Surface wind variability due to convection over land

Numerical weather prediction models predict the grid-averaged wind vector. However, the average wind speed is needed to calculate the turbulent surface fluxes. The average wind speed can be obtained from the magnitude of the average wind vector and the wind gustiness. The wind gustiness is due to subgrid-scale variability of the wind vector. For the clear convective boundary layer, the wind gustiness is proportional to the convective velocity scale. Wind gustines is also produced by deep precipitating convection, and has been parameterized as a function of convective rainfall rate. However, such parameterizations have been developed and tested only for tropical oceanic deep convection. We are using observations of wind gustiness from the Oklahoma Mesonet to evaluate the applicability of such oceanic parameterizations for use over land. We are also using the Oklahoma Mesonet observations to evaluate the ability of a cloud-resolving model to reproduce the observed wind gustiness for the DOE Atmospheric Radiation Measurement program's Summer 1997 Single-Column Model Intensive Observation Period at the Southern Great Plains site in Oklahoma.

Knowledge of other aspects of surface wind variability are required for fire weather forecasting (e.g., wind vector variability) and for estimating peak gusts. Our longer-term goal is to compile a climatology of surface wind variability statistics from the Oklahoma Mesonet observations to provide a test bed for evaluating parameterizations of subgrid-scale wind variability for such applications.