Representation of Microscale Surface Turbulent Fluxes in the Planetary Boundary Layer: The Case of the Complex Heterogeneous Terrain of the Arctic Tundra

Microscale surface fluxes respond to turbulent forces of mechanical and buoyancy origin. Such fluxes can be measured by eddy-covariance towers but they express an extremely localized version of the land-surface atmosphere interaction, on the order of a hundreds m² or less. This representation is critical; especially in complex heterogeneous terrain and under variable atmospheric flow like the case in the basins populating the Arctic tundra. On the other hand, turbulent fluxes at the PBL scale are the result of the space-time convolution of surface radiative properties (albedo, soil moisture, vegetative gradients) and surface aerodynamic characteristics, as well as radiative fluxes and surface atmospheric flow forcing. Therefore, the PBL-fluxes represent more accurately the land-surface atmosphere interaction. Here, the problem we seek to understand relates to the representation of the eddy-covariance fluxes in the PBL-flow and we seek to investigate if it is possible to dynamically upscale such fluxes on the basis of parameters characterizing the atmospheric surface layer turbulent state and the dynamic structure of the PBL.

To investigate this problem, we conducted an experiment on three consecutive summers from 2009 to 2011 in the Imnavait Basin at the foothills of the Brooks Range and the headwaters of the Kuparuk River Basin in the Alaskan Arctic tundra. In these experiments we have determined surface turbulent fluxes at two locations based on eddy covariance towers separated by about 1 km north to south and, we complemented these observations with boundary layer heat fluxes determined by a Large Aperture Scintillometer (LAS). To provide a comprehensive view of the flux integration across the heterogeneous landscape, the outputs of the Weather Research Forecasting (WRF) model surface fluxes at a 1 km scale are used in combination with the multiscale fluxes determination.

As a summary result of this experiment, we proposed two parameterization mechanisms to upscale fluxes: one of them is based on the Obukhov length and the other is based on the retrieval of the PBL-height. These mathematical representations of the dynamic upscaling are critically discussed and assessed in this contribution.