Evaluation of a Compact DIAL for Atmospheric Water Vapor Profiling at the ARM SGP Site

The lack of continuous observations of boundary layer thermodynamic profiles is a critical measurement gap that currently impedes the improvement of mesoscale predictive models. The U.S. National Research Council (NRC) has identified the need for a national observation network to fill this gap, with vertical humidity profiles being one of the highest priority measurements. In this study, we present field campaign results that demonstrate the performance of a compact Water Vapor Differential Absorption Lidar (DIAL) system prototype incorporating a ceilometer-type telescope design. The instrument uses semiconductor laser sources in the near infrared wavelength region and is designed to continuously report water vapor mixing ratio profiles in unattended all-weather operation. The purpose of the study is to assess the performance of the DIAL against several other measurement platforms in humid convective conditions. The evaluation was conducted at the U.S. Department of Energy’s Atmospheric Radiation Measurement (ARM) Southern Great Plains (SGP) atmospheric observatory in Oklahoma from 15 May to 12 June 2017. We compare the height-resolved water vapor mixing ratio measurements from the DIAL to those of a Raman Lidar, an Atmospheric Emitted Radiance Interferometer (AERI), and 83 radiosonde launches. During the campaign, the boundary layer mixing ratio varied between 4 g kg^-1 and 16 g kg^-1. There were around 27 distinct rain events, including five heavy rain events with maximum precipitation rates greater than 10 mm hr^-1. The DIAL system operated unattended and failure-free throughout the campaign. The maximum height for valid DIAL measurements was found to be about 1200 m, based on 50 % data availability. Comparing the DIAL to all radiosondes over the height range of 0 m to 1500 m, we found the DIAL measurements to be essentially unbiased (-0.01 g kg^-1) with a standard deviation of 0.68 g kg^-1, and linear correlation coefficient of 0.98. The comparison to Raman Lidar was also essentially unbiased with random uncertainty of 0.61 g kg^-1, and a correlation coefficient of 0.99.