A Model

Document Sample
A Model Powered By Docstoc
					Lecture 4. GFDL Terrestrial Carbon Cycling Model




           Elena Shevliakova & Chip Levy
 Ocean Biogeochemical and Dynamic Land
         Carbon Cycle Modeling
     (the GFDL Earth System Model)


John Dunne               GFDL/NOAA
(John.Dunne@noaa.gov)

Ron Stouffer             GFDL/NOAA
Elena Shevliakova        Princeton University
(Elena.Shevliakova@noaa.gov)

Sergey Malyshev          Princeton University
Chris Milly              USGS
Steve Pacala             Princeton University
Hiram levy               GFDL/NOAA
GFDL’s earth system model (ESM) for coupled carbon-climate



                    Atmospheric circulation and radiation
  Climate
  Model           Sea Ice                         Land Ice
                                           Land physics
                  Ocean circulation
                                           and hydrology



                    Atmospheric circulation and radiation
Earth System       Chemistry – CO2, NOx, SO4, aerosols, etc
   Model
                  Sea Ice                           Land Ice
                  Ocean ecology and       Plant ecology and
                  Biogeochemistry             land use
                                            Land physics
                  Ocean circulation
                                            and hydrology
               The Zero Order View of the Carbon Cycle
         (Integrated Assessment Models - IAMs for example)




Fossil Fuels
                           Atmosphere                   4 yr
                  CO2 = 280 ppmv (560 PgC) + FF
                 90                       60±
                 ±
                 Ocean Circ.             Biophysics
100-103 yr       + BGC                   + BGC         100-102 yr
                                          2000 Pg C
               37,400 Pg C + FF



                                  …equilibrium takes 103-104 yrs
Resolution



             • Cubed sphere, Lat x Lon
                – 2 ° x 2.5 ° [atm.]
                – 1° x 1 ° [land]

             • Time step:
               ~30 Minutes

             • Simulation:
               centuries
             Current land processes represented in
                      GFDL’s current ESM
      in collaboration with Princeton U., U. New Hampshire and USGS
                         (Schevliakova et al., 2009)

• Plant growth
   – Photosynthesis and respiration – f(CO2, H2O, light, temperature)
   – Carbon allocation to leaves, soft/hard wood, coarse/fine roots, storage

• Plant functional diversity
   – Tropical evergreen/coniferous/deciduous trees, warm/cold grasses

• Dynamic vegetation distribution
   – Competition between plant functional types
   – Natural fire disturbance – f(drought, biomass)

• Land use
   – Cropland, pastures, natural and secondary lands
   – Conversion of natural and secondary lands and abandonment
   – Agricultural and wood harvesting and resultant fluxes
                Land Model Forcings
Lands use changes




             Climate
                                      Canopy and
             Soil/snow                canopy air       Atmosphere




                  Energy and moisture balance


                         Photosynthesis
                    Plant and soil respiration
                                                                     t~ 30 min
                                                                                                         and nitrogen exchange
                                                                                                        Land energy, water, carbon




                  C & N uptake and release



                       roots
                                                                                 Climate statistics




                       Plant type
                                                       Carbon gain




                       LAI, height,




leaching
           C & N allocation and growth, t ~ 1 day
                                                                                                                                       Dynamic Land Model LM3




                                 Phenology, t~ 1 month


             Mortality, natural and fire t ~ 1 year
                                                                                                                 Vegetation dynamics




                                                                     {
                     Biogeography, t ~ 1 year


                  Land-use management, t ~ 1 year
                                                                                                      Humans




           wood     sapwood           labile   fine    leaves
                                               roots
LM3 structure: sub-grid heterogeneity
Vegetation structure in LM3
                              5 vegetation types
                              • C3 and C4 grasses
                              • temperate deciduous
                              • evergreen coniferous
                              • tropical

                              5 vegetation C pools
                              2 or 4 soil C pools

                              Sub-grid land use
                              • 4 land-use types
                              • up to 15 tiles for different
                              ages

                              Natural mortality and fire

                              Land and atmosphere are
                              on the same grid
                              • 2°x 2.5° in ESMs
                              • Cube-sphere in CM3
Hydrology
Now For Some Detail
Biosphere-atmosphere exchange: photosynthesis and respiration




    Photosynthesis: 6 CO2 + 12 H2O + light → C6H12O6 + 6 O2 + 6 H2O
      Carbon Dioxide + Water + Light energy → Glucose + Oxygen + Water

                Respiration: C6H12O6 + 6O2 → 6CO2 + 6H2O
Response to drying, lower CO2: C4 photosynthesis evolves in plants
photosynthesis
                               CO2+H2O --> CH2O (C3) +O2
                 CO2           or
     C3



                               CH2O+O2 --> CO2+H2O

                         The enzyme Rubisco catalyzes both reactions.
                         Oxidation increases at lower CO2.
photosynthesis




                                                          C4      C4 --> C3+CO2
                              C3+CO2 --> C4
                 CO2
                              (CO2 molecule is
     C4




                              loosely bound to C3                 CO2+H2O --> CH2O+O2
                                                          C3
                              compound                            or
                                                                  CH2O+O2 --> CO2+H2O

•        Advantages of C4 photosynthesis
          • Higher CO2/O2 ratio where Rubisco catalyzes photosynthesis, less CH2O oxidation
          • Plants can take up CO2 at night, when humidity is high, and not lose water
•        Consequence: C4 plants do better at low CO2, dry climates
•        C3 plants - trees and some grasses
•        C4 plants - other grasses, grains (corn, sorghum, millet)
  Carbon fixation




C3, C4 and CAM plants
                          Carbon Engine – Photosynthesis Model
                             (Farquhar et al. 1980, Collatz et al. 1992, Leuning 1990)

                              m  An _ pot
 g s _ pot                                               , Eq 1
             (Ci   )  (1  ( qsat (Tl )  qca ) d 0 )
             g s _ pot
 An _ pot              (Cca  Ci ), Eq 2
              1.6
                                                                                                  
                                               Ci                                              
                               J E  a 3 Q               ,                                       
                                              Ci  2*                                            
                                                                                                  
 for C 3 : A                                                        Ci  
                n _ pot  min J C  Vm (Tl )                                                      ,   Vm (Tl ),   Eq 3C 3
                                                               pref                        pref 
                                               Ci  K c (Tl )        (1  O2  K o (Tl )     ) 
                                                                 p                           p
                                                                                                   
                                      Vm (Tl )                                                    
                              Jj                                                                 
                                          2                                                       
                               J E  a 4Q,                   
                                                              
 for C 4 : An _ pot  min  J C  Vm (Tl ),                      Vm (Tl ), Eq 3C 4 ,
                              J                               
                               CO 2  18000 Vm (Tl ) Ci 
 A system of three equations with three unknowns, the stomatal conductance; gs,the intercellular
 concentration of CO2, Ci (mol/mol); and the net photosynthesis, An(mol CO2/m2s), defines the
 plant uptake of CO2 and the rate of non-water-stressed transpiration for a thin canopy layer dLAI’
 at a temperature Tl(K)receiving an incident photosynthetically active radiation flux Q(LAI’)
 (Einstein/m2s) and surrounded by canopy air with vertically uniform specific humidity qca(kg/kg)
 and CO2 concentration Cca(mol/mol):
                                  Parameters

a is the leaf absorptance of photosynthetically active radiation,

α3 and α4 are the intrinsic quantum efficiencies,

Vm is the maximum velocity of carboxylase in molCO2/m2s,

Γ*= αco2 [O2] KC /(2KO) is the compensation point,

KC and KO are the Michaelis-Menten constants for CO2 and O2,

[O2] is the atmospheric oxygen concentration,

pref=105 Pa is the reference pressure and p is an atmospheric pressure.

The temperature dependence of the Michaelis-Menten constants, the maximum
velocity of carboxylase, and the compensation point are described by Arrhenius
function where T is the temperature (˚K) and E0 is a temperature sensitivity factor
(Foley et al. 1996)
Equation 13 gives the leaf stomatal conductance for vegetation if the soil water is not
limiting. It links the rate of stomatal conductance for water gs to the net photosynthesis
(An), intercellular concentration of CO2 (Ci), and humidity deficit between intercellular
space and external environment (qsat(Tl) - qca). This equation is a simplification of
Leuning’s (1985) empirical relationship assuming that contribution of cuticular
conductance is negligible.

Equation 14 is a one-dimensional gas diffusion law The factor of 1.6 is the ratio of
diffusivities for water vapor and CO2. We assume that the diffusion of CO2 is mostly
limited by stomatal conductance and not by leaf boundary layer conductance.

Equations 15C3 and 15C4 are based on the mechanistic model of photosynthesis by
Farquhar et al. (1980) and its extensions by Collatz et al. (1991, 1992).

The net photosynthesis is the difference between the gross photosynthesis and leaf
respiration. The gross photosynthesis for C3 plants is the minimum of three limited
rates: the light limited rate JE, the Rubisco limited rate JC, and the export limited rate of
carboxylation Jj. Similarly, in Collatz et al. (1992) the gross photosynthesis rate for C4
plants is the minimum of the light limited rate JE, the Rubisco limited rate JC, and the
CO2 limited rate JCO2. Leaf respiration is computed as Rleaf = γVm(Tl). Although the
formulation of Collatz et al (1991) is widely used in dynamic vegetation and land
surface models, it requires computationally expensive iterative solutions. The
simplifying assumption made in equation 13 that cuticular conductance is negligible,
allows an analytical solution for the three unknowns.
                        Present Day Simulated Vegetation and Soil C pools
                       Veg C                                 Soil C
model potential veg,
  current climate
observation-based
    estimates




             LM3 generates present-day spatial distribution of vegetation and soil carbon
LM3 is designed to diagnose and predict the land use sink
      LM3 Predicted Carbon Loss Due To Land Use Change
                                                                                          total


                                                                                      natural




  Carbon Loss from 1700 to 2000                                       crops   secondary
                                                                                                  pasture
                                                  natural

                                       kg   /m2


   Total carbon loss: 228 Gt                                  Ecosystem carbon, kg




 Current Pasture Fraction                                   Current Crop Fraction


Current pasture area: 3.1 billion ha                        Current crop area: 1.4 billion ha
                           Why secondary vegetation is important ?



                                                          No wood harvesting
                                                                                              HYDE

                                                                                            SAGE/HYDE
C flux, GT C/yr




                                                                                             HYDE

                                                                                          SAGE/HYDE




                  Land-use scenarios from Hurtt et al. 2006
                  Stand alone LM3V forced by the atmospheric data from the GFDL AM2 model, CO2=350 ppm
                                                                                        Shevliakova et al. 2009
Why do we need a model of vegetation dynamics and C cycle?
Current Models Predict a Big Sink From CO2 Fertilization




                                            2050
Uncertainty about the magnitude of CO2 fertilization is the key factor
   determining whether vegetation is a net carbon source or sink

                     Change in Vegetation Biomass, kgC/m2
        No CO2 fertilization                   CO2 Fertilization at 700 ppm




   -460Pg                                 +200 Pg



  GFDL Slab-Ocean Climate Model SM2.1coupled to Dynamic Land model LM3V
                  Atmospheric CO2 concentration: 700 ppm

                                                                 Shevliakova et al. 2006
Transient land C flux and storage, Historic and A1B Future (ESM2.1)
                  phot_286



                  phot_hist                A1B_phot_fert
         GtC




                      A1B_phot_286




  • Even under assumption of CO2 fertilization, C storage declines after 2100

  •Under assumption of “no CO2 fertilization ” land biosphere will undergo a
  catastrophic loss of C
                                                                                    Terrestrial Sink

                                              Cumulative Emissions Reductions Necessary to Stabilize at 500 PPM

                                     170
Fossil Carbon Emissions Reductions




                                     145

                                     120                                                                                                   Hypotheses for the
                                                                                                                                           terrestrial sink:
                                      95
               (Gt/y)




                                                                                                                  CO2 Fertilization Sink
                                                                                                                  Land Use Sink            1. CO2 Fertilization
                                      70

                                                                                                                                           2. Climate Change
                                      45

                                      20
                                                                                                                                           3. Land Use

                                      -5
                                       2000       2010         2020          2030       2040         2050
                                                                      Year




                                     Solving the carbon problem is twice as hard if the missing sink is caused by
                                                        land use instead of CO2 fertilization.
                Ocean processes represented in
                     GFDL’s current ESM

• Coupled C, N, P, Fe, Si, Alkalinity, O2 and clay cycles

• Variable Chl:C:N:P:Si:Fe stoichiometry

• Phytoplankton functional groups
   – Small (cyanobacteria) / Large (diatoms/eukaryotes)
   – Calcifiers and N2 fixers

• Herbivory - microbial loop / mesozooplankton (filter feeders)

• Carbon chemistry/ocean acidification

• Atmospheric gas exchange/deposition and river fluxes

• Water column denitrification

• Sediment N, Fe, CaCO3, clay interactions
              Coupled elemental cycles in the
             GFDL global biogeochemical model

Carbon           Oxygen      Phosphorus
                                               DOM cycling

                                               Particle sinking

                                               Atm. Deposition
CaCO3 only
                                               Gas exchange
Nitrogen          Iron      Alkalinity/CaCO3
                                               Solubility pump

                                               Loss from system

                                               River Input
                  Silicon     Lithogenic
                                               Sediment Input

                                               Scavenging
            Ocean ecology in the GFDL global
                biogeochemical model




             N2-fixers


            Small phyto.       Protists
Recycled
nutrients                                            Filter feeder

                                DOM

New             Large phyto.
nutrients                                 Detritus
CO2 Flux   Observations (present)




           Model (pre-industrial)
20 Yr Time Series of Southern Hemisphere CO2 Flux
                (total and oceanic)
 Where are we in the ESM model development process?

Plan to use ESMs for next IPCC (AR5)
    – Thousands of years to spin-up and 300-yr runs into the future
    – 4 new future scenarios of GHGs and land-use change
    – ~40 experiments planned




            Atm CO2 anomaly in a control integration
CO2 Anomaly Time Series
           Scientific Questions For The Land Model
• How did recent changes in climate, CO2 and land use shape
  the present day distribution of land carbon and nitrogen
  sources and sinks?
• What are the influences of land cover changes on
  continental precipitation and runoff?
• What are the implications of climate change for the
  distribution and functioning of terrestrial vegetation? This is
  particularly important for agriculture. (Why?)
• What are the terrestrial biosphere feedbacks on climate?
• What is the role of plant diversity in the global
  biogeochemical cycles and climate system?
       Scientific Questions For The Earth System Model
• How will climate interactions with the Land Model influence
  CO2 levels in the atmosphere over the short term?

• How will land use interactions with the Land Model influence
  CO2 levels in the atmosphere over the short term?

• What role will CO2 fertilization play in controlling CO2
  levels?

• Will ocean biogeochemistry control the long-tem level of
  CO2 in the atmosphere and what will it be?

• Two longer-term land wild cards: soil C release in a warmer
  Arctic; CH4 release in a warmer Arctic
The End
                            Summary

• The GFDL land model:
   – represents a range of biosphere-climate interactions and
      feedbacks;
   – captures effects of both climate change and land use on
      vegetation dynamics and structure;
   – simulates historic and future distribution of Carbon sources
      and sinks;
   – Will characterize coupled Carbon-Nitrogen dynamics in plants
      and soils.
• Upcoming improvements include increased biodiversity, seasonal
  fire, N and P cycles, and ecological data assimilation for formal
  parameter estimation.
• Currently there is considerable uncertainty about the magnitude
  of climate effects on biosphere and its feedbacks.
                         GFDL LM3 Functionality

•    Land surface parameterization:
    – energy, water, and momentum exchange
•    Hydrological processes:
    – River flow, water resource development and use, extreme
        events
•    Ecological processes:
    – vegetation functioning, structure, distribution, disturbance*
        (natural and anthropogenic), and succession*
•    Carbon cycling
    – CO2 fluxes, vegetation and soil carbon pool
•    Land use and management*

      * These are relatively unique features
                     A Computer Model is:



• a theoretical/numerical construct that represent a set of
  particular processes and phenomena
   – a set of variables – input, output, state, parameters
   – a set of logical and quantitative relationships between
      them
   – a set of assumptions

• Idealized logical framework to test hypotheses and to ask
  scientific questions
Historic C emissions from anthropogenic pools simulated by LM3




                                                     Malyshev et al., 2009
Above Ground Biomass (AGM) vs Annual mean Temperature




                                         Lichstein et al. in prep (2009)
             AR5 RCPs
       (van Vuuren et al. 2008)
• New scenarios are developed for the next IPCC
  assessment




(pre-ind. to present day +2.3 W/m2, IPCC AR4)

				
DOCUMENT INFO
Shared By:
Categories:
Tags:
Stats:
views:5
posted:10/7/2011
language:English
pages:44