Adjustment of CO2 flux measurements due to the bias in the EC150 infrared gas analyzer

During the Regional Approaches to Climate Change (REACCH) program, eddy covariance monitoring over agricultural fields were used to estimate annual carbon and water budgets in the inland Pacific Northwest. Here, we assess the effect of a bias in the high-frequency CO2 concentration measurements using the Campbell Scientific EC150 infra-red gas analyzer on the CO2 fluxes and field-scale carbon balances. The bias stems from using a lower frequency temperature measurement to calculate the CO2 density, which misses higher frequency temperature fluctuations. To generate the bias adjustment, data were collected over four similar agricultural sites as part of the Long-Term Agroecosystem Research network for multiple months using the same four instrument sets used in the REACCH project. The difference between the high-frequency and low-frequency CO2 fluxes were regressed against the kinematic heat fluxes to generate a correction equation for each instrument set, which were applied to the historical REACCH data to determine the effect of the bias on the measured and gap-filled flux values. The re-calculated positive biases in the measured fluxes were 40 gC-CO(2)m(-2) yr(-1) to 126 gC-CO(2)m(-2) yr-1, indicating greater losses to the atmosphere than initially estimated. Once gap-filled, three out of fourteen site-years switched from weak carbon sinks to weak carbon sources. When the carbon exported via harvest was included in the budget calculation the bias correction still impacted the source/sink strength but did not change the sign of the carbon balance. Overall, the total net ecosystem exchange decreased between 300-470 gC-CO(2)m(-2) per site (29-46%) over the 4 crop-years from the bias adjustment process.