Vegetation demographics in ACME: Capturing structural ...€¦ · Jennifer Holm1, Ryan Knox1,...
1
Solution Attempts For additional information, contact: Jennifer Holm Research Scientist Lawrence Berkeley National Laboratory (510) 495-8083 [email protected] climatemodeling.science.energy.gov/acme Issues concerning dynamic vegetation modeling Solution Attempts Ideas and future evaluation & testing PFT1% PFT1% PFT2% PFT2% PFT3% PFT2% PFT3% PFT1% PFT2% PFT3% PFT2% PFT3% Spa+ally% average% the%direct% and%diffuse% radia+on% transmi=ed%onto% the%understory% layer.%% Reflectance,% absorp+on% and% transmi= ance% calculated% for%mul+ple% leaf% layers% for%each%PFT.%% Direct% Diffu e% Light% transmi=ed%onto% soil/snow% reflected%back%up% through% canopy% (itera+ve% solu+on)%% Individual tree stomata response (FLCgs) (prev timestep) Individual tree canopy-level transpiration (E) (current timestep) Plant traits: wood + leaf economics (WD, LMA, N L , P L ) Tree size: height, total leaf area HOST LAND SURFACE MODEL – FATES PLANT HYDRAULICS MODULE TRAIT TRADE-OFFS FOR PFT PARAMETERIZATION (from meta-analysis) Model prediction Field data Mortality Rate (yr -1 ) DBH Size Class (cm) (1) Testing radiative transfer schemes: Discretized Perfect Plasticity Approximation (PPA) (4) Processes underlying demographic structure Manaus Case Study – testing current plant mortality • ED2 = over-estimates mortality (4.2%) • FATES = over-estimates mortality (3.2%) • ZELIG-TROP = similar to observed (1.2%) (2) Testing competition for water/plant hydraulics and integration with trait based forest model: Examples of near-term development priorities for FATES and progress towards demographic ESMs: FATES model (Functionally-Assembled Terrestrial Ecosystem Simulator) 2 • Carbon pools, fluxes, allocation; litter fluxes; phenology; regeneration, growth, mortality represented by ED (Ecosystem Demography Model). • Canopy physics, soil BGC, land surface hydrology, photosynthesis, respiration represented by ALM. • Incorporates discretized PPA for canopy structure. • ‘Some’ current development foci: - Introduction of plant hydrodynamics and competitive plant water uptake - Librarification of ED code to allow multi-model compatibility (ACME/CESM/ARCOS) - Sensitivity analysis to input parameters - Multi-assumption photosynthesis module testing - Mechanistic mortality algorithms vs. static turnover Vegetation demographics in ACME: Capturing structural forest dynamics, plant co-existence, and plant functional shifts Jennifer Holm 1 , Ryan Knox 1 , Charlie Koven 1 , William Riley 1 , Chonggang Xu 2 , Elias Massoud 2 1 Lawrence Berkeley National Laboratory, 2 Los Alamos National Laboratory The inclusion of vegetation demography into Earth System Models (ESMs) will better represent plant ecology, and vegetation processes that govern fluxes of carbon, energy, water. • However, incorporating dynamic vegetation demography poses huge challenges owing to their increased model complexity. • Current issue = global application of dynamic vegetation in ALM Global vegetation demography developments and science impacts: • Investigate current and future outcomes of FATES as a result of various drought scenarios and climate change in California, over the 21 st century. • We hypothesize that under drought conditions in California the mortality of all trees will increase, but there will be higher mortality for large trees. • Secondly, we hypothesize that changing climates in the 21 st century will cause the climate of southern California to migrate to northern California. V c,max25 Interactions All others parameters V c,max25 Growth respiration fraction Target storage carbon GPP NPP Total LAI Total Biomass Leaf maintenance respiration Leaf allometry Target storage carbon V c,max25 Specific leaf area Stem allometry coef c Target storage carbon V c,max25 Leaf maintenance respiration Target storage carbon (3) Sensitivity analysis of 66 input parameters into FATES 3 • FATES has >200 parameters • Single site testing in Brazil • Using Fourier Amplitude Sensitivity Testing (FAST) = variance based sensitivity analysis. • Repeatedly found to be important for carbon dynamics = Vcmax,25; target storage carbon, stem allometry coef. 0 10 20 30 40 50 60 70 80 10-15 15-20 20-30 30-40 40-50 50-60 60-70 70-80 80-90 90-100 >100 Basal Area (m2/ha) Old parameter file Total BA = 84.2 m2/ha 100 yrs 200 yrs 0 2 4 6 8 10 12 10-15 15-20 20-30 30-40 40-50 50-60 60-70 70-80 80-90 90-100 >100 Basal Area (m2/ha) New parameter file Total BA = 30.6 m2/ha 100 yrs 200 yrs 300 yrs 400 yrs 500 yrs DBH Size Class (cm) DBH Size Class (cm) Changes to FATES parameter file The role of ecosystem heterogeneity and diversity 1 : Aggregated “big-leaf” ecosystem vs. demographic; structured ecosystem means ability to capture differences in biomass with dry season length. ED2 – Ecosystem Demography ED2- BL (“Big-leaf” analog) The Ecosystem Demography (ED) model vegetation structure, the basis for ALM-FATES. Tracks age and size of tree “cohorts”, incorporates disturbance , and dynamic turnover. But these processes can lead to more model variability. Testing variations in allometry equations for Western US evergreen trees 4 (currently 1 global allometric equation for all PFTs) References = 1. Levine NM, Zhang K, Longo M et al. (2016) Ecosystem heterogeneity determines the ecological resilience of the Amazon to climate change. Proceedings of the National Academy of Sciences, 113, 793-797. 2. Fisher, R. A., et al. Taking off the training wheels: the properties of a dynamic vegetation model without climate envelopes, CLM4.5(ED), Geosci. Model Dev., 8, 3593-3619, 2015. 3. Massoud, E.C., C. Xu, et al. Identification of key biological controls in forest dynamics based on a sensitivity analysis to the Community Land Model with Ecosystem Demography, CLM4.5(ED), JGR-Biosciences, Submitted. 4. Ter-Mikaelian, M. T. and M. D. Korzukhin. Biomass equations for sixty-five North American tree species. Forest Ecology and Management 97(1): 1-24., 1997. 0 1 2 3 4 5 6 1 30 59 88 117 146 176 Basal Area (m2/ha) Years Default CA_FATES Needleleaf, evergreen Broadleaf, deciduous Testing default FATES in Sierra Forest, CA, USA. (Single point) Very low basal area, biomass, and stem density. Competitive exclusion of needleleaf evergreen trees AGB = 18 MgC/ha; Density = 87 stems/ha
Transcript of Vegetation demographics in ACME: Capturing structural ...€¦ · Jennifer Holm1, Ryan Knox1,...
Solution Attempts
For additional information, contact:
Jennifer Holm Research Scientist
Lawrence Berkeley National Laboratory
(510) 495-8083 [email protected]
climatemodeling.science.energy.gov/acme
Issues concerning dynamic vegetation modeling
Solution Attempts
Ideas and future evaluation & testing
PFT1% PFT1%
PFT2% PFT2%PFT3%
PFT2%
PFT3%
PFT1% PFT2% PFT3% PFT2% PFT3%
Spa+ally%average%the%direct%and%diffuse%radia+on%transmi=ed%onto%the%understory%layer.%%
Reflectance,%absorp+on%and%transmi=ance%calculated%for%mul+ple%leaf%layers%for%each%PFT.%%
Direct% Diffu
s
e %
Light%transmi=ed%onto%soil/snow%reflected%back%up%through%canopy%(itera+ve%solu+on)%%
Individual tree stomata response (FLCgs) (prev timestep)
Individual tree canopy-level transpiration (E)
(current timestep)
Plant traits: wood + leaf economics
(WD, LMA, NL, PL)
Tree size: height, total
leaf area
HOST LAND SURFACE MODEL – FATES
PLANT HYDRAULICS MODULE
TRA
IT TRA
DE-O
FFS FOR
PFT PA
RA
METER
IZATIO
N
(from
me
ta-analysis)
Model prediction
Field data M
ort
alit
y R
ate
(yr-1
)
DBH Size Class (cm)
(1) Testing radiative transfer schemes: Discretized Perfect Plasticity Approximation (PPA)
(4) Processes underlying demographic structure
Manaus Case Study – testing current plant mortality • ED2 = over-estimates mortality (4.2%) • FATES = over-estimates mortality (3.2%) • ZELIG-TROP = similar to observed (1.2%)
(2) Testing competition for water/plant hydraulics and
integration with trait based forest model:
Examples of near-term development priorities for FATES and progress towards demographic ESMs:
FATES model (Functionally-Assembled Terrestrial Ecosystem Simulator)2
• Carbon pools, fluxes, allocation; litter fluxes; phenology; regeneration, growth, mortality represented by ED (Ecosystem Demography Model).
• Canopy physics, soil BGC, land surface hydrology, photosynthesis, respiration represented by ALM. • Incorporates discretized PPA for canopy structure. • ‘Some’ current development foci: - Introduction of plant hydrodynamics and competitive plant water uptake - Librarification of ED code to allow multi-model compatibility (ACME/CESM/ARCOS) - Sensitivity analysis to input parameters - Multi-assumption photosynthesis module testing - Mechanistic mortality algorithms vs. static turnover
Vegetation demographics in ACME: Capturing structural forest dynamics, plant co-existence, and plant functional shifts
Jennifer Holm1, Ryan Knox1, Charlie Koven1, William Riley1, Chonggang Xu2, Elias Massoud2
1 Lawrence Berkeley National Laboratory, 2 Los Alamos National Laboratory
The inclusion of vegetation demography into Earth System Models (ESMs) will better represent plant ecology, and vegetation processes that govern fluxes of carbon, energy, water.
• However, incorporating dynamic vegetation demography poses huge challenges owing to their increased model complexity.
• Current issue = global application of dynamic vegetation in ALM
Global vegetation demography developments and science impacts:
• Investigate current and future outcomes of FATES as a result of various drought scenarios and climate change in California, over the 21st century.
• We hypothesize that under drought conditions in California the mortality of all trees will increase, but there will be higher mortality for large trees.
• Secondly, we hypothesize that changing climates in the 21st century will cause the climate of southern California to migrate to northern California.
Sensitivity outputs-overall dynamics
6
Vc,max25 Interactions All others parameters
Vc,max25 Growth respiration fraction Target storage carbon
GPP
NPP
Total LAI
Total Biomass
Leaf maintenance respiration
Leaf allometry Target storage carbon Vc,max25 Specific leaf area
Stem allometry coef c Target storage carbon Vc,max25 Leaf maintenance respiration
Target storage carbon
(3) Sensitivity analysis of 66 input parameters into FATES3
• FATES has >200 parameters • Single site testing in Brazil • Using Fourier Amplitude Sensitivity Testing
(FAST) = variance based sensitivity analysis. • Repeatedly found to be important for
carbon dynamics = Vcmax,25; target storage carbon, stem allometry coef.
0
10
20
30
40
50
60
70
80
10-1
5
15-2
0
20-3
0
30-4
0
40-5
0
50-6
0
60-7
0
70-8
0
80-9
0
90-1
00
>10
0
Basal A
rea (
m2/h
a)
Old parameter file Total BA = 84.2 m2/ha 100 yrs
200 yrs
0
2
4
6
8
10
12
10-1
5
15-2
0
20-3
0
30-4
0
40-5
0
50-6
0
60-7
0
70-8
0
80-9
0
90-1
00
>10
0
Basal A
rea (
m2/h
a)
New parameter file Total BA = 30.6 m2/ha
100 yrs
200 yrs
300 yrs
400 yrs
500 yrs
DBH Size Class (cm) DBH Size Class (cm)
Changes to FATES parameter file
The role of ecosystem heterogeneity and diversity1: Aggregated “big-leaf” ecosystem vs. demographic; structured ecosystem means ability to capture differences in biomass with dry season length.
ED2 – Ecosystem Demography ED2- BL (“Big-leaf” analog)
The Ecosystem Demography (ED) model vegetation structure, the basis for ALM-FATES. Tracks age and size of tree “cohorts”, incorporates disturbance , and dynamic turnover. But these processes can lead to more model variability.
Testing variations in allometry equations for Western US evergreen trees4 (currently 1 global allometric equation for all PFTs)
References = 1. Levine NM, Zhang K, Longo M et al. (2016) Ecosystem heterogeneity determines the ecological resilience of the Amazon to climate change. Proceedings of the National Academy of Sciences, 113, 793-797. 2. Fisher, R. A., et al. Taking off the training wheels: the properties of a dynamic vegetation model without climate envelopes, CLM4.5(ED), Geosci. Model Dev., 8, 3593-3619, 2015. 3. Massoud, E.C., C. Xu, et al. Identification of key biological controls in forest dynamics based on a sensitivity analysis to the Community Land Model with Ecosystem Demography, CLM4.5(ED), JGR-Biosciences, Submitted. 4. Ter-Mikaelian, M. T. and M. D. Korzukhin. Biomass equations for sixty-five North American tree species. Forest Ecology and Management 97(1): 1-24., 1997.
0
1
2
3
4
5
6
1 30 59 88 117 146 176
Bas
al A
rea
(m2
/ha)
Years
Default CA_FATES
Needleleaf, evergreen
Broadleaf, deciduous
Testing default FATES in Sierra Forest, CA, USA. (Single point) Very low basal area, biomass, and stem density. Competitive exclusion of needleleaf evergreen trees
AGB = 18 MgC/ha; Density = 87 stems/ha
