Impact of Detection Anomalies on Cloud Depolarization Ratios from the Raman Lidar

Description

The ARM Raman lidars include multiple channels for the detection of elastic and inelastic (i.e. Raman) backscatter from the atmosphere. The elastic detection channels include a wide-field-of-view (WFOV) unpolarized channel, a narrow-field-of-view (NFOV) co-polarization channel, and a NFOV cross-polarization channel. For each detection channel, simultaneous measurements are made of both analog photomultiplier current and photon counts using Licel data recorders. The analog signal exhibits good linearity in the presence of strong backscatter, but poor sensitivity in the weak signal regime. The photo counting signal, on the other hand, exhibits excellent sensitivity in the weak signal regime, but poor linearity due to pulse pile-up effects. The so-called MERGE value-added-product (VAP) optimally combines the analog and photon counting signals into a single signal with improved dynamic range through a process referred to as “gluing”. All other Raman lidar data products (e.g. water vapor, temperature, aerosol backscatter, etc …) are computed using the output from the MERGE VAP. The gluing process works well when the relationship between analog (A) and photon counting (PC) sample pairs is well behaved. This is usually the case for all of the inelastic channels and for the elastic channels under clear-sky conditions. Under cloudy conditions, however, A-PC sample pairs from the elastic channels exhibit substantial scatter. Currently, the MERGE VAP makes no effort to identify and flag anomalous A-PC sample pairs. In this poster, we describe the anomalies and examine their impact on estimates of cloud depolarization ratio. It is proposed that a quality control (QC) flag variable be incorporated into the MERGE VAP to identify invalid photon counting data, and that downstream processes (e.g. water vapor, temperature, aerosol backscatter, etc…) be modified to use the new QC flag.