Coupled Climate Modeling

1
Coupled Climate Modeling

• Andrew Wittenberg Michael Winton (CMDT
  co-chairs)?
• Geophysical Fluid Dynamics Laboratory
• 25 March 2008

2
AR5 strategy at GFDL

• Emphasizes comprehensiveness
• New processes feedbacks Jan '09 deadline
• Mitigate development risk via parallel
  development tracks, CM2.1 foundation
• Could merge tracks if AR5 deadline slackens
• CM2.1-like resolution (computing)?

3
Development Pathway
2008
CM3 AM3 LM3 CM2.1 ICE/OCEAN
5yr/day on 500PEs 2400 CPUhr/yr
ESM2.1 CM2.1 LM3V OBGC
CM2M CM2.1 with MOM
ESM2M
CM2G CM2.1 with GOLD
ESM2G
8yr/day on 140PEs 425 CPUhr/yr
10yr/day on 120PEs 300 CPUhr/yr
4
Why two ocean models?

• z-coordinate better in weakly stratified regions
• ?-coordinate better on sloping bottom

Role of ocean in transient climate change?
5
LM2 -gt LM3 Land Model

• Dynamic vegetation
• Subgrid land-use heterogeneity
• Distinct treatment of ground, vegetation,
  canopy air
• Soil water dynamics (liquid frozen)?
• Multilayer snow pack
• River network with capacity for tracers and
  temperature

6
Land ice modeling a gap

• IPCC could not estimate sea level change arising
  from altered land ice dynamics
• Land ice modeling still in its infancy conducted
  by a small community of scientists (Jan 07
  meeting, EOS article)?

7
What we can do about marine ice sheets (from the
ocean side)?

• Make a static marine ice slab (with sub-ice
  cavity)?
• Develop ice cavity ocean simulation (coordinate
  issues)?
• High res. Southern ocean downscaling to simulate
  heat transport to the cavity
• Ocean-ice shelf heat/water transfer for a
  transient simulation

Smethie Jacobs (DSR 2005)?
8
Coupling infrastructure

• Needed for parallel component development
• Recent enhancements ESM tracers,
  cubed-sphere/tripolar coupling, conservation
  diagnostics
• Future e.g. ice shelves, grounding line movement
  will require restructuring
• Like other components, model coupling
  infrastructure requires continuous development

9
Simulation analysis a crucial part of coupled
model development

• Richer simulation -gt need new evaluation metrics

• Model-specific analysis adds value -- e.g.
  CM2.1's projected severe Sahel drying is an
  outlier, but GFDL analysis indicates we can't yet
  rule it out

10
Need long runs to characterize model behavior
e.g. ENSO
11
Understanding ENSO sensitivities e.g. convective
momentum transport (CMT)?
12
ENSO period vs. y-width/longitude of zonal stress
anomalies
Capotondi, Wittenberg Masina (Ocean Modelling,
2006)?
corr(T, Tp) 0.82 /- 0.15
Tp 3.05 (Ly-14)/9.6 (C-184)/30
13
Coupled Model DevelopmentBalances

• resolution comprehensiveness
• components
• construction analysis
• process-level details global realism
• immediate needs long-lead research

14
(No Transcript)
15
New capabilities of LM3