ChEAS 2005 D.S. Mackay June 1-2, 2005 Reference canopy conductance through space and time: Unifying properties and their conceptual basis D. Scott Mackay.

ChEAS 2005 D.S. Mackay June 1-2, 2005

Reference canopy conductance through space and time:Unifying properties and their conceptual basis

D. Scott Mackay1 Brent E. Ewers2 Eric L. Kruger3

Jonathan Adelman2 Mike Loranty1 Sudeep Samanta3

1SUNY at Buffalo 2University of Wyoming 3UW-Madison

NSF Hydrologic Sciences EAR-0405306EAR-0405381EAR-0405318


Problem

• Prediction of water resources from local to global scales requires an understanding of important hydrologic fluxes, including transpiration

• Current understanding of these fluxes relies on “center-of-stand” observations and “paint-by-numbers” scaling logic

• Spatial gradients are ignored, but this is an unnecessary simplification

• New scaling logic is needed that includes linear or nonlinear effects of spatial gradients on water fluxes


Why is canopy transpiration important to hydrology?W

ate

r F

lux

(Sa

p fl

ux

or

12

2m

WL

EF

) (m

m/d

ay)

0.0

1.0

2.0

3.0

4.0

5.0

2001

VPD (KPa)

0.0 0.5 1.0 1.5 2.00.0

1.0

2.0

3.0

4.0

5.0WLEF 122m eddy covarianceAggregated sap flux

R2=0.88

R2= 0.65

2000

R2 = 0.79

R2 = 0.91

Average annual precipitation:800 mm

Growing season precipitation:300-500 mm

Growing season evapotranspiration:350-450 mm

Canopy transpiration (forest):150-200 mm

Canopy transpiration (aspen):300 mm

Ewers et al., 2002 (WRR)Mackay et al., 2002 (GCB)


Assumptions

• Transpiration is too costly to measure everywhere, and so appropriate sampling strategies are needed

• The need for parameterization (e.g., sub-grid variability) will never go away

• Both forcing on and responses to transpiration are spatially related (or correlated), but this correlation is stronger in some places

• Human activities may increase or decrease this correlation


Transpiration[mm (30-min) –1]

0 .05 .1 .15 .2

What if we increase edge effects?

Center-of-StandBasis

Spatial Gradient Basis


Why is Transpiration a Nonlinear Response?

RelativeResponse

Relative water demand

StomatalConductance (Jarvis, 1980;

Monteith, 1995)

Transpiration (No stomata) “hydraulic failure”

ReferenceConductance

Transpiration (With Stomata) “prevents hydraulic failure”

Prevention of hydraulic failure is a key limiting factor for carbon gain and nutrient use by woody plants.


Conceptual Basis of Spatial Reference Conductance

GS = GSref – mlnD m = 0.6GSref

(Oren et al., 1999)

GSref1 2 3

0

30

60

90

120

150

D

GS

Gsref

m = d

d ln

- G

D

S

m

EnvironmentalGradient

Canopy stomatal controlof leaf water potential

Hydraulic“Universal” line

Mapping from spatial domain into a linear parameter domain


0

0.4

0.8

1.2

0 0.5 1 1.5 2

G Sref (mm s-1)

m [mm s-1

ln(kPa)-1]

red pine

aspen

sugar maple

alder

cedar

slope = 0.6

y = 0.601x - 0.022

R2 = 0.96

0

0.4

0.8

1.2

0 0.5 1 1.5 2

G Sref (mm s-1)m

[m

ms

-1 ln

(kP

a)-1

]

Mackay et al., 2003 (Advances in Water Resources)


Hypothesis 1

• GSref varies in response to spatial gradients within forest stands, but the relationship between GSref and m remains linear

• Note that 1/D 1- 0.6ln(D) for 1 ≤ D ≤ 3 kPa; error is maximum of 16% at 2 kPa

• Thus many empirical stomatal conductance models are applicable, but discrepancies will occur at moderate mid-day D when it is hydrologically most relevant


ChEAS, HC 2001

Q0

(m

ol m

-2s-1

)

0

500

1000

1500

2000

D (

kPa)

0.0

0.5

1.0

1.5

2.0

2.5Q0

D

Julian Day

188 190 192 194

EC (

mm

30-

min

-1)

0.00

0.05

0.10

0.15

0.20ModelMeasured

A. saccharumA

B


ChEAS, WC 2001

Q0

(m

ol m

-2s-1

)

0

500

1000

1500

2000

D (

kPa)

0.0

0.5

1.0

1.5

2.0

2.5Q0

D

Julian Day

194 196 198 200

EC (

mm

30-

min

-1)

0.00

0.05

0.10

0.15

0.20ModelMeasured

A. saccharumA

B


Julian Day

194 196 198 200

EC (

mm

30-

min

-1)

0.00

0.05

0.10

0.15

0.20ModelMeasured

ChEAS, SV 2001

Q0

(m

ol m

-2s-1

)

0

500

1000

1500

2000

D (

kPa)

0.0

0.5

1.0

1.5

2.0

2.5Q0

D

A. saccharumA

B


S

SS QQ

QDgg 1max

Hydraulic constraint

Light sensitivity

Some model realizations follow hydraulic theory

Best dynamic response

Agricultural and Forest Meteorology (in review)


These models preserve plant hydraulics and represent the regional variability for Sugar maple

Agricultural and Forest Meteorology (in review)


Aspen flux study, northern Wisconsin

Wetland

Transition

Upland

X – sample pointX - Aspen

Funded by NSF Hydrological Sciences


ChEAS, Aspen, August 3, 2004

0

20

40

60

80

100

120

0 20 40 60 80 100 120 140 160 180

Simulated G Sref (mmol m-2 s-1)

Sim

ulat

ed m

(m

mol

m-2

s-1

kP

a-1)

WetlandTransitionalUplandm = 0.6Gsref

Aspen Restricted Simulations

Funded by NSF Hydrological Sciences


Lodgepole pine study, Wyoming

A1, riparian zone

Row 4, lower slope

Row 5, mid-slope

Row 6, mid-slope

Row 7, mid-slope

Row 8, upper slope

X – sample pointX – Lodgepole pine


Wyoming, Lodgepole Pine, August 25, 2004

0

10

20

30

40

50

60


Sim

ula

ted

m (

mm

ol m

-2 s

-1 k

Pa

-1)

A1Row 4Row 5Row 6Row 7Row 8m = 0.6 Gsref

0

10

20

30

0 10 20 30 40 50 60 70 80 90


So

il m

ois

ture

(%

)

A1, riparian zone

Row 4, lower slope

Row 5, mid-slope

Row 6, mid-slope

Row 7, mid-slope

Row 8, upper slope

Basal area crowding

Lodgepole Pine Restricted Simulations


ReferenceCanopyConductance

Water availability Indexlow

high

low

high

xeric mesic

HydraulicConstraintIndex

Summary of Ecohydrologic Constraints

high low


Hypothesis 2

• Variation in leaf gS within and among species and environments is positively related with leaf nitrogen content and leaf-specific hydraulic conductance

• The relative response of gSmax to light intensity (Q) is governed in large part by leaf, and this dependence underlies stomatal sensitivity to D

– Corollary i: gS will increase with increasing Q until it reaches a limit imposed leaf, which for a given leaf is mediated primarily by D

– Corollary ii: The limit imposed on relative stomatal conductance (g/gSmax) by leaf (relative to the threshold linked to runaway cavitation, crit) is consistent within and among species


Hypothesis 3

• The model complexity needed to accurately predict transpiration is greater in areas of steep spatial gradients in species and environmental factors

• Model complexity (e.g. number of functions, non-linearity) should be increased when absolutely necessary, and it should subject to a penalty

• We should gain new knowledge whenever we are forced to increase a model’s complexity



ChEAS 2005 D.S. Mackay June 1-2, 2005 Reference canopy conductance through space and time: Unifying properties and their conceptual basis D. Scott Mackay.

Documents