User-driven downscaling: advances in data apportioning and analysis to augment adaptation planning User-driven downscaling: advances in data apportioning and analysis to augment adaptation planning E.P. Maurer 1 , L.D. Brekke 2 , T. Pruitt 2 , K. White 3 , E. Ochs 3 , P. Duffy 4 , and E.H. Girvetz 5 1 Santa Clara University 2 U.S. Bureau of Reclamation 3 U.S. Army Corps of Engineers 4 Climate Central 5 The Nature Conservancy 90th American Meteorological Society Annual Meeting Joint Session 8 New challenges for applied meteorology and climatology January 21, 2010
20
Embed
User-driven downscaling: advances in data apportioning and ...emaurer/papers/maurer_2010-1-21_ams_final.pdf · 21/01/2010 · User-driven downscaling: advances in data apportioning
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
User-driven downscaling: advances in data apportioning and analysis to
augment adaptation planning
User-driven downscaling: advances in data apportioning and analysis to
augment adaptation planningE.P. Maurer1, L.D. Brekke2, T. Pruitt2, K. White3, E. Ochs3, P. Duffy4, and E.H. Girvetz5
1Santa Clara University2U.S. Bureau of Reclamation3U.S. Army Corps of Engineers4Climate Central5The Nature Conservancy
90th American Meteorological Society Annual MeetingJoint Session 8
New challenges for applied meteorology and climatologyJanuary 21, 2010
Global Climate is ChangingGlobal Climate is Changing
• Temperatures are increasing globally
• Most recent warming attributed to human-driven GHG emissions
• Some impacts already evident and attributable to warming
Source: U.S. Global Change Research Program (USGCRP)
Projections of Global ChangeProjections of Global Change
• Range of ‘likely warming’ by end of 21st
century variable
• By mid-21st
century most differences smaller
2010
1.8° 3.4° 4.0°2.4° 2.8°
Regional ChangesRegional Changes
• Projected changes non-uniform
• Impacts also non-uniform
Median runoff change, 2041-2060 minus 1901-1970
Greater water scarcityMore wildfiresAccelerating invasive speciesTourism, recreation impactsAgricultural vulnerability
Extreme urban heat eventsWorsening air quality episodes
Ocean fishery migrationIncreased severe flooding events
Source: U.S. Global Change Research Program (USGCRP)
Adapted from Cayan and Knowles, SCRIPPS/USGS, 2003
2. Global Climate Model4. Land surface
(Hydrology) Model
3. “Downscaling”
5. Operations/impacts
Models
IPCC AR4 GCM SimulationsIPCC AR4 GCM Simulations20th century through 2100 and beyond>20 GCMsMultiple Future Emissions Scenarios
http://www-pcmdi.llnl.gov/
Selecting GCM runs for “Bookend”Studies
Selecting GCM runs for “Bookend”Studies
• Brackets range of uncertainty• Useful where impacts models are
complex• Downscale output from few GCMs
Projected Impacts: Loss of SnowProjected Impacts: Loss of Snow• Snow water in reserve on April 1• Change (Sacramento-San Joaquin basin, 2
GCMs, 2 emissions scenarios):-12% to -42% (for 2035–2064) (up to 1 Lake Shasta)
-32% to -79% (for 2070–2099) (up to 2 Lake Shastas)
Ref: Luers et al., 2006, CEC-500-2006-077 GFDL CM2.1 results
How many GCMs to select?How many GCMs to select?• Ensemble mean provides better skill• Little advantage to weighting GCMs according to skill• Most important to have “ensembles of runs with enough
realizations to reduce the effects of natural internal climate variability” [Pierce et al., 2009]
Gleckler, Taylor, and Doutriaux, Journal of Geophysical Research (2008) as adapted by B. Santer Brekke et al., 2008
Impact Probabilities for PlanningImpact Probabilities for Planning
Snow
wat
er e
quiv
alen
t on
Apr
il 1,
mm
Point at:120ºW, 38ºN
2/3 chance that loss will be at least 40% by mid century, 70% by end of century
• Combine many future scenarios, models, since we don’t know which path we’ll follow (22 futures here)
• Choose appropriate level of risk
Mass Production of Downscaled Climate Projections
Mass Production of Downscaled Climate Projections
• PCMDI CMIP3 archive of global projections• New archive of 112 downscaled GCM runs• gdo4.ucllnl.org/downscaled_cmip3_projections• Allows quick analysis of multi-model ensembles
Use of U.S. Data ArchiveUse of U.S. Data Archive• Over 500 users downloaded >4 TB of data• From across US and outside• Uses for Research (R), Management & Planning
(MP), Education (E)
Using Archive to Select Specific GCM Runs
Using Archive to Select Specific GCM Runs
Bivariate probability plot shows correlation between ∆T, ∆ P
Identify Change Range: 10 to 90 %-tile ∆T, ∆ P
Select bounds based on:•risk attitude•interest in breadth of changes•number of simulations desired
Brekke et al., 2009
What is missing from downscaled data archive?
What is missing from downscaled data archive?
Global BCSDGlobal BCSD• Similar to US archive• Allows probabilistic representation of