assimila'on*of*forest*remote*sensing*biomass*and* height...
TRANSCRIPT
E.#Joetzjer,#J.#Chave,#P.#Ciais,#B.#Poulter,#P.#Peylin##
LSCE/EDB###
26/09#
Assimila'on*of*forest*remote*sensing*biomass*and*height*data*into*process5based*models:*
ORCHIDEE*in*the*tropics**##
#• Forests#regulate#biogeochemical#cycles#(at#global#scale)##
#e.g.#carbon#sequestraHon:#1/3#of#anthropogenic#CO2#emission#(Le#Quere#et#al.,2017)##• Signs#of#forest#carbon#sink#weakening#(Nabuurs#et#al.,#2013,#Brienen#et#al.,#2015)#in#
response#to#global#anthropogenic#disturbances##
#########
2#
CONTEXT#
#• Forests#regulate#biogeochemical#cycles#(at#global#scale)##
#e.g.#carbon#sequestraHon:#1/3#of#anthropogenic#CO2#emission#(Le#Quere#et#al.,2017)##• Signs#of#forest#carbon#sink#weakening#(Nabuurs#et#al.,#2013,#Brienen#et#al.,#2015)#in#
response#to#global#anthropogenic#disturbances#
• PotenHal#biophysical#&#biochemical#feedback#on#climate###eg.#Amazon#dieback#(Cox#et#al.,#2000)##
##########
3#
Brienen#et#al.,#2015#312#undisturbed#plots#(Amazon#forest)###
CONTEXT#
##
Tools#to#monitor#forest#ecosystem#(ObservaHons#and#process#based#models)#
Physiology#param##e.g.#VCMAX,#turgor#loss#point#
Field#inventory###DBH,#specie,#tree#height#
FluxTOWER#(GPP,#LE)# Remote#sensing##Canopy#height,#AGB,#LAI,##
4#
##
3#
##Stochas'c*individual*based*models*
*e.g.*TROLL,*FORMIND*
Cohorts*based*models**e.g.*ORCHIDEE5CAN,*FATES*
“Big*Leaf”*land*surface*models**
e.g.*ORCHIDEE,*CLM**
Tools#to#monitor#forest#ecosystem#(ObservaHons#and#process#based#models)#
Benchmarking#ParameterizaHon#IniHalizaHon####
##
3#
##Stochas'c*individual*based*models*
*e.g.*TROLL,*FORMIND*
Cohorts**e.g.*ORCHIDEE5CAN,*FATES*
“Big*Leaf”*land*surface*model**e.g.*ORCHIDEE5TRUNK,*CLM**
Tools#to#monitor#forest#ecosystem#(ObservaHons#and#process#based#models)#
ORCHIDEE5TRUNK*
ORCHIDEE:# ##• exchange#Carbon#H20#&#energy#between#atmosphere#and#surface#• mulH`scale#model#(1#pixel#to#global)##
ORCHIDEE#
ORCHIDEE:# ##• exchange#Carbon#H20#&#energy#between#atmosphere#and#surface#• mulH`scale#model#(1#pixel#to#global)#–#can#be#ONLINE#(feedbacks)#• several#versions###
##
ORCHIDEE*
##
ORCHIDEE*
New*genera'on*DGVM*–*forest*structure*–*open*new*possibili'es*to*combine*OBS*and*Models*examples#of#using#field#inventory#and#remotely#sensed#data#into#ORCHIDEE`CAN##*
Allometric*Height5Diameter*calibra'on#data#from#Afrisar#&#Tropisar#*Tree#Height=A*Diameterb#####
##*
10 20 50 100 200
10
20
30
40
50
Selected model : log1
D (cm)
H (m
)
DataModel selected
10 20 50 100
5
10
20
50
Selected model : log1
D (cm)
H (m
)
DataModel selected
10 20 50 100
5
10
20
50
Selected model : log1
D (cm)
H (m
)
DataModel selected
10 20 50 100 200
10
20
50
Selected model : log1
D (cm)H
(m)
DataModel selected
10 20 50 100 200
10
20
50
Selected model : log1
D (cm)
H (m
)
DataModel selected
10 20 50 100
5
10
20
50Selected model : log1
D (cm)
H (m
)
DataModel selected
LOP # # ###MAB # # #######MON# # ############NOU# # # #PAR ###############################RAB#
Linear#fit#used#in#for#CAN#configuraFon#set#up#
CALIBRATION*site*specific*e.g.*ALLOMETRY***
Allometric*Height5Diameter*calibra'on#data#from#Afrisar#&#Tropisar#*Tree#Height=A*Diameterb#####
##*
10 20 50 100 200
10
20
30
40
50
Selected model : log1
D (cm)
H (m
)
DataModel selected
10 20 50 100
5
10
20
50
Selected model : log1
D (cm)
H (m
)
DataModel selected
10 20 50 100
5
10
20
50
Selected model : log1
D (cm)
H (m
)
DataModel selected
10 20 50 100 200
10
20
50
Selected model : log1
D (cm)H
(m)
DataModel selected
10 20 50 100 200
10
20
50
Selected model : log1
D (cm)
H (m
)
DataModel selected
10 20 50 100
5
10
20
50Selected model : log1
D (cm)
H (m
)
DataModel selected
LOP # # ###MAB # # #######MON# # ############NOU# # # #PAR ###############################RAB#
Linear#fit#used#in#for#CAN#configuraFon#set#up#
Sites* CODE*H/D*allometric*equa'on*
A*(m)* b* Nbtrees*
LOPE# LOP* 44.148# 0.550# 499#
MABOUINE# MAB* 41.767# 0.531# 592#
MONDAH# MON* 40.671# 0.506# 510#
NOURAGUES# NOU* 46.538# 0.534# 3154#
PARACOU# PAR* 44.608# 0.561# 1092#
RABI# RAB* 38.515# 0.470# 1060#●
●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●
●●●●●●●●●●●●
●●●●●●●●●●●●
●●●●●●●●●●●●
●●●●●●●●●●●●●
●●●●●●●●●●●●●●
●●●●●●●●
0 50 100 150 2000
10
20
30
40
50
60
70
D (cm)
H (m
)
CANLOPMABMONNOUPARRAB
CALIBRATION*site*specific*e.g.*ALLOMETRY***
Labriere#et#al.,#2018#
BENCHMARKING*(forest*structure*from*field*inventory*and*pseudo*obs)***
13#
0 20 40 60 80 100
0
50
100
150
200
Above Ground Biomass
years
AGB
(tC/h
a)
REFTRUNKCANCAN−RS
0 20 40 60 80 100
0
10
20
30
40
Basal Area
years
BA (m
2/ha
)
REFCANCAN−RS
Forest*establishment**
Joetzjer#et#al.,#2018#
PSEUDO`OBS:#Bayesian#modelling#to#assimilate#30yrs#of#forest#inventory#and#simulate#long#term#C#accumulaHon#(Chave#et#al#in#prep)#
BENCHMARKING*(forest*structure*from*field*inventory*and*pseudo*obs)***
14#
●
●
●●
●
●
●
●●
● ●●
●
●
●● ●● ● ●
●
●
●●
●
●
●●
●
● ● ●
●●
●
●
●
●
●
●
●●
●
●●
●●●
●
●
●●
●
●●●
●
●
●
●
●
●
●
● ●●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●●
●
●●
●● ●
●
●●
●
●
●●
●
●●
●
●
●●
●
●
●
●
●
●
●
●
●
●
● ●●●
●
●●●
●
●●
●●
● ●
●
●●
●
●●
●
●
●
●
●
●●
●
●
● ●●
●●
●
●
●●
●
●
●
●
●
●●
●
●
●
● ●
● ●
●●●●
●●●
●
●●
●
●
●
●●●
●●●
●●●
●
●
●
●
● ●●
●
●●
●
●
●
●
●
●
●
●●●
●
●
●●
●●●
● ●
●
●●
●
●
●●
●●
●●
●
●●
●●●
●
●●
●
●
●
●
●
● ●●
●
●●
●
●●●
●
●
●●●
●
●●
●
●
●●
●
●● ●●
●
●
●●
●
●
●●
●
●
●●●●
●●
●
●
●
●
●
●●
●●●
●
●●●●
●
●●
●
●
●
●●●
●●
●●●●
●
●●
●●
●
●
●●
●
●
●
●
●
●●●●
●●●
●
●
●●
●
●
●
●●
●●
●
●
●
●
●
●
●
●
●
●●
●
●
●●
●
●
●
●
●
●
●●
●
●
●
●
●
●● ●
●
●
●●
●
●
●
●●
●
●●
●
●
●
●●
●
●
●
●
●●
●
●
●●
●●
●●
●
●
●
●●●
●
●
●●●
●●
●
●
●●●
●
●
●
●
●
●
●
●
●
●
●
●
●●
● ●
●
●
●
●
●
●●●
●
●
●
●
●
●
●
●
●●
●
●
●
●●●
●
●
●●●
● ●
●
●
●
●
●
●● ● ●
●
●
●●●
●●
●●
●
●
●●
● ●
●●
●●
●
●●
●●●
●
●
●●●
●
●
●
●
●
●●
●
●●
●
●
●
●●
●
●●
● ●
●
● ●
●
●
●●
●
●
●●
●
●
●●●
●●
●
●
●
●
●
●
●
●
●
●
●●
●
● ●●
●
●●●
●
●
●
●
●
●●
●
●●
●
●
●
●●●
●●
●●
●●
●●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●●
● ●●
●
●
●
●
●
●
●●●●
●
●●●
●
●●●
●
●
●●
●●
●●
●
●
●●
●●
●
●
●●●
●●
●●
●
●
●
●
●●
●
●
●●
●
●●
●
●
●
●
●
●●
●●
●●●
●
●
●
●
●
●
●
●●
●
●●
●
●
●
●
●
●
●
●
●
●
●● ●
●
●
●●
●
●
●●
●
●
●
●●●
●
●
●
●
●●
●●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●●
●●
●
●
●
●
●●●
●
●
●●●●
●
●
●
●
●
●
●
●
●
●
●●●
●
●
●●●
●
●●
●●
●
●
●
●
●
●●● ●
●●
●●
●●●
●
●
●
●
●
●
●
●
● ●●●
●
●
●
●●●●●●●
●
●
●● ●
●●●●●
●●●
●
●
●
●
●
●
●●
●
●
●●●●
●●
●
●●●
●
●●●
●
●
●
●
●
●
●●
●
● ●
●●●●
●
●●
●●
●●
●
●
●
●
●
●
●
●●
●
●●●
●●
●
●●●
●
●●●
●
●●●●
●●●
●
●●
●●●
●●
●
●●
●●●●●●●●
●
●
●
●
●
●
●●●●●●
●●
●●●●
●●
●●●●
●●●●●●●●●●●●
●●
●●●●●
●
●●●
●●
●●
●
●
●●●●
●
●●●●●
●●●●
●●●●●●●●
●
●●●
●●
●
●●
●
●
●
●●●●
●
●●
●
●
●● ●
●
●●
●●●
●●
●
●●
●●●●●
●●
●●
●
●
●
●
●
●
●●
●
●
● ●●●
●
●●
●
●● ●
●
●
●
●
●
●●
●●
●
●●●
●
●
●●●●
●
●
●
●
●●
●
●
●
●
●●●●
●
●
●
●
●
●
● ●
●
●
●●●●
●
●
●●
●●
●●
●
●
●●
●●
●
●●
●
●
● ●
●
●
●
●
●
●
●
●
●
●●
●
● ● ●
●
●
●
●
●
●
●
●
●
●
●●
●
●●
●●
●
●
●
●
●
●
●●
●
●●
●
●
●
●
●●●
●
●
●●●
●
●
●
●
●●
●●●
●
●
●
●
●
●●
●
●
●●
●
●
●
●
●
●
●
●●●
●
●
●
●
●
●
●
●
●●
●●
●
●
●
●
●
●
●●
●
●
●
●●
●●
●●●
●
●●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●●
●●
●
●●
●
●●
●
●●
●
●●
●
●
●
●
●
●●
●
●
●
●
●●
●
●
●●
●●
●
● ●
●
●
●
●●
● ●● ●
●
●●
●●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●●
●●
●
●
●● ●●
●
●●
●
●
●
●
●
● ●
●
●
●
●
●●
●
●●
●●●●
●
●
●
●
●
●●●
●
●
●●
●
●
● ●
●●●●
●●
●
●
●
●●●
●
●
●●
●
●●
●
●
●
●
●●●
●
●
●●
●●●
●
●
●
●
●
●
●●
●● ●
●●
●
●●
●
●
●●
●
●●
●●
●●
●●
●●●
●
●●●●
●
●
●●●
●
●●●
●
●●●●
●
●
●●●●
●
●
●
●●●
●
●●●
● ●●●
●
●●●
●
●
●●●●●●
●●
●
●●●
●
●
●
●
●
●●
●
●●●
●●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●●
●
●●●●
●
●
●●
●● ●●
●
●●
●●
●●
●
●●● ●●
●
●
●
0 50 100 150 200
0
10
20
30
40
50
60
height vs. diameter
diameter (cm)
tree
heig
ht (m
)
forest inventoryCAN−RSallometric equation
a) diameter distribution
diameter (cm)
Ntre
es p
er h
ecta
re
0
100
200
300
400
500
600
0 20 40 60 80 100 120
Forest inventoryCAN−RS
b)
0 20 40 60 80 100
0
50
100
150
200
Above Ground Biomass
years
AGB
(tC/h
a)
REFTRUNKCANCAN−RS
0 20 40 60 80 100
0
10
20
30
40
Basal Area
years
BA (m
2/ha
)
REFCANCAN−RS
Forest*establishment**
Equilibrium*state*
Joetzjer#et#al.,#2018,#Chave#et#al#in#prep#
#######95p#CANOPY#HEIGHT##
285000 286000 287000 288000
5810
0058
2000
5830
0058
4000
Longitude E (meters UTM 22N)
Latit
ude
N (m
eter
s U
TM 2
2N)
10
20
30
40
Paracou assimilated 50th percentile height (m)
ORCHIDEE`CAN#
APPLICATION*Data*“assimila'on”*assuming*“perfect*model”**
#######95p#CANOPY#HEIGHT##
285000 286000 287000 288000
5810
0058
2000
5830
0058
4000
Longitude E (meters UTM 22N)
Latit
ude
N (m
eter
s U
TM 2
2N)
10
20
30
40
Paracou assimilated 50th percentile height (m)
ORCHIDEE`CAN#
APPLICATION*Data*“assimila'on”*assuming*“perfect*model”**
#######95p#CANOPY#HEIGHT##
285000 286000 287000 288000
5810
0058
2000
5830
0058
4000
Longitude E (meters UTM 22N)
Latit
ude
N (m
eter
s U
TM 2
2N)
10
20
30
40
Paracou assimilated 50th percentile height (m)
ORCHIDEE`CAN#
0 20 40 60 80 100
0
50
100
150
200
Above Ground Biomass
years
AGB
(tC/h
a)
REFTRUNKCANCAN−RS
0 20 40 60 80 100
0
10
20
30
40
Basal Area
years
BA (m
2/ha
)
REFCANCAN−RS
ORCHIDEE`CAN#
APPLICATION*Data*“assimila'on”*assuming*“perfect*model”**
18#
#######95p#CANOPY#HEIGHT##
285000 286000 287000 288000
5810
0058
2000
5830
0058
4000
Longitude E (meters UTM 22N)
Latit
ude
N (m
eter
s U
TM 2
2N)
10
20
30
40
Paracou assimilated 50th percentile height (m)
ORCHIDEE`CAN#
0 20 40 60 80 100
0
50
100
150
200
Above Ground Biomass
years
AGB
(tC/h
a)
REFTRUNKCANCAN−RS
0 20 40 60 80 100
0
10
20
30
40
Basal Area
years
BA (m
2/ha
)
REFCANCAN−RS
ORCHIDEE`CAN#
285000 286000 287000 288000
5810
0058
2000
5830
0058
4000
Longitude E (meters UTM 22N)
Latit
ude
N (m
eter
s U
TM 2
2N)
50
100
150
Paracou above ground biomass (MgC/ha)
ORCHIDEE`CAN#products#(exemples)##
285000 286000 287000 288000
5810
0058
2000
5830
0058
4000
Longitude E (meters UTM 22N)
Latit
ude
N (m
eter
s U
TM 2
2N)
10
15
20
25
30
Paracou mean crown area of dominant trees (m2)AGB# Crown#size#
285000 286000 287000 288000
5810
0058
2000
5830
0058
4000
Longitude E (meters UTM 22N)
Latit
ude
N (m
eter
s U
TM 2
2N)
1
2
3
4
Paracou litter (MgC/ha)Lijer#
APPLICATION*Data*“assimila'on”*assuming*“perfect*model”**
Joetzjer#et#al.,#2017#
19#
*Method*RelaHvely#fast#and#easy#(CalibraFon#with#sparse#forest#inventory)##Provide#Detailed#esHmates#of#forest#stocks#(mulHple#carbon#pools)#
APPLICATION*Data*“assimila'on”*assuming*“perfect*model”**
20#
*Method*RelaHvely#fast#and#easy#(CalibraFon#with#sparse#forest#inventory)##Provide#Detailed#esHmates#of#forest#stocks#(mulHple#carbon#pools)#*Limits:*assimila'on*of*PLEAIDES*canopy*height*
APPLICATION*Data*“assimila'on”*assuming*“perfect*model”**
21#
*Method*RelaHvely#fast#and#easy#(CalibraFon#with#sparse#forest#inventory)##Provide#Detailed#esHmates#of#forest#stocks#(mulHple#carbon#pools)#*Limits:*assimila'on*of*PLEAIDES*canopy*height*
APPLICATION*Data*“assimila'on”*assuming*“perfect*model”**
New#generaHon#of#data#coming##BIOMASS*(CNES5ESA)*�[Le*Toan*et*al.,*2011];*GEDI*(NASA)*�[Stysley*et*al.,*2015]**
ABG#(comp.#Mitchard#et#al.,#2014#forest#inventory#RAINFOR)#
PERSPECTIVES*on*how*to*constrain*AGB*using*RS*data*
#TRUNK#undersHmates#AGB#(background#tree#mortality)#CAN#overesHmates#AGB#self`thinning#parameters#(and#other)#calibrated#using#sites#in#the#east.####CAN#and#TRUNK#failed#to#reproduce#the#observed#SE`NW#AGB#gradient##(as#most#models,#Johnson#et#al.,#2016)##
23#
ABG#(comp.#Mitchard#et#al.,#2014#forest#inventory#RAINFOR)#
#Constant#mortality#in#TRUNK##Lijle#variability#in#CANs#even*with*forest*structure*##How*can*we*constrain*Carbon*Residence*Time*?*(AGB/NPP)*
PERSPECTIVES*on*how*to*constrain*AGB*using*RS*data*
●
0 5 10 15 20
0
50
100
150
200
250
300
AMAZON AGB vs. NPP mean 1981−2016
NPP (tC ha yr)
AGB
(tC
ha)
TRUNKCANCAN−RS
●
0 1000 2000 3000 4000
0
50
100
150
200
250
300
AMAZON AGB vs. P mean 1981−2016
P (mm yr)
AGB
(tC
ha
yr)
TRUNKCANCAN−RS
Above*ground*biomass*(key#forest#feature#monitored)#
PERSPECTIVES*on*how*to*constrain*AGB*using*RS*data*
Inspired#from#Mahli#et#al.,#2012#
Above*ground*biomass*(key#forest#feature#monitored)#
Well#constrained#in#models####`#trait#observaHons#(e.g.#Vcmax)##`#benchmark#(e.g.#fluxtower)#
As#important#as#photosynthesis#quanHfy#ABG#and#AGB#variaHon##But#poorly#constrained#
PERSPECTIVES*on*how*to*constrain*AGB*using*RS*data*
Above*ground*biomass*(key#forest#feature#monitored)#
Well#constrained#in#models####`#trait#observaHons#(e.g.#Vcmax)##`#benchmark#(e.g.#fluxtower)#
As#important#as#photosynthesis#quanHfy#ABG#and#AGB#variaHon##But#poorly#constrained#
PERSPECTIVES*on*how*to*constrain*AGB*using*RS*data*
PERSPECTIVES*on*how*to*constrain*AGB*using*RS*data*
Carbon#residence#Hme#underesHmated#in#Global#climate#model#compared#to#OBS#(Carvalhais#et#al.,#2014)###
CRT#in#model#is#more#sensiHve#to#T#than#P##=>#C#feedbacks#in#future#may#be#more#sensiHve#to#hydrological#changes#than#temperature##
• BIOMASS*data*=>#Carbon#residence#Hme#map#(PFT#dependent#parameter)#that#we#can#use#as#an#input#in#ORCHIDEE#using#e.g.#NPP#from#FLUXNET#(Jung#et#al.2017)#• SpaHal#gradient#of#AGB##• Succession?##• Disturbance?#
• Model*development:*mortality*processes**#
PERSPECTIVES*on*how*to*constrain*AGB*using*RS*data*