"3.:*67.8=4+438&3&"3.:*67.8=4+438&3&
(-41&6#4607&8"3.:*67.8=4+438&3& (-41&6#4607&8"3.:*67.8=4+438&3&
92*6.(&1!*66&)=3&2.( .291&8.436495
9'1.(&8.437
92*6.(&1!*66&)=3&2.( .291&8.436495
*(*38(1.2&8*&3)@6*).7896'&3(*.25&(8743'46*&1&3)&6(8.(*(*38(1.2&8*&3)@6*).7896'&3(*.25&(8743'46*&1&3)&6(8.(
*(47=78*2564)9(8.:.8=*78.2&8*)97.3,&7&8*11.8*'&7*)*(47=78*2564)9(8.:.8=*78.2&8*)97.3,&7&8*11.8*'&7*)
8*66*786.&1(&6'43A9<24)*18*66*786.&1(&6'43A9<24)*1
$43,-43,$.
"3.:*67.8=4+438&3&.77491&
4-3 .2'&11
"3.:*67.8=4+438&3&.77491&
9(&743*7
41+*.(-1*
&2&06.7-3&*2&3.
**3*<85&,*+46&)).8.43&1&98-467
4114;8-.7&3)&)).8.43&1;4607&8-88577(-41&6;4607928*)9387,%59'7
*897034;-4;&((*77848-.7)4(92*38'*3*@87=49
*(422*3)*).8&8.43*(422*3)*).8&8.43
$.$ .2'&1143*7*.(-1**2&3.&3)&6,41.7*(*38(1.2&8*&3)
@6*).7896'&3(*.25&(8743'46*&1&3)&6(8.(*(47=78*2564)9(8.:.8=*78.2&8*)97.3,&7&8*11.8*'&7*)
8*66*786.&1(&6'43A9<24)*1*45-=7*7.4,*47(.?)4./,6,
!-.768.(1*.7'649,-884=49+46+6**&3)45*3&((*77'=8-*92*6.(&1!*66&)=3&2.( .291&8.436495&8
(-41&6#4607&8"3.:*67.8=4+438&3&8-&7'**3&((*58*)+46.3(197.43.392*6.(&1!*66&)=3&2.( .291&8.43
64959'1.(&8.437'=&3&98-46.>*)&)2.3.786&8464+ (-41&6#4607&8"3.:*67.8=4+438&3&46246*.3+462&8.43
51*&7*(438&(87(-41&6;4607274928*)9
JOURNAL OF GEOPHYSICAL RESEARCH: BIOGEOSCIENCES, VOL. 118, I-I 7 , doi:I0.I002/jgrg.20053, 2013
Recent climate and fire disturbance impacts on boreal and arctic
ecosystem productivity estimated using a satellite-based terrestrial
carbon flux model
Yonghong Y i / ’^ John S. K im b all/’^ Lucas A. Jo n e s/’^ Rolf H. Reichle,^
Ramakrishna Nemani,"^ and Hank A. Margolis^
Received 24 September 2012; revised 19 M arch 2013; accepted 23 M arch 2013.
[i] Warming and changing fire regimes in the northem (>45°N) latitudes have
consequences for land-atmosphere carbon feedbacks to climate change. A terrestrial
carbon flux model integrating satellite Normalized Difference Vegetation Index and
burned area records with global meteorology data was used to quantify daily vegetation
gross primary productivity (GPP) and net ecosystem CO2 exchange (NEE) over a
pan-boreal/Arctic domain and their sensitivity to climate variabiiity, drought, and fire
from 2000 to 2010. Model validation against regional tower carbon flux measurements
showed overall good agreement for GPP (47 sites: R = 0.83, root mean square
difference (RMSD)= 1.93gCm ^^d^^) and consistency for NEE (22 sites: R = 0.56,
RMSD = 1.46 g C m^^ d^^). The model simulations also tracked post-fire NEE recovery
indicated from three boreal tower fire chronosequence networks but with larger model
uncertainty during early succession. Annual GPP was significantly (p < 0.005) larger in
warmer years than in colder years, except for Eurasian boreal forest, which showed greater
drought sensitivity due to characteristic warmer, drier growing seasons relative to other
areas. The NEE response to climate variabiiity and fire was mitigated by compensating
changes in GPP and respiration, though NEE carbon losses were generally observed in
areas with severe drought or burning. Drought and temperature variations also had larger
regional impacts on GPP and NEE than fire during the study period, though fire
disturbances were heterogeneous, with larger impacts on carbon fluxes for some areas and
years. These results are being used to inform development of similar operational carbon
products for the NASA Soil Moisture Active Passive (SMAP) mission.
Citation: Yi, Y., J. S. Kimball, L. A. Jones, R. H. Reichle, R. Nemani, and H. A. Margolis (2013), Recent climate and fire
disturbance impacts on boreal and arctic ecosystem productivity estimated using a satellite-based terrestrial carbon flux
model, J. Geophys. Res. Biogeosci., 118, doi: 10.1002/jgrg.20053.
1. Introduction have experienced generally greater warming than other
global areas in recent decades [Serreze and Francis, 2006].
[2] Northem boreal and arctic ecosystems are purported to
be a considerable land sink for atmospheric CO2 [Euskirchen
A lengthening growing season has been linked to vegetation
greening and enhanced productivity in the northem latitudes
eta/., 2006; McGwire eta/., 2009], though sensitivity of the . Nemani eta l 2003'Chen et al IQQC Beck and Goetz
.] .. .] . . . . . . y C - f i l U H l ’ c - f C t t . , IC -r I c - f C t t . , ^ U c - L A - L U l L l \ J
regional carbon cycle to a warming climate is uneertam W in u • • . . a
n T,, 2011 . However, increasing vegetation water stress and
McGwtreeta/., 2009;//ayex eta/., 2011 . The mgnlatitudes , j u rr 1
2 > J to enhanced autumn respiration may be ottsettmg the potential
benefits of longer growing seasons and decreasing regional
'F lathead Lake Biological Station, The U niversity o f M ontana, Poison, Carbou sequestration \Angert et al., 2005, Piao et al., 2008,
Montana, USA. Zhang et al., 2008]. Boreal forests have had more fre-
^Numerical Terradynarmc Sim ulation Group, The University o f queut and widespread wildfirc and inscet disturbances [Rowt/-
M o n t o a MissoUa, M ontana U SA ,, , o Law/ierty ct a/., 2007; Kwrz et a/., 2008], which may alter vcg-
Global Modeling and A ssim ilation Office, NASA Goddard Space . . . , ^ . , . , , ,
Flight Center, Greenbelt, Maryland, USA. Composition and function and associated growth and
respiration processes [Amiro et al., 2010; CoursoUe et ah,
Faculte de foresterie, geographic et de 2012]. Moreover, a large portion of soil organic carbon stored
geom atique, Um versite Laval, Quebec, Quebec, Canada. northem borcal forcst and tuudra areas is potentially vuluer-
Corresponding author: Y. Yi, Flathead Lake Biological Station, The able to soil warming and likely more fi-cqucnt bumiug due to
University o f M ontana, Poison M T 59860, USA. climate waimiug \Turetsky et ah, 2010; Mack et ah, 2011].
(yonghong.yi@ ntsg.um t.edu) [3] diverse response of northem ecosystems to regional
©2013. American G eophysical Union. A ll Rights Reserved. Warming and drought has bccu reported from limited field
2l69-8 953/l3/l0.l002/jg rg.2 00 53 experiments [e.g., Chen et a l, 2006; Welp et ah, 2007;
YI ET AL.: DROUGHT AND FIRE IMPACT ON HIGH LATITUDE
Schwalm et al., 2010; Yi et al., 2010b; Peng et al., 2011],
while extension of these findings to the larger pan-boreal/
Aretie region is eonstrained by the limited extent of these
studies and a sparse regional measurement network. A reeent
deeline in northem eeosystem produetivity, espeeially in the
boreal forest, has been reported from satellite measurement
reeords [e.g., Angert et al., 2005; Goetz et al., 2007; Beck
and Goetz, 2011]. However, many of these studies are based
on satellite-derived vegetation “greenness” indiees (Vis)
ineluding the EVI (Enhaneed Vegetation Index) and NDVI
(Normalized Differenee Vegetation Index) that do not dis
tinguish underlying gross primary produetivity (GPP) and
respiration proeesses or their environmental drivers. Mean
while, the impaet of wildfire on the northem earbon eyele is
inereasing, while the assoeiated effeets on regional vegetation
and soil earbon reeovery have not been investigated using
eeosystem models until reeently [Balshi et a l, 2007; Yi
et a l, 2010a]. Therefore, regional applieations of eeosystem
models are desirable to elarify reeent impaets from elimate
variability and disturbanee on the northem earbon eyele.
[4] Satellite remote sensing offers a diverse set of land
parameter retrievals that ean serve as eritieal inputs to
regional eeosystem models for estimating land-atmosphere
earbon fluxes and assoeiated elimate impaets to the terrestrial
earbon budget [e.g., Nemani et al, 2003; Potter et al, 2003;
Zhang et al., 2008; Kimball et al., 2009]. The Moderate
Resolution Imaging Speetroradiometer (MODIS) provides
eontinuous, well ealibrated, and relatively long-term global
reeords that are sensitive to photosynthetie eanopy eover
[Running et al, 2004]; global monitoring Irom these sensors
ean also deteet abmpt disturbanee-related vegetation ehanges
[Mildrexler et al., 2007; Giglio et al., 2010]. Satellite miero-
wave sensors provide synergistie information on land surface
moisture and temperature variations owing to strong micro
wave sensitivity to assoeiated ehanges in land surface dieleetrie
properties and emissivity [Ulaby et al, 1982]. Land surface
retrievals at longer microwave wavelengths also have reduced
sensitivity to solar illumination effeets, clouds, and atmo
spheric aerosol contamination relative to optical sensors; these
properties have been exploited for determining daily landscape
freeze/thaw (FT) status and nonfrozen season variability, which
provides an effective surrogate for frozen temperature con
straints to vegetation produetivity and the potential growing
season in boreaPAretie regions [Kimball et al, 2004; Kim
et al., 2012]. The planned NASA Soil Moisture Active Passive
(SMAP) mission will provide global measurements of surface
soil moisture and FT status, with improved spatial resolution
(<10 km) and enhaneed L-band microwave sensitivity to soil
proeesses relative to current satellite microwave sensors
[Entekhabi et al, 2010]. The SMAP land parameter retrievals
will inform higher level land model simulations including a
planned level 4 earbon (L4_C) product that will provide regular
global estimates of terrestrial earbon fluxes and underlying
environmental drivers [Kimball et al, 2012]. These new
measurements and geophysical products are intended to
improve understanding of proeesses linking terrestrial water,
energy, and earbon cycles, quantify the net earbon flux in
boreal landscapes, and reduce uneertainties regarding the
piuported missing earbon sink on land
[Entekhabi et al, 2010].
[5] In this study, we applied a terrestrial earbon flux (TCP)
model partially driven by satellite-derived FPAR (Fraction
of Photosynthetieally Active Radiation absorbed by
vegetation), FT, and burned area inputs to estimate daily
GPP, net eeosystem CO2 exchange (NEE), and surface
(<10 cm depth) soil organic earbon (SOC) stocks over all
northem vegetated land areas. The TCP model used for this
study is similar to the L4_C algorithm being developed for
the SMAP mission. Our primary objectives were to use the
TCP projections to examine how reeent elimate variability
and fire disturbanee have affected northem GPP and NEE
earbon sink activity during the 11 year satellite record
(2000-2010). We hypothesized that potential produetivity
gains from regional warming still outweigh produetivity
losses caused by reeent drought stress and wildfire distur
banee in the northem latitudes, while prediction of NEE
is more uncertain due to similar, compensating GPP and
eeosystem respiration (Reco) responses to these factors. These
results were also used to test the initial algorithm and model
performance for the planned SMAP L4_C product.
[e] The following sections include descriptions of the
TCP model equations and methods used for the model sim
ulations, validation and uneertainty assessment (section 2);
presentation of model validation results relative to indepen
dent GPP and NEE estimates from northem tower eddy
covariance CO2 flux measurement sites and SOC data from
soil inventory reeords (section 3.1); model assessment of
regional drought and fire impaets on the northem earbon
eyele (section 3.2); discussion of model and observation
uneertainties (section 4); and the implications of the study
results for informing development of similar earbon model
simulations for the SMAP mission (section 5).
2. Methods
[7] This study extends a previous TCP model development
effort that used satellite [Advanced Microwave Scanning
Radiometer-EOS (AMSR-E)] microwave soil moisture and
temperature retrievals, and MODIS-derived GPP inputs within
a three-pool soil decomposition model to estimate NEE over a
regional network of northem temperate grassland, boreal
forest, and tundra tower sites [Kimball et a l, 2009]. A light
use elfleieney (LUE) algorithm [Running et al, 2004] and
the soil decomposition model were combined in this study to
estimate surface SOC stocks and earbon fluxes under dynamic
steady state conditions during the study period. A synthetic
approach integrating information from fire ehronosequenee
tower earbon flux observations, satellite Vis (NDVI, EVI),
and bumed area products was also used to account for
nonsteady state post-flre reeovery effeets on the TCP calcula
tions. The following sections describe the model equations
(section 2.1), primary data sets used for model inputs and
validation (section 2.2), the model parameterization scheme
(section 2.3), and a Monte Carlo based model uneertainty
assessment (section 2.4).
2.1. Model Description
2.1.1. TCP Equations
[s] NEE (g C m “ ^ d~^) is computed on a daily basis as the
residual differenee between GPP and Reco defined as the sum
of autofrophie (R^) and heterofrophie (R*) components:
N EE ^ {Ra + R h) - GVV
(1)
where positive (-t) and negative (—) NEE fluxes denote the
terrestrial loss or uptake of CO2, respectively. A LUE
YI ET AL.: DROUGHT AND FIRE IMPACT ON HIGH LATITUDE
approach similar to the MODIS (MOD 17) productivity
algorithm [Running et al., 2004] was used to estimate GPP:
G P P = £ X F P A R X P A R
(2)
where s (g C M J“ ^) is the LUE coefficient converting
absorbed photosynthetieally active solar radiation (APAR,
MJm~^d~^) to vegetation biomass and FPAR defines the
fraction of incident PAR (MJm~^d~^) absorbed by the
vegetation eanopy. PAR is estimated as a constant proportion
(0.45) of incident shortwave solar radiation at the surface,
which is derived from reanalysis data. The LUE coefficient £
is derived from a maximum LUE coefficient (e„„, g C MJ“ ^)
prescribed for each land eover class and reduced for sub-
optimal environmental conditions:
£ ^ m x ^ L n « _ s c a l a r ^ ^ P D s c a l a r ^ F T g c a l a r (3)
Ra = { l - C U E ) X G P P
(4)
R h — A ,netU m e t T (1 R s tr} ^ A g trC str T .^ r ec G r e
(5)
where Cmet, Qtr, and C^ec represent metabolic, straetnral,
and recalcitrant SOC pools (g C m “ ^), respectively, and
ifmet, ^str, and ifj-ec are the corresponding decomposition rate
parameters (day~^). The metabolic and straetnral SOC pools
represent plant litter with relatively short (e.g., <5 years)
turnover periods, while the recalcitrant or slow pool repre
sents more physically and chemically protected SOC with
a longer turnover time. In
Kimball et al. [2009], a fixed
proportion of earbon from the straetnral SOC pool was
transferred to the recalcitrant pool, which was found to be
effective at annual time scales. In this study, we assume that
a fixed proportion (iXstr) of the respiration flux from the strae
tnral pool is transferred to the slow pool on a daily basis:
d C m et/d t — Cfi-actNPP .^metCmet
d C str /d t = (1 - Cfracf)N PP - ZatrCstr
dC^ec/dt A^ecCrec
(6)
(V)
(8)
where scalar and VPDscaiar are dimensionless rate scalars
for snb-opfimal temperature and moisture conditions repre
sented by respective daily minimum air temperature and
atmospheric vapor pressure deficit inputs. These rate scalars
are defined as simple linear ramp functions [Heinsch et al.,
2006] and vary according to prescribed minimum and
maximum environmental constraints defined for different
global biome types (T„„ mm and max, VPDmin and
VPDmax, Table 1). An additional FTgcaiax term defines the
frozen temperature constraints to landscape water mobility
and GPP as determined from regional comparisons between
tower observation based GPP and daily FT retrievals from
satellite microwave remote sensing [Kimball et al., 2004;
Kim et al., 2012]. In this study, FTgcaiar is set to 0 (fully
eonstrained) if the FT retrievals indicate frozen landscape
conditions, 1 (no constraint) under nonfrozen conditions, and
0.5 (partially constrained) for transitional FT days defined by
midday thawing and nighttime freezing.
[9] The Ra term (g C m“^ d“ ^) in equation (1) represents the
sum of vegetation growth and maintenance respiration and is
estimated as a fixed proportion of GPP for each biome type
based on an assumption of conservatism in vegetation earbon
use efficiency (CUE) within similar plant functional types
[Waring et al., 1998; Gifford, 2003] as follows:
where CUE is the dimensionless ratio of vegetation net primary
production (NPP) to GPP, and NPP (gC m “^d“ ^) represents
the differenee between GPP and
R„. While the assumption of
CUE conservatism provides a key simplification for a remote
sensing based algorithm, the proportion of plant photosynthesis
devoted to biophysical growth and maintenance may vary
under changing environmental conditions and over the course
of vegetation development [DeLucia et al., 2007].
[10] The Rh term (g C m “ ^d~^) in equation (1) is com
puted as the sum of variable soil decomposition and respira
tion rates from three distinct earbon pools as follows:
where the Cg^act term defines the rate in which litterfall from
NPP is allocated to Cmet and varies for different biome types
and assoeiated litterfall chemistry [Ise and Moorcroft, 2006].
Estimated annual litterfall is evenly distributed on a daily
basis over each annual eyele for all biome types. This finer
daily temporal allocation of litterfall allows for better repre
sentation of the seasonal dynamics of the two fast SOC
pools, which generally have faster (< 1 year) turnover rates,
espeeially under warmer conditions. This approach ignores
the characteristic variable litterfall seasonality in deciduous
ecosystems but facilitates regional application by avoiding
more complex eanopy phenology representation, model
parameterization and input requirements, and assoeiated
uncertainty [White et al., 2000].
[11] The soil decomposition rate is derived as the product
of a theoretical maximum rate constant 0.0301 day~^)
[Ise and Moorcroft, 2006] and dimensionless multipliers for
soil temperature (7)nuit) and moisture (ILmuit) constraints to
decomposition under prevailing elimate conditions:
Amet ~ K,,ix X TLult ^ frmult (9)
where Tmuit and ILmuit vary between 0 (frilly constrained)
and 1 (no constraint). The decomposition rate parameters
for Cstr and Qec, i-£-, ^str and ifj-ec, are estimated as 40%
and 1% of ifmet ttnd Moorcroft, 2006]. The soil decom
position rate response to temperature follows an Arrhenius
type function [Lloyd and Taylor, 1994]:
Tmuif = exp{308.56[l/(r„pt + 46.02) - l / ( r , + 46.02)] } (10)
where Topt and are the respective reference and
input surface soil temperatures (°C) for Ts<To^x. Above the
reference temperature (Topt), soil decomposition is no longer
limited by temperature; under these conditions, soil moisture
is expected to deeline with warmer soil temperatures and
becomes the primary constraint to decomposition.
[12] The soil decomposition and Rh response to soil
moisture varies according to multiple factors including soil
texture, elimate, and vegetation type but generally has
optimum rates at intermediate soil moisture levels and is
increasingly inhibited at lower or higher soil water contents
according to site observations and laboratory incubation
studies [Davidson et al., 2000]. For this investigation, the
soil moisture constraint on soil decomposition under unsatu
rated conditions (SM < SMopt) is defined as follows:
ILnuif = [1 + aexp(Z) X SMopt)]/[l + aexp(Z) x SM)| (11)
where a and b are empirical fitting parameters that define the
decomposition rate response to soil moisture variability and
![Figure 2. (a) Spatial pattems of estimated surfaee (~< 10 em depth) soil organie earbon (SOC, kg C m“ )̂ from TCF simulations and IGBP and NCSCD soil inventory reeords for the northem domain (>45°N); (b) regional mean and standard deviation (indieated by error bar) for tundra, boreal forest, and other regions (mainly grasslands and eroplands) of surfaee SOC for North Ameriea (NA) and Eurasia portions of the domain. There were no NCSCD SOC reeords available for the European portion of the domain at the time of this investigation. A eonstant (0.2) eonversion eoeffieient was used to estimate surfaee SOC from the original 1 m depth IGBP and NCSCD SOC data [Jobbdgy and Jackson, 2000].](/figures/figure-2-a-spatial-pattems-of-estimated-surfaee-10-em-depth-2try2579.webp)