How much gas is emitted? It seems like a simple question, and answering it concretely is the foundation for many climate governance meetings, ranging from local municipalities to global Conference of the Parties (COP) summits of the United Nations Framework Convention on Climate Change (UNFCCC). For the southeastern Brazilian state of Minas Gerais, answering this question was also seemingly simple. In a landmark 2008 report by state climate scientists, it was estimated that Minas Gerais emits 122,950 gg of CO₂eq per year (the equivalent of 135 million tons CO₂). At COP summits, greenhouse gas (GHG) emission estimates are one of the key starting points for setting Nationally Determined Contributions (NDCs), the state-centered goals for climate change adaptation, mitigation, and emissions reduction. In principle, accurate estimates of GHG emissions allow the negotiators to establish national responsibility and set reduction targets. It would seem that accurate estimated GHG emissions is a vital piece of information necessary for global climate negotiations.

GHGs are largely invisible and amorphous, but they are ultimately physical objects that are finite and quantifiable. It is easy for climate negotiators to refer to GHG estimates as if they are concrete reality and provide a firm anchoring point for the complex negotiations taking place. However, my fieldwork with the team of climate scientists that created Minas Gerais’ GHG report revealed complex epistemological politics that locally construct, and potentially inhibit, global climate governance.

There are two main ways that ambiguity enters into GHG inventories. First, the category of “greenhouse gas” encompasses a number of different substances. Carbon dioxide is the most plentiful, but there are others that can have the same effects. Climate analysts must decide what gases are counted in an GHG inventory. Second, GHG emissions cannot be directly measured as they are emitted because there are too many sources to monitor. As such, climate analysts had to rely on emission reports from a wide range of both governmental and non-governmental partners to craft the Minas Gerais GHG inventory. The conceptual challenge of defining what gases count as GHGs and the political challenge of procuring emissions reports introduced doubt to the seemingly straightforward emissions report.

In popular discussions, references to GHGs might appear to be an unnecessarily complicated way to refer to carbon dioxide. GHGs are measured by their relative equivalence to carbon dioxide’s thermal properties. For example, methane is a less common but more effective GHG with a CO₂eq of approximately 84. Using carbon dioxide equivalence as a unit and referring almost exclusively to carbon dioxide when discussing GHGs gives the impression that GHGs are a relatively simple entity. However, there are many gases which contribute to the greenhouse effect, each with their own dynamics for how they get into the atmosphere and how they behave once there.

The Mineiro GHG inventory included five gases. Along with carbon dioxide and methane, it included nitrous oxide, tetrafluoromethane, and hexafluoroethane. While these gases are significant contributors to the greenhouse gas effect, it is important to note that this list did not include other prominent greenhouse gases like ozone and water vapour (Manabe 2019). These other gases were not excluded arbitrarily. For example, the role of water vapor in precipitation, cloud formation, and other weather phenomena means that it stays in the atmosphere for much less time than other GHGs and its effect on global temperatures is much more complex. Mineiro climate analysts felt confident in simplifying their work by removing water vapor from the assessment, even if water could justifiably qualify as a GHG.

In their study of polychlorinated biphenyls, Michelle Murphy cautions scholars not to imagine chemicals as isolated, abstract entities (2017, 495). The creation of the state GHG inventory exemplified their suggestion. Climate analysts necessarily encountered the networks surrounding GHGs as they worked with industries, municipal monitoring stations, and international scientific communities to collect sufficient and accurate emissions data.

To manage the complexity of greenhouse gases, the climate analysts identified four “sectors” of emitters: energy, industrial processes and products, land use, and waste management. Each of these four sectors further broke down into component parts, many of which are themselves composites of different types of emissions. The taxonomy can continue seemingly indefinitely. For example, the “energy” sector breaks down into three components: the consumption of energy in the production of more energy, “fugitive” methane emissions from natural gas, and energy use. That last category breaks down into five more subsections: transport, industry, residential, agriculture, and commercial and public energy use. Of those sections, industrial energy use was further divided into at least 10 different types of industry.

For each of these subsections (or sub-subsections, etc.), climate analysts needed to find a source of reliable data on the amount of greenhouse gases emitted. The data could be self-reported by various industries or calculated by climate analysts based on the chemical processes involved in the sector. After hundreds of data sources were collected, verified, synthesized, and analyzed, they were finally ready to announce the deceptively simple quantity of GHG emissions.

The complexity of the emission estimates provided skeptics with an opportunity to derail actions. As more data were collected and new methodologies emerged, the emissions estimate changed slightly. In 2014, Mineiro climate scientists revised their 2008 estimate from 122,950 gg of CO₂eq to 124,167.3 gg CO₂eq. The 2014 estimate is less than one percent higher. However, when Brazil’s national NDC called for Minas Gerais to reduce emissions by 38 percent by 2025, opponents of climate action seized on the discrepancy. Did the 38 percent apply to the older number which had been used during the negotiations? If so, Minas Gerais needs to eliminate 76,229 gg CO₂eq of emissions. However, if they targeted 38 percent of the new estimate, Minas Gerais would need to eliminate 76,529 gg CO₂eq. They were quibbling over a difference of only a quarter of a percent of the state’s total emissions, but the mere existence of doubt stalled the negotiations.

This incident shows how overestimating the significance of ambiguity can serve as a means of stalling effective climate action. Epistemic perfectionism, demanding that all doubts must be absolutely resolved before moving forward with proposed actions, is incompatible with the realities of the climate crisis. Doubt is built into the process of forming GHG estimates through the choices of gases to measure in complex process of gathering data. Doubt is unavoidable, but in the words of the Czech-Brazilian philosopher Vilém Flusser, thinking proceeds “through the path of doubt” (2014, 5). Rather than treat doubt as an unavoidable barrier for climate negotiation, ambiguity must be recognized as a central animating feature of climate politics.

[1] The term “industry” appears twice in this diagram. To avoid confusion, the first instance, as a subsection of “energy,” refers to the emission of GHGs in producing electricity to power industry. The second instance, as a category for “industrial processes and product use,” refers to GHGs that are the outcome of the chemical reactions in various industries.