Measuring carbon dioxide levels in outdoor air is both a simple and complex exercise. One of the main problems with talking about regulating global CO2 levels is that scientists aren’t all in agreement on how to measure it.
At the simplest level, the monthly average global CO2 levels has been measured since 1958 at the Mauna Loa observatory in Hawaii. At Mauna Loa, the remote location, undisturbed air, and minimal influences of human activity and vegetation are ideal for monitoring CO2 levels over time.
The data collected at Mauna Loa indicates a continual rise in global CO2 levels from less than 320ppm (parts per million) when measurements were first recorded to approximately 395ppm today.
However, when it comes to measuring CO2 at the local level (or by country), the science becomes more complicated. Carbon dioxide mixes quickly in air, and any CO2 sensor left outdoors will record many different CO2 levels depending on the time of day, the time of year, and even local weather conditions.
To solve the problem of measuring CO2 levels on a local level, scientists have developed a 2-pronged approach. The first approach is for researchers to predict all the CO2-creating objects in a given area, then extrapolate the amount of CO2 that is likely to be produced. For example, by adding up all the cars, buildings and power plants in an area, a theoretical model can be made for total CO2 produced over a given time period. This kind of theoretical model was popularized in a 2011 video by Dickinson and Tenorio entitled "Inventory of New York City Greenhouse Gas Emissions."
The second approach is for researchers to actually measure CO2 levels at the micro scale. This involves laying down grids of dozens or even hundreds of CO2 sensors, then creating a real-time map of CO2 levels. One of the most ambitious projects like this has been undertaken by the Global Institute of Sustainability at Arizona State University.
Project like this are completed or in process in dozens of major urban areas worldwide where most CO2 emission occurs. Sensors like the K33 ELG 1% Data Logging Sensor or the COZIR Ambient CO2 Sensor have been used in many of these projects due to their ability to record CO2 levels over long periods of time.
To date, the theoretical and measured models show as much as a 15% discrepancy between the results, too high an error to claim the CO2 measurement problem “solved.” Yet with continued testing in this area, an accurate model of CO2 levels seems to be within reach. It will then be up to governments to decide what to do with this information.