Back to search

Georgia Institute of Technology

Multi-scale modeling of ice crystal formation in clouds

Abstract

dc:description.abstract

Understanding human impact on climate is the foremost challenge of the 21st century. In particular, significant work remains to explain how anthropogenic emissions affect cloud properties and how cloud properties, in turn, affect climate. This thesis addresses this need by studying ice formation via primary nucleation and mechanical, secondary production at various model scales. A gradient model of several ice nucleation codes is constructed with automatic differentiation tools to understand the sensitivity of in-cloud ice crystal number to different input variables. The sensitivities output from this gradient model are used to investigate nucleation efficiency and regime disparities between various formulations. More efficient and frequent heterogeneous ice nucleation is predicted by laboratory data-based formulations. The sensitivities from this gradient model are also used for attribution analysis to identify the inputs that control temporal variability in output nucleated ice crystal number. We determine that input vertical velocity is a crucial factor in global models that parameterize turbulence and recommend that it be better constrained with additional measurements. After development of these tools with global-scale insight, work shifts focus to a smaller scale, and a zero-dimensional parcel model and mesoscale parameterizations for secondary ice production are constructed. The parcel model describes the processes of breakup upon graupel collision and rime splintering with six hydrometeor classes and coupled, time-delay hydrometeor number tendencies. Simulations are run to explore the impact of initial conditions, graupel non-sphericity, and updraft stochasticity on secondary ice production processes. Then empirically constrained codes are written to describe breakup upon ice hydrometeor collision and frozen droplet shattering in a more sophisticated, mesoscale meteorological model. We use these new parameterizations to investigate how additional in-cloud ice production may affect surface precipitation intensity in a cold frontal rain band case study. Finally ice nucleation and secondary production are brought together with estimates of the number of preexisting ice crystals needed to trigger secondary production. We find that a very low bound, suggesting that initial droplet formation may be more important to subsequent cloud phase partitioning. An outlook is included with ideas to evaluate cloud parameterizations with information theory and to construct more complete cloud-hydroclimate feedbacks.

Degree

thesis:*
Level thesis:degree_level
Doctoral
Department dc:contributor.department
Chemical and Biomolecular Engineering
Grantor dc:publisher
Georgia Institute of Technology
Year dc:date.issued
2017

Author and committee

dc:creator, dc:contributor.*
Author dc:creator
  • Sullivan, Sylvia Camille
Advisor dc:contributor.advisor
  • Nenes, Athanasios
Committee members dc:contributor.committeemember
  • Rousseau, Ronald W.
  • Hoose, Corinna
  • Deng, Yi
  • Styczynski, Mark P.

Subjects

dc:subject × 4

Rights

Language dc:language.iso
en_US

Identifiers

dc:identifier.*
Handle dc:identifier.uri
http://hdl.handle.net/1853/60136

Chain of custody

source
Harvested from
Georgia Tech
Base URL
repository.gatech.edu/server/oai/request
Last updated
2026-11-19
Source record
OAI-PMH GetRecord
citation

Sullivan, Sylvia Camille. Multi-scale modeling of ice crystal formation in clouds. Doctoral thesis, Georgia Institute of Technology, 2017. http://hdl.handle.net/1853/60136