ALGORITHM THEORETICAL BASIS DOCUMENT (ATBD) FOR A-FeatureMask-KNMI

This algorithm is referred to as A-FeatureMask-KNMI. It deals with the lidar-only feature mask based on signal to noise ratios and neighbouring pixel information using image reconstruction techniques.
The feature mask is an important step in the retrieval of higher order products like extinction and target classification. In the flow-diagram [Figure 3-1] the rational of the ATLID only retrievals is presented from L1b data to the retrieval of cloud and aerosol properties.
The A-Featuremask is the first algorithm to be applied to the L1b ATLID data and retrieves a feature mask on the single shot grid. The single shot mask is read in by the extinction retrieval algorithm together with the original L1b data. The combination of the signals, retrieved extinction and backscatter and feature mask can finally be used as an input to the higher order retrievals: target classification and cloud and aerosol properties. The following sections will refer only to the A-FeatureMask-KNMI.