| Summary: | In atmospheric LIDAR (Light Detection and Ranging) remote sensing, the dynamic range of the return power signals can span up to five orders of magnitude. Modern acquisition systems such as those based on LicelTM transient recorders combine a dual acquisition mode in which the return signal is recorded simultaneously in both analog (analog-to-digital (AD) conversion) and photon-counting (PC) modes. Although both data records can be analyzed separately, their combination obtained through gluing gives the advantage of the high linearity of the AD conversion for high light-level signals (especially in the near range) and the high sensitivity of the PC mode for low light-level signals (in the far range). Recently, eruptions of volcanoes such as Eyjafjalla (Iceland) in 2010, Grímsvötn (Iceland Puyehue (Argentina) Nabro (Eritrea) and Mount Lokon (Indonesia) in 2011 have yielded emergency situations with a strong economical cost due to human evacuation and/or air traffic interruption. Volcanic aerosols are first injected in the troposphere and often reach the stratosphere where they can reside for several years. The mixed analog/photo-counting acquisition approach is particularly suitable for the detection of volcanic aerosols in both the troposphere (near/mid range) and the stratosphere (far range). In this line, a case study showing volcanic aerosols from the Nabro volcano (Eritrea, 2011) in the stratosphere and simultaneously Saharan dust in the troposphere over the multi-spectral Barcelona lidar station during the period 27 June – 1 July 2011 is presented. The benefits of the proposed gluing technique will be shown through the comparison of analog, PC, and glued backscatter-coefficient time series. Existing gluing algorithms solve the fitting coefficients by matching both analog and PC data over a predefined spatial range (or equivalently, over a predefined upper and lower counting rate) depending on which kind of photodetector (usually a photomultiplier) is being used. In this work, an enhanced data-gluing formulation is presented. The method automatically finds the spatial range where both analog and PC signals are more similar based on Euclidian distance minimization over piece-wise range intervals along the whole acquisition spatial range. Peer Reviewed Postprint (published version)
|
|---|