Reclamation/DWR Project to Assess Climate Change Risks for CVP/SWP Operations
1 March 2006
Pacific Grove, CA
Levi Brekke, Reclamation D-8520
Knowing potential impacts is useful…
builds concern…
but water planners need to know what will likely happen…
Risk Assessment
1. Survey a spectrum of Scenarios2. Analyze Scenario-specific Impacts3. Estimate Scenario-specific Probabilities4. Integrate Scenario Impacts & Probabilities to
assess RISK
Probability depends on a number of uncertainties…
1. CO2Emissions Scenario
Adapted from Cayan and Knowles, SCRIPPS/USGS, 2003
?
2. Global Climate Model4. Hydrologic
Models
3. Global-to-Local “Climate Downscaling”
5. Operations Models
?
?
?
?
Climate Change Risk Assessment should represent the range of both
model options and emission possibilities
IPCC Third Assessment Report, 2001
About the Project
• Develop and demonstrate a risk-assessment methodology– Case Study: CVP/SWP– Develop information for Reclamation MP & CA-DWR
• Distribute methodology within Reclamation– Publish peer-review articles– Share methods beyond Reclamation/DWR
Acknowledgments
• Reclamation– Research Office, FY2006 funding for project #X6253– Mid-Pacific Region Division of Planning
• Collaborators:– CA DWR (Climate Change Work Team)– Ed Maurer (Santa Clara University)– Mike Dettinger (USGS/SCRIPPS)
• Software & Data:– NOAA/NWS CNRFC (E. Strem, P. Fickenscher)
Tasks
FY06 (10/1/2005 – 9/30/2006)1) Select an Ensemble of Global Climate Projections
– USGS, SCU, USBR-TSC, CA DWR-Flood
2) Spatially Downscale Global Projections to Basin-Scale Shifts in Climatological Precip & Temp
– SCU
3) Estimate Climate Projection “Distribution Function" over Northern California (based on Task 2 output)
– USGS, SCU, USBR-TSC
Tasks
FY06 (10/1/2005 – 9/30/2006)4) Assess Ensemble Impacts - Headwater Runoff
– USBR-TSC, NOAA-NWS CNRFC
5) Assess Ensemble Impacts - CVP/SWP Water– USBR-TSC, CA-DWR
Tasks
FY07 (10/1/2006 – 9/30/2007)6) Assess Ensemble Impacts– Other Conditions
– E.g., Hydropower, Aquatic Temps, Delta Salinity/Levels– USBR-TSC, CA-DWR
7) Assess Risk– Metric-specific– construct risk distribution by using Task 3 assignment of
scenario probabilities to guide resampling from the impacts ensemble– USBR-TSC, CA-DWR, USGS/SCRIPPS, SCU
Tasks
FY07 (10/1/2006 – 9/30/2007)8) Evaluate Risk Mitigation Options
– Analyze non-infrastructure strategies to manage risk– USBR-TSC; CA-DWR
9) Share Results – Reports, Journal Publications – USBR-TSC, CA DWR, USGS/SCRIPPS, SCU
Task 1 – Scenarios Selection
• Ensemble #1:– For Assessing Impacts (Tasks 4-6)
• Size N, 20 to 30 members– Inclined to proceed with SCU 22-member ensemble– Membership: simulates 20C3M, SRESA2, SRESB1
• Ensemble #2:– For Assessing Probabilities (Task 3)
• Size M, larger than Ensemble #1– Membership: exploring rationale (e.g., rank models?)
On Ranking Models
• Consider 20th Century Simulation (20C3M)– Adopt Climate Datums
• Kaplan SST, NCEP/NCAR Reanalysis– Compare Model Output to Datums:
• ENSO variability?• Sacramento Valley climate norms?
Model Name 20C3M SRESA2 SRESA1B (3) SRESB1 Candidate? (1)
BCC-CM1, China 4 2 0 2 yesBCCR-BCM2.0, Norway 1 1 0 1 yesCCSM3, USA 9 5 7 8 yesCGCM3.1(T47), Canada 5 5 5 4 yesCGCM3.1(T63), Canada 1 0 1 1 noCNRM-CM3, France 1 1 1 1 yesCSIRO-Mk3.0, Australia 3 1 1 1 yesECHAM5/MPI-OM, Germany 3 3 3 3 yesECHO-G, Germany/Korea 5 3 2 3 yesFGOALS-g1.0, China 3 0 3 3 noGFDL-CM2.0, USA 3 1 1 1 yesGFDL-CM2.1, USA 3 1 1 1 yesGISS-AOM, USA 2 0 2 2 noGISS-EH, USA 5 0 4 0 noGISS-ER, USA 9 1 5 1 yesINM-CM3.0, Russia 1 1 1 1 yesIPSL-CM4, France 2 1 1 1 yesMIROC3.2(hires), Japan 1 0 1 1 noMIROC3.2(medres), Japan 3 3 3 3 yesMRI-CGCM2.3.2, Japan 5 5 5 5 yesPCM, USA 4 4 4 4 yesUKMO-HadCM3, UK 2 1 1 1 yesUKMO-HadGEM1, UK 1 1 1 0 no
Number of Runs
Available Projections on 1/30/06 (at LLNL PCMDI, http://www-pcmdi.llnl.gov/ipcc/data_status_tables.htm)
Task 2 – Spatial Downscaling• Method
– Statistical Approach (Wood et al 2004)
• Algorithm (from SCU)– Imports GCM data from PCMDI– Interpolates GCM data to common 2-deg grid– Downscales 2-deg circulation to 1/8-deg weather
• Precip amount• Temp max/min/avg
Task 3 – Projection Uncertainty• Given Ensemble #2, what’s the relative
probability of CA climate change:– annual-average precipitation?– annual-average daily air temperature?– joint-changes?
• Estimation Method?– Several likely to be used…
Task 4 – Runoff Impacts• Tool
– (Software) NWS RFS from NOAA/NWS– (Applications) Calibrated Basin Models from CNRFC
• Method (Miller et al 2003)– Upscale 1/8 degree projected weather to basin-scale– Simulate two ~30yr periods (base & projected future)– Compare results to get mean-monthly changes
• Preliminary Results (courtesy of Tom Pruitt)
Trinity Res Basin (TRN) - EXAMPLE
Sac Riv Trib to Shasta (SAC)
Feather Riv Middle Fork (FEA)
American Riv North Fork (AMER)
Yuba Riv Inflow, NBB Res (YUBA)
Stanislaus Riv, NM Res (STAN)
Tuolumne Riv, HH Res (TUOL)
Merced Riv, Pohono (MERC)
San Joaquin Riv, Friant (SJR)
30 40 50 60 70 80
44
46
48
50
52
54
56
58
Total-Annual-Prec [in]
Mea
n-A
nnua
l-Tem
p [F
]
TRIN
Base2011-20402041-20702071-2100
Trinity Basin, Climate Change from Ensemble #1 (N = 22)
Base: 1963-1992
… distributions of mean-annual runoff
Base 2011-2040 2041-2070 2071-2100100
200
300
400
500
600
700
800
900
1000
1100To
tal-A
nnua
l-Run
off [
cmsd
]
Climate Period
TRIN
… distributions of mean-monthly runoff
O ND J F MA M J J A S0
20
40
60
80
100
120
140
160
180
200
Run
off [
cmsd
]
BaseOND J FMA M J J A S
0
20
40
60
80
100
120
140
160
180
200
TRIN
2011-2040OND J FMA M J J A S
0
20
40
60
80
100
120
140
160
180
200
2041-2070O ND J F MA M J J A S
0
20
40
60
80
100
120
140
160
180
200
2071-2100
… distributions of Apr-Jul Volume runoff
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 10
50
100
150
200
250
300
350
400TRIN
Cumulative Probability
Tota
l-AM
JJ-R
unof
f [cm
sd]
2011-20402041-20702071-2100
Base: 1963-1992
Task 5 – CVP/SWP Water Impacts
• Issues:– CALSIM II Setup
• Adjust Water Demands (CU Analysis)?• Adjust Seasonal Runoff Forecasts?• Reclassify Hydrologic Year-Type?• Choice of regulatory environment?
– “WQCP/B2/JPOD” vs “D1641 only”?
FY2006 Schedule
• Spring 2006 – Task 1 – Ensemble Selection – Task 2 – Downscaling– Task 4 – Runoff Analysis
• Summer 2006– Task 3 – Climate Projection Dist’n Function– Task 5 – CVP/SWP Water Ops Analysis