Inferring gas fluxes from point or lineaveraged concentrations

1
Inferring gas fluxes from point or line-averaged
concentrations

• Tom Denmead
• Fellow, CSIRO Land and Water University of
  Melbourne
• Ozflux Conference, 4 February 2008

2
A backward Lagrangian stochastic (bLs) dispersion
model

• The model traces particles backwards from sensor
  to origin using a Lagrangian dispersion model
• Surface fluxes calculated from number of
  touchdowns inside and outside source area in many
  simulations
• (C/Q)sim (1/N) S 2/w0
• C is downwind concentration
• Q is the surface flux
• N is the number of trajectories commonly, 50,000
  
• w0 is the vertical velocity of particles at
  touchdown
• Q (C-Cbackground) / (C/Q)sim

Point concentration sensor
Source area
Micromet.
wind
Touchdowns
3
A backward Lagrangian stochastic (bLs) dispersion
model

• Suitable for point, line or area sources (any
  shape)
• Inputs
• geometry of source area height and location of
  sensor, wind speed and direction, atmospheric
  stability,
• gas concentrations upwind and downwind
• Uses a software package called WindTrax to
  calculate surface fluxes from concentration and
  micrometeorological data

Point concentration sensor
Source area
Micromet.
wind
Touchdowns
4
Point concentration measurements an example from
grazing (315 dairy cows)
Ammonia concentrations measured with passive
samplers
5
WindTrax map
2 adjoining pasture bays grazed in 6 sessions,
one-third of a bay at a time Sensors located at
heights of 1.4 and 2m on 12 masts on the corners
of each grazed section
Chemical sensors
Meteorological Sensors 2 anemometers Wind
vane Atmos. stability Background concentration unk
nown
Grazed sections
6
Sensor numbers measuring NH3 emissions after N
fertiliser applied to the whole bay
Average fluxes (µgNH3-N m-2 s-1), 0900-1800,
using different sensor combinations wind
direction 170o
If background unknown, need 2 sensors If gt2
sensors, problem is over-determined model
returns least-squares, best-fit background and
flux
2.66
2.33
2 sensors, one upwind one downwind, each at 1.4m
1.55
2.05
24 sensors, 2 to each mast, at 1.4 and 2m
7
Multiple source areas (using 16 sensors)
Average fluxes, 0800-1730, µgNH3-N m-2 s-1
Grazed yesterday ?
0.14
Grazed today ?
0.30
-0.02
Ungrazed ?
8
An example result emissions from one grazed
section

• Before grazing small NH3 uptake
• Continuous NH3 emission during after grazing
• Large NH3 emissions after fertilizing
• Emissions cease after irrigation

9
Line-averaged concentrations laser and Fourier
Transform Infrared (FTIR) systems

• Lasers measure line-averaged gas concentrations
  up to 1km, FTIR less
• Lasers tripod-mounted, stand alone,
  battery-operated units FTIR requires mains power
• Suitable for point, line and small area sources

Open-path FTIR (CO2,CH4, N2O, NH3)
Open-path laser (CO2, CH4, NH3)
10
Tests releases and recoveries
Daisy our virtual cow

• CH4, N2O, NH3 released from cylinders through
  mass-flow controllers
• Tests conducted of recoveries from point source
  and plane source emissions

40m x 15m grid of permeable pipes
40m x 15m grid of permeable pipe
11
Tests releases and recoveries_ point sources

• Average NH3 concentrations measured by a laser
  instrument at 1.5m height along a line of 123m,
  10m downwind of a point source of ammonia 0.5m
  above ground.

12
Tests releases and recoveries_ areal sources

• Top
• Recovery by laser of NH3 released from ground
  level grid, 25m x 25m
• Laser 2m downwind of grid
• Path 128m
• NH3 released at 5L min
• Bottom
• Recovery by 2 lasers and FTIR of CH4 released
  from ground level grid, 40m x 15m
• Path 140m

13
Example application of open-path systems CH4
emission from a feedlot with 14,000 cattle
WindTrax map of feedlot layout
Laser paths
Micromet. tower
14
Strengths and weaknesses

• bLs technique WindTrax represent a powerful new
  tool for measuring gas emissions from
  well-defined source areas
• Main advantage fluxes determined from just one
  concentration measurement and knowledge of the
  background concentration turbulence statistics
• Both closed and open-path measuring systems
  possible
• Path lengths of up to 1 km possible, but 100 to
  300m seem more reliable
• Open path systems
• Lasers tuned to individual gases CO2, CH4, NH3
  and H2O
• FTIR units measure many of the gases of interest
  in the context of landscape-atmosphere exchanges
  simultaneously CO2, CH4, NH3, H2O, N2O and CO
• The main disadvantage of the bLs technique may be
  in its parameterisation of turbulent transport,
  but many tests have shown that with appropriate
  precautions, gas emissions can be measured with
  acceptable accuracy (Flesch et al., 2004 McBain
  and Desjardins, 2005 Laubach et al., 2008).

15
Acknowledgements

• Collaborators
• University of Melbourne
• Deli Chen, Debra Turner, Yong Li, Zoe Loh,
  Julian Hill
• University of Wollongong
• David Griffith, Mei Bai, Glenn Bryant, Travis
  Naylor
• DPI Victoria
• Kevin Kelly, Frances Phillips
• Charlton Feedlot
• Sandalwood Feedlot
• Funding
• Australian Greenhouse Office
• Meat and Livestock Australia

16
Thank you
CSIRO Land and Water and University of
Melbourne Tom Denmead Fellow Phone [phone]
5568 Email tom.denmead_at_csiro.au Web www.csiro.au