ESM Developments at GFDL

1
ESM Developments at GFDL

• Ronald J Stouffer
• GFDL/NOAA

2
An Earth System Model (ESM) closes the carbon
cycle
Atmospheric circulation and radiation
Land physics and hydrology
Climate Model
Sea Ice
Ocean circulation
Atmospheric circulation and radiation
Allows interactive CO2
Earth System Model
Plant ecology and land use
Sea Ice
Ocean ecology and biogeochemistry
Land physics and hydrology
Ocean circulation
3
Vegetation structure in the LM3/LM3V land models

• 5 vegetation types
• 5 vegetation C pools
• (Cl, Csw, Cw, Cr, Clv)
• 2 soil C pools
• (Cfs and Css)
• 4 land-use types
• Up to 15 tiles of different ages per grid-cell
• Natural mortality and fire

Shevliakova et al., 2009
4
LM3V generates present-day distributions of
vegetation and soil C
Vegetation C, Kg/m2
Soil C, Kg/m2
model potential,current climate
observation-based estimates
5
LM3V simulates the land-use sinks sources,
including forestry
Simulated historical wood harvests compare well
with the FAO-based estimates
The models estimate of the 90s land use flux,
1.1-1.3 PgC/a, is about half of previous
estimates and implies a smaller missing sink
Shevliakova et al., 2009
6
LM3 Hydrological Component

• Research Areas
• Water availability, variability, and scale
• Vegetation, climate, and hydrologic change
• Water development as driver of climate change
• Aquatic environment
• Seasonal soil freezing and permafrost
• Sea-level variability

6
7
Tracers Of Phytoplankton with Allometric
Zooplankton (TOPAZ) simulates the mechanisms that
control the ocean carbon cycle
Biogeochemistry
Phytoplankton ecology
DOM cycling
Particle sinking
N2-fixer
Filter feeder
Atm. Deposition
Small phyto.
Protist
Recycled nutrients
Gas exchange
River Input
Removal
DOM
Large phyto.
New nutrients
Sediment Input
Detritus
Scavenging
Carbon
Oxygen
Phosphorus
Nitrogen
Iron
Alkalinity
Lithogenic
Silicon
CaCO3
Dunne et al (2005, 2007)
8
HISTORICAL (CORE forced) simulation reproduces
SeaWiFS chlorophyll variability and puts it in a
multidecadal context
log(SeaWiFS satellite Chlorophyll)
log(Model Chlorophyll)
Simulated expansion of oligotrophic gyres (ocean
deserts) is similar to SeaWIFS (Polovina et al,
2008) and within a large, multidecadal envelope.
N Atlantic
N Pacific
Indian
S Atlantic
S Pacific
9
We have developed three new ESMs
All models use a 2? atmosphere and 1? ocean
10
ESM Current Status

• ESM2.1 - running
• Backup for ESM2M and ESM2G
• Useful for science projects
• Historical Future projection
• Last Glacial Maximum (LGM)
• ESM2M and ESM2G complete (FINALLY!)
• Starting spin-ups for IPCC runs
• Land use changes will also be included
• Use of 2 different ocean components allows
  investigation of heat and carbon uptake
  uncertainties

11
ESM2.1 has produced a successful simulation of
climate change

• Long 1860 control
• Atm pCO2 drift less than 10ppm per 100 yrs
• Climate similar to CM2.1
• Historical Future (A1B A2) runs
• Purpose document model, evaluate model
  performance, investigate impacts, feedbacks and
  uncertainties
• Note No land use included potential vegetation
  only

12
ESM2.1 Shows Similar Global Surface Air
Temperature Response to CM2.1

• Response slightly smaller than CM2.1 in
  historical period
• Temperature increase similar magnitude in future

Surface air temperature change (oC)
1900
1950
2000
2050
2100
13
ESM2.1 warming pattern and magnitude very
similar to CM2.1
Surface air temperature change (K)
ESM2.1
CM2.1
2081-2100 minus 100 yr average in control
14
ESM2.1 ocean CO2 uptake response to atmospheric
CO2 and climate forcing similar to previous work
Significant reduction in anthropogenic CO2 uptake
due to climate change (Sarmiento and Le Quere,
1996 Sarmiento et al., 1998)
Overall response
Sabine (2005)
Little climate change effect on natural CO2
partitioning
15
Net Carbon Flux Land vs Ocean
Black control CO2 constant for bio and
radiation Green CO2 increases for bio, CO2
constant radiation Blue CO2 increases for bio,
CO2 increases for radiation Red CO2 constant for
bio, CO2 increases for radiation
16
Model captures natural CO2 cycling and
anthropogenic CO2 invasion
CO2 Flux out of ocean (mol m-2 yr-1) Takahashi
(2006)
Anthropogenic CO2 Inventory (mol m-2) Sabine et
al (2004) Key et al. (2004)
1.4 PgC/yr
99-137 PgC
Observed
5 4 3 2 1 0 -1 -2 -3 -4 -5
2.3 PgC/yr
104 PgC
historical
2.0 PgC/yr
99 PgC
ESM2.1
17
ESM2.1 projects ocean acidification changes with
severe implications for CaCO3 cycling
Acidification from 1860 to 2100 ( H change)
Resulting 52 decrease in calcite export from 100m
1860
2000
2100
Atlantic
This reduction in calcite export similar to
previous work (Ridgewell et al. 2006)
This reduction in calcite export reduces the
efficiency of sinking organic material to 1000 m
by 16
Pacific
18
ESM2.1 projects modest decrease in total
productivity and major decrease in large
phytoplankton
Projected change in global productivity from 1860
()
8 4 0 -4 -8 -12 -16
Little change in N2 fixation
2 decrease in total productivity
Total Productivity Large Phytoplankton
Productivity N2 Fixation
11 decrease in large phytoplankton productivity
winners and losers, but more losers than winners
13 increase in oligotrophic gyre (ocean desert)
area
19
Uncertainty about CO2 fertilization is the key
factor for future land C uptake
GFDL Slab-Ocean Climate Model (SM2.1-LM3V) Atmosph
eric CO2 concentration 572 ppm in both
experiments
No fertilization, photosynthesis at 286 ppm
Fertilization, photosynthesis at 572 ppm
Equilibrium changes in land C from preindustrial
levels
The difference in the resulting land carbon
storage suggests that the magnitude of the CO2
fertilization uncertainty can be as large as that
associated with clouds or oceanic heat uptake
Shevliakova et al., subm.
20
Transient changes in land C storage depend on
CO2 fertilization (ESM2.1)
CO2 fertilization Carbon storage increases in
21st century and then declines No CO2
fertilization Catastrophic loss of land carbon
2400
Photosynthesis uses atmospheric concentrations
2200
Land Carbon, GtC
2000
1800
Photosynthesis fixed at 286
1600
1900
2000
2100
2200
2300
Historical CO2
SRES A1B CO2 concentrations
21
Research plans include a suite of investigations
of the earth system

• Land use impact on carbon changes
• Increasing CO2 impacts on ecosystems
• Oceanic heat and carbon uptake
• Carbon cycle feedbacks
• Climate Fisheries interactions
• Water availability changes
• Paleo-climate

22
Improving earth system model realism

• Close additional biogeochemical cycles
• N, P, CH4, Fe, etc.
• Improve representation of biodiversity
• Improve components
• Dust, sea salt, fire, land use
• Migrate to CM3-based ESM
• Aerosol-cloud interactions, strat-trop
  interactions
• Investigation of high resolution
• Coastal processes
• Seasonal-decadal scale variability
  analysis/prediction

23
Toward marine resource prediction at interannual
to centennial time-scales

• Incorporate enhanced ecosystem dynamics into
• Historical ocean/ice simulations forced with
  atmospheric reanalysis
• Century-scale climate simulations
• Interannual and decadal scale "initial value"
  simulations
• Develop capacity to link with fisheries foodweb
  models
• Engage NOAA fisheries laboratories through
  postdocs and workshops to develop new and
  innovative applications of GFDL models

24
Future ESM Developments ISSUES

• Boundaries
• How much of human decision making should go into
  ESMs?
• Have fire models/land use in ESMS now
• Where does it stop?
• How to define signal/noise?
• Uncertainties
• CO2 fertilization of land plants
• Southern Ocean
• Coastal progresses How important to global
  response?
• Resolution implications!
• Processes What all is needed? What can be
  simplified?
• Land use How fine a grid needed? What is
  important?
• CH4 and other chemical cycles, peat, land ice,
  aerosols

25
(No Transcript)