ESTIMATION OF THE ERROR CAUSED BY SMOOTHING ON DIAL MEASUREMENTS OF STRATOSPHERIC OZONE

Several ozone retrievals from DIAL (DIfferential Absorption Lidar) measurements have been proposed. Each retrieval method has its own definition of the range resolutions and therefore one cannot easily compare the ozone concentration errors resulting from the use of these different methods. In this paper, assuming the atmospheric model quoted from the U.S. Standard Atmosphere 1976 and the system constants of our lidar system, we numerically compared the differences between some representative DIAL methods by means of estimating the error caused by smoothing. Initially we defined a common range resolution as a scale. Using the common scale, we focused on the distortions arising from smoothing and the uncertainty of statistical signal errors. These estimates revealed that some algorithms reduce statistical error but make substantial distortions at higher altitude. It was possible to distinguish for each algorithm both the statistical errors and the systematic errors.