HSRL measurements of Lidar ratios in the presence of oriented ice crystals

HSRL measurements of Lidar ratios in the presence of oriented ice crystals.

Ed Eloranta, Ilya Razenkov, Joseph Garcia, University of Wisconsin—Madison

eloranta@wisc.edu

Aerodynamic forces can create preferred orientations in falling ice crystals; often producing dramatic effects on the observed lidar return[1,2,3]. Microphysical cloud models must consider orientation because crystals take on an orientation that maximizes aerodynamic drag. This alters the evolution of ice particle size, shape and precipitation rate as lower fall speeds extend crystal residence times in the cloud.

This paper reports High Spectral Resolution Lidar (HSRL)[4] measurements of backscatter cross-section, extinction cross-section, and depolarization made while scanning within 20 degrees of the zenith. These observations are sensitive to particle orientation and provide direct measurements of the lidar ratio. The HSRL is particularly suited to this task because it provides robustly calibrated values of particulate scattering separated from molecular scattering and these observations are independent of extinction below the measurement altitude.

The upper panels of figure 1 show the high backscatter cross-section and low depolarization feature observed at zenith when oriented crystals are present. The low depolarization observed at zenith makes it likely that ice clouds will be misidentified as water. For this reason, most cloud sensing lidars are pointed slightly off vertical (The University of Wisconsin HSRL’s are typically operated at 4-degrees off-zenith). In figure 1, the depolarization remains well below typical (~35%) ice cloud values throughout the scan range. This implies that much of the return signal is a result of specular reflection. Because specular reflection occurs from crystal faces oriented perpendicular to the lidar; some of the crystals must be oscillating or tumbling. The small values of depolarization may cause misidentification of the cloud phase even when the lidar is pointed off-zenith.

Figure 2 shows the backscatter peak caused by specular reflection from horizontally oriented ‘plate like’ crystals. Backscatter variations with angle are not confined to the narrow peak at the zenith; the cross-section continues to decrease throughout the scan range. The oriented faces that direct light back to the lidar when it is pointing at the zenith, direct light away from the lidar at off-zenith angles and decrease the measured cross-section. The thin black line shows the extinction cross-section computed as a function of angle; this is somewhat noisy due to short integration times. We expect the extinction cross-section to vary with the projected area of the crystals. This should show a cosine dependence on zenith angle that produces little variation over a 20-degree zenith angle range. As expected, the extinction cross-section shows little angle dependence and it can be replaced with the median value plotted in green.

Multiple scattering contributions cancel out to first order in HSRL measurements of the backscatter cross-section. However, extinction cross-section values derived from the attenuation of the molecular HSRL signal will be biased low. The left-hand axis of figure 3 shows the lidar ratio computed from the direct measurement of molecular attenuation. The right-hand axis shows value computed from a multiple scattering corrected[6] attenuation. Ice particles sizes of 1 mm were assumed for the multiple scatter correction because smaller particles maintain nearly constant orientation [5] and would not exhibit low depolarization for off-zenith angles. Using larger assumed sizes would not significantly change the correction; it becomes nearly constant as the size is increased beyond 1 mm.

It is noteworthy, that many cirrus clouds and nearly all ice crystal virga layers and snowfall events show some evidence of orientation. Small regions with very strong orientation effects are often embedded in regions of reduced effect. Thin layers with off-zenith lidar ratios greater than 100 were occasionally observed. In layers with strong zenith backscatter peaks, the depolarization is often suppressed well below the typical values of ~35%. This suppression existed at all angles less than our maximum scan angle of 20-degrees off-zenith.

References:

1. Platt, C., Abshire, N., McNice, G.,1978, Some Microphysical Properties of an Ice Cloud from Lidar Observation of Horizontally Oriented Crystals, App. Met., 17,1220-1224
2. Sassen and S. Benson, 2001: A midlatitude cirrus cloud climatology from the Facility for Atmospheric Remote Sensing: II. Microphysical properties derived from lidar depolarization, J. Atmos. Sci. 58, 2103–2112.
3. Noel, V. G. Roy,L. Bissonnette , H. Chepfer and P. Flamant (2002): Analysis of lidar measurements of ice clouds at multiple incidence angles, Res. Let., vol. 29, no. 9, 1338
4. Eloranta, E., 2005: High Spectral Resolution Lidar in Lidar: Range–Resolved Optical Remote Sensing of the Atmosphere, C. Weitkamp, ed. (Springer), pp. 143–164.
5. Cheng, P. K. Wang, T. Hashino, 2015: A Numerical Study on the Attitudes and Aerodynamics of Freely Falling Hexagonal Ice Plates, J. Atmos. Sci., 72, 3685–3698.
6. Eloranta, E. 1998: A practical model for the calculation of multiply scattered lidar returns, Appl. Opt. 37, 2464 –2472 .