The free air CO₂ enrichment (FACE) methodology for ecosystem research has been viewed by some as a 'real-world' approach. It has been argued that this approach is the best test for the effect of the impending CO₂ enrichment on agricultural and on natural ecosystems. The pros and cons of FACE methodology are presented below at some length because of the uncertainty of the usability of FACE methodology in ecosystem research.

   The FACE approach to CO₂ enrichment is to apply a network of vertical vent pipes (McLeod et al., 1983), or other release system (Allen, 1975), near the ground in order to provide elevated CO₂ to the ambient air of the plants. The object is to avoid the need for an enclosure or chamber around the plants. The major differences between FACE and either outdoor con- trolled mini-greenhouses or ecocosms or open-top chambers are that FACE eliminates the following chamber effects: (1) reduction of the solar radiation, and (2) unnatural wind flow, turbulence, and micrometeorological patterns; further, they potentially permit the study of many ecosystem phenomena.

   FACE arrays that have been used include the US Environmental Protection Agency Zonal Air Pollution System (ZAPS) plots with dimensions of 73 m by 85 m for a prairie grassland (Lee and Lewis, 1978). The US Department of Energy used air pollution exposure plots with dimensions of 29 m by 27 m for a soybean crop at Argonne National Laboratory (Miller et al. , 1980). The UK Central Electricity Research Laboratories used a circular plot array of 27 m in diameter for a wheat crop exposed to air pollutants (McLeod et al., 1983). This latter system was modified for use over cotton fields for CO₂ release in Mississippi in 1987 and 1988, and in Arizona in 1989-1991 (Hendrey et al., in progress).

   Large, uncontained sample areas are an advantage when the system is heterogeneous or diverse vegetatively. Ecological studies of effects of elevated CO₂ on cycles of litter production, organic matter accumulations, soil respiration, nutrient cycling, above-ground competition, and phenology require a large area of uniform exposure and treatment. When replication of sample plots is multiplied by the size and number of plots required to conduct open-area studies, however, logistics and financial requirements may become limiting.

   The concentration of CO₂ in a large area supplied through a network of pipes will depend inversely upon wind speed, directly upon the release rate of CO₂ (Allen, 1975; McLeod and Fackrell, 1983) and inversely with vegetation height when mass consistency is taken into account (Hanna et al., 1982). To hold CO₂ concentration constant on the average, the delivery rate must be increased at higher wind speeds, and this requires a feedback mechanism to be included in the FACE design. Nevertheless, it will be very difficult to maintain constant CO₂ under all weather conditions. Under most conditions only the very center of a circular design will have a uniform horizontal distribution of concentration.

   The FACE system being used in Arizona by Hendrey and colleagues is composed of an array of 32 individually valved vertical pipes, each containing multiple gas injection ports connected to a 22 m diameter toroidal distribution plenum chamber (Fig. 15.7). A high-volume blower injects air containing variably elevated levels of CO₂ into the plenum torus. The number and location of open vent pipes (Fig. 15.8) is based on both wind direction and speed. The amount of gas metered into the air stream entering the plenum is based on wind speed and real-time measurements of the CO₂ concentration at the center of the array. An empirically derived proportional, integrative, differential control algorithm adjusts the supply of CO₂ and the number of vent pipes releasing gas to maintain the desired concentration within the FACE array.

   The gas control system is designed to maintain CO₂ releases that are always upwind of the center of the FACE array. This is achieved by measurement of wind direction and controlling the vertical vent pipe (VVP) ball valves so that the upwind valves are open. At wind velocities lower than a threshold wind speed, all of the VVPs are opened and the system operates without directional control.

   It may be possible to provide heating of air in the FACE array to analyze possible effects of climate warming by burning methane and injecting the combustion gas import of the FACE air intake. Starting with the heat of combustion of CH₄, 12 kcal/g, and the specific heat of air, 7 cal/K per mole of air, 0.1036g of CH₄ will heat 1 m³ air 4K and will add 144ppm CO₂. Horizontal air flux through a 22 m diameter, 2 m high FACE array at 1 m/s air velocity is 2840 m³/min. This can be heated 4K by 294 g CH₄/min. On a yearly basis, with a 150-day growing season, 64 tons of CH₄ would be required to enhance ambient air by 144 ppm CO₂ and 4K across the FACE array.

   Disadvantages of the FACE approach include the fact that carbon dioxide concentration is variable across the ring due to natural atmospheric turbulence. Although CO₂ is also somewhat variable in nature on a smaller scale, the atmospheric reservoir that replenishes CO₂ is already 'premixed' to a relatively fixed concentration. When CO₂ is injected in a FACE system it enters the wind flow field which has a wide range of speeds and directions. Using vertical standpipes and computer controlled injection provides a partial solution to the problem, but significant divergence from programmed concentrations will occur. The technical difficulty of establishing and maintaining a sophisticated electronic system of sensors, controllers, and computers in the field is difficult. The cost is also a disadvantage.

   In spite of the disadvantages described above, FACE may be the preferable technique to use in mid-size to large-stature ecosystems. Open tops have not been applied in these systems and ecocosms have not yet been developed for any system.