Some infrared CO2
sensors, such as GMD20 and GMT222
(VAISALA), are widely used for soil CO2 efflux measurements despite
the fact they have a slow response rate. The output signal is delayed both from
diffusion processes in the sample cell and internal averaging calculations
necessary for stable data output. For accurate estimations of CO2
efflux, we therefore need to know the actual increase in CO2
concentration in a chamber without composite delays. To parameterize these
delays, we conducted laboratory experiments to determine the response
characteristics of sensors under diffusion and flow-through conditions. Next,
we developed a backward calculation method for estimation of the actual CO2
concentration increase using the delayed sensor output (BCDC: Backward
calculation for delay compensation). The results showed that the slow response
of sensors caused large estimation errors in CO2 efflux
measurements. In the case of GMT222,
a 10% underestimation was suggested when the soil CO2 efflux was
calculated with non-corrected data using a nonlinear regression method with
sampling intervals of 300 seconds. Thus, correction of the sensor response with
a backward estimation might be effective. We also calculated and evaluated the CO2
efflux from a forest floor in a deciduous forest employing the BCDC method.
Author: Y. Mizoguchi, Y. Ohtani, T. Watanabe, and Y. Yasuda (pop128 at ffpri dot affrc dot go dot jp)
Filesize: 72.28 Kb