Variation in mass-dimensional relationships for mixed phase Arctic clouds

Open Access
Author:
Van Der Horn, Jennifer Lauren
Graduate Program:
Meteorology
Degree:
Master of Science
Document Type:
Master Thesis
Date of Defense:
April 01, 2014
Committee Members:
  • Jerry Y Harrington, Thesis Advisor
Keywords:
  • mass dimensional relationships
  • mixed phase clouds
  • parcel model
Abstract:
Discerning if a simple relationship exists between the mass-dimensional (m-D) coefficients found by fitting in situ ice crystal measurements and those needed in ice growth models is explored. Models use m-D relationships as a method of relating changes in mass to changes in size. The m-D relationships in models often come from observational studies. The relationship is a power law that relates the mass to the maximum dimension of the crystal with two coefficients, am and bm, often treated as constants. The relationships from in situ measurements include the entire growth history of the crystals and have large standard deviations in the fit coefficients. The leading coefficient, am, depends on the density of the crystal and therefore has a higher standard deviation and range than the power coefficient, bm. The bm coefficient depends on the temperature of the crystal and the crystal’s habit. Ice crystals were collected in Barrow, Alaska during two separate campaigns in the fall and spring of 2012. An image-processing algorithm was developed in order to automate the crystal analysis and classification from digital photographs. Over 30,000 crystals were classified by habit and analyzed in order to create m-D relationships. Meteorological influences, specifically average cloud temperature and thickness, were investigated for possible influences on the m-D relationships and the standard deviations of the coefficients, but none were found. Separating crystals by habit provided the highest correlation in the power law fits to the data. The spring and fall provided two starkly different cases in which the former had mostly pristine crystals while the latter had significant riming. This difference between the cases allowed for the significance of riming on m-D relationship coefficients to be examined. To explore the physical connections between m-D coefficients determined from fitting and the coefficients for individual ice particles required by growth theories, a model that predicts mass, aspect ratio, density, and size was used. This ice growth model was embedded in a Lagrangian parcel model and then utilized to examine the evolution of the m-D coefficients. The parcel model was forced with input derived from 2000 trajectories derived from Large Eddy Simulation output. This provided relatively realistic parcel trajectories. Consequently, the m-D coefficients can be predicted in time with the model. The m-D coefficients were calculated in time and compared to the coefficients derived from fitting the model similarly to what is done with observations. The model yielded predicted coefficients that were higher than those seen in several observational studies and those derived from fitting the model output, especially for the leading coefficient, am. When the results of the mass and crystal maximum dimension produced by the model were plotted and then fitted to determine the m-D relationship, results of the coefficients were similar to the observational cases. No simple relation was found between the model predicted and fitted coefficient. Further studies will need to determine whether predicted and fitted m-D coefficients can be related to each other. Because the leading coefficient depends on density, finding an analytical method to relate the density to particle size may provide the necessary link.