Currently, levels of CO2 in the atmosphere are at an uncomfortable 380 parts per million—a level that hasn’t been seen in over 20 million years. Now, more than ever, it is becoming necessary to know where CO2 is coming from and where it is going. A new project does exactly that, and may be the best tool yet in charting the fate of our feverish world.

In essence, CarbonTracker is a CO2 data assimilation system. It takes atmospheric measurement data from hundreds of places around the globe, charts emissions and uptake in real time, and inserts them into a complicated predictive model. Over the seven years the project has been operational, the system has collected over 36,000 air measurements from the Earth’s surface. As project leader Dr. Wouter Peters, who works for the National Oceanic & Atmospheric Administration, explains:

“The atmosphere sees everything that is going on—the atmosphere won’t lie.”

Already, the project is giving startling results. A highball estimate of future carbon emissions made in the year 2000 is now seen as a low estimate compared to the data collected so far. There is a slight silver lining, for us Canadians at least: it turns out that Canada’s vast boreal forests are a carbon sink (a region taking in more CO2 than it gives off). This is no license to burn, however, as this uptake represents only a third of North America’s total carbon output.

Global climate change is a nuanced and complex problem. A key question is how human-driven landscape alteration, including widespread deforestation, will affect carbon uptake and output. Recently, and with great alarm, it has been noted that the permafrost has been thawing in some areas (the prefix ‘perma-’ can be misleading). If this de-icing continues unabated, the huge amounts of carbon that are currently sequestered could be released into the atmosphere. It would be equivalent to burning all the coal reserves in the United States—at once.

Annual variations in carbon movement can be visualized with the CarbonTracker model. Every summer, areas where corn is growing in the Midwestern United States can be clearly seen due to their uptake of atmospheric carbon. Conversely, in drought years the amount of carbon taken up by affected ecosystems across North America drops drastically. With less plant matter growing, less atmospheric CO2 is taken up by photosynthesis. This large reduction in carbon uptake in drought years (by as much as half, compared to nondrought years) was a surprising result to the team.

Peters is clear about the effect humans are having and the need for greater accountability: “Although we can’t blame everything on fossil fuels, we can at least blame everything on humans.”

His model, designed with 15 other collaborators, is open-source, meaning that anyone can download the model’s source code and see its inner workings. Mostly a testament to the unwillingness of American politicians to address climate change, the fact that the project is open to the public allows for maximum transparency, and results that can be completely scrutinized.

High altitude atmospheric measurements and other predictive models have already validated CarbonTracker’s results. By basing its predictions on numerous real observations, the system offers accurate predictions and a clear picture of the movement of carbon around North America. With further expansion of the project and more measurements, the model will eventually be able to give a clear picture of the global carbon cycle, which remains one of the least understood issues in global climate change science. We can only hope a greater understanding of the ecological disaster will force governments and citizens to take action in a meaningful way. A high fever is a sign that something is wrong—and we’d better do something about it soon.