Spatial and temporal footprint analysis for latent energy flux mapping

Latent energy (LE) flux map, derived from lidar data, estimated by Monin-Obukhov similarity assume an ergodic turbulent field of infinite extent. In practice however, the inherent assumptions required for similarity theory are not met in the field. In order for flux estimates to be made using similarity theory, time and space averages of atmospheric quantities are made with the understanding that the fluxes originate from discrete regions on the surface. The size of these regions, footprints, are determined in part by the temporal and spatial averaging of the atmospheric parameters. Lidar derived flux maps are dependent upon proper sampling size; this work evaluates how vertical and horizontal sampling size affects the quality of lidar derived LE fluxes.

Lidar spatial sample size was evaluated by finding the optimal spatial lag via flux discrepancy. The flux discrepancy is a non-dimensional value computed as the ratio of the lidar estimated LE flux to that of eddy covariance tower LE observations. The lidar estimated flux discrepancy was computed with spatial averaging distances ranging from 5 m by 5 m to 300 m by 300 m horizontally and from 0.5 m to 20 m vertically. When the flux discrepancy is minimized, the averaging distance in either horizontal or vertical dimensions is considered optimal. Footprint analysis using analytical and/or Lagrangian models is also presented and compared to the "optimal" spatial averaging distances derived from lidar data.