New Methods for Measurements of Photosynthesis from Space
New
Methods
for
Measurements
of
Photosynthesis
from
Space
Study
start
date:
August
26,
2012
Study
end
date:
August
31,
2012
Final
Report
submission
date:
April
1,
2013
Team
Leads
Joseph
A.
Berry
Department
of
Global
Ecology
Carnegie
Institution
of
Washington
jberry@dge.stanford.edu
Christian
Frankenberg
Jet
Propulsion
Laboratory
California
Institute
of
Technology
Christian.Frankenberg@jpl.nasa.gov
Paul
Wennberg
California
Institute
of
Technology
wennberg@caltech.edu
New
Methods
for
Measurements
of
Photosynthesis
from
Space
ii
Acknowledgments
The research described
in this paper was spo
nsored by the Keck
Institute for Space Studies
(KISS) and was carried out in pa
rt at the Jet Propulsion Labora
tory, California Institute of
Technology, under a contract with
the National Aeronautics and
Space Administration.
This report is a joint effort of
all the people and organizatio
ns who participated in the KISS‐
sponsored study, and we greatly
acknowledge ever
yone’s involvem
ent at the workshop and
the report. This workshop would have been—speaking euphemistica
lly—very different
without the support of the KISS
team, most impor
tantly Michele
Judd and Tom Prince.
Owing to Michele’s organizational and herding skills, the team
leads could focus on the
workshop itself and not worry about logistics.
Further, we thank Pat Rawlings f
or creating the professional co
ver art and Susan Foster for
her tremendous help (and patience) in editing and getting the f
inal report together.
© 2013. All rights reserved.
New
Methods
for
Measurements
of
Photosynthesis
from
Space
iii
Table
of
Contents
Acknowledgments ...............................................
.................................................................................................... ii
Acronyms and Abbreviations ...................................................................................................
......................... iv
1. Executive Summary .........................................................................................................
........................... 1‐1
2. Introduction ...............................................
...............................................................
..................................... 2‐1
2.1 Scientific motivation and opportunities ................................................................................ 2
‐1
2.2 Technical motivation and opportunities ............................................................................... 2‐2
2.3 Scope of study ............................................
...............................................................
......................... 2‐2
3. Components of the study:
schedule and organization .........
....................................................... 3‐1
4. Outcome of the study .......................................
...............................................................
........................... 4‐1
4.1 Basics of chlorophyll fluore
scence across spatial scales (m
olecular, leaf level,
canopy, mixed vegetation) ....................................................................................................
....... 4‐1
4.1.1 Introduction ...........................................................................................................
............ 4‐1
4.1.2 Modeling fluorescence and photosynthesis from the bottom
up ............... 4‐6
4.1.3 Equations for interpreting SIF ..........................
....................................................... 4‐12
4.2 Global carbon cycle modeling (GPP estimates, source/sink in
versions) .............. 4‐20
4.2.1 Application to the global carbon cycle ..................
............................................... 4‐20
4.2.2 Case studies supporting the added value of fluorescence .
.......................... 4‐25
4.2.3 How one imagines implementing SIF in global carbon cycle
analysis ... 4‐27
4.3 Retrieval of chlorophyll flu
orescence from ground and space
.................................. 4‐29
4.3.1 Introduction ...........................................................................................................
.......... 4‐29
4.3.2 Retrieval concept .......................................
...............................................................
..... 4‐30
4.3.3 Relation to other reflectance‐based remote sensing ......
............................... 4‐30
4.3.4 Current suite of satellites
capable of retrieving fluorescence .................... 4‐32
4.3.5 Future suite of satellites
capable of retrieving fluoresc
ence ...................... 4‐32
4.3.6 The orbiting carbon observatory prospects for fluorescenc
e ................... 4‐32
4.3.7 Validation Strategies ..................................................................................................
.. 4‐33
5. Future plans and development ...............................
...............................................................
............... 5‐1
5.1 Roadmap for technical development ...................................................................................... 5
‐1
5.2 Recent and planned papers .................................
...............................................................
......... 5‐1
5.2.1 Published papers ........................................
...............................................................
...... 5‐1
5.2.2 Planned papers ..........................................
...............................................................
........ 5‐2
5.3 How team will continue to move work forward ...............
................................................. 5‐2
5.4 Lessons learned ...........................................
...............................................................
...................... 5‐2
6. Conclusions ...............................................................................................................
..................................... 6‐1
Appendix A: Workshop participants .............................
...............................................................
...............
A‐1
Appendix B: Workshop agendas .................................................................................................................... B‐1
Appendix C: References ........................................
...............................................................
.............................. C‐1
New
Methods
for
Measurements
of
Photosynthesis
from
Space
iv
Acronyms
and
Abbreviations
AGCM
atmospheric general circulation models
APAR
absorbed photosynthet
ically active radiation
ASD
ASD Inc.
Cab
chlorophyll content
CarbonSat
Carbon Monitoring Satellite
CASA
Carnegie, Stanford, Ames Approach
CEFLES
CarboEurope, FLEX, and Sentinel 2
CLN
cropland
CP
chlorophyll‐protein complex
CSM
climate systems model
DBFN
deciduous broadleaf forest in northern latitudes
DGVM
dynamic global vegetation models
DNFN
deciduous needleleaf forest in northern latitudes
EBFS
evergreen broadleaf for
est in southern latitudes
ECMWF
European Centre for Medium‐Range Weather Forecasts
ENVISAT
Environm
ental Satellite
ESA
European Space Agency
EVI
enhanced vegetation index
FAPAR
fraction of absorbed photosynthetically active radiation
FLEX
FLuorescence EXplorer
FLUXNET
a network of regional
flux tower site networks
FPAR
fractional absorptance of sunlight
fs
femptoseconds
FTS
Fourier Transform Spectrometer
FWHM
full width at half maximum
GCC
global carbon cycle
GLN
grasslands in northern latitudes
GOME
Global Ozone Monitoring Experiment
GOSAT
Greenhouse Gases Observing Satellite
GPP
gross primary production
HyPlant
Hyperspectral Plant Imaging Spectrometer
IGBP
International Geosphe
re‐Biosphere Programme
IPCC
Intergovernmental Pa
nel on Climate Change
JAXA
Japan Aerospace Exploration Agency
New
Methods
for
Measurements
of
Photosynthesis
from
Space
v
KISS
Keck Institute for Space Studies
KM
Kubelka‐Munk
LAI
leaf area index
LHCII
light‐harvesting complexes
LUE
light‐use efficiency
MERIS
MEdium Resolution Imaging Spectrometer
MODIS
Moderate Resolution Imaging Spectroradiometer
MOE
Ministry of the Environment (Japan)
MPI‐BGC
Max Planck Institute for Biogeochemistry
NCAR
National Center for Atmospheric Research
NDVI
normalized difference index
NDVI
normalized difference vegetation index
NEE
net ecosystem exchange
NEON
National Ecosystem Observatory Network
NIES
National Institute for Environmental Studies (Japan)
NPP
net primary production
NPQ
nonphotochemical quenching
OCO
Orbiting Carbon Observatory
PAM
Pulse Amplitude Modulated
Pg C
petagrams (one billion metric tonnes) of carbon
PROSPECT
A radiative transfer mo
del describing the optical prop
erties
of plant leaves from 400 nm to 2500 nm
PSI
photosystem 2
PSII
photosystem 1
RCI
reaction center 1
RCII
reaction center 2
Reco
ecosystem respiration
Rh
soil heterotropic respiration
RT
radiative transfer
SAIL
Scattering by Arbitrarily Inclined Leaves, a vegetation
canopy reflective model
SCIAMACHY
SCanning Imaging Absor
ption SpectroMeter for Atmosphe
ric
CHartographY
SCOPE
Soil‐Canopy Observation of Photosynthesis and Energy
SiB
Simple Biosphere Model
SIF
sun‐induced fluorescence
New
Methods
for
Measurements
of
Photosynthesis
from
Space
vi
SpecNet
Spectral Network
SVAT
soil‐vegetation‐atmosphere
SVN
Savannas
SZA
sun zenith angle
TCCON
Total Carbon Column Observing Network
TOA
top of atmosphere
TOC
top of canopy
VPD
vapor pressure deficit
ZEA
zeaxanthin
New
Methods
for
Measurements
of
Photosynthesis
from
Space
1
‐
1
1.
Executive
Summary
Our ability to close the Earth's carbon budget and predict feed
backs in a warming climate
depends critically on knowing wh
ere, when, and how carbon dioxi
de (CO
2
) is exchanged
between the land and atmosphere. In particular, determining the
rate of carbon fixation by
the Earth's biosphere
(commonly referred to as gross primary pr
oductivity, or GPP) and the
dependence of this productivity
on climate is a central goal. H
istorically, G
PP has been
inferred from spectral imagery
of the land and ocean. Assessmen
t of GPP from the color of
the land and ocean requires, how
ever, additional knowledge of t
he types of plants in the
scene, their regulatory mechanis
ms, and climate variables such
as soil moisture—just the
independent variables of interest!
Sunlight absorbed by chlorophyll in photosynthetic organisms is
mostly used to drive
photosynthesis, but some can als
o be dissipated as heat or re‐r
adiated at longer wavelengths
(660–800 nm). This near‐infrared l
ight re‐emitted from illumina
ted plants is termed solar‐
induced fluorescence (SIF), and
it has been found to strongly c
orrelate with GPP. To advance
our understanding of SIF and its relation to GPP and environmen
tal stress at the planetary
scale, the Keck Instit
ute for Space Studies
(KISS) convened a w
orkshop—held in Pasadena,
California, in August 2012—to focus on a newly developed capaci
ty to monitor chlorophyll
fluorescence from terrestrial veg
etation by satellite. This rev
olutionary approach for
retrieving global observations of
SIF promises to provide direc
t and spatially resolved
information on GPP, an ideal bott
om‐up complement to the atmosp
heric net CO
2
exchange
inversions.
Workshop participants leveraged
our efforts on previous studies
and workshops related to
the European Space Agency’s FLuor
escence EXplorer (FLEX) missio
n concept, which had
already targeted SIF for a possible satellite mission and had d
eveloped a vibrant research
community with many important pu
blications. Thes
e studies, most
ly focused on landscape,
canopy, and leaf‐level interpret
ation, provided the ground‐work
for the workshop, which
focused on the global carbon cyc
le and synergies with atmospher
ic net flux inversions.
Workshop participants included key members of several communiti
es: plant physiologists
with experience using active fluo
rescence methods to quantify p
hotosynthesis; ecologists
and radiative transfer experts w
ho are studying the challenge o
f scaling from the leaf to
regional scales; atmospheric sc
ientists with experience retriev
ing photometric information
from space‐borne spectrometers; and carbon cycle experts who ar
e integrating new
observations into models that de
scribe the exchange of carbon b
etween the atmosphere,
land and ocean. Together, the pa
rticipants examined the link be
tween “passive” fluorescence
observed from orbiting spacecraf
t and the underlying photochemi
stry, plant physiology and
biogeochemistry of the land and ocean.
This report details the opportun
ity for forging a deep connecti
on between scientists doing
basic research in photosynthetic
mechanisms and the more applie
d community doing
research on the Earth System. To
o often these connections have
gotten lost in empiricism
associated with the coarse scale
of global models. Chlorophyll
fluorescence has been a major
tool for basic research in photosynthesis for nearly a century.
SIF observations from space,
although sensing a large footprin
t, probe molecular events occu
rring in the leaves below.
New
Methods
for
Measurements
of
Photosynthesis
from
Space
1
‐
2
This offers an opportunity for di
rect mechanistic insight that
is unparalleled for studies of
biology in the Earth System.
A major focus of the workshop was to review the basic mechanism
s that underlie this
phenomenon, and to exp
lore modeling tools th
at have been develo
ped to link the biophysical
and biochemical knowledge of pho
tosynthesis with the observable
—in this case, the
radiance of SIF—seen by the sate
llite. Discussion
s led to the i
dentification of areas where
knowledge is still lacking. For example, the inability to do co
ntrolled illumination
observations from space limits t
he ability to fully constrain t
he variables that link
fluorescence and photosynthesis.
Another focus of the workshop explored a “top‐down” view of the
SIF signal from space.
Early studies clearly identified
a strong correlation between t
he strength of this signal and
our best estimate of the rate of photosynthesis (GPP) over the
globe. New stud
ies show that
this observation provides improv
ements over conventional reflec
tance‐based remote
sensing in detecting seasonal an
d environmental (particularly d
rought related) modulation
of photosynthesis. Apparently SI
F responds much more quickly an
d with greater dynamic
range than typical greenness indi
ces when GPP is p
erturbed. How
ever, discussions at the
workshop also identified areas w
here top‐down an
alysis seemed t
o be “out in front” of
mechanistic studies. For example,
changes in SIF based on chang
es in canopy light
interception and the light use e
fficiency of the canopy, both o
f which occur in
response to
drought, are assumed equivalent in the top‐down analysis, but t
he mechanistic justification
for this is still lacking from the bottom‐up side.
Workshop participants considered
implications of these mechanis
tic and empirical insights
for large‐scale models of the carbon cycle and biogeochemistry,
and also made progress
toward incorporating SIF as a si
mulated output in land surface
models used in global and
regional‐scale analysis of the c
arbon cycle. Comparison of remo
tely sensed SIF with model‐
simulated SIF may open new possi
bilities for model evaluation a
nd data assimilation,
perhaps leading to better modeli
ng tools for analysis of the ot
her retrieval from GOSAT
satellite, atmospheric CO
2
concentration. Participants als
o identified another applicatio
n for
SIF: a linkage to the physical climate system arising from the
ability to better identify
regional development of plant wa
ter stress. Decreases in transp
iration over large areas of a
continent are implicat
ed in the development
and “locking‐in” of
drought conditions. These
discussions also identified areas
where current land surface mo
dels need to be improved in
order to enable this research. S
pecifically, the radiation tran
sport treatments need dramatic
overhauls to correct
ly simulate SIF.
Finally, workshop participants e
xplored approaches for retrieva
l of SIF from satellite and
ground‐based sensors.
The difficulty of res
olving SIF from the
overwhelming flux of reflected
sunlight in the spectr
al region where fluore
scence occurs was o
nce a major impediment to
making this measurement. Placement of very high spectral resolu
tion spectrometers on
GOSAT (and other greenhouse gas–
sensing satellites) has enabled
retrievals based on in‐
filling of solar Fraunhofer lines
, enabling accurate fluorescen
ce measurements even in the
presence of moderately
thick clouds. Perhaps the most interesti
ng challenge here is that
there is no readily portable gro
und‐based instrumentation that
even approaches the
capability of GOSAT and other pl
anned greenhouse gas satellites
. This strongly limits
New
Methods
for
Measurements
of
Photosynthesis
from
Space
1
‐
3
scientists’ ability to conduct gr
ound‐based studies to characte
rize the footprint of the GOSAT
measurement and to conduct studi
es of radiation transport neede
d to interpret SIF
measurement.
The workshop results represent a snapshot of the state of knowl
edge in this area. New
research activities have sprung f
rom the deliberations during t
he workshop, with
publications to follow. The intro
duction of this new measuremen
t technology to a wide slice
of the community of Earth System
Scientists will
help them unde
rstand how this new
technology could help solve prob
lems in their research, address
concerns about the
interpretation, identify future research needs, and elicit supp
ort of the wider community for
research needed to suppo
rt this observation.
Somewhat analogous to the origin
al discovery that vegetation in
dices could be derived from
satellite measurements originally intended to detect clouds, th
e GOSAT observations are a
rare case in which a (fortuitous
) global satellite dataset beco
mes available before the
research community had a consoli
dated understanding on how (bey
ond an empirical
correlation) it could be applied t
o understanding the underlyin
g processes. Vegetation
indices have since changed the w
ay we see the gl
obal biosphere,
and the workshop
participants envision that fluor
escence can perform the next in
dispensable step by
complementing these me
asurements with independent estimates tha
t are more indicative of
actual (as opposed to potential)
photosynthesis. Apart from the
potential FLEX mission, no
dedicated satellite missions are
currently planned. OCO‐2 and ‐
3 will provide much more
data than GOSAT, but will still
not allow for regional studies
due to the lack of mapping
capabilities. Geostationary obse
rvations may even prove most us
eful, as they could track
fluorescence over the course of
the day and clearly identify st
ress‐related down‐regulation of
photosynthesis. Retrieval of fluo
rescence on the global scale s
hould be recognized as a
valuable tool; it can bring the
same quantum leap in our unders
tanding of the global carbon
cycle as vegetation
indices once did.
New
Methods
for
Measurements
of
Photosynthesis
from
Space
2
‐
1
2.
Introduction
The invention of oxygenic photos
ynthesis by cyan
obacteria more
than 2 billion years ago
remade the Earth. In changing th
e redox chemistry of the ocean,
atmosphere, and land,
oxygenic photosynthesis complete
ly altered Earth’s geological a
nd biological evolution. The
quantity of photosynthesis takin
g place on planet Earth places
an ultimate limit on the
quantity and activity of organis
ms that can be supported by Ear
th's biosphere.
The growth of the human populati
on has been made possible by ap
propriating an ever‐
increasing fraction of Earth’s p
roductivity for human use (Vito
usek et al., 1986). Over the
short term it can be argued that
some things man has done—such
as improved agricultural
practices and increased nitrogen
fertilization—have increased p
roductivity, while other
things—such as de‐forestation, t
op soil loss, and climate chang
e—may lead to decreasing
global photosynthesis. As we contemplate the transition to a su
stainable population size and
economic model, it is of importa
nce to have an answer to key qu
estions: What is the
photosynthetic productivity of Ea
rth? Is it changing? However,
current technology limits our
ability to answer thes
e questions directly.
The
New
Methods
for
Measurements
of
Photosynthesis
from
Space
workshop centered on a
new technique to quantify photos
ynthesis; namely, using solar‐i
nduced chlorophyll
fluorescence as a direct probe i
nto the photosynthetic process
itself. The study leveraged
from decades of fluorescence res
earch in the laboratory and on
the leaf‐level as well as
preparatory studies for the Europ
ean FLuorescence EXplorer (FLE
X) satellite mission
proposal. In addition, chlorophyll fluorescence data on a globa
l scale became recently
available from the Japanese Green
house Gases Observing Satellit
e (GOSAT) usin
g a technique
that was previously not thought
possible (using solar absorptio
n lines to derive fluorescence
estimates). The workshop focused
on how chloroph
yll fluorescenc
e can inform research on
the global carbon cycl
e, especially on how it may benefit us by
providing estimates of actual
photosynthetic rates, as opposed
to estimates of potential phot
osynthesis that can be derived
using classical remote sensing t
echniques. As with all new tech
niques, initial skepticism in
the larger community n
eeds to be overcome and a critical mass o
f researchers reached to
support this new concept. In this workshop, a diverse group of
researchers—ranging from
laboratory‐scale plant physiologists to remote sensing expert t
o global carbon cycle
modelers—discussed the potential
and shortcomings of the curren
t data. Photosynthesis is
pivotal for Earth’s budgets of c
arbon, energy, and water. Fluor
escence now provides a highly
credible opportunity to exploit
a by‐product of photosynthesis
for global studies. We must
not miss this opportunity.
2.1
Scientific
motivation
and
opportunities
The primary scientific stimulus f
or this study was the sudden a
vailability of chlorophyll
fluorescence observations from the GOSAT satellite as well as t
he potential of the upcoming
second Orbiting Carbon Observatory (OCO‐2) mission to do the sa
me but with much higher
spatial resolution and a 100‐fold
increase in available data. H
istorically, the GOSAT and
OCO‐2 satellite community is str
ongly linked to atmospheric sci
entists who use atmospheric
greenhouse gas abundances to inve
rt spatially resolved net flux
es of CO
2
between the Earth’s
surface and atmosphere. If the s
erendipitous fluorescence retri
eval can be used as a spatially
and temporally expli
cit constraint on the gross uptake of CO
2
by terrestrial vegetation, then if
New
Methods
for
Measurements
of
Photosynthesis
from
Space
2
‐
2
may be possible for net flux inv
ersions to disentangle uptake f
rom respiration—a
prerequisite for a process‐based
understanding of the global ca
rbon cycle and its feedback to
global warming.
2.2
Technical
motivation
and
opportunities
Similar to the scientific motiva
tion, the technical motivation
also had its origin in the GOSAT
fluorescence retrievals: GOSAT d
emonstrated that high spectral
resolution enables
chlorophyll fluorescen
ce retrievals that are not affected by at
mospheric interferences
because retrievals are based on i
n‐filling of solar absorption
features (Fraunhofer lines). This
is a paradigm shift from the traditional application of oxygen
lines, which work very well if
the distance between the sensor
and the fluorescence emitter is
small. Atmospheric
scattering, however, can be detr
imental to this technique, espe
cially from a satellite.
The new method, however, also ha
s drawbacks, mainly related to
the required spectral
resolution and high single‐measu
rement noise. However, neither
GOSAT nor OCO‐2 were
optimized for fluorescence retr
ievals, and there are opportunit
ies to substantially improve
on these sensors. The lack of current ground‐based instrumentat
ion with the high spectral
resolution of GOSAT and OCO‐2 al
so warrants further investigati
on in optimized detector
design for long‐term monitoring of fluorescence at fixed locati
ons such as flux‐tower sites.
2.3
Scope
of
study
The scope of the study centered
on principal themes with respec
tive organizing questions;
namely
1.
Basics
of
chlorophyll
fluorescence
across
spatial
scales
(molecular,
leaf
level,
canopy,
mixed
vegetation)
What
are
the
biophysical
mechanisms
of
fluorescence
and
its
relation
to
gross
primary
production
(GPP)?
Is
there
adequate
knowledge
of
fluorescence
principles
to
relate
emission
to
GPP?
Where
are
the
main
uncertainties
(canopy
radiative
transfer
or
relation
of
fluorescence
yield
to
photosynthesis
yield)?
How
can
the
scale
gap
from
leaf
‐
scale
measurements
to
the
satellite
footprint
be
bridged?
2.
Global
carbon
cycle
modeling
(GPP
estimates,
source/sink
inversions)
Why
is
there
such
a
large
spread
in
current
GPP
model
estimates?
How
can
fluorescence
be
implemented
in
terrestrial
vegetation
models
(
towards
carbon
cycle
data
assimilation)
Can
fluorescence
and
CO
2
net
flux
inversions
derived
from
atmospheric
CO
2
data
be
used
synergistically?
3.
Retrieval
of
chlorophyll
fluorescence
from
ground
and
space
What
are
the
advantages
and
disadvantages
of
the
new
technique?
What
would
an
optimal
fluorescence
sensor
look
like
based
on
the
new
knowledge?
New
Methods
for
Measurements
of
Photosynthesis
from
Space
3
‐
1
3.
Components
of
the
study:
schedule
and
organization
The workshop was a bas
ic, intensive 1‐week K
eck Institute for S
pace Studies (KISS)
workshop designed to bring toget
her a diverse community of scie
ntists and engineers to
work on a new interdisciplinary
research field. On the Sunday p
rior to the core workshop, we
held an open short‐course with t
he aim of developing a common l
anguage and introducing
the main ideas to all workshop participants as well as to inter
ested researchers from the
California Institute of Technolo
gy (Caltech), Jet Propulsion La
boratory (JPL), and other
universities. The short‐cour
se lectures were as follow:
1.
The global carbon cycle, an over
view (Ian Baker, Colorado State
University)
2.
A Primer into Photosynthesis and
Chlorophyll Fluorescence (Jose
ph Berry,
Carnegie Institution for Science, Stanford)
3.
Retrieval of Chlorophyll Fluores
cence from Space (Christian Fra
nkenberg, JPL)
Videos of the short‐course are available at
http://www.kiss.caltech.edu/wo
rkshops/photosynthesis2012/schedu
le.html
.
The workshop itself was based on
introductory talks for each ma
jor topic with adequate time
allotted for discussions. The ove
rall schedule was kept flexibl
e in case some t
opics needed
additional attention. Towards the end of the workshop, even mor
e free discussion time was
available in order to stimulate
open discussions and “digestion
” of the ideas that came up in
the beginning. Generous lunch and
coffee breaks as well as dinn
ers and group events were
essential for both community bui
lding and scientific discussion
s in subgroups.
New
Methods
for
Measurements
of
Photosynthesis
from
Space
4
‐
1
4.
Outcome
of
the
study
4.1
Basics
of
chlorophyll
fluorescence
across
spatial
scales
(molecular,
leaf
level,
canopy,
mixed
vegetation)
4.1.1
Introduction
The possibility that climate chan
ge will affect crop production
while population continues to
increase has prompted several institutions such as the World Ba
nk to warn of an impending
food crisis. While we have excellent infrastructure for reporti
ng crop yield in
many countries,
a reliable method to assess crop
health across broad agricultur
al areas to anticipate or
diagnose problems with crop production is needed (Lobell and Fi
eld, 2007). Similarly, our
ability to close the Earth's carb
on budget and predict feedback
s in a warming climate
depends critically on knowing wh
ere, when, and how carbon dioxi
de (CO
2
) is exchanged
between the land and atmosphere (
Le Quéré et al., 2009).A metho
d is needed to study how
climate‐driven variability in bi
ological processes control this
net flux (Friedlingstein et al.,
2006) as the future trajectory of atmospheric CO
2
depends on the response of plants to
climate change. Even our ability
to understand our weather and
climate system is tied up
with the productive activities o
f plants. For example, transpir
ation of water vapor from
plants is directly linked to pho
tosynthesis and moistens the at
mosphere over the continents,
thereby moderating the cli
mate (Lee et al., 2005).
All of the above and more are li
nked to the photosynthetic acti
vities of the plants that cover
the continents, and it follows that we need to have the best to
ols possible to measure and
monitor this key process of the b
iosphere. The recent demonstra
tion that chlorophyll
fluorescence can be monitored fr
om a satellite platform provide
s a novel and possibly
breakthrough tool for studies of photosynthesis. Up to this poi
nt our knowledge of vegetation
dynamics has been obtained from a
nalysis of sunlight reflected
by leaves and other objects at
the ground surface. The high spe
ctral resolution of the GOSAT s
ensor has enabled us to see
the light that plant leaves emit
as chlorophyll fluorescence. T
his is an entirely new light
arising from within the photosyn
thetic machinery
of plants. Onl
y plants conducting
photosynthesis emit this light.
In this section we examine how
this light is linked to the
photosynthetic process.
4.1.1.1
Terrestrial
Photosynthesis
While photosynthesis is normally
defined as the use of energy f
rom absorbed light to
accomplish the uphill synthesis of sugars from CO
2
, it is useful here to
think of this in the
context of plant growth in a te
rrestrial environment. Photosynt
hesis on land requires that
the plant replace the water that
inevitably escapes from its le
aves when CO
2
is taken up from
the dry atmosphere. Plants also
require a supply of nutrients i
n addition to the water
exchanged for carbon, since the
ultimate product of photosynthe
sis is not only carbohydrates
but also new plant tissue. There
fore, we will refer to photosyn
thesis synonymously with
gross primary production (GPP). Physiological and developmental
mechanisms operate to
adjust the rate of GPP to the availability of resources. For ex
ample, a recent an
alysis (Beer et
al., 2009) of CO
2
and water vapor exchange measured by eddy covariance from a wi
de range
of ecosystems indicates a loss of about 200±63 mol of water for
each mol of C taken up as
GPP by the canopy at a sampling
of flux sites. T
his close meter
ing of water use is due in part
New
Methods
for
Measurements
of
Photosynthesis
from
Space
4
‐
2
to physiological regulation of t
he stomata, which are valves on
the leaf surface that permit
the exchange of gases between the
leaf interior and the atmosph
ere. When thes
e close both
water loss and GPP are restricted
In addition, there are import
ant indirect effects of water
supply on the rate of leaf area
expansion and th
e allocation of
new growth to roots vs. shoots
that tends to balance the plant’
s demand for water with its abi
lity to obtain water over its
growing season. Similarly, def
iciency of an essential nutrient
tends to cause plants to
suppress expansion of leaf area in favor of forming more roots.
Field et al. (1995) point out
that evolutionary processes have
tuned the physiology and devel
opmental programs of
plants to reduce the impact of s
ingle limiting factors such tha
t growth can be
co‐limited by
several factors. This is the strategy that would make most effi
cient use of the specific suites of
resources available in different
locations, and it has importan
t implications for the way that
we view productivity of terrestrial plants.
John Monteith (1972) proposed an
equation that has become the p
aradigm for
understanding GPP:
GPP = PAR ⋅FPAR ⋅
ε
p
(4‐1)
It is given as proportional to the incident short wave radiatio
n, the fractional absorption of
that flux (FPAR) and the efficien
cy with which th
e absorbed rad
iation is converted to fixed
carbon,
ε
p .
There has been a tendency to emp
hasize one term or the other of
this equation. In
England, crop physiologists focu
sed on the PAR term that explai
ns the seasonal growth of
crops and year‐to‐year variation in yield. Many in the remote s
ensing community have
focused on the FPAR term (Sellers
et al., 1985), and more recen
tly on the light
‐use‐efficiency
term as the arbiter of productiv
ity—particularly in strongly se
asonal and nutrient‐limited
forests (Coops et al 2010). From t
he ecological perspective abo
ve, we could argue that much
of the long‐term spatial variati
on in productivity is likely to
reside in the FPAR term, as it
reflects differences between sit
es in the average availability
of resources for plant growth.
Variation in
ε
p
; however, is likely to be signi
ficant over shorter time frames
when water or
temperature stress develops. The
take home message is that this
simple equation sits atop a
great deal of biological and bio
physical complexity. Researcher
s have develop
ed models of
GPP that deal with this comp
lexity in different ways.
4.1.1.2
Estimating
photosynthesis
at
global
scales
With the development of a global
infrastructure for weather for
ecasting and satellite remote
sensing of surface reflectance,
it has become po
ssible to speci
fy the environmental conditions
that plants experience and the d
ensity of plant cover over the
continents, and to use these
data to drive models of the land
surface. Checking these models
has always been
problematic.
Traditionally, GPP has been estim
ated by measurement of net pri
mary production at a plot
scale and correcting for respirat
ory losses (about half) in con
verting sugars to new plant
material (Field et al.
, 1995). The developm
ent of eddy correlat
ion as a method for quantifying
the carbon, water, and energy ba
lance over so‐called “flux site
s” has given us a wealth of
observational data to test and t
une models; but these measure n
et CO
2
exchange, the sum of
ecosystem respiration, and GPP. S
everal approaches are used to
estimate GPP (Desai et al.,
2008), but these are difficult to verify. In neither case is th
ere sufficient density of sampling
to get regional or continental scale GPP. This is the domain of
models. Three general types of
models are in use:
New
Methods
for
Measurements
of
Photosynthesis
from
Space
4
‐
3
1
Light
‐
use
‐
efficiency
models
, exemplified by the MODIS (
Moderate Resolution Imaging
Spectroradiometer)‐GPP product (Zhao et al., 2005), make direct
use of remote sensing
to estimate the flux of absorbed
light on a grid over the land
surface. These models
apply Eqn. 4‐1 to estimate GPP.
This approach uses meteorologic
al data from reanalysis
products to modulate
ε
P
but does not explicitly model t
he biophysical environment. The
CASA (Carnegie, Stanford, Ames Approach) model (Field et al., 1
995) has a similar
productivity model that is coupl
ed to a multi‐pool carbon cycle
module to provide a
gridded net CO
2
flux as a product.
2
Process
models
of the Earth system include a m
uch more complete representatio
n of the
biophysical environment that the
plant experiences and simulate
the plant’s
physiological responses to the e
nvironment. In particular, thes
e models integrate a
number of state variables, such
as the quantity of water stored
in the soil, the soil
temperature, and the leaf area i
ndex from time step to time ste
p. These models have
been developed from climate models, in which the focus is on th
e exchange of energy
between the land surface and the atmosphere. Several models of
this type—for
example, the National Center for
Atmospheric Research’s Climate
Systems Model (NCAR
CSM)—are now being used in Interg
overnmental Panel on Climate C
hange (IPCC)
studies to predict the course a
nd impacts of climate change.
3
Diagnostic
models
do not attempt to represent the
mechanisms occurring on the la
nd
surface, but instead use empiric
al studies to calibrate GPP to
potentially limiting
resources; for example, can GPP
be statistically related to the
quantity of precipitation,
temperature, leaf‐area index (LAI) and other observations? The
classic Miami model,
which was the first to estimate
global productivity, is such a
model: Beer et al. (2009)
have made use of a machine learn
ing approach to analyze over 10
00 site years of flux
observations to calibrate a glob
al model of GPP, and they provi
de a global GPP estimate
that they claim is accurate to ±5%.
Each of these model types has a
niche where it e
xcels. For exam
ple, the MODIS‐GPP model is
directly linked to remote sensing and weather forecast products
and can provide near–real
time information on pr
oductivity and the influence of anomalies
such as droughts. However,
it can only be run retrospectively; hence, it is not appropriat
e for “what if” questions.. The
process models are designed to b
e predictive and are widely use
d for studies of the carbon
cycle and how it will respond to
climate change
and for studies
of carbon cycle climate
feedbacks. The diagnos
tic models are able to take advantage of
the large pool of
measurement data, so t
hey have a much stronger statistical basi
s for estimating GPP.
However, since the model does no
t represent mechanisms, the res
ults should be viewed
more as a “climatology” of GPP th
an as a predictor of the respo
nse of GPP to future climate.
The modeling approache
s also have sp
ecific limitations. Studies
with the MODI
S‐GPP product
(Heinsch et al., 2006) highlight
its ability to correctly predi
ct observed fluxes at tower sites,
but also draw attention to the u
ncertainty in the MODIS vegetat
ion indices due to cloud and
aerosol contamination problems,
errors in the re‐analysis meteo
rology, and difficulty
constraining the light‐use‐effic
iency term. The process models,
while faithful to the
mechanisms, have great difficulty with calibration and accumula
tion of errors in the
simulated state variables. For e
xample, an error in calibration
of the effective storage
capacity for water in the soil can lead to errors in site hydro
logy impacting the water
New
Methods
for
Measurements
of
Photosynthesis
from
Space
4
‐
4
available for photosynthesis—eve
n if it was perfect in other re
spects. Currently process
models can only be calibrated at flux sites, and there is very
limited capability to assess or
correct for errors in state variab
le integration. The diagnosti
c models do very well at the flux
sites, but there is a lot of space between these that is poorly
constrained. Remote sensing
approaches that could help fill
these gaps would improve the ac
curacy of both the process
models and the diagnostic models
. Models inter‐comparison studi
es (Huntzinger et al., 2012)
show differences by a factor of nearly 2 in simulated GPP of No
rth America using the same
input data. Clearly, there is room for improvement.
One of the primary causes for this large uncertainty in GPP is
the fact that it cannot be
measured directly on a
geographically relevant scale. At the le
af scale, GPP can be
determined as the gross CO
2
uptake (i.e., the sum of net CO
2
uptake in the light and
respiratory CO
2
release in the dark). As the sc
ale of the measurement goes up,
respiration
becomes an increasingly importan
t component of the carbon balan
ce, and it becomes more
difficult to measure GPP
independently of net CO
2
uptake. At the regional or global scale, the
two processes—GPP and ecosystem
respiration (Reco)—are nearly b
alanced, but neither is
strongly constrained by the observed net CO
2
flux. It is ironic that one of the most important
processes of the terrestrial b
iosphere is hidden from us.
4.1.1.3
Chlorophyll
fluorescence
With this background, we will no
w turn to fluorescence and what
it might do for us.
Chlorophyll fluorescence has been
used in laboratory‐scale stud
ies of photosynthesis for
several decades (Krause and W
eis, 1987) and has been used in st
udies of the effect of
nutrient stress on marine produc
tivity (Beherenfeld et al., 201
2). Technical difficulties
relating to the variable reflect
ance of terrestrial vegetation
in the band where chlorophyll
fluorescence resides h
as inhibited the use of this approach for
studies of photosynthesis on
the land. Recently, it was found
that chlorophyll fluorescence
can be retrieved from high‐
resolution spectra around 757 nm r
ecorded by GOSAT (Joiner et a
l., 2011; Frankenberg et al.,
2011a,b). The spectral
channel used for this retrieval was main
ly intended for correcting
scattering effects in atmospheric
greenhouse gas retrievals. Ty
pically about only 1% of the
absorbed photons are re‐emitted as fluorescence. This re‐emitte
d light mixes with sunlight
and is difficult to detect, but
the signal can be resolved usin
g high‐resolution spectrometer
instruments, by observing the in
‐filling of solar Fraunhofer li
nes (Joiner et al., 2011;
Frankenberg et al., 201
1). The GOSAT satellite has been making
measurement of sun‐induced
fluorescence (SIF) since 2009. Thi
s signal is a distinct "glow"
from plants at wavelengths
between 690 nm and about 800 nm that is quite specific for the
presence of green plants. It
reports on the flux density of p
hotons absorbed by chlorophyll
molecules and on the
processing of these photons by photosynthetic reaction centers
at the time of satellite
observation (approximately noon o
n clear days). The footprint o
f the observation is a circle
about 10 km in diameter. The ret
rieval is also very insensitive
to atmospheric scattering and
clouds (Frankenberg et al., 2012),
which is in contrast to conv
entional reflectance
spectroscopy. It is also not inf
luenced by the reflective prope
rties of soil or other materials
that may be present in the scene.
As a first approximation, the fl
ux of SIF detected by a radiome
ter looking down on the land
surface can be expressed by an eq
uation that is analogous to th
e expression for GPP,
New
Methods
for
Measurements
of
Photosynthesis
from
Space
4
‐
5
SIF = PAR ⋅ FPAR ⋅
ε
f
(4‐2)
where
ε
f
is the yield of fluorescence photons at the top of the canopy
and PAR ⋅ FPAR is the
flux of absorbed light. Figure 4‐
1 shows monthly mean of GOSAT‐
measured SIF aggregated
by biome regressed against FPAR (
from MODIS) times the cosine o
f the solar zenith
(cos(SZA)) which is essentially t
he flux of absorbed light at t
he time of GOSAT overpass.
While SIF correlates well with ex
isting absorbed PAR products,
it is important to recognize
that it is an independent measur
ement linked to a specific comp
onent of APAR: that
absorbed by chlorophyll. As such
it may provide insight into th
e way we use vegetation
indices. In addition, there is evidence (see section 4.2) that
SIF is more dynamic than
greenness, indicating that there
may be additional control by
ε
f.
It is interesting that this expression can be combined with Eqn
. 4‐1 and rearranged to
eliminate the parallel
dependence of both processes on APAR to
yield
GPP = SIF ⋅
ε
p
/
ε
f
(4‐3)
It is well established that
ε
p
varies with the level of physiol
ogical (water or low temperatur
e)
stress. This begs the question, what happens to
ε
f
under stress? If
ε
f
and
ε
p
respond in parallel
to stress, then both GPP and SI
F will decline under stress; and
measurements of SIF could
also provide a proxy for variation
that occurs when stress (in addition
to light harvesting) restricts
photosynthesis. Research inspired by
the European Space Agency’s
FLuorescence EXplorer (FLEX)
mission concept (Meroni et al., 2009,
Moya et al., 2006) provides clear
evidence for an effect of stress on
ε
f
.
For example, leaf‐sc
ale studies show
that physiological effects of drought
that lead to a decre
ase of light‐use
efficiency (LUE) for photosynthesis
(
ε
p
) are associated with decreases in
fluorescence yield (
ε
f
) (Flexas et al.,
2002). In a field experiment,
measurements of fluorescence
during an episode of drought
(Figure 4‐2 Daum
ard et al., 2010)
demonstrate that fluorescence
declines, wherea
s normalized
difference vegetation index (NDVI)
(and presumably light interception)
in that experiment remained
constant (Daumard et al., 2010).
Therefore, it seems reasonable to
expect that changes in SIF may
Figure
4
‐
1
.
Linear
regressions
of
GOSAT
SIF
vs.
SIF*cos(SZA)
where
SZA
is
the
solar
zenith
angle
at
the
time
of
overpass
(Guanter
et
al.,
2012).
Each
symbol
represents
one
month
(2009–2011).
Biomes
follow
the
International
Geosphere
‐
Biosphere
Programme
(IGBP)–based
land
cover
classes:
[DBF]=Deciduous
Broadleaf
Forest,
[EBLN(S)]=Evergreen
Broadleaf
Forest
in
the
northern
(southern)
hemisphere,
[NF]=Needleleaf
Forest,
[CLN(S)]=Cropland
in
the
northern
(southern)
hemisphere,
[GL]=Grasslands,
[SVN(S)]=Savannas
in
the
northern
(southern)
hemisphere
(from
Guanter
et
al.,
2012).
New
Methods
for
Measurements
of
Photosynthesis
from
Space
4
‐
6
indicate changes in GPP associated with
episodes of stress, and that SIF might
sense the development of stress before
significant changes in FPAR occur.
Satellites are ideally suited for change
detection, and deviations of SIF from the
expected correlation with absorbed PAR
may be a powerful indicator of
physiological stress.
There is strong empirical evidence to
support this assertion (see Section 4.2). In
the remainder of this section, we consider
the mechanistic basis for linking
photosynthetic rate with satellite
observations of SIF.
4.1.2
Modeling
fluorescence
and
photosynthesis
from
the
bottom
up
These topics have received abundant
attention from researchers interested in
basic photosynthetic mechanisms, but the
need to explain SIF presents a ne
w challenge that require chang
es in existing models. The
Soil‐Canopy Observation of Photosynthesis and Energy (SCOPE) ba
lance model (van der Tol
et al., 2009a,b) is presented her
e as an example of a model und
er development for predicting
the flux of SIF at the top of a p
lant canopy (what the satellit
e observes) from the mechanisms
that occur in the chloroplasts a
nd leaves of the canopy. This i
s by necessity a highly detailed
presentation, and many readers may wish to skip over it or refe
r back to it for information on
specific mechanisms.
4.1.2.1
Fluorescence
emission
at
the
molecular
scale
Chlorophyll molecules are very efficient in absorbing visible l
ight, especially in the blue and
red regions. This property is re
sponsible for the green color o
f chlorophyll and generally of
leaves. We begin by considering
the interactions of chlorophyll
dissolved in an organic
solvent. Chlorophyll‐a presents two absorption peaks around 430
nm (blue) and 662 nm
(red), whereas chlorophyll‐b absorption peaks are slightly gree
n‐shifted at around 453 and
642. Upon absorption, the energy
carried by a photon of blue li
ght is able to excite one of the
chlorophyll molecule electro
ns from the ground state (S
0
) to the second molecular orbital
(S
2
), whereas a photon of red light is able to bring the electron
to the first molecular orbital
(S
1
). From S
2
, the energy is rapidly lost as heat through internal conversio
n and radiationless
decay within a few picoseconds (10
−12
s) and the electron relaxes to the first molecular
orbital (S
1
) (Gobets and Grondelle, 2001; C
legg, 2004). It is from the fir
st orbital (S
1
) that the
excitation energy can take different pathways. Intrinsically, a
n isolated chlorophyll molecule
has three main de‐excitation pat
hways; namely, the electron can
relax to the ground state via
internal conversion, it can be r
e‐emitted as a photon of light
(fluorescence), or it can undergo
intersystem crossing and form a
chlorophyll triplet state. We c
an express the fluorescence
yield as:
Figure
4
‐
2
.
A
plot
of
SIF
as
a
function
of
PAR
(the
slope
is
the
apparent
ε
f
)
made
from
a
radiometer
on
a
tower
above
a
rain
fed
sorghum
crop
(adapted
from
Daumard
et
al.,
2010).
The
authors
note
that
there
was
no
change
in
LAI
or
NDVI,
indicating
that
absorption
of
light
was
constant
over
this
interval.
Thus,
ε
f
appears
to
decrease
with
water
stress.
Photosynthesis
was
not
measured
in
these
studies,
but
there
is
every
reason
to
expect
that
GPP
also
declined,
leading
us
to
expect
that
SIF
may
be
used
as
a
proxy
for
stress
effects
on
GPP.
New
Methods
for
Measurements
of
Photosynthesis
from
Space
4
‐
7
Φ
F
(chl) =
k
f
/(
k
f
+
k
D
+
k
isc
)
(4‐4)
where
k
f
is the first‐order rate constant of fluorescence,
k
D
the rate constant associated to the
constitutive thermal deactivatio
n (radiationless decay), and
k
isc
the rate constant
representing the process of inte
rsystem crossing to the triplet
state, which is orders of
magnitude smaller to that of
k
f
and
k
D
(Butler and Kitajima, 1975) for excitation in S
1
. Under
these conditions, when
chlorophyll is isolat
ed and diluted in a
solvent, the fluorescence yield
can be very high, e.g., 0.33 or 0
.16 for chlorophyll‐a and chlo
rophyll–b, respectively (Latimer
et al., 1956; Brody and Brody,
1962; Rabinovitch and Govindjee,
1965) Yet, the
chlorophyll‐a
fluorescence yield in vivo (in th
e leaf) is more than one order
of magnitude sm
aller because
much of the absorbed
energy is trapped to do useful work.
4.1.2.2
Fluorescence
emission
at
the
scale
of
chloroplast
membranes
During evolution, primordial bac
teria, algae—and, later on, hig
her plants—have evolved
structures to efficiently and co
st‐effectively capture light, a
nd to move its associated energy
(exciton) from chlorophyll to ch
lorophyll and ultimately to a r
eaction center where the
exciton is used to effect a photo
chemical reactio
n. Fluorescenc
e yield is strongly suppressed
and may be viewed a “leak” from the system rather than a major
pathway for de‐excitation.
These structures are what we now
call photosynthetic antennas,
a complex multiunit matrix
of proteins that bind pigments a
nd that are carefully arranged
and distributed to efficiently
supply excitation energy to trap
s, where the excitation energy
is converted into more stable
form of energy.
There are at least three types of
traps for excitons in chlorop
last membranes:
a.
Reaction center 2 (RCII) of phot
osystem 2 (PSII), where photoch
emical reaction (charge
separation) occurs and molecular
oxygen is formed from water. T
he electron liberated
in this reaction is passed to el
ectron carriers and photosystem
1 (PSI).
b.
Reaction center 1 (RCI) of PSI,
where photochemical reaction (c
harge separation) also
occurs and can be described as the second photo act that uses e
xciton energy to move
the electron produced by PSII to
a higher energy level (redox p
otential) needed to
reduce CO
2
.
c.
In addition to PSI and PSII, there
may be non‐photochemical que
nching (NPQ) traps
that dissipate excitons harmlessl
y to heat. The population of t
hese traps at PSII is
related to the concentration of
a specific carotenoid, zeaxanth
in, that is formed when
the supply of electrons from PSII
exceeds the capacity to use t
hem for reducing CO
2
.
NPQ traps are analogous to pressure relief valves on a steam bo
iler; they only open as
much as necessary, and they are
specifically associated with PS
II.
All of these traps use excitons from the S
1
excited state of chloroph
yll. Excitons formed from
absorption of blue light
are converted to the S
1
state by internal conversion. Thus, a photon
of blue and a photon of red light have the same inherent probab
ility of driving
photosynthesis; that is why the
term “quantum e
fficiency” is wi
dely used in photosynthesis.
PSII
.
The supramolecular structure of
PSII of higher plants is compo
sed of the following: (1) a
core that includes the
reaction center chlorophyll pair P680 an
d two protein complexes
(CP43 and CP47) that bind togethe
r 38 chlorophyll‐a molecules a
nd a number of
carotenoids; (2) a peripheral antenna that connects the core to
the outer antennas and that is
New
Methods
for
Measurements
of
Photosynthesis
from
Space
4
‐
8
composed of (a) three different
chlorophyll‐protein complexes (
CP26, CP24, and CP29), each
binding chlorophyll‐a, chlorophy
ll‐b, and carotenoids, and (b)
the main outer antenna that is
composed of large prot
ein complexes known as light‐harvesting c
omplexes (LHCII), where
the gross of the chlorophyll‐a an
d chlorophyll‐b is found (Vass
iliev and Bruce, 2008). A PSII
unit has 180 to 400 chlorophyll molecules, slightly more than f
ound in PSI, a
lthough with
PSI’s having less chlorophyll‐b.
This number is
approximate bec
ause the structure of the
photosystem and its different com
plexes has not yet been resolv
ed in vivo, where
photosystems with different antenna sizes coexist.
Binding of chlorophyll molecules
to a protein and subsequently
to a reaction center has a
number of important implications
. First, unlike in solution, bi
nding of a chlorophyll molecule
to a specific site in a protein w
ill affect its electronic prop
erties and thus its absorption and
emission (fluorescence) spectra.
As a result, the antenna is co
mposed of chlorophylls with
different spectral forms that tend to form an energetic gradien
t for transfer of excitons from
the peripheral antenna towards t
he reaction cent
er (Vassiliev a
nd Bruce, 2008;
Novoderezhkin and Grondelle, 2010). Second, the location and or
ientation of the pigment
binding sites in proteins has b
een highly optimized during evol
ution to promote fast and
efficient energy transfer between
neighboring chlorophylls as w
ell as between different
protein complexes by means of “li
nk” chlorophylls. Energy trans
fer in the antenna takes
place through different mechanism
s: during the first hundreds o
f femptoseconds (fs); after a
chlorophyll absorbs a
photon, the excitation energy appears to
be delocalized among
adjacent chlorophylls in what ha
s been defined as a quantum coh
erent state (Engel et al.,
2007; Ishizaki and Fleming, 2009).
This quantum coherence opera
tes at the level of antenna
subunit, where neighboring pigmen
ts are highly coupled, allowin
g the excitation to sample
the most optimal route for the e
xcitation to be passed downstre
am. Subsequently, energy
transfer takes place through Förster resonance energy transfer
(Förster, 1955;
Novoderezhkin and Grondelle, 2010); this type of energy transfe
r would be responsible for
the migration of excitation ener
gy from the antenna to the reac
tion center.
As a result, because chlorophyll
s are now connec
ted to a reacti
on center via the antenna,
energy transfer (leading to phot
ochemistry) will also compete w
ith fluorescence for de‐
excitation of energy at the leve
l of chlorophyll molecule. In t
urn, if we consider the energy
partitioning at the level of PSI
I antennas, and also take into
account the regulatory quenching
mechanisms by NPQ traps (see item c above) we can update Eqn. 4
‐4 and express the
fluorescence yield in
the photosystem as
Φ
F
(PSII) =
k
f
/ (
k
f
+
k
D
+
k
P
+
k
NPQ
)
(4‐5)
where
k
P
is the rate constant of the pho
tochemical process by which exc
itation energy is
bound chemically and
k
NPQ
is the rate constant associated to the regulated thermal
dissipation mechanisms (NPQ). Fig
ure 4‐3 is a schematic diagram
of photon processing in
PSII.
In vivo, most of the fluorescenc
e comes from PSII. By comparing
Eqn. 4‐4 and Eqn. 4‐5, it is
obvious that the yield of chloro
phyll‐a fluorescence in vivo wi
ll be lower than that of
chlorophyll in solutio
ns because we now have two more processes
competing for excitation
energy with fluorescence; namely,
k
P
and
k
NPQ
. Yet, if we estimate the
fluorescence yield upon
a saturating light pul
se that momentarily “blocks” electron tra
nsport and photochemistry
New
Methods
for
Measurements
of
Photosynthesis
from
Space
4
‐
9
(
k
P
=0) and in dark‐acclimated
samples that do not present any
NPQ (
k
NPQ
=0), the yield fluorescence
at this so called F
m
level is still
smaller than that obtained in
solution. One of the reasons for the
lower yield in vivo
is the
reabsorption of red chlorophyll‐a
fluorescence by
the high local
concentration of chlorophyll in the
membranes. The spectra of emitted
fluorescence extends from 650 nm
to well beyond 800 nm; thus,
fluorescence in the range 650 nm to
700 nm can be re‐absorbed by
chlorophylls associated to PSI and
PSII and used in photochemistry.
For this reason, in addition to the
factors expressed in Eqn. 4‐5, the
yield of fluorescence in the red
region (650 nm to 700 nm) will also
depend on the chlorophyll content
of the leaf. In fact, this phenomenon
has been exploited as a means to
estimate leaf chlorophyll conten
t and has been s
hown to correla
te very well with the
F(red)/F(far‐red) ratio (
Gitelson et al., 1999).
PSI
. Similarly, the emission of flu
orescence in the far‐red region
occurs from PSI. The
fluorescence yield in PSI is generally assumed to be much lower
than that of PSII, and it is
also assumed to remain constant
given that the reaction center
of PSI is an effective
excitation trap independently of its redox state. For this reas
on, fluorescence from PSI does
not exhibit the dynamics of PSII,
and it has typically been con
sidered as something to be
corrected from the otherwise “photosynthesis‐sensitive” signal
coming from PSII. The
fluorescence yield in PSI could be expressed as
Φ
F
(PSI) =
k
f
/(
k
f
+
k
D
+
k
P700
)
(4‐6)
where
k
P700
is the rate constant of energy trapping by the PSI reaction ce
nter P700 (named
P700 after the chlorophyll in its
reaction center that absorbs
light at 700 nm, in contrast to
PSII reaction center chlorophyll
P680, which absorbs at 680 nm)
. The contribution of PSI
fluorescence to total fluorescence has been estimated in differ
ent species and using different
methods, and it has been found t
o be as high as 30% to 50% in t
he far‐red regions (>700 nm)
(Genty et al., 1991; Pfündel et al
., 1998, and Franck et al., 2
002). The contribution to Fo
fluorescence by PSI and PSII var
ies with wavelength (Figure 4‐4
). Much of the work on
chlorophyll fluorescence has ign
ored the contribution of PSI be
cause it does not contribute to
dynamic changes in fluorescence
yield. However, it is a signifi
cant, albeit rather constant,
Figure
4
‐
3
.
A
schematic
representation
of
the
processing
of
absorbed
photons
(excitons)
in
the
chlorophylls
associated
with
a
PSII
reactions
center
(PSII)
and
a
non
‐
photochemical
trapping
center
(NPQ).
The
absorbed
photons
can
be
lost
as
radiationless
decay
(
k
D
),
re
‐
emitted
as
a
fluorescent
photon
(
k
f
),
used
for
photochemistry
(
k
P
),
or
quenched
by
NPQs
(
k
NPQ
),
or
as
given
by
Eqn.
4
‐
5.
The
concentration
of
Zeaxanthin
(Zea)
modulates
the
level
of
NPQ.