Evaluating the value of enhanced atmospheric measurements and models to improve interpretation of flux data

Gaps in trace-gas fluxes measured via eddy-covariance occur for many reasons. One of these reasons is that many sites are equipped with flux measurements at only one height above the canopy. The optimal height for detecting fluxes, however, changes in time with the environmental conditions at the site. In some instances, a measurement such as friction velocity (u*) can indicate when data should be eliminated. When u* is high however, data are not eliminated but sensors may not actually be detecting fluxes from the ecosystem of interest. This latter scenario often occurs at night when the surface layer of the atmosphere is stably stratified and may be lower than the tower measurement height. Here, we propose that increased information about atmospheric structure near the surface and the processes occurring within the surface layer can inform an improved interpretation of fluxes as measured at a single point above the canopy. We adjusted and tuned the Advanced Canopy-Atmosphere-Soil Algorithm (ACASA) for modeling a California mediterranean oak savanna, the Tonzi AmeriFlux Site near Ione, CA, USA. We fused datasets of wind profiles, temperature profiles, and fluxes from standard eddy-covariance measurements, radiosonde launches, and upward-facing LIDAR measurements to determine surface layer depth and then drive ACASA from the top of the surface layer. By combining modeled profiles of the surface layer with flux measurements above and below the canopy, we were able to better interpret when flux signals were true indications of canopy processes, and the sources of flux anomalies.