Quantitative comparison of in situ soil CO2 flux measurement methods
Development of reliable regional or global carbon budgets requires accurate measurement of soil CO2 flux. We conducted laboratory and field studies to determine the accuracy and comparability of methods commonly used to measure in situ soil CO2 fluxes. Methods compared included CO2 traps; a multichamber open dynamic system (OC) with an infrared gas analyzer; and a headspace analysis system, with gas chromatograph with circulating fan (GCF) and gas chromatograph without circulating fan (GC), measuring changes in headspace CO2 concentration using a gas chromatograph. In the laboratory, we compared NaOH, OC, GCF, and GC using sand-filled cylinders to simulate a soil system. Three concentrations of standard CO2 gas (representing low, medium, and high soil-CO2 flux rates) entered the cylinders through individually monitored flow meters. Flux rates measured using the OC method equaled the actual CO2 flux at all three CO2 concentrations. Flux rates determined with GC and GCF were < 55 percent of the actual flux rate. The percent of actual CO2 flux measured increased with increasing CO2 concentrations. At all concentrations, NaOH collected between 40 and 47 percent of CO2 entering the system. A field study was conducted to verify laboratory results and allow comparison with the soda lime trap (SODA) method. In laboratory and field studies, all methods detected significant differences in flux rates among the standard CO2 concentrations and field sites. Regression analyses showed good relationships between NaOH, SODA, and GC methods with flux rates measured using the OC methods (r2 >= 0.78). Slope values for these regression equations ranged from 0.34 for NaOH to 0.54 for GC and SODA. These results suggest that data collected using the other methods could be standardized to OC flux rates. However, because methodological differences significantly affect CO2 flux measurements, care should be used in applying these relationships.