Velocity retrieval for coherent Doppler lidar

Coherent Doppler light detection and ranging (lidar) measures a component of atmospheric wind motion in the direction of the lidar beam. However, velocity vectors are needed in order to obtain more complete knowledge of flow fields in the atmospheric boundary layer (ABL). Several methods such as the velocity azimuthal display (VAD) and dual-Doppler approaches exist to retrieve two-dimensional vector fields, but these methods tend to lose local information about the flow field due to the averaging involved. A two-dimensional vector retrieval method has been adapted from radar data assimilation techniques. The retrievals are performed on data from the Joint Urban 2003 (JU2003) experiment and are shown to retain the local velocity information which is lost in a method like the VAD. A comparison with dual-Doppler retrievals shows reasonable agreement. Furthermore, radial velocities computed using the projections of the retrieved vectors from one lidar and the look direction from the other lidar compare well with the measured radial velocity from the other lidar.