Process Studies in
Photosynthesis to Improve
Representation of Vegetation
in Models
Next-Generation Ecosystem
Experiments (NGEE Arctic)
Project
Alistair Rogers ([email protected])
Environmental & Climate Sciences Department
Brookhaven National Laboratory
Model Representation of Arctic Photosynthesis
Goal
• Quantify model sensitivity and target critical areas where new
data will reduce model uncertainty.
• Link measurement of key parameters to spectral signatures that
enable scaling.
• Test and inform models iteratively.
Leaf
Spectra
GPPmodeled
Tram
Scaling
Complexity
Airborne
Spaceborne
UQ
2
Respiration
Photosynthesis
Early NGEE-Arctic example - Photosynthetic capacity
Barrow, AK
Autotrophic
Heterotrophic
𝐴 = min 𝑤𝑐 , 𝑤𝑗
Γ∗
1−
− 𝑅𝑑
𝑐𝑖
Farquhar et al. (1980)
3
Two key variables for modeling CO2 uptake are the maximum
carboxylation rate -photosynthetic capacity- (Vc,max) and the electron
transport rate (J)
Rubisco
(dark reactions)
RuBP regeneration
(light reactions)
𝑤𝑐 =
𝑉𝑐,𝑚𝑎𝑥 𝑐𝑖
𝑐𝑖 + 𝑘𝑐
𝑂
1+
𝑘𝑜
𝐽𝑐𝑖
𝑤𝑗 =
4.5𝑐𝑖 + 10.5Γ ∗
4
Farquhar et al. (1980)
CO2 Assimilation Rate
These key parameters can be estimated from an “A-ci curve”
vc,max
Jmax
Petasites frigidus, Barrow, AK
CO2 Concentration
5
Photosynthesis
Autotrophic
Heterotrophic
Respiration
Vc,max drives estimation of photosynthesis in Earth System Models
(ESMs). Through multipliers it is also used to estimate other parameters,
including Jmax and autotrophic respiration.
Parameterization of Vc,max
in ESMs impacts Net
Primary Production
through two of the three
major CO2 fluxes.
6
Barrow, AK
Models are extremely sensitive to parameterization of Vc,max, e.g. in CLM
two of the parameters used to determine Vc,max (red arrows) have a large
impact on model outputs
Sensitivity analysis of the impact of 80 CLM inputs (x-axis) on five key outputs (yaxis). Model inputs with darker colors have a greater impact on model outputs.
Sargysan et al. (2013)
Te
NE mp
NDT Bera
BEBE T orete
T T T Bor al
T r e
BDBD em opi al
T T T per cal
T r a
BE B em opi te
p c
BDS TDT er al
S emBo ate
Te p rea
Tu
nd B m era l
ra DS pe te
ve B rat
g or e
C3eta eal
t
C4 Gr ion
a
C3Grass
C4 Cr ss
Crop
op
NE
T
-2
-1
Vc,max (µmol m s )
Photosynthetic capacity is poorly represented in models – especially in
the Arctic (highlighted)
200
160
120
AVIM
BETHY
Biome-BGC
CLM
CTEM
Hybrid
IBIS
JULES
O-CN
ORCHIDEE
80
40
0
Plant Functional Type
Rogers (2014)
Current model representation of photosynthetic capacity in Arctic Plant
Functional Types is based limited data and unsupported assumptions
AVIM (Beijing Climate Center Model) uses Vc,max to
tune their model to match remotely sensed GPP and
site specific Eddy Covariance data
𝑉𝑐,𝑚𝑎𝑥 = 𝑖𝑣 + 𝑠𝑣 . 𝑁𝑎
1000
𝑉𝑐,𝑚𝑎𝑥 = 𝑁𝑎 . 𝑛2 . 𝑛𝑓 . 𝑀
1
. 𝐹𝐿𝑁𝑅 . 𝐹𝑁𝑅
𝐿 .𝑆𝐿𝐴
𝑉𝑐,𝑚𝑎𝑥 = 𝐶𝑁
𝑁
. α𝑅25
BETHY (JSBACH) uses a data from a single from an
undefined source
Hybrid uses an arbitrary multiplier (nf tundra = 0.75nf
deciduous forest + 0.25nf tropical forest)
CLM4.0 uses parameters derived from datasets lacking
representation of Arctic species
Rogers (2014)
9
Measuring Photosynthetic capacity in Barrow
Mosquito clogged cooling fan
Measurements aimed to capture dominant vegetation, but also
species representing major plant functional types
Chapin et al. (1996)
11
D
st
i
up s la
tif
o
Ar ntia olia
ct
op fish
Er
io
ph Ca hilia eri
or
r
um ex fulv
a
a
an qu
Pe gu atil
ta stif is
si
te oliu
s
m
Sa frig
lix idu
pu s
lc
hr
a
AV
I
BE M
T
C HY
LM
-C
N
H
yb
rid
ro
Ar
ct
ag
-1
150
-2
125
vc,max25 (µmol m s )
Photosynthetic capacity is underestimated in the Arctic
175
100
Mean +/- SE (n=9-21)
Grass
Sedge
Forb
Shrub
ESM
75
50
25
0
Rogers Unpublished
12
Vc,max is very sensitive to temperature so low values at 25oC are really
low at observed temperatures
CLM-CN Vc,max25 = 35
50
-2
-1
Vc,max (µmol m s )
60
40
30
20
10
0
0
5
10
15
Tleaf (oC)
20
25
30
Bernacchi et al. (2001)
13
Modeling photosynthesis reveals flaws in ESM estimates of
photosynthetic capacity
25
ESM
20
Salix pulchra
Tleaf = 6.6oC
15
PAR = 500 µmol m-2 s-1
NGEE-Arctic
5
0
Salix pulchra
20 T = 15.6oC
leaf
-2 -1
15 PAR = 1700 µmol m s
10
5
0
Petasites frigidus
20 T = 16.1oC
leaf
-2
-1
15 PAR = 1700 µmol m s
10
5
od
el
ed
ed
M
O
bs
er
v
yb
rid
H
N
-C
Y
LM
C
TH
BE
IM
0
AV
-2
-1
Photosynthesis (µmol m s )
10
Rogers Unpublished
14
These data are also providing key inputs that inform the incorporation of
N partitioning into new model frameworks
Light
Absorption
Electron
Transport
Light harvesting
Carboxylation
Photosynthetic
Respiratory
Growth
Storage
Functional
Structural
Whole Plant Nitrogen
Xu et al. (2012)
15
Future Direction
• Develop and validate spectral signatures for Arctic plant
physiological traits.
• Scale leaf level process knowledge using near surface,
airborne and satellite sensors.
Leaf
Tram
Scaling
Complexity
Serbin et al. (2012)
Airborne
Spaceborne
16
Interagency links with NASA-ABoVE
Prototype Vc,max Map (NASA HyspIRI)
Page 95 Photosynthetic
Capacity (Vc,max) is
classified
as
a
provisional
remotely
sensed data product
requiring resources for
further refinement.
Serbin et al. unpublished
We aim to provide a spatially and temporally resolved trait database
for key physiological traits, and an independent method to ground
truth prognostic state variables generated by new models.
17
The Next-Generation Ecosystem Experiments
(NGEE Arctic) project is supported by the Office of
Biological and Environmental Research in the DOE
Office of Science.
18