Abstract:

Cirrus clouds play a significant role in the Earth’s radiation budget, and act as important controls on the distribution of water vapor in the upper troposphere and lower stratosphere. However, ice microphysical processes are observationally challenging to constrain. One major challenge in constraining ice growth models against observations is that experimental measurements are often challenging to interpret, and require assumptions about the underlying physical processes. This lack of clear physical understanding limits our ability to model ice crystal habits and growth rates in atmospheric models and to interpret in situ observations. While there has been an increasing availability of experimental and observational data from laboratory and in situ observations, different methods provide disparate sources of information. In situ observations from aircraft and balloon platforms provide detailed local information, but only at single snapshots in time, in limited sampling conditions, and without growth time-scales. Laboratory experiments allow ice growth to be monitored over time, but previous studies on depositional ice growth have found significant discrepancies between single crystal experiments and larger scale cloud chamber experiments that more realistically simulate atmospheric conditions. I will describe work from several recent studies investigating how observations from the AIDA Aerosol and Cloud Chamber can be used to derive constraints on ice microphysical models using cirrus simulation experiments in the pure ice regime (180 – 230 K) with both homogenous and heterogenous ice nuclei. I will discuss how methods such as physics-informed machine learning can be used to reduce parametric and structural uncertainty in ice microphysical process rates, without a priori assumptions about the ice growth models. I will also discuss how these experiments are used to characterize isotopic water as a tracer for cirrus processes, and the potential of these measurements for constraining atmospheric processes. 

Bio: Kara Lamb is a research scientist at Columbia University and at the NSF Learning the Earth with Artificial Intelligence and Physics (LEAP) Center at Columbia. Her current research lies at the intersection of observations (from laboratory and field studies) and high-resolution modeling, with the goal of better understanding how aerosols and clouds impact the climate. She combines traditional process-based approaches with data science and machine learning. She obtained her PhD in physics from the University of Chicago, and previously worked as a research scientist at CIRES/NOAA, where she was on the science team for the NASA KORUS-AQ and ATOM aircraft campaigns and the NOAA FIREX Firelab study.