IEEE GEOSCIENCE AND REMOTE SENSING LETTERS, VOL. 5, NO. 2, APRIL 2008 285 Validating Mixed-Phase Cloud Optical Depth Retrieved From Infrared Observations With High Spectral Resolution Lidar David D. Turner and Edwin W. Eloranta Abstract—Single-layer mixed-phase clouds are prevalent in the Arctic atmosphere. The properties of mixed-phase clouds, includ- ing the optical depth of both the liquid and ice components, can be retrieved from spectrally resolved infrared radiance observations that are made in both the 8–13-μm and 17–24-μm windows. The accuracy of the retrieved properties from this algorithm has been established in single-phase clouds (i.e., clouds that contain only liquid or only ice) but not in mixed-phase clouds. A polarization- sensitive high spectral resolution lidar (HSRL) was deployed to the Atmospheric Radiation Measurement Program’s Barrow, Alaska site during the fall of 2004. The HSRL measures optical depth di- rectly, and the phase can be discriminated using the depolarization ratio measured by the lidar. Comparisons of the infrared retrieved optical depths with the optical depths directly observed by the li- dar in clouds that consist of supercooled liquid layers precipitating ice are in good agreement, with the slope and correlation being 1.055 and 0.65 for the ice portion of the mixed-phase cloud and 0.954 and 0.82 for the liquid portion. Index Terms—Atmospheric measurements, clouds, infrared measurements. I. I NTRODUCTION C LOUDS are an important modulator of the energy budget of the planet. The impact of clouds on the radiative fluxes, both at the surface and top of the atmosphere as well as the redistribution of the radiant energy in the atmosphere, depends first on whether a cloud is present, second on the fraction of the sky that is covered by clouds, and third by the phase and optical depth of the clouds. While there are numerous approaches to characterizing the optical depth of ice- or liquid-only clouds from ground-based remote sensors ([1] and [2], respectively), there are relatively few ground-based techniques for mixed- phase clouds [3]. Furthermore, these mixed-phase techniques are considerably younger than the single-phase techniques, and detailed evaluations of them are required. Here, we investigate the accuracy of the mixed-phase cloud retrieval algorithm (MIXCRA [4]), which retrieves the op- Manuscript received October 1, 2007; revised November 19, 2007. This work was supported by the U.S. Department of Energy, Office of Science, Office of Biological and Environmental Research, Environmental Sciences Division as part of the ARM program under Grants DE-FG02-06ER64167 and DE-FG02- 06ER64187. Construction of the HSRL was supported by the National Science Foundation under Grant OP-9910304. The authors are with the Space Science and Engineering Center, Univer- sity of Wisconsin—Madison, Madison, WI 53706 USA (e-mail: dturner@ ssec.wisc.edu). Digital Object Identifier 10.1109/LGRS.2008.915940 tical depth of both ice and liquid portions of single-layer mixed-phase clouds from spectrally resolved infrared radiance observations from the Atmospheric Emitted Radiance Inter- ferometer (AERI) using observations from a state-of-the-art high spectral resolution lidar (HSRL). This comparison was performed using the data collected during the Mixed-Phase Arctic Cloud Experiment (M-PACE [5]) at the Atmospheric Radiation Measurement (ARM) North Slope of Alaska (NSA) climate research facility in Barrow, AK [6], in the fall of 2004. The AERI is a facility instrument at the NSA site, and thus, the MIXCRA retrievals are available for multiple years from this site. The HSRL was only at the NSA site during the M-PACE. Thus, this validation effort will characterize the accuracy of the MIXCRA ice and liquid optical depths in mixed-phase clouds and, thus, provides confidence in the long record of MIXCRA values. Accurate long time series of cloud properties are critical in order to understand cloud-radiation feedback mechanisms, the impact of changing atmospheric and cloud properties on surface properties, and how the clouds in the Arctic are changing with time. II. METHODS A. MIXCRA The MIXCRA algorithm uses an optimal estimation-based approach to retrieve the optical depth of the liquid and ice components, along with the effective radius of the liquid and ice particles, from infrared radiance observations in single- layer mixed-phase clouds [4]. The differences in the absorption coefficients of ice and liquid water across the infrared spectrum, where ice is more absorbing than liquid at 12 μm and the opposite is true at 18 μm, allow the algorithm to discriminate between the two phases of the cloud [4], [7]. The primary input to the MIXCRA retrieval algorithm is the infrared radiance observations from the AERI. The AERI at the NSA site is a passive hardened interferometer that measures downwelling infrared radiance at 1-cm -1 resolution from 400 to 3000 cm -1 (25–3.3 μm). Two well-characterized blackbod- ies, as well as corrections for nonlinearity of the detector and instrument self-apodization, yield radiance observations that are accurate to better than 1% of the ambient radiance [8], [9]. During the M-PACE, the AERI provided a 12-s averaged sky radiance spectrum every 25 s, with periodic gaps of less than 1 min when the instrument was viewing the calibration 1545-598X/$25.00 © 2008 IEEE