The Lagrangian time scale for turbulent transport in forest canopies, determined from measured fluxe

1
The Lagrangian time scale for turbulent transport
in forest canopies, determined from measured
fluxes and concentrations and modelled source
distributions

• Vanessa Haverd, Ray Leuning, David Griffith,
• Eva van Gorsel, Matthias Cuntz
• February 06, 2008

2
Lagrangian Dispersion in Plant Canopies

• Dispersion Matrix, D
• Calculated using Lagrangian Dispersion theory
  (e.g. LNF, Raupach 1989)
• Inputs are turbulence statistics
• Standard deviation of vertical wind velocity sw
  (measured)
• Lagrangian Timescale TL (parameterised)

3
Lagrangian Dispersion in Plant Canopies
Applications

4
Parameterisations of Lagrangian Timescale, TL
5
Approach(1)
6
Approach(2)
7
SVAT model, applied to Tumbarumba

• Multilayer canopy model (Wang and Leuning, 1998)
• Sunlit/ shaded leaf model coupling stomatal
  conductance, photosynthesis and energy
  partitioning
• Radiation distribution in the canopy (3
  wave-bands)
• Recent additions
• Soil and biomass respiration rates (Keith et al.
  2008)
• Heat storage fluxes in the canopy air and biomass
  (Haverd et al. 2007)
• Multilayer soil model with coupled heat/moisture
  fluxes and litter

8
Measurements

• 70 m tower in 40 m temperate Eucalyptus forest
  (Tumbarumba Ozflux site)
• Continuous measurements at 70 m
• eddy fluxes (H, lE, CO2)
• Rnet
• upward and downward solar radiation fluxes
• met data
• Continuous temperature and water vapour profiles
• 2 week campaign (Nov 2006)
• 7 tower inlets to 2 Fourier transform infrared
  spectrometers
• Hourly measurements of CO2, water vapour
• Ground-based Lidar measurements of direct beam
  transmission probability (Pgap) and hence foliage
  density profile. (Jupp et al. 2008)
• Array of 3D sonic anemometers ? sw profile,
  Eulerian timescale.
• Chamber measurements of CO2 and CH4 soil fluxes
  (Fest et al. 2008)

9
Measured Concentration Profiles
10
Foliage Area Volume Density and Pgap
Jupp et al., 2008
11
Hourly Net Fluxes at 70 m
12
Vertical source/sink distributions
13
Hourly flux partitioning between ground and
vegetation
14
Turbulence Statistics from Measurements
15
Functional form of TL
c1 ?8 c2 0.32 c1 0 c2 0.32 c1
7.32 c2 0.32
16
Measured and Predicted Concentrations using prior
TL (Styles et al. parameterisation)
17
Measured and Predicted Concentrations using
optimised TL (Styles et al. parameterisation)
18
Massman and Weil parameterisation (1999)
19
Measured and Predicted Concentrations using
optimised TL (Styles et al. and MW
parameterisations)
20
Prior and Optimised Profiles of TL
Styles et al. Massman and Weil
Piecewise linear
21
Summary

• TL required to calculate dispersion matrix
  linking in-canopy source/sinks to concentrations.
• High uncertainty in TL because it cannot be
  measured directly
• We have estimated the profile of TL using
• Hourly vertical profiles of ?, CO2, H2O
• SVAT model predictions of source/sink
  distributions and uncertainites, constrained by
  meaurements
• Net fluxes above canopy
• Chamber measurements of CO2 ground fluxes
• Pgap from ground-based Lidar
• Result
• with c1 8.43.2 c2 0.490.05
• decreases with canopy depth
• 1.7 times higher than prior estimate based on TE

22
Thank you
CMAR Vanessa Haverd Phone 02 6246
5981 Emailvanessa.haverd_at_csiro.au