Diapositive 1

1
Vienna International Conference 19-22 April
2004 Modeling Forest Production
Modelling the subregional impacts of climate
change on French plains forests.D Loustau,
INRA Bordeaux.
2
Vienna International Conference 19-22 April
2004 Modeling Forest Production

• So far, most analyses of climate change impacts
  - ignore the interaction with site
  characteristics (site index). - are based on
  climate data with coarse spatial resolution
• - some predictions compound the climate and
  changing age effects
• ? questions
• How can be assessed the interaction between
  climate change effect and site fertility ?
• Is there an interaction between forest management
  and climate effects ?
• Can we disentangle the climate change effects
  from the effects of changing forest age
  structure?

3
Approach used
Data
Forest Inventory
Soil Map
Climate Scenario (1960-2100)
Regional matrices
(SW N) 3 Soil N
3 Soil Water retention
Capacity 4 Grid points
3 Management scenarios
Prediction 1. Canopy annual C and H2O fluxes
1960-2100 Prediction 2. Complete rotations at
four periods 1980, 2015, 2045, 2080
Model of Forest Production 2-20 layers
canopy Beer, Farquhar, Jarvis, Rutter, 3-PG
allocation scheme Respiration process-based Pheno
logy on local data C-Reserves for broadleaf
species. N effects on LUE, allocation,
respiration, Rh.
X

4
Summary of the process model
Water balance
Tree Growth
Carbon balance
Rainfall
Radiative Transfer (Beer s law)
Foliage
Branches
Light conversion (LUE) GPP
others
Evapotranspiration interception, soil
evaporation, canopy transpiration
Stem wood
Respiration
soil water
Bark
NPP
drainage
Distribution of NPP among individuals
Coarse roots
Allocation
Fine roots
5
What s next ?
Overview of the climate scenario, North South
Long term shift in potential production
species and location effects Management
scenarios, Pinus, Quercus, Fagus Rotations
under three management options in 1980, 2015,
2045, 2080. overview of some raw
results Synthesis
6
Climate scenario. Arpege meteorological
forcast model predictions 1960-2100 IPCC B2
scenario (IS92a), Spatial grid 60 km, 0.25 day
time resolution.
Northern transect (Quercus and Fagus)
N
Southern transect (Pinus pinaster)
100 km
7

• Climate scenario. Annual averages 1960-2100
• Southern transect

1960 2100 sliding 10y average
8
Climate scenario. Annual averages 1960-2100 2.
Northern transect
-2080-
-1980-
-2015-
-2045-
1960 2100 sliding 10y average
Climate periods
9
Climate scenario. Summary

• The change in climate is not monotonous over the
  period 1960-2000.
• Significant variations in climate prediction
  occurs at the subregional scale
• along a West- to- East gradient
• between northern and southern France

10
Prediction 1 annual values of C and H2O fluxes.

• GOAL assessment of the climate change effects on
  canopy fluxes.
• Model runs repeated for one year duration.
• typical mature stand with fixed canopy structure
  (no growth or mortality)
• Driven by ARPEGE meteorological data for
• 1960, 1961, 1962,... /..., 2100 (n140).
• 6 grid points, 2 Soil Water Capacity (SWC), 5
  species

11
Results 1. Long term shift in annual C and H2O
fluxes
Pine South
Fagus NW
Fagus NE
Quercus N
-1980-
-2015-
-2045-
-2080-
12
Results 1. Long term shift in annual C and H2O
fluxes

• We interpret annual canopy fluxes of a fixed
  canopy as a climate-driven production potential.
• The annual canopy fluxes change according to
  climate
• Discontinuities occur over the period 1960-2000.
• Significant variations at the subregional scale
• along a West- to- East gradient
• between Northern and Southern France
• Climate effects interact with species, e.g.
  Quercus canopy being more responsive than Fagus.

13
Prediction 2 complete rotations under four
climates

• GOALS
• assessment of climate effects over complete
  rotations
• analysis of interaction with management and site
  characteristics
• Model runs over complete rotations (growth,
  thinnings, mortality) at four dates
• meteorological data sets
• 1980 1960-2000 ARPEGE data
• 2015 2001-2030
• 2045 2031-2060
• 2080 2061-2100
• 3 species, 8 grid points (so far), 2-3 management
  options
• 2-3 N levels (N) , 3 soil water holding
  capacities (SWC)

14
Management scenarios 1. MARITIME PINE
biomass
Extensive
Thresholds for thinnings as biomass Thinning
intensity as of biomass.
Business as usual
Intensive
age
15
Management scenarios 2. OAK (BEECH not shown but
similar)
Standard
Extensive
Thresholds - thinnings age - final harvest
age Thinning intensity fixed of stems number.
16
Results 2 Complete rotations. MAI (m3.ha-1.a-1)
- PINE
Pine SE
Pine SW
High SWR
Low SWR
?Interaction CC x management
? - x
geographical location
? -
x SWR
17
Results 2 Complete rotations. MAI (m3.ha-1.a-1)
- PINE
Pine SE
Pine SW
High N
Low N
? Interaction with soil fertility
18
Results 2 Complete rotations. MAI (m3.ha-1.a-1)
- BEECH
BEECH NW
BEECH NE
High SWR
Intensive
Standard
Low SWR

• Change with geographical location
• Little interaction with N and SWC

19
Results 2 Complete rotations. MAI (m3.ha-1.a-1)
- BEECH
BEECH NW
BEECH NE
High N
Low N
20
Results 2 Complete rotations MAI (m3.ha-1.a-1)
- OAK
OAK NW
OAK NE
High N
Low N
21
Results 2 Complete rotations - MAI
(m3.ha-1.a-1) - OAK
OAK NW
OAK NE
High SWR
Low SWR
22
Results 2 Complete rotations. Biomass C Stocks
SE
200
20 0
23
Results 2 Complete rotations. PINE
SW
200 0
24
Results 2 Complete rotations. PINE
S centre
200 0
25
Synthesis 1/3

• CC includes opposite effects of increased soil
  and atmospheric water deficits versus CO2
  fertilisation.
• The effects of CC are inverted from northern to
  southern France site index will increase in
  North and decrease in South whereas the West-
  to-East gradient is reversed.
• Site conditions and species interacts with CC
  effects
• N and to a lesser extend SWR amplify the climate
  impacts, both negative or positive
• There is a strong interaction CC x management
  partly explained by age and rotation length.

26
Synthesis answers to question 2/3

• - only for the CC impacts mediated through
  well known processes
• - YES ! Mainly explained by age-related effects
• - For sure ! (and we must do it !) Thats why
  process models are made for !

• Can we assess the interaction between climate
  change effects and subregional variation in
  climate and site characteristics?
• Is there an interaction between forest management
  and climate effects ?
• Can we disentangle the climate change effects
  from the effects of changing forest age
  structure?

27
Synthesis next steps 3/3

• To apply this approach in a retrospective mode
  for predicting past and present changes in forest
  productivity and carbon balance.
• Include N and SOM dynamics in the
  model (volunteers to contact loustau_at_pierroton.i
  nra.fr)
• Assessing the overall atmospheric impact of the
  forest sector through connecting the forest
  production to wood products life cycle.

28
Vienna International Conference 19-22 April
2004 Modeling Forest Production
Co Authors Bosc Alexandre, Colin Antoine,
Bakker Mark, Meredieu Céline Arrouays Dominique,
Le Bas Christine, Saby Nicolas Dhôte
Jean-François Pignard Gérôme, Hamza
Nabila, Davi Hendrik, Soudani
Kamel, François Christophe, Dufrêne
Eric, Déqué Michel, Cloppet Emmanuel
29
Vienna International Conference 19-22 April
2004 Modeling Forest Production
Acknowledgements
The CARBOFOR project was co-funded by the GICC
(MEDD) and GIP Ecofor (MAP) and involves 14
laboratories from INRA, CIRAD, CNRS, IFN, CEA,
Universities of Paris XI and Orléans and
Météo-France.
http//www.carbofor.fr.st