Date of Award

Spring 1-1-2013

Document Type


Degree Name

Doctor of Philosophy (PhD)


Atmospheric & Oceanic Sciences

First Advisor

Linnea M. Avallone

Second Advisor

John J. Cassano

Third Advisor

Martina Kraemer

Fourth Advisor

Shelly L. Miller

Fifth Advisor

Darin W. Toohey


Ice clouds are known to be major contributors to radiative forcing in the atmosphere, yet understanding and describing their microphysical properties remains challenging. The ice water content (IWC) of cirrus clouds is of particular interest because it can be directly related to a number of other radiatively important variables such as extinction and effective radius. This dissertation explores data obtained from numerous airborne research campaigns to elucidate the dominant factors that determine the IWC and crystal concentration (Ni) characterizing cirrus clouds.

The first part of this study uses a new merged dataset to obtain a representation of the IWC-temperature relationship using probability distribution functions (PDFs), which allows for the variability of IWC within a temperature band to be visualized. The IWC-PDFs are observed to be bimodal over the whole cirrus temperature range. This bimodality is also found in Ni-PDFs and might be attributed to different cirrus formation mechanisms such as heterogeneous and homogeneous freezing.

Next, two midlatitude datasets (MACPEX and MidCiX) with differing IWC and Ni distributions are used to explore how cloud formation processes contribute to the resulting cirrus characteristics. Starting with large-scale processes, a case study analysis demonstrates the link between synoptic-scale dynamics and the differences observed between the MACPEX and MidCiX datasets. However, that link is not always clear, indicating that additional information is required to explain it. This is particularly evident when considering the Ni observations from MACPEX, which appear unaffected by factors such as vertical velocity. This result is unexpected considering the relationship between vertical velocity and nucleation processes.

The final component of this analysis explores the cirrus formation processes at the nucleation level, while continuing to examine why the data in MACPEX and MidCiX are different from each other. A comparison to model simulations reveals that heterogeneous nucleation and subsequent homogeneous nucleation suppression are responsible for the Ni distribution in MACPEX. Conversely, the MidCiX IWC and Ni datasets exhibit a compatibility with model results showing heterogeneous nucleation followed by a secondary homogeneous nucleation event. Thus, it is likely that the differences observed in MACPEX and MidCiX mostly stem from differences in ice nuclei concentrations.