Horizontal diffusion by submeso motions in the stable boundary layer

摘要

Four networks of wind data are used to construct the first systematic estimates of the horizontal diffusivity from observations of submeso motions on scales often unresolved in numerical models. Currently, the horizontal diffusivity in numerical models is specified mainly for numerical reasons without observational support. The data analysis in this study emphasizes the stable boundary layer although results are briefly presented for the unstable boundary layer. The horizontal diffusivity is estimated from the horizontal gradient and the observed flux. Horizontal gradients of scalars are generally difficult to directly estimate from observations with sufficient accuracy for much of the data. As an alternative, simulated particles with conservative properties are introduced into the observed wind field in order to estimate the horizontal diffusivity for submeso motions. The sensitivity of the horizontal diffusivity to details of the method is examined. The horizontal diffusivity increases with the range of time and space scales that are included in the evaluation. The horizontal diffusivity is much larger with significant topography and may increase with wind speed, depending on the site location. The coarse station spacing or the small domain size is found to be a major limitation to the analysis.