Multiwavelength lidar: challenges of data acquisition and processing

Abstract

LIDAR operates by transmitting light pulses of few nanoseconds width into the atmosphere and receiving signals backscattered from different layers of aerosols and clouds from the atmosphere to derive vertical profiles of the physical and optical properties with good spatial resolution. The Data Acquisition System (DAS) of the LIDAR has to handle signals of wide dynamic range (of the order of 5 to 6 decades), and the data have to be sampled at high speeds (more than 10 MSPS) to get spatial resolution of few metre. This results in large amount of data to be collected in a short duration. The ground based Multiwavelength LIDAR built in Space Physics Laboratory, Vikram Sarabhai Space Centre, Trivandrum is capable of operating at four wavelengths namely 1064, 532, 355 and 266 nm with a PRF of 1 to 20 Hz. The LIDAR has been equipped with a computer controlled DAS. An Avalanche Photo Diode (APD) detector is used for the detection of return signal from different layers of atmosphere in 1064 nm channel. The signal is continuous in nature and is sampled and digitized at the required spatial resolution in the data acquisition window corresponding to the height region of 0 to 45 km. The return signal which is having wide dynamic range is handled by two fast, 12 bit A/D converters set to different full scale voltage ranges, and sampling upto 40 MSPS (corresponding to the range resolution of few metre). The other channels, namely 532, 355 and 266 nm are detected by Photo Multiplier Tubes (PMT), which have higher quantum efficiency at these wavelengths. The PMT output can be either continuous or discrete pulses depending upon the region of probing. Thick layers like clouds and dust generate continuous signal whereas molecular scattering from the higher altitude regions result in discrete signal pulses. The return signals are digitized using fast A/D converters (upto 40 MSPS) as well as counted using fast photon counters. The photon counting channels are capable of counting upto 200 MHz with a spatial resolution of few metres. The LIDAR data generated comes in burst mode and gets transferred to computer system. Pulse to pulse averaging is done rangebinwise for SNR improvement. The range normalized signal power is computed and the vertical profiles of backscatter and extinction coefficients are derived. This paper describes the intricacies in the design of the high resolution DAS developed in-house to obtain the scientific data. The optimization methodology used for handling the data is also described.