Modeling theModeling theEarth System at INPEEarth System at INPE
Domingos Urbano & Paulo Nobre
National Institute for Space Research – INPE
Center for Weather Forecasting and Climate Studies – CPTEC
Center for Earth System Sciences – CCST
SAMOC-3
Rio de Janeiro/Niterói, 11-13 May 2010
outline
• Climate Change research programs in Brazil
• Global climate model development
• Near future enhancements on seasonal to interannual climate predictions at INPE
Climate Change Research Programs in BrazilClimate Change Research Programs in Brazil
• Rede CLIMA: Brazilian network on Climate Change Research (CCR)– 50 research groups on 10 Focal points– Grant: US$ 15 million over 5 years
• INCT-MC: National Institute on Science and Technology for CC – 26 research projects covering the science of climate change,
impact-adaptation-vulnerability studies, and mitigation; technological developments in climate modeling and geosensors
– 90 research groups in Brazil and 8 countries; 400 researchers and students involved
– Grant: US$ 3.7 M for 3 years
• FAPESP program on CCR– New 15 TF supercomputer facility at INPE– Brazilian Model of the Global Climate System (ESM)– Grant of US$ 5 M per year during the next 10 years
“Rede CLIMA” Focal Points
Water ResourcesUFPE, Recife
Regional DevelopmentUnB, Brasília
Coastal ZonesFURG, Rio Grande
Biodiversity and EcossistemsMPEG, Belém
Cities and IndustryUnicamp, Campinas
AgricultureEMBRAPA, Campinas
Renewable EnergyCOPPE / UFRJ, RJ
Human HealthFiocruz, RJEconomy of Climate Change
USP, São PauloClimate ModelingINPE CST
GLOBAL CLIMATE SYSTEM MODEL
INCT - MC
INPE’s Supercomputer for Climate Change Research
Sustained Throughput
15 Tflops
Main Memory 20 TBytes
Primary Storage 3 PBytes
AquisitionInstallation Early 2010
Total budget US$ 25 M
...
and will make it possible to run global climate model simulations at high spatial resolutions to grid sizes of 20 km !
Development of a Global Climate System Model
• A single proposal will be selected
• Consortium of research institutions lead by São Paulo State institution (but engaging institutions from all over Brazil and from abroad)
• Build-up of existing experience and models developed in Brazil
• Applied to climate change research
• No “black-box” solution accepted
• Aims to developing capacity in all components of Global Climate System Modelling
• Strongly associated to PhD training
• Four years duration; US$ 1.5 M
Brazilian Model of the Global Climate System
FAQ 1.2, Figure 1
Mid 2000s
Atmosphere
Land Surface
Ocean & Sea-Ice
Sulphate
Aerosol
Non-sulphate
Dynamic
Vegetation
Atmospheric
Chemistry
Aerosol
Carbon Cycle
Atmosphere
Land Surface
Ocean & Sea-Ice
Sulphate
Aerosol
Non-sulphateAerosol
Carbon Cycle
Atmosphere
Land Surface
Ocean & Sea-Ice
Sulphate
Aerosol
Atmosphere
Land Surface
Ocean & Sea-Ice
Atmosphere
Land Surface
Atmosphere
Around 2000Late 1990sEarly 1990sMid-1980sMid-1970s
The Climate System
Figure 1 - The Global Climate System. A modified "Bretherton Diagram"highlighting linkages between biogechemical and physical climate systems. (Guy P.Brasseur, NCAR)
CPTEC.2.0 T213L64
MOM4, Global, 1/8 x 1/8 deep tropics, L50
SSiB + NCAR’s IBIS
GFDL's FMSGFDL's FMS
INPE’s Climate Forecast INPE’s Climate Forecast Coupled Suite GoalCoupled Suite Goal
Data AssimilationData Assimilation
AGCM
OGCM
AtmosOBS
T-S
IC
Coupled ForecastCoupled Forecast
AtmosFCST
daily/hourly
SFCfluxes
SST
OGCM
AGCM
6 hourlyRegional
Model
3 hourly
HydrologyModel
CropModel
HealthModel
WaveModel
INPE’s CGCM FMS implementation
Flux Ocean to IceUpdate Ice Model up
Update Atmos ModelAtmos Loop
Atmos Loop End
Flux Ice to OceanUpdate Ice Model down
Ocean Loop
Ocean Loop EndUpdate Ocean Model
Coupled Loop
Coupled Loop
Coupler End
Coupler Init
Slo
w L
oo
pS
low
Lo
op
Fas
t F
ast
Fas
t F
ast
Land Loop
Land Loop EndUpdate Land Model
Fas
t F
ast
Flux Ice to LandUpdate Ice Model down
Component models…
• Atmos GCM: – CPTEC.2.0 mpi/open_mp, – Semi-Lagrangian, – Resolution T62L64; T126L64; T213L64
• Increased PBL and Stratosphere vertical resolution– RAS/Grell deep cumulus convection
• Improved stratus parameterization scheme– atmospheric chemistry & aerosols
• Land Surface Model: IBIS/INLAND– Dynamic vegetation– Carbon Cycle– Fire Model– Improved hires land surface hidrology
Component models…
• OGCM: – MOM4, – Global, 1/4 x 1/4 deep tropics, – L50, 10m spacing upper 250 m, – Philander and Pakanowski vertical mixing– free surface, – fresh water flux, – river inflow;– Dynamical ice model (SIS)– Biogeochemistry model (Topaz, Bling)
• GFDLs FSM coupler– Up to 3-hourly coupling interval (lim. atmos radiation)
INPE developments on the atmospheric chemistry modeling
Coupled Chemistry-Aerosol-Tracer Transport model to the Brazilian Coupled Chemistry-Aerosol-Tracer Transport model to the Brazilian developments on the RAMSdevelopments on the RAMS
CATT
280tot
parcel env
al
Th kWm
rw
T
230 total
parcel envh kWm
T T
rw
2
230
80
w h kWm
w h kWm
Injection layerInjection layer
plume top lower bound
plume top upper bound
1D plume-rise model for vegetation fires 1D plume-rise model for vegetation fires Biome: Biome: ForestForestTime duration: 50 mnFire size: 20 haHeat flux: 80 kWm-2 / 30 kWm-2
Including plume rise mechanism trough
”super-parameterization” concept
aero-dynamics
energybalance
waterbalance
ATMOSPHERE(prescribed atmospheric datasets)
nitrogen cycling
growth of leaves,stems & roots
allocationand turnover
mortality &disturbance
VEGETATION DYNAMICS MODULE
VEGETATION PHENOLOGY MODULE
canopy physics
L A N D S U R F A C EM O D U L E
gross photosynthesis,foliage respiration, &optimal C:N ratios foreach plant functional type
daily LAI
vegetation structureand biomass
temperature,photosynthesis
soil physics
canopy nitrogenallocation
t ~ minutes to hours t ~ days to weeks t ~ years
senescencebudburst dormancy
decomposition of litter& soil organic matter
soilrespiration
carbon cycling
BELOWGROUND CARBON & NITROGENCYCLING MODULE
litterfall
nitrogen supply
IBIS Structure
nitrogenmineralization nitrification denitrification
energybalance
waterbalance
photosynthesis& leaf respiration
stomatalconductance
plant physiology
gross primaryproduction
totalrespiration
net primaryproduction
INPE developments on the surface modelingINLAND
River flow and flooded areasRiver flow and flooded areas
• Aplications over: – Amazônia
– Pantanal
– Araguaia
• Implications:– Flux exchanges
between sfc and atmos– Hydrology– Carbon cycle
Costa et al., in press
HAND
3 Classes
HAND
Processes Resolved:Processes Resolved:River inflow effects on salinityRiver inflow effects on salinity
MOM4with river discharge
MOM3w/o river discharge
Ocean biogeochemistryOcean biogeochemistry MOM4p1: OCMIP-2 (diagnostic) + 3
functional groups prognostic Challenges:
Dynamic Green Ocean Models (Le Quéré & al. 2005): 10 functiona groups
Emergent Biogeography (Follows & al. 2007): emerging functional groups
Ice Cover SimulationINPE COUPLED O-A GCM
OGCM Grid
INPE-CPTEC CGCM V.2.0 T213 L64, Kuo, 4 x daily coupling
30 days avrg spinup SST
INPE-CPTEC CGCM V.2.0 T213 L64, Kuo, 4 x daily coupling
30 days avrg spinup Precip
Few results from INPE’s CGCM V.1.0
Nobre et al., 2009Journal of Climate, vol. 22
27
Vegetation Scenarios
Nobre et al. (2009)
28
1st jEOF SST-Wind Stress200 yrs ctrl run
Nobre et al. (2009)
Defor3
CTRL
ENSO like mode of variabilityreproduced by CGCM and amplified by AM deforestation
29
Long term departures average (experiment – control)
20 yrs & 10 member
• Local rainfall reduction & temperature increase over AM (> CGCM)• Precipitation and temperature increase over eastern OP (is a coupled response -- def.)
Amazon Deforestation Experiment: Increased El Niño
Conditions
PRECIPITATION
TEMPERATURE
Nobre et al. (2009)
Statistically significant departures are shaded
31
Zonal Velocities over the Pacific Ctrl (contours); Deforest – Ctrl (shades)
Nobre et al. (2009)
32
Pacific Thermocline DepthDeforest – Ctrl (shades)
Nobre et al. (2009)
33
mensagemensage• Amazon deforestation induces global rainfall, Amazon deforestation induces global rainfall,
atmospheric, and oceanic circulation pattern atmospheric, and oceanic circulation pattern changes which resemble ENSO-like states;changes which resemble ENSO-like states;
• There seems to exist a positive feedback There seems to exist a positive feedback between Amazon rainfall reduction associated between Amazon rainfall reduction associated with tropical rainforest clearing and ENSO with tropical rainforest clearing and ENSO conditions in the coupled model results;conditions in the coupled model results;
• Fully coupled ocean-atmosphere modeling is Fully coupled ocean-atmosphere modeling is central to produce the global-scale rainfall and central to produce the global-scale rainfall and circulation pattern changes due to tropical circulation pattern changes due to tropical deforestation.deforestation.
Planned enhancements of Seasonal and Interannual climate
predictions at CPTEC
• Enhance AGCM resolution to T213L64 op• Enhance CGCM resolution to global
– AGCM T213L64– OGCM 1/8 degree lat-lon, 50 levels
• Increase FCST time to 12 months• Increase ensemble size to 30 per model• Multi-model: RAS, Grell, Kain-Fritsch,
Kuo
PIRATA PBR-XI e SWE-IVPIRATA PBR-XI e SWE-IV
PART 2
CNPq 38/2009:48 XCP (1200m)2 LADCP (bottom)2 PCD+Tide Gauge
UCTD
Top Related