Winds on horizontal scans from Doppler lidar during T-REX

A study of the strong wind events over an alpine valley (Terrain Induced Rotor Experiment - TREX) is conducted using newly adopted statistical interpolation method to retrieve wind velocity vectors from Doppler lidar. This vector retrieval method has been adapted from radar data assimilation techniques. Since Doppler lidars measure only the radial velocity, obtaining wind vectors requires additional information, either through assumptions (such as in the Velocity Azimuthal Display, or VAD method) or through complimentary data (such as in dual-Doppler approaches). In the statistical interpolation approach, the two-dimensional non-isotropic form of the error covariance functions is derived for the radial and tangential components of the background error on the conical surfaces of a low elevation lidar scan. The conventional innovation method is used to estimate these functions statistically. Review of relevant characteristics of such methods and how they compare to the new method will be given, particularly with regard to their ability to preserve local information about the flow field. This study is meant to compliment previous wind retrievals in which vortices were educed and tracked on a vertical cross-barrier plane. In this study horizontal lidar scans will be studied to educe the horizontal structure of the flow field and independent data from other instruments will be used to evaluate our results.