Use of radar Doppler spectra in Arctic mixed-phase cloud studies

Open Access
Yu, Guo
Graduate Program:
Doctor of Philosophy
Document Type:
Date of Defense:
September 09, 2013
Committee Members:
  • Johannes Verlinde, Dissertation Advisor
  • Johannes Verlinde, Committee Chair
  • Eugene Edmund Clothiaux, Committee Member
  • Jerry Y Harrington, Committee Member
  • Kultegin Aydin, Committee Member
  • Runze Li, Committee Member
  • mixed-phase cloud
  • phase partitioning
  • radar Doppler spectra
Stratiform clouds across the globe frequently contain both liquid- and ice-particles within the same region of the atmosphere; that is, they are often mixed-phase clouds. Modeling the processes of mixed-phase stratiform clouds remains a challenge. Progress is slow in part because of the dearth of observations that shed light on critical processes in these clouds. Recent availability of high quality millimeter-wave cloud radar Doppler spectra is an important new source of information on the micro-physical and dynamical processes within these clouds. In the current study a radar Doppler spectrum simulator is used to produce in-cloud radar Doppler spectra from outputs of large eddy simulations with detailed bin microphysics. These model-simulated radar Doppler spectra are compared with observed radar Doppler spectra for the case study period via comparisons of reflectivity, mean Doppler velocity, and spectrum width, as well as estimates of volume-mean vertical air velocities and hydrometeor fall speeds extracted from the spectra. The results indicate that there is a mismatch in the important processes governing model and observed mixed-phase clouds, including cloud vertical air motion structures, ice nucleation, ice aggregation, and how the model increases ice particle fall speeds as particles gain mass. In order to improve mixed-phase cloud parameterizations in models, specific observations to constrain model parameters are required. Retrieving and quantifying the radiatively important liquid-phase particles within mixed-phase clouds remains a challenge because the radar signal is frequently dominated by the returns from the ice particles within these clouds. The ice masks the small reflectivity contributions from the liquid phase. To extract these small liquid-phase contributions a spectral deconvolution algorithm is developed that separates them from MMCR/KAZR Doppler-velocity spectra in which ice-particle returns dominate. In this approach spectra are first decomposed using a continuous wavelet transform. The resulting coefficients are then used to identify regions in the spectra where cloud-liquid drops contribute; Gaussian distributions are subsequently fit to these regions. Results from this algorithm indicate that this approach is capable of separating the cloud-liquid drop and ice-particle contributions to the radar Doppler spectra. In the process volume-mean vertical air velocities, turbulent broadening and mean ice-particle fall speeds are also extracted. Statistics of the macro-physical, micro-physical, thermodynamical, and dynamical properties of mixed-phase clouds are characterized using a month of ground-based measurements collected during October 2011 at Barrow, Alaska. Cloud layer temperature is the primary factor that determines mixed-phase cloud properties. One-hour mixed-phase cloud events are classified into cold and warm cases. Clouds in category 1 (warm) were lower, thinner and had weaker and liquid-dominated precipitation, while clouds in category 2 (cold) were higher, thicker and had more intense and ice-dominated precipitation. Precipitation reflectivities and reflectivity-weighted mean precipitation fall speeds are found to increase with vertical air motion. Ice particles generally form, grow and fall out of the cloud layer in updraft regions. Ice precipitation processes also play a role in determining the macro-physical properties and liquid microphysics of these clouds. The newly developed spectral deconvolution algorithm enables us to investigate relationships between cloud macro-physical, micro-physical, thermodynamical and dynamical properties based on long-term observations of mixed-phase clouds. Studies on these relationships derived from long-term observation will strength our understanding of the physical processes in Arctic mixed-phase clouds, which is a necessary to develop more sophisticated parameterizations to partition cloud phases in climate models.