Visualization of the wind drift of turbulent inhomogeneities

We suggest a method for filtration of incoherent optical images of objects in the atmosphere, which allows one to observe cross shifts of turbulent air inhomogeneities at the observation path. It is shown that the characteristic size of anisoplanatic distortions in images observed through the atmosphere is mainly determined by the distance to the layer of turbulent inhomogeneities that cause these distortions. An algorithm is suggested for the filtration of turbulent distortions by their characteristic size in order to mark out the effect of turbulent medium layers specified. The capabilities of the algorithm for cross wind retrieval from the analysis of a video sequence of short-exposure images at surface horizontal paths are shown. The main advantage of the suggested method for cross wind retrieval from the analysis of two neighbor frames of video sequence is its high speed since there is no need to accumulate statistics of the analyzed parameters. The retrieved wind speed is compared with data of acoustic anemometers located along an observation path 500 m long. The good accuracy of retrieval of the cross wind at path segments close to an observer, as well as of the average wind speed over the entire observation path, is demonstrated.