Using the Optimal Estimation Method (OEM) for Retrieval of Stratospheric Ozone Profiles from DIAL lidar measurements
Abstract
The Optimal Estimation Method (OEM) is an inverse method that allows the retrieval of parameters based on measurements and a forward model of the measurement. A complete uncertainty budget on a profile to profile basis, plus the vertical resolution of the measurements as a function of height can be found by this method. We use OEM for the first time to retrieval ozone profiles from a DIAL ozone lidar. The retrievals will be used on measurements from the CANDAC Stratospheric Ozone Differential Absorption Lidar located in Eureka, Canada. We will show results for simulated measurements using one Rayleigh channel. The synthetic pro?les are similar to 3 hours of the real measurements. The ozone is retrieved at 900 m vertical resolution from 7 km to 55 km altitude. The averaging kernel shows essentially no contribution from the a priori below 40 km; above this altitude the response of the averaging kernel drops to 0.8 around 45 km; above which height the retrieval becomes less sensitive to the measurements. The percentage error between the true and retrieved profiles varies between 0.5% to 2% in the region where the retrieval is valid and is less than the statistical uncertainties. In this pilot study background counts are also retrieved. A constant background was used to make synthetic measurements, however, due to the Signal Induced Noise (SIN), the background counts are not constant. We will include the effect of SIN offset on the background counts in the near future. The retrieval is currently being extended to use both of the lidar data channels. Using the two channels, we are planning to retrieve the ozone density profile, the aerosol extinction coefficient, deadtime of the detectors, and the lidar constants. We will then validate the method using measurements of ozone from other instruments, as well as against the traditional DIAL ozone analysis.