Using Atmospheric Lidar to Interpret CO2 Fluxes in Complex Terrain at Three AmeriFlux Towers

Lower levels of the planetary boundary layer (PBL) tend towards stable conditions at night with turbulent transfer of mass and energy exchange driven solely by mechanically forced turbulence, either from frictional forces near the ground or at the top of the plant canopy or from “top-down” forced intermittent bursts of turbulence generated from wind shear aloft. The nighttime atmosphere is further complicated by variable topography which creates complex wind patterns near the surface including strong, along-valley-axis flows and gravity-driven downslope flows. These features challenge some of the major tenants of eddy covariance theory including the assumption of homogenous flat terrain and dominant vertical turbulence transfer. However, our need for understanding the terrestrial carbon budget requires that these types of measurements be made in places with non-ideal conditions.

Flux towers are usually equipped with eddy covariance instrumentation only to the top of the plant canopy or a few tens of meters above to measure the net transfer of mass and energy between the canopy and atmosphere. At almost every flux site, as the atmosphere becomes more stable at night, inadequate or intermittent turbulence often results in anomalous or erroneous CO2 fluxes, e.g., negative fluxes which indicate photosynthesis at night. This leads to measurement uncertainty and the need for extensive gap-filling of nighttime fluxes. Here, we test the hypothesis that by extending atmospheric flow measurements to well above the canopy we can better understand the atmospheric drivers behind some anomalous CO2 flux signals, with particular focus on nighttime hours and flow features caused by complex terrain. We present over 2000 hours of wind flow observations above three AmeriFlux sites, including a tall evergreen old-growth forest in mountainous Washington State (Wind River AmeriFlux), a two-layered oak savannah in the California Sierra Foothills (Tonzi AmeriFlux), and an annual grassland in the Altamont Hills of California (Diablo AmeriFlux). Measurements of wind speed, direction and turbulence were collected with a vertically-profiling Laser Detection and Ranging (lidar) instrument up to 300 m above the surface at 10 m resolution.

Although there are limitations to calculating accurate turbulence from lidar, as will be discussed, the instrument does provide observations of 3-D wind flow well above the canopy for identifying drivers of “top-down” forced turbulence as well as nearer to the ground (down to 10 m a.g.l.) for identifying katabatic flows. Here, we show evidence of intermittent turbulent bursts and identify times when they penetrate the plant canopy and influence CO2 measurements. We also show evidence of advective flows driven by the surrounding complex terrain. The contribution of advective terms to CO2 exchange is often ignored in the carbon flux budget although we provide evidence that it should be considered even at sites with more gentle local terrain.

This work performed under the auspices of the U.S. Department of Energy by Lawrence Livermore National Laboratory under Contract DE-AC52-07NA27344 and funded by a Laboratory Directed Research and Development Grant (12-ERD-043).