP
PNW SSJ CRB 1916-1946 mean (mm) 574.7 443.9 174.7 variance 88.8 100.1 30.6 CV 0.15 0.23 0.17 lag 1auto corr -0.15 0.06 0.11 trend (% per decade) -1.1 6.9 -3.5 1947-1976 mean (mm) 640.3 477.1 168.6 variance 84.4 99.3 34.0 CV 0.13 0.21 0.20 lag 1auto corr -0.42 0.12 -0.29 trend (% per decade) 1.5 2.8 3.8 1977-2003 mean (mm) 594.3 488.1 190.8 variance 126.2 141.9 50.8 CV 0.21 0.29 0.27 lag 1auto corr 0.22 0.12 0.15 trend (% per decade) 4.2 2.4 -9.7
Summary Statistics for Regionally Averaged Cool Season
-3
-2
-1
0
1
2
319
16
1920
1924
1928
1932
1936
1940
1944
1948
1952
1956
1960
1964
1968
1972
1976
1980
1984
1988
1992
1996
2000
Std
An
om
alie
s R
elat
ive
to 1
961-
1990
PNW
CA
CRB
Regionally Averaged Warm Season Precipitation Anomalies
PRECIP
-2
-1
0
1
2
3
1917
1919
1921
1923
1925
1927
1929
1931
1933
1935
1937
1939
1941
1943
1945
1947
1949
1951
1953
1955
1957
1959
1961
1963
1965
1967
1969
1971
1973
1975
1977
1979
1981
1983
1985
1987
1989
1991
1993
1995
1997
1999
2001
Sys
tem
Wid
e H
ydro
po
wer
P
rod
uct
ion
(S
td A
no
mal
ies)
CRB
SSJ
PNW
Correlation:CRB-SSJ = 0.07CRB-PNW = 0.08SSJ-PNW = 0.36
Correlation:CRB-SSJ = 0.14CRB-PNW = -0.14SSJ-PNW = 0.06
Correlation:CRB-SSJ = 0.73CRB-PNW = 0.51SSJ-PNW = 0.65
Simulated Changes in System Wide Energy Production in the Western U.S.
DJF
Avg
Tem
p (
C)
20-year Flood for “1973-2003” Compared to “1916-2003” for a Consistent Late 20th Century Temperature Regime
X20 ’73-’03 / X20 ’16-’03
X20 ’73-’03 / X20 ’16-’03
Hamlet A.F., Lettenmaier D.P., 2007: Effects of 20th Century Warming and Climate Variability on Flood Risk in the Western U.S., Water Resour.
Res., 43, W06427
Are the changes in variability that have been observed in the last third of the 20th century consistent with normal patterns
of variability?
Long-Term Comparison of Annual Flow Records from Observations and Paleo Reconstructions
PNW:Observed (naturalized) annual flow in the Columbia River at The Dalles, OR
1858-1877 (reconstructed from observed peak river stage)1878-2003 (naturalized from observed monthly records)
CA:Reconstructed combined annual flow in the Sacramento/San Joaquin basin from tree-ring records.
(Overlapping period 1858-1977)
(Meko, D.M., 2001: Reconstructed Sacramento River System Runoff From Tree Rings, Report prepared
for the California Department of Water Resources, July)
Colorado River Basin:Reconstructed annual flow in the Colorado River at Lees Ferry, AZ from tree ring records.
(Overlapping period 1858-1977)
(Woodhouse, C.A., S.T. Gray, and D.M. Meko, 2006: Updated Streamflow Reconstructions for the Upper
Colorado River Basin, Water Resources Research, Vol. 42, W05415)
-1.5
-1
-0.5
0
0.5
1
1.5
1868
1874
1880
1886
1892
1898
1904
1910
1916
1922
1928
1934
1940
1946
1952
1958
1964
1970
1976
1982
1988
sum lag1 correl
sum inter-regional correl
sum CV
Changes in Streamflow Variability from Long-Term Observations and Paleo Reconstructions (1858-1977)
-1.5
-1
-0.5
0
0.5
1
1.5
1868
1874
1880
1886
1892
1898
1904
1910
1916
1922
1928
1934
1940
1946
1952
1958
1964
1970
1976
1982
1988
sum lag1 correl
sum inter-regional correl
sum CV
Changes in Streamflow Variability from VIC Simulations of Annual Flow (1916-2003)
-1.5
-1
-0.5
0
0.5
1
1.5
1868
1874
1880
1886
1892
1898
1904
1910
1916
1922
1928
1934
1940
1946
1952
1958
1964
1970
1976
1982
1988
sum lag1 correl
sum inter-regional correl
sum CV
Changes in Streamflow Variability from Combined Paleo Reconstructions and VIC Simulations of Annual Flow (1916-2003)
All three metrics high together
What about changes in ENSO and PDO as possible explanations?
150000
200000
250000
300000
350000
400000
450000
1900
1910
1920
1930
1940
1950
1960
1970
1980
1990
2000
Ap
r-S
ept F
low
(cfs
)
Natural Flow Columbia River at The Dalles
Patterns of ENSO Related Variability About a Shifting Long-Term Mean Seem to be Robust in the 20th Century
Could ENSO explain the lag1 and interregional metrics being anti-correlated?
What about the most recent behavior?
In periods of especially strong (weak) controls on cool season storm track behavior associated with ENSO (i.e. strong or weak NW/SW bipole), both interregional and lag1 autocorrelation would tend to be LOW (HIGH) at the same time.
The data, however, show that typically lag1 autocorrelation and interregional correlation are anti-correlated for the West as a whole. So it would seem that variations in the strength of the ENSO related NW/SW dipole does not provide an explanation of the typical behavior over most the record.
Coupled with the fact that there is little compelling evidence to suggest a systematic change in ENSO telleconnections, it seems that both the explanation for the typical behavior and the most recent changes in variability must lie elsewhere.
-1
-0.8
-0.6
-0.4
-0.2
0
0.2
0.4
0.6
0.8
119
16
1920
1924
1928
1932
1936
1940
1944
1948
1952
1956
1960
1964
1968
1972
1976
1980
1984
1988
1992
1996
2000
Std
An
om
alie
s R
elat
ive
to 1
961-
1990
(sm
oo
thed
)
PNW
CA
CRB
GB
PDO
Cool Season Precipitation Anomalies Compared to the PDO(Pattern is not robust)
-0.845-0.264-0.438-0.053
(Regional to PDO Correlation R2 )
A working hypothesis:
The most recent changes suggest:
1) Increasingly unstable storm track in cool season (increased interregional correlation)
2) Increased lag1 autocorrelation and variation in storm intensity at the scale of the Pacific Rim
Are the changes in cool season precipitation variability in the 20th
century consistent with GCM projections for the PNW?
IPCC AR4 “A2”GCM
Simulations
Large-ScaleBias Correction at
GCM Grid
UpscalingTo PNW
Overview of GCM Data Processing
CDFs Match Observations for the
Training Period1915-1964
Simple Aggregation of GCM cells over
the PNW
-3.5
-3
-2.5
-2
-1.5
-1
-0.5
0
0.5
118
65
1875
1885
1895
1905
1915
1925
1935
1945
1955
1965
1975
1985
1995
2005
2015
2025
2035
2045
2055
2065
2075
2085
2095
Oc
t-M
ar
Av
g T
em
p (
C)
-2.5
-1.64
Change in Pacific Northwest winter temperatures for HadCM3 between 1970-2000 and 2030-2060
Change = + 0.86 C *Signal to noise ratio is high*
500
550
600
650
700
75018
65
1875
1885
1895
1905
1915
1925
1935
1945
1955
1965
1975
1985
1995
2005
2015
2025
2035
2045
2055
2065
2075
2085
2095
Oc
t-M
ar
Av
g T
ota
l Pre
cip
(m
m)
Change in Pacific Northwest winter precipitation for HadCM3 between 1970-2000 and 2030-2060
Change = - 2%
628614
*Signal to noise ratio is low*
500
520
540
560
580
600
620
640
660
68018
65
1875
1885
1895
1905
1915
1925
1935
1945
1955
1965
1975
1985
1995
2005
2015
2025
2035
2045
2055
2065
2075
2085
2095
Oc
t-M
ar
Av
g T
ota
l Pre
cip
(m
m)
Change in Pacific Northwest winter precipitation for ECHAM5 between 1970-2000 and 2030-2060
Change = + 5.4%
571
602
*Signal to noise ratio is low*
Lessons Learned from 20th Century Observations
-4
-3.5
-3
-2.5
-2
-1.5
-1
-0.5
0
1903
1910
1917
1924
1931
1938
1945
1952
1959
1966
1973
1980
1987
1994
Oct
-Mar
Avg
Tem
per
atu
re (
C)
cgcm_3.1
echam5
ipsl_cm4
cnrm_cm3
giss_er
miroc_3.2
ccsm3
csiro_mk3
hadcm
pcm1
obs
ensemble mean
Linear (ensemble mean)
Linear (obs)
Comparison of 20th century winter temperature observations and 10 bias-corrected IPCC AR4 GCM simulations for the
Pacific Northwest
A2 Emissions Scenarios
500
550
600
650
700
750
800
1903
1909
1915
1921
1927
1933
1939
1945
1951
1957
1963
1969
1975
1981
1987
1993
Oct
-Mar
To
tal
Pre
cip
itat
ion
(m
m)
cgcm_3.1
echam5
ipsl_cm4
cnrm_cm3
giss_er
miroc_3.2
ccsm3
csiro_mk3
hadcm
pcm1
obs
ensemble mean
Comparison of 20th century winter precipitation observations and 10 bias-corrected IPCC AR4 GCM simulations for the
Pacific Northwest
A2 Emissions Scenarios
Evaluating Precipitation ChangesUsing a GCM Super Ensemble Approach
500
550
600
650
700
750
1903
1909
1915
1921
1927
1933
1939
1945
1951
1957
1963
1969
1975
1981
1987
1993
Oct
-Mar
To
tal
Pre
cip
itat
ion
(m
m) cgcm_3.1
echam5
ipsl_cm4
cnrm_cm3
giss_er
miroc_3.2
ccsm3
csiro_mk3
hadcm
ensemble mean
270 years1930-1959
270 years1970-1999
A super ensemble approach applied to nine GCM simulations of PNW winter precipitation for two different 30-year periods.
400
500
600
700
800
900
1000
11000
0.06
0.11
0.17
0.22
0.28
0.34
0.39
0.45 0.
5
0.56
0.61
0.67
0.72
0.78
0.83
0.89
0.95
Probability of Exceedence
Oct
-Mar
Pre
cip
itat
ion
(m
m)
1970-1999
2010-2039
2030-2059
2060-2089
Sample Size = 270 years
Super ensemble CDFs of PNW winter precipitation for four 30 year time slices from nine GCM simulations
Conclusions•Cool season precipitation is a major driver of annual river flow, hydropower production, and flood risk in the West.
•Substantial and persistent changes in cool season precipitation variability have emerged over the West since about 1975, including increased CV, within-region persistence, and inter-regional correlation.
•Long-term streamflow reconstructions show that the current changes in variability are very unusual in the context of natural variations over the last 150 years or so, and the changes are broadly consistent with GCM projections of cool season precipitation in the PNW.
•Are these systematic changes? Can they be related to changes in circulation associated with greenhouse-forced warming?
Thoughts on Planning Implications:
Even if the current precipitation regime in the West is not a systematic change, it is clear that this is something that can emerge suddenly and persist for a long time. I.e. we can expect that there may be analogous periods in the 21st century that we should be prepared to cope with.
Given the relative performance of GCMs in predicting precipitation and the inherently greater noise that is present in precipitation records, I think it is doubtful that we will have any conclusive information about whether these observed changes are related to greenhouse forcing or not.
This suggests to me that flexible approaches based on monitoring may be the only workable options.