Smoke test showing distribution of CO2 across the ring
The OzFACE facility is now fully operational at our tropical savanna site at Yabulu,
about 20km north of Townsville. Construction commenced in December 2000 and was
completed in April 2001. Campbell Scientific Australia, with some assistance from
Campbell Scientific in the US, designed and constructed the system. The FACE system
consists of six rings, each 15 m in diameter. The vegetation enclosed by two rings
is exposed to ambient CO2 (370ppm), two rings are exposed to a CO2 level of 450 ppm
and the remaining two rings are exposed to CO2 concentrations of 550 ppm. Within
these plot areas, one third of the area is left intact, one third is clipped to
simulate grazing and one third has nutrients added to simulate higher fertility
savanna locations. Local eucalypt and acacia seedlings have been planted to study
woody-grass dynamics, which is a critical issue in savannas, in terms of current
land management practices and the impacts of future climate. It was not possible to
include mature eucalypts or acacias within the FACE rings.
Measurements include net primary productivity and species abundances, plant basal
area, forage quality, a range of plant physiological measures, plant defence
compounds, plant phenology, litter decomposition, soil respiration, soil moisture
(NMM and TDR monthly), soil fauna, soil organic carbon. Through collaboration we
hope to look at root turnover using mini-rhizotrons.
The FACE design is an octagonal system with small blowers attached to each of the
eight 100mm PVC pipes in each ring. We have a two stage CO2 control process. For
each ring we have four solenoids of different capacities controlling overall release
of CO2 to the ring and finer level control is provided for each segment with 2
solenoids (the segment most perpendicular to the wind releases at proportionately
1.0 and the two side segments at 0.7). CO2 is injected into the plenums immediately
in front of the blowers. We have the IRGAs in the middle of each ring to reduce
delay time in CO2 sampling.
We have been testing the system for the last month or so and at this stage we are
within 10% of the setpoint 80% of the time (one minute averages) during the day.
However, at night we are within 10% of the setpoint only about 65% of the time
because of almost zero wind. We expect further improvements as we fine tune the PID
algorithm. Currently our CO2 use is about 1.5 tonnes from 6am to 6pm and about
0.7 tonnes between 6pm and 6am. Wind gusts in the afternoon can reach 15 m/s and
often average 6 m/s. We have constructed a portable spatial sampling system and
spatial control looks pretty good so far though we can only test one ring at a time.
