• Field Notes
  • Posts
  • Science update: a new tool for catching polluters

Science update: a new tool for catching polluters

This past week, a new scientific paper I led came out in the journal Geoscientific Model Development (read here for free). The paper itself isn't very important, but I think the tool it documents — CHEEREIO* — is pretty cool.CHEEREIO is an open-source, totally free software that uses observations of the atmosphere to catch polluters. Critically, it works on any pollutant, including greenhouse gases like carbon dioxide and methane or air quality hazards like nitrogen oxides.  The math behind CHEEREIO is complicated, but the idea is simple. Environmental scientists estimate the emissions of various pollutants from around the world, a process called inventory building. They do this by painstakingly going through terrible spreadsheets, like country-level reports to the UN. The work is necessary, but the resulting emissions estimates are very uncertain. People fudge the numbers, but even in ideal cases it's really hard to know true emissions. CHEEREIO puts emissions inventories to the test, using a model called GEOS-Chem to simulate what the atmosphere would look like if those emissions were correct. CHEEREIO then compares the simulated atmosphere to the real atmosphere as observed by satellites or equipment on the Earth’s surface. CHEEREIO uses the difference between the model simulation and the real world to update our emissions maps, using a form of Bayesian optimization called the localized ensemble transform Kalman filter. Here's a demo with methane, the second most important greenhouse gas after carbon dioxide. At the top left are emissions estimates from the inventories, and at the top right are the emissions after CHEEREIO corrects them. These are just test results; fully validated estimates are coming soon.

Methane emissions, adjusted by CHEEREIO with satellite data

Right now, I'm working on using CHEEREIO to create global, real-time maps of methane emissions. Already we can use satellites to see things like leaks from space, even if there are no monitors on the ground. My hope is that real-time maps of emissions will give us a better idea of where to look. Some collaborators at the University of Toronto are using CHEEREIO to look at carbon monoxide from Canadian wildfires. Hopefully, others will follow.More science and writing soon, and by soon I mean 2024. I am slow! 

*CHEEREIO stands for the CHEmistry and Emissions REanalysis Interface with Observations, but that doesn't matter. What matters is my advisor says that to be funded, scientists need to (1) come up with a cute acronym, and (2) misspell it so it comes up first on Google.