Atmospheric constraints on plant
water use efficiency in JULES
Margriet Groenendijk and Peter Cox
JULES general science meeting
17 December 2012
Water Use Efficiency?
Agriculture - Conference in London: “More Crop Per
Drop - Raising Water Use Efficiency”
Why CO2 is good: “Rising CO2 Boosts Plant Water Use
Efficiency” (http://www.plantsneedco2.org)
Water Use Efficiency?
WUE controls the relationship between the
ecosystem water and carbon balance
Improving understanding of vegetation responses
to climate variations, responses of carbon and
hydrological cycle
When the WUE is known, GPP can be calculated
from hydrological cycle variables
s
a
and internal partial pressures of CO2 [Pa] and es and ea the saturated
he ambient and internal partial pressures of CO2 [Pa] and es and ea the satu
ure [Pa]. [See Katul papers for conversions between units and why
apour pressure [Pa]. [See Katul papers for conversions between units and
quation]
ate in this equation]
Hypothesis
Ecosystem
WUE
eco derived from fluxes is equal to
water
and
carbon
fluxes
model
developed
for
the
leaf-scale
are
used
ystem-scale
water and carbon
fluxes
model
developed
for the leaf-scale are
atmospheric
WUEatm
derived
from
meteorological
s, 1980).
impliesThis
thatimplies
the WUE
atthe
theWUE
ecosystem-scale
can be estimated
with
variables
atthat
the
leaf
levelat the ecosystem-scale
can be estimated
dequal:
should be
equal:
This
is valid on larger scales as well
U Eatm
GP P
W U Eeco =W U Eeco = GP P
ET
ET
Ca Ci Ca CC
f ) Ca (1 f )
a (1
i
=W U Eatm ==
=
1.6D
1.6q
RH)
sat (1 1.6q
1.6D
RH)
sat (1
(5)
(6)
y production,
flux, Ca and
CiCare
theCambient
gross
primary
ET the transpiration
flux,
a and
i are the am
f = 1 production,
-ET
f0 the transpiration
f0 CO
:vapour
stomatal
(Ciq/C
a)deficit,
CO
deficit,
the saturated
pressure
ressures
of
theresistance
vapour
pressure
qsat thevapour
saturated
vapour pre
2 , D the
sat
2 , D pressure
Chumidity.
is meaningless
and therefore
be
Ci at the ecosystem-scale
is meaningless
andassumed
thereforetoassumed
i at the ecosystem-scale
Global Water Use Efficiency
Ecosystem WUE
Atmospheric WUE
Examples from HadCM3 QUMP ensemble
l:
W U Eeco
m
GP P
=
ET
Which fluxes to (5)
use?
Ca Ci
Ca (1 f )
=
=
1.6D
1.6qsat (1 RH)
(6)
oduction, ET the transpiration flux, Ca and Ci are the ambient
D the vapour pressure deficit, qsat the saturated vapour pressure
the ecosystem-scale is meaningless and therefore assumed to be
pplied in many model formulations [Refs].
3
er and carbon fluxes model developed for the leaf-scale are used
m-scale water and carbon fluxes model developed for the leaf-scale are used
plies that the WUE at the ecosystem-scale can be estimated with
80). This implies that the WUE at the ecosystem-scale can be estimated with
al:
ould be equal:
The same?
GP P
GP P
W U Eeco =
(5)
WU
Eeco =
(5)
ET
ET
Ca Ci
Ca (1 f )
Ca (1 f )
=Ca Ci
(6)
tm =
W U1.6D
Eatm = 1.6qsat (1
= RH)
(6)
1.6D
1.6qsat (1 RH)
roduction,
ET the transpiration
flux, Ca and
theC ambient
i are
primary production,
ET the transpiration
flux,CC
and
are the ambient
a
i
, D the
vapour
pressure deficit, qsat the saturated vapour pressure
ures
of CO
2 , D the vapour pressure deficit, qsat the saturated vapour pressure
the ecosystem-scale
is meaningless
and therefore
assumed assumed
to be to be
midity.
Ci at the ecosystem-scale
is meaningless
and therefore
in many model formulations [Refs].
Cpplied
a , as it is applied in many model formulations [Refs].
3
3
59
and ambient water vapour pressure [Pa]. [See Katul papers for co
60
vpd is a good estimate in this equation]
Increasing CO2
61
To simulate the ecosystem-scale water and carbon fluxes model d
62
(e.g., Farquhar et al., 1980). This implies that the WUE at the eco
63
both eqs. 3 and 4 and should be equal:
W U Eeco =
64
W U Eatm =
GP P
ET
Ca Ci
Ca (1
=
1.6D
1.6qsat (1
65
where GP P is the gross primary production, ET the transpirati
66
and internal partial pressures of CO2 , D the vapour pressure defici
67
and RH the relative humidity. Ci at the ecosystem-scale is meani
68
a constant fraction f of Ca , as it is applied in many model formul
3
Increasing CO2
Average annual
WUE for 9 sites
Ecosystem
Air temperature
Surface
temperature
Increasing CO2
Relative change
of WUE for 9
sites
Ecosystem
Air temperature
Surface
temperature
qual:
d should be equal:
U Eatm
GP P
W U Eeco = W U Eeco = GP P
ET
ET
Ca Ci Ca CC
f ) Ca (1 f )
a (1
i
= W U Eatm =
=
=
1.6D
1.6q
RH)
1.6D
RH)
sat (1 1.6q
sat (1
(5)
(5)
(6)
(6)
primary production,
ET the transpiration
flux,CC
are the ambient
ygross
production,
ET the transpiration
flux, Ca and
theCiambient
a and
i are
pressures
of vapour
CO2 , Dpressure
the vapour
pressure
qsat the saturated
vapour pressure
CO
deficit,
qsatdeficit,
the saturated
vapour pressure
2 , D the
i
humidity.
Ci at the ecosystem-scale
is meaningless
and therefore
at the ecosystem-scale
is meaningless
and therefore
assumed assumed
to be to be
Why
are
there
still
differences?
of
C
,
as
it
is
applied
in
many
model
formulations
a
s applied in many model formulations [Refs]. [Refs].
Soil moisture limitation?
3
3
VPD
sensitivity of f0?
Dark respiration?
Averaging?
Global WUE
BIG QUESTION reproducible with observations over last
century?
Only growing season?
Validation: Fluxnet and ...
What is best observational data?
(Surface) Temperature
Humidity
CO2
NEW perturbing land
surface in a large ensemble
MOSES2 coupled to HadAM3P
climateprediction.net
Perturbations:
Model parameters
Land use maps from different datasets
Soil type maps