Terrestrial carbon fluxes and pools simulated with diurnal

Terrestrial carbon fluxes and pools simulated with diurnal variability in both photosynthesis and respiration
Bakr Badawy*, Vivek Arora**, Joe Melton**, Ray Nassar*
*
**
Climate Research Division, Environment Canada, Toronto, Canada
Climate Research Division, Environment Canada, Victoria, Canada
The World Weather Open Science Conference, August 16-21, 2014
Montreal
Introduction: Global Carbon Cycle
Pre-industrial
Atmosphere
Reco
120
GPP
120
Land
GPP
Reco
: Gross Primary Production
: Ecosystem Respiration
Based on IPPC-AR4 2007Units: PgC/year (Pg =1 x 1015 g)
70
70
Ocean
Introduction: Global Carbon Cycle
Industrial
Atmospheric Growth of CO2
4.1 ± 0.1
Land-use
1.1 ± 0.7
Fossil Fuel
7.7 ± 0.5
Reco
120
Atmosphere
GPP
120
2.4
±1.0
Calculated as
the residual of
all other flux
components
Land
GPP
Reco
: Gross Primary Production
: Ecosystem Respiration
90
90
2.3 ± 0.4
Average of 5
models
Ocean
Residual carbon sink ???
Based on IPPC-AR4 2007Units: PgC/year (Pg =1 x 1015 g)
Anthropogenic CO2 emission (2000-2009) from Global Carbon Project 2010; Updated from Le Quéré et al. 2009
EC‐CAS

Residual carbon sink (budget, spatio-temporal variability)?
 Measurements ( i.e. Eddy flux measurements, Forest Inventory)
 Models ( i.e. Bottom-up, Top-down )
Courtsy: Saroja Polavarapu
To learn/understand
model errors and uncertainties
 Process understanding Land Model
The Canadian Terrestrial Ecosystem Model (CTEM)
The Canadian Terrestrial Ecosystem Model (CTEM)
 It is a dynamic vegetation model that grows vegetation from bare ground
 It provides time-varying vegetation structural attributes to CLASS
 It provides net fluxes of CO2 between the land and the atmosphere
Courtesy: Vivek Arora
CTEM and CLASS coupling
Courtesy: Vivek Arora
Current vs Target
Current
Target
Resolution
~3.75o
0.9o
Forcing
CRU+NCEP
GEM
Output
- GPP (30 mins)
- Re (daily)
Sub-daily NEE (3 hrs)
(30min or 3 hrs)
R
Rhh
Rdd
Tmin
Tmean
Tmax
T
Simulating sub‐daily fluxes Switch
Half‐hourly Tair
and Tsoil
Half‐hourly GPP & Re
Spin‐Up Simulations
 Spin-up simulation (total of 400 years) to equilibrium
 Began with globally uniform CO2 of 286.37 ppm (pre-industrial year 1861)
 Vegetation fractional coverage corresponding to year 1861
 Forced with 1901-1940 CRU-NCEP climate, used repeatedly
 Two simulations performed:
1.
CTEM-dd: respiratory fluxes are modeled at daily time step
2.
CTEM-hh: respiratory fluxes are modeled at 30-min time step
Fluxes &
carbon pools 400 years
Geographical patterns of the simulated fluxes
Rh (gC/m2/yr)
Ra (gC/m2/yr)
CTEM‐dd
CTEM‐hh
Diff.
Values are 40‐year average at the end of the pre‐industrial spin‐up
GPP (gC/m2/yr)
Geographical patterns of carbon pools
Soil C mass (KgC/m2)
Veg. biomass (KgC/m2)
CTEM‐dd
CTEM‐hh
Re‐tuning of model parameters is required
Diff.
Values are 40‐year average at the end of the pre‐industrial spin‐up
Max. LAI (m2/m2)
Re‐tuning of model parameters Re‐tuning the respiration rate parameters for (leaf, stem, root, litter, and soil) After Tuning: Geographical patterns of Carbon pools
Soil C mass (KgC/m2)
Veg. biomass (KgC/m2)
CTEM‐dd
CTEM‐hh
Diff.
Values are 40‐year average at the end of the pre‐industrial spin‐up
Max. LAI (m2/m2)
After Tuning: Geographical patterns of the simulated fluxes
Rh (gC/m2/yr)
Ra (gC/m2/yr)
CTEM‐dd
CTEM‐hh
Diff.
Values are 40‐year average at the end of the pre‐industrial spin‐up
GPP (gC/m2/yr)
transient climate simulations
 Time-varying global mean CO2 concentration
 Time varying climate data (1861-2009 CRU-NCEP)
 Vegetation fractional coverage corresponding to year 1861
 Two simulations performed:
1.
CTEM-dd: respiratory fluxes are modeled at daily time step
2.
CTEM-hh: respiratory fluxes are modeled at 30-min time step
Preliminary results
Geographical patterns of the simulated fluxes
Rh (gC/m2/yr)
Ra (gC/m2/yr)
CTEM‐dd
CTEM‐hh
Diff.
Values are 40‐year average at the end of the simulation period
GPP (gC/m2/yr)
Half‐hourly fluxes (2009)
Land Total
Equator
NH Land
SH Land
NEE
Reco
GPP
Seasonal Cycle of NEE (2009)
Mean NEE (PgC/year)
CT
= -4.54
SiB
= -0.06
CTEM_hh = -3.75
CTEM_dd = -3.67
Interannual Variability of NEE
Mean NEE (PgC/year)
CT
= -5.771
CTEM_hh = -3.715
CTEM_dd = -3.733
Conclusions
 CTEM is modified, so both photosynthesis and respirations sub-modules operate at a
time step of 30 min
 Simulating respiration at a 30-min time step changed the equilibrium states of primary
carbon pools and fluxes  re-tuning of model parameters. The same situation will
happen with changing climate forcing or grid resolution
 Simulating respiration with half-hourly temperature or daily mean temperature (ignoring
the nonlinear response of respiration to temperature) almost the same. BUT might not be
the case for regional simulation with higher spatial resolution
 NEE seasonal cycle from CTEM is shifted by 1 month  check phenology module
Thank you
Daily Rh (2009)
(1) NDL‐EVG
(4) BDL‐EVG‐CLD
(3) BDL‐EVG
(5) BDL‐EVG‐DRY
Fluxes vs Climate (JJA) for BDL‐EVG
Rh
Ra
GPP
SDPRM-inv
Total Flux = Fland + Fff + Focean
Introduction
Objectives
Ocean flux
: Takahashi et al (2009)
Fossil Fuel emission: EDGAR version 4.0 (2009)
Model
Framework
SDPRM
Equations
SDPRM
Results
NEE = Reco − GPP
Conclusion
Part1
SDPRM-inv
Equations
SDPRM-inv
Results
Conclusion
Part2
Outlook
21/28
FNEE  (R pri 

Pi ( R0 ,E,K ,LAI )
R
P
i

P )  (GPPPFTpri 
i i

P j (  PFT ,VPD 0,Tmin )
GPP
P
j

j
24 parameters are optimized (time-independent)


3 parameters / 7 Plant Functional Type (PFT)
3 parameters (globally)
A-priori fluxes with small temporal-scale variability  process-understanding
P )
j
Simulated global values of primary carbon pools and fluxes
from a pre-industrial spin-up
Values are 40‐year average at the end of the pre‐industrial spin‐up
Modeled 40-year zonally-averaged fluxes
Values are 40‐year average at the end of the pre‐industrial spin‐up
Modeled values of primary daily fluxes for the last 20 years of the spin-up run
Amazon
broadleaf
evergreen
W-EU
Crop C3
E-US
broadleaf cold
deciduous
SDPRM: Results – Climate limitations – IAV of GPP
GPP = f (VPD, Radiation, Temperature)
Introduction
Objectives
Model
Framework
SDPRM
Equations
SDPRM
Results
Conclusion
Part1
VPD limitation
Temperature limitation
SDPRM-inv
Equations
SDPRM-inv
Results
Conclusion
Part2
Outlook
16/28
Radiation limitation
Geographic distribution of potential climatic constraints to plant
growth derived from long-term climate statistics
SDPRM: Results – Climate limitations -- Reco
Introduction
Objectives
Model
Framework
SDPRM
Equations
SDPRM
Results
Temperature contribution to the IAV of Reco
Conclusion
Part1
SDPRM-inv
Equations
SDPRM-inv
Results
Conclusion
Part2
Outlook
18/31
Precipitation contribution to the IAV of Reco