Results of a campaign on the use of Lidar in an industrial area in the city of Cubatão, Brazil, are presented. Scattered light signal at different heights was correlated with air quality data monitored at ground level using multivariate techniques, aimed at identifying in a quantitative basis the similarities in the behavior of groups of variables. By using neural networks as a non-linear association method, a clear correlation was obtained between the Lidar scattered light signal and hourly-averaged ground-level ozone concentration.