The sky outside was ominously dark, threatening to burst into rain at any minute — providing a perfect backdrop for a discussion about thunderstorms with Marcus van Lier-Walqui via Zoom.
Van Lier-Walqui studies clouds and precipitation at Columbia Climate School’s Center for Climate Systems Research and NASA’s Goddard Institute for Space Studies. He focuses on the complex interactions of tiny particles and microphysics inside of clouds that can lead to rain, hail, thunderstorms, and tornadoes.
On stormy days like this one, he still feels a visceral excitement — “I still think, ‘Oh cool, lightning!’” he says — but he also keeps an eye on the weather apps. A tablet computer displaying radar data makes a brief appearance in the Zoom screen.
Although he typically works with climate and weather models on a computer, van Lier-Walqui recently got back from fieldwork near Houston, Texas, where he is a co-investigator on a project funded by the Department of Energy that is deploying radar, lidar, radiometers, aerosol measurements, drones, and a lightning mapping array to study thunderstorms. The project, called TRacking Aerosol Convection interactions ExpeRiment (TRACER), aims to understand how thunderstorms form, and how particles from air pollution may affect their strength and lifecycle, in order to improve weather and climate forecasting.
Some research has suggested that air pollution makes storms stronger, but it’s a source of contentious debate among scientists.
In the following Q&A, which has been edited for length and clarity, van Lier-Walqui shares more about the project and how he’s involved.
What are the big questions TRACER is trying to answer?
We’re looking at how do aerosols — so, dust and pollution particles — affect how thunderstorms behave and evolve? The Houston region is a great place to study this, because you have this relatively clean air flowing from the Gulf, and you have a huge pollution source in Houston’s urban area and all of the companies that refine petrochemicals there, so we can compare storms that form in polluted versus less polluted air.
Our project is geared around getting a more detailed sense of what’s happening inside real thunderstorms, and trying to come up with the best model possible to compare that to. And then looking at where the differences are and how that might tell us what aspects we don’t understand as well as we think we do, and where we can get improvement.
Why don’t we understand the role of aerosols already?
A lot of advanced radar work has focused on tornadic storms, because they cause so much damage, as opposed to simpler and gentler thunderstorms, and how their vertical motions develop and interact with cloud and precipitation particles. There’s been less focus on just getting observational insight into the nitty gritty details of how ice particles interact as they grow and collide with other particles in the atmosphere and how that affects, say, lightning and electrification and the subsequent evolution of the storm system.
Could studying regular thunderstorms aid our understanding of tornadoes as well?
Absolutely, there should be that knock-on impact. The better we can understand microphysics — basically how cloud droplets grow, how they collide to form rain, how liquid particles start to freeze — the better we can make our forecast models. And ideally, the better we can make severe weather warnings.
One of the things about microphysics is that when you get condensation of water vapor into liquid, you release energy. When you get freezing of liquid into ice, you release energy. And it’s that energy that drives these storms and drives the strong circulations associated with them. So you absolutely have to get these things right in the models.
What’s your role in the TRACER project?
Together with Eric Bruning and Kelcy Brunner from Texas Tech University, we proposed to deploy additional lightning mapping array sensors. These antennas basically detect the electrical signal that’s produced by lightning discharges. You have a number of these located all around a region and they can pinpoint the exact location in space and time that the discharge is happening. Even when it’s obscured by clouds, you can actually see through the clouds and see exactly how that branch arced out.
Lightning is tied to the microphysics inside the storm — what size ice particles are being produced, how many are being produced — which relates to the charging rate of the atmosphere and how much lightning is actually produced.
Some people have had this hypothesis that aerosols will make thunderstorms much stronger. There’s been a lot of contentious debate about this topic. My colleagues found out that although the aerosols don’t actually affect the strength of the updrafts, they do affect details about the microphysics of the updraft — the size of the cloud and rain particles that are produced within it change. And what that means for lightning is that the size of the ice crystals that might be produced there — the ice crystals that then interact with supercooled water to produce these charging effects — could change significantly. And so even if we don’t see an effect of aerosols on the strength of thunderstorms, we’re sort of in a good position to look at the effects on the detailed microphysics that might affect lightning and other properties that we observe with the radar.
We received funding to add antennas in the existing lightning mapping array in the Houston region, which will improve the accuracy of the network. There’s also a lightning modeling component that’s being led by Toshi Matsui at Goddard Space Flight Center in Maryland. Most cloud simulations don’t include microphysical charging processes, but this one actually looks at where the charging is happening and the simulated cloud carries that charge around, and then will discharge it to simulate lightning.
What did you do when you were at the field site in June?
I participated a little bit in the forecasting efforts. On any given day, the team decides whether or not to release additional sounding balloons to get the conditions of the atmosphere, and also whether to switch the radars into this tracking mode that basically tracks thunderstorms as they move. That decision-making needs to be informed by forecasting. So we run our model every day to help out with the forecasting effort.
Besides that, I was basically a tourist, learning about the different aspects of the project. This is my first time being on a field campaign, so it was really exciting for me to just go into the field and see what other groups were doing. I visited the groups that were doing the unmanned aerial vehicle launches. One group out of University of Colorado launched an airplane that flew back and forth along a track to see the horizontal differences in atmospheric conditions. There was also a coptersonde that went up and down in the atmosphere to get the details of the lower atmosphere.
What happens when a storm comes in?
There weren’t many storms when I was there, but I did see some images of one of the groups sheltering beneath one of their tents when they got hit by rain. They have to shut down the copter and the UAV flights when that happens, to protect their equipment and avoid flying in low visibility. Things like the radar, they just keep on scanning nonstop.
What comes next for the TRACER project?
The field campaign runs until the end of September, and then we’ll start to sift through the mountains of data we’ve collected, running simulations and trying to make sense of the observations. Specifically what we’re interested in are the cases where we captured a thunderstorm from its very beginning all the way to when it dissipates. If we can get a certain number of those in relatively clean and polluted conditions, that would be really great. So we really need to just dig through all this data and see what’s there, and then set up the simulations.
What else are you working on right now?
A big part of the work I do is developing a new approach to modeling the microphysics or clouds and precipitation. Most weather models include certain assumptions about how liquid droplets behave and how their populations evolve. Certain things are kind of hard baked into the model, and it makes progress a little bit difficult because in some cases, when you build an approximation or assumption into a weather or climate model, you then lose track of it. You kind of forget you did that, but the uncertainties remain, and the effects of those uncertainties remain. So it’s kind of coming back and saying, Okay, where did this come from? Can we quantify those uncertainties and can we reduce them in a systematic way?
In the place of assumptions, we’re going to let observations directly constrain the physics. We put in the bare minimum of a priori information and are trying to build it from the ground up to try to get rid of some of the structural errors that we think are holding back the science. This has been funded by the Department of Energy to put into their climate model. We’ve also been funded by NASA to put it into two of their climate models, and the National Center for Atmospheric Research Community atmospheric model.
It’s important to get these things right in climate models. For example, stratocumulus clouds are incredibly important to climate. They are purely liquid clouds that are strongly affected by what’s going on microscopically. I saw one recent paper that suggested that these clouds might disappear with a certain degree of warming, which would be one of those climate feedbacks that increases the pace of warming. If you remove them, you remove one of the Earth’s strongest cooling mechanisms. And if that’s true, then that’s really bad news. We need to make sure our models of cloud microphysics are as good as possible, and try to better quantify the remaining uncertainties, in order to refine our expectations for future impacts of global warming.