dave lawrence2, rosie fisher , pierre gentine daniel kennedy1, … · 2018. 2. 9. · ψ...
TRANSCRIPT
Implementing Plant Hydraulics for CLM5
Daniel Kennedy1, Sean Swenson2, Keith Oleson2, Dave Lawrence2, Rosie Fisher2, Pierre Gentine1
1: Columbia University, Department of Earth & Environmental Engineering2: NCAR, Climate & Global Dynamics Lab
slide courtesy: Dave LawrenceIntroduction - 1
slide courtesy: Dave LawrenceIntroduction - 1
Can we represent vegetation water potential in CLM?
Model Vegetation Water Potential
Introduction - 2
Model Vegetation Water Potential
Hydraulic Redistribution
Root Water Extraction
Water Stress
Introduction - 2
...
ψsunlit-leaf
ψshade-leaf
ψroot
ψstem
Esun
Eshade
ks,1
ks,2
ks,n
k2
k1,a
k1,b
kr,1
kr,2
kr,n
ψsoil,2
ψsoil,n
ψsoil,1
Plant Hydraulic Stress: “PHS”● Solve for vegetation water
potential values that:○ satisfy flow continuity○ match supply & demand
● Water supply○ Darcy’s Law○ q = k( ) A Δ
● Water demand○ E = Emaxf( leaf)○ Emax via Medlyn
Model Description - 1
Each timestep PHS solves for● root, ● stem, ● shade, and ● sun leaf water potential
Find the set of water potentials that balances● water supply
○ q = kAΔ ● water demand
○ E = Emaxf( )
Model Description - 2
PHS
Model Description - 3
SMS
Medlyn Stomatal Conductance Model
CLM5 default All previous versions
PHS
Model Description - 3
SMS
Medlyn Stomatal Conductance Model
Attenuation based on leaf water potential
Attenuation based on soil water potential
CLM5 default All previous versions
STRESS
PHS
Model Description - 3
SMS
Medlyn Stomatal Conductance Model
Attenuation based on leaf water potential
Attenuation based on soil water potential
Distributed among soil layers based on
hydraulic gradient
Distributed among soil layers based on BTRAN heuristic
CLM5 default All previous versions
STRESS
SINK
Caxiuana, Brazil
Experiment Description
PHS SMS
Throughfall Excluded
Throughfall Ambient
Throughfall Ambient
Throughfall Excluded
Experiment Description
Caxiuaña, Brazil
Hydraulic Gradient Root Water Extraction
...
ψsunlit-leaf
ψshade-leaf
ψroot
ψstem
Esun
Eshade
ks,1
ks,2
ks,n
k2
k1,a
k1,b
kr,1
kr,2
kr,n
ψsoil,2
ψsoil,n
ψsoil,1
Hydraulic Gradient Root Water Extraction
...
ψsunlit-leaf
ψshade-leaf
ψroot
ψstem
Esun
Eshade
ks,1
ks,2
ks,n
k2
k1,a
k1,b
kr,1
kr,2
kr,n
ψsoil,2
ψsoil,n
ψsoil,1
Hydraulic Gradient Root Water Extraction
...
ψsunlit-leaf
ψshade-leaf
ψroot
ψstem
Esun
Eshade
ks,1
ks,2
ks,n
k2
k1,a
k1,b
kr,1
kr,2
kr,n
ψsoil,2
ψsoil,n
ψsoil,1q = k A Δ
Hydraulic Gradient Root Water Extraction
...
ψsunlit-leaf
ψshade-leaf
ψroot
ψstem
Esun
Eshade
ks,1
ks,2
ks,n
k2
k1,a
k1,b
kr,1
kr,2
kr,n
ψsoil,2
ψsoil,n
ψsoil,1q = k A ( soil,i- root)
PHS: prognostic water potential
Hydraulic Gradient Root Water Extraction
...
ψsunlit-leaf
ψshade-leaf
ψroot
ψstem
Esun
Eshade
ks,1
ks,2
ks,n
k2
k1,a
k1,b
kr,1
kr,2
kr,n
ψsoil,2
ψsoil,n
ψsoil,1q = k A ( soil,i- root)
PHS: prognostic water potential
SMS: constant, parameter
Hydraulic Gradient Root Water Extraction
...
ψsunlit-leaf
ψshade-leaf
ψroot
ψstem
Esun
Eshade
ks,1
ks,2
ks,n
k2
k1,a
k1,b
kr,1
kr,2
kr,n
ψsoil,2
ψsoil,n
ψsoil,1q = k A ( soil,i- root)
PHS
SMS
Hydraulic Gradient Root Water Extraction
...
ψsunlit-leaf
ψshade-leaf
ψroot
ψstem
Esun
Eshade
ks,1
ks,2
ks,n
k2
k1,a
k1,b
kr,1
kr,2
kr,n
ψsoil,2
ψsoil,n
ψsoil,1q = k A ( soil,i- root)
PHS
SMS
Root Water Extraction
● Caxiuana 2003● Bars show median conductance: k [s-1]● Lines span interquartile range
PHS
SMS
AMB TFE
Hydraulic Redistribution
cartoon credit: https://nature.berkeley.edu/dawsonlab
Hydraulic Redistribution - 1
PHS onAmbient throughfall● total HR = 1.12m● total ET = 4.34m
Hydraulic Redistribution - 1
PHS onAmbient throughfall● total HR = 1.12m● total ET = 4.34m
Hydraulic Redistribution - 1
PHS onAmbient throughfall● total HR = 1.12m● total ET = 4.34m
Hydraulic Stress
cartoon credit: Fields et al., Acta horticulturae
Hydraulic Stress - 1
PHS SMS
● Dry season (SON-2003) mean diurnal cycle● PHS based on leaf matric potential● SMS based on soil water potential
Hydraulic Constraint vs. Soil Moisture Stress
Hydraulic Stress - 2
wettest
driest
AMB
TFE
constrained to points
with FSDS>400 W/m2<425 W/m2
TFEbegins
Simulation Results - 1
Monthly mean ● Stress● Photosynthesis● Transpiration
Image: http://nptel.ac.in/courses/104103020/module7/lec5/4.html
Conclusion - 1
Modeling vegetation water potential in CLM ...
● Yields a more physical model
Image: http://nptel.ac.in/courses/104103020/module7/lec5/4.html
Conclusion - 1
● Yields a more physical model● Expands interface with
○ hydraulic community○ remote sensing
Modeling vegetation water potential in CLM ...
Thanks!
● Entire NCAR land model working group● Columbia Water Center● Slides online: goo.gl/Mqoozb● Contact: [email protected]
extra slide
AMB TFE
SON
FMA