A project aims to improve how ice cloud microphysics is represented in models

A wall of clouds at the ARM Southern Great Plains observatory during the MC3E field campaign, a main source of data for an ongoing ASR project on how to better represent ice cloud microphysics in models.
A wall of clouds at the ARM Southern Great Plains observatory during the MC3E field campaign, a main source of data for an ongoing ASR project on how to better represent ice cloud microphysics in models.

The clouds we see every day are composed of water and ice. Microphysical processes govern the many ways cloud-making water drops and ice particles grow, shrink, stick together, break up, and fall. In turn, cloud microphysics influence the atmospheric dynamics that create clouds and storms.

These “micro” processes have a “macro” influence on storm systems, weather patterns, and—collectively, and by extension—on Earth’s environment.

Ice in clouds is very important, but also presents special research challenges.

“Ice particles can strongly affect long-wave radiation from the Earth and incoming solar radiation,” says Hugh Morrison, a scientist at the National Center for Atmospheric Research (NCAR) in Boulder, Colorado. Yet the dynamism and variability of the ice in clouds are not well known, nor well represented in earth system models, he says, in part “because of all the shapes ice particles can take.”

Compared to water droplets, ice particles often have a complex morphology. They take on a "fun house" variety of shapes that reflect and absorb light in hard-to-measure ways.

A Project on Ice

Morrison is principal investigator in a three-year project to improve the way ice microphysics is represented in cloud and earth system models. The title, for climate initiates, sums it up: “Developing and Testing a Novel Stochastic Ice Microphysics Parameterization for Cloud and Climate Models Using ARM Field Campaign Data.”

“Stochastic” refers to processes that seem to change in random ways that cannot be exactly predicted, but can be statistically analyzed.

In his NCAR office, Hugh Morrison works on candidate stochastic schemes for ice microphysics parameterization. Image courtesy of Hugh Morrison.
In his NCAR office, Hugh Morrison works on candidate stochastic schemes for ice microphysics parameterization. Image courtesy of Hugh Morrison.

The project is funded by ASR, a U.S. Department of Energy (DOE) program designed to address uncertainties in earth system simulations. “ARM” stands for the Atmospheric Radiation Measurement (ARM) Climate Research Facility, a DOE user facility that oversees numerous data-generating fixed and mobile atmospheric observatories.

One of them is the 9,000-square-mile Southern Great Plains (SGP) observatory in Oklahoma and Kansas. It is the main source of data for the ASR project, which for now is focused on midlatitude convection. In particular, the project researchers will test their new stochastic schemes with data from the 2011 Midlatitude Continental Convective Clouds Experiment (MC3E), which was sited at the SGP.

“Ultimately, we want to look at all of the ARM in situ observations,” says co-investigator Greg McFarquhar of the University of Oklahoma.

Adding data from different conditions and locations, he says, will help researchers determine how cloud parameters vary with environmental parameters like temperature, vertical velocity, humidity, and cloud depth.

Joining Morrison and McFarquhar in the project are co-investigators Adam Varble and Edward Zipser, University of Utah; Wojciech Grabowski, NCAR; and Junshik Um, University of Oklahoma.

There are postdoctoral fellows taking part too, along with NCAR scientist Judith Berner, a collaborator who has done extensive work on stochastic frameworks in earth system models.

Why a Scheme for Ice

By 2019, the project will produce a stochastic framework for ice microphysics parameterization ready to be plugged into a model. That includes the conceptual idea behind such a framework, including algorithms, math, and code, says Morrison. “What we’re doing here is pretty new, and we’d like to get the word out to the scientific community.”

Before long, the project will also engender peer-reviewed papers (three are underway), though the project is just 18 months old—early for published work.

These planned endpoints put the project in alignment with one ASR mission, which is “to improve the scientific understanding and treatment of clouds in atmospheric models,” says McFarquhar.

An observationalist from the start, Greg McFarquhar has spent a lot of time in the field. Here he is in Hobart, Australia in 2017, during the run-up to the Measurements of Aerosols, Radiation, and Clouds over the Southern Ocean (MARCUS) field campaign.
An observationalist from the start, Greg McFarquhar has spent a lot of time in the field. Here he is in Hobart, Australia in 2017, during the run-up to the Measurements of Aerosols, Radiation, and Clouds over the Southern Ocean (MARCUS) field campaign.

The work will bring models closer to the important (but little understood) place ice occupies in clouds.

“Ice plays a critical role in weather and climate models,” he says. “But its representation is highly uncertain—in large part because of the wide range of ice particle characteristics controlled by parameters that are typically held fixed in models.”

Other researchers are looking at stochastic microphysics parameterization, says Morrison, but the ASR project is the only one using in situ aircraft observations “to constrain what this parameter variability is.”

The goal is to have parameters for ice microphysics that will be “allowed to vary over a range of possible values,” says McFarquhar, by applying novel analytical techniques to the MC3E data.

For five years or more, he adds, his own research group has developed techniques to represent variable uncertainties in cloud microphysical properties. Past schemes have allowed parameters like particle mass and fall velocity to remain fixed—not vary, as observations show.

“We know it’s real,” says Morrison. “We know this variability exists.”

How big a difference will capturing this parameter variability make to modeling? Both Morrison and McFarquhar caution that their proposed work will have to be finished before they really know. Candidate modeling programs are being developed, tested, refreshed with data, and redeveloped in an ever-refining loop of collaborative effort.

Ice and Earth Systems

Meanwhile, there is certainty on one score: that ice microphysical properties wield important influences on earth systems.

Morrison earlier mentioned effects on solar and earth radiation.

He adds that ice microphysics influences how precipitation is generated, making these processes important for the hydrological cycle.

“A large fraction of the Earth’s precipitation starts as ice,” says Morrison.

Ice microphysics also affects storm dynamics, including the intensity of winds and storms. The details of midlatitude deep convective systems rolling over the SGP are captured in MC3E data.

“Ice has different thermodynamics than liquid drops,” says Morrison. “And also, ice particles have weight, causing ‘drag’ in the clouds.”

High up in storms the water mass is dominated by ice condensate. That affects cloud buoyancy, weighing clouds down and dampening the vertical velocity within.

As the grid size of earth system models gets smaller, the resolution of those grids gets higher and more precise. All the more reason, says Morrison, “to represent ice microphysics well in models.”

As the project glides along, with periodic check-ins via Skype, every team of experts has an interlocking role. McFarquhar’s group assesses observational data. The NCAR group develops candidate variations of the stochastic scheme. And Varble’s Utah group runs and tests them.

“The expertise,” says Morrison, “fits together like a puzzle.”

# # #

This work was supported by the U.S. Department of Energy's Office of Science, Office of Biological and Environmental Research as part of the Atmospheric System Research Program.