Just over a year since we "opened for business," the Yeung Lab has published its first data! It is a paper I'm quite proud of, and I'm glad it ended up in the Journal of Geophysical Research (we even got an Editor's Highlight!) It is the first of hopefully many fruitful collaborations with my colleague Lee Murray, an atmospheric modeler at the University of Rochester. It lays the foundation for many years of ice-core work in our group, and new insights into our atmospheric machinery over the past million years or so.
In the paper, we argue that we can trace ozone, temperature, and (to some extent) the circulation of the ancient atmosphere using the abundance of isotopic "clumps" in atmospheric oxygen, e.g., the number of 18O18O molecules in O2. Before I tell you how we extract this information, though, let me first tell you why one would even want to know these things.
The atmosphere is incredibly dynamic
If you look at the main components of the atmosphere today—oxygen, nitrogen, argon, water, and carbon dioxide—they are actually pretty unreactive. As such, they don't drive very much chemistry in our atmosphere. Instead, our atmosphere's chemistry is mainly driven by a panoply of highly reactive species that are present in tiny concentrations, often less than a part per million. The influence of these species greatly outweighs their modest abundances, however.
Consider, for example, ozone. You may have heard of it. Between 20-25 km altitude, it is often called "good" ozone, as it absorbs much of the sun's ultraviolet light that can damage living tissue. Lower in the atmosphere, it is called "bad ozone" because it is pollutant that negatively affects human health and agriculture. Yet it is also instrumental to the chain reactions that rid the atmosphere of its pollutants. It plays a complex and important role in the Earth system, all at concentrations of several parts per million or less. Its past variations are therefore fundamental to our understanding of the atmosphere.
Catching lightning in a bottle, indirectly
Humans have greatly affected the atmosphere's ozone levels, both high up and closer to the surface. But by how much? Because it is so reactive, ozone doesn't stick around long enough in our usual archives of the ancient atmosphere (e.g., gas bubbles in ice cores) for us to analyze. We have to look at the marks it leaves on species that do stick around.
It turns out that the best candidate is actually good ol' O2. We believe this because ozone molecules (O3) are made by adding a single O atom to O2. This reaction happens everywhere in the atmosphere, all the time; sometimes the addition is successful, and you get an ozone molecule. At other times (well, actually, most of the time), the reaction fails, and you are left with O and O2 again.
Ah, but the beauty is in the details: Even during these "failed" reactions, you can shuffle the atoms around. It goes something like this:
The red O atom often gets incorporated into the new O2 molecule. In fact, these O + O2 reactions probably shuffle the atoms around in the entire atmospheric O2 reservoir in under ten years. Crazy, right?
Our new tracer, 18O18O, tracks these atom-shuffling reactions: When you have more ozone, you get more atom shuffling (or isotope exchange, as it is called in the paper). Their patterns, in fact, match the patterns of ozone in the atmosphere.
There's one more thing you need to know: It's warmer at the surface than it is high in the atmosphere, generally speaking, and the number of 18O18O "clumps" varies inversely with temperature. So you get slightly more 18O18O when atom-shuffling happens high in the atmosphere, and slightly less when it happens lower in the atmosphere.
We see the balance between high- and low-altitude chemistry when we sample O2 in air, either directly or from ancient bubbles trapped in glacial ice. More "bad" ozone near the surface (e.g., air pollution) means there is slightly less 18O18O in these archives, and vice versa.
The ice-core work is ongoing, and we are seeing some very neat results in the preliminary data. We think we will be able to quantify the increase in surface ozone since humanity started burning fossil fuels with abandon around 1850.
About that map image...
One of the coolest things to emerge from the work, in my opinion, is the possibility of tracing atmospheric circulation back in time. Specifically, I think we'll see signals related to vertical aspects of the circulation. Knowing how air is circulated vertically on long timescales tells us a lot about how heat is distributed within the atmosphere in different climates. It's a vital part of the atmosphere's machinery that we don't know much about.
The cover image of this blog post shows this idea in action. The red shading tells you where the atom-shuffling reactions are most strongly expressed. Notice how these patterns shift with season: toward the Northern hemisphere in April, and the Southern hemisphere in October. They wax and wane in lock-step with global circulation patterns!
Moreover, atom-shuffling is most strongly expressed in at moderate altitudes between 4-6 km. Changes in circulation patterns will alter these patterns, and ultimately also the balance of atom shuffling at warm and cool temperatures. It's a bit complicated to interpret, as you might surmise, but ultimately worth it in my opinion. We want to keep pushing the boundaries of what's possible to learn about the ancient atmosphere.