Development of an onsite computation scheme of eddy-covariance fluxes

Increasingly, agriculture and related industries rely on real-time measurements of surface fluxes and device control based on flux measurements. We have developed a program that computes the eddy-covariance fluxes of CO2 and H2O in real-time on a commercial data logger model. The calculated flux values are targeted for device control. The performance of the program was compared with those of published software programs on maize field data. The calculated fluxes were consistent among the programs. Because practical onsite and real-time quality control procedures of flux measurements were unavailable, we developed a quality control procedure for our program that utilizes sensor disability signals. Filtering the sensor-disabled signals greatly improved the detection of anomalous fluxes during rainy periods. However, filtering the instationary fluxes by the standard deviation criterion was of limited practical use. By slightly modifying the source code, we can adapt the framework of our program to other combinations of sonic anemometer and gas analyzer.