As I have noted previously, there are significant disparities in the carbon footprints associated with various food products. Beef and lamb, in particular, generate substantially higher greenhouse gas emissions compared to pork, chicken, or plant-based alternatives.
This data indicates that the most impactful strategy for lowering the climate footprint of your diet is to reduce overall meat consumption, with a specific focus on dairy and red meat.
In this piece, I intend to examine whether these findings rely on the specific metric used to measure greenhouse gas (GHG) emissions. One could argue that red meat and dairy have larger footprints because their emissions are primarily composed of methane—a greenhouse gas that is significantly more potent than carbon dioxide, though it remains in the atmosphere for a shorter duration. To date, methane emissions have been responsible for a substantial portion of global warming, with estimates suggesting they account for between 23% and 40% of the total.
In the box at the end of this article, I explore the debate surrounding emission metrics and the handling of methane in greater depth. For now, I will keep it brief:
Because there are numerous greenhouse gases, researchers often combine them into a single unit of measurement to facilitate comparisons. The most prevalent method for doing this is to use a metric known as “carbon dioxide-equivalents.” This is the standard adopted by the Intergovernmental Panel on Climate Change (IPCC) and serves as the official metric for reporting and target-setting under the Paris Agreement.
“Carbon dioxide-equivalents” (CO2eq) consolidate the effects of all greenhouse gases into one metric by utilizing “global warming potential.” Specifically, this uses the global warming potential over a 100-year period (GWP100), a timeframe that reflects a mid-to-long-term outlook for climate policy.
To determine CO2eq, one must multiply the quantity of each greenhouse gas emission by its GWP100 value—a figure intended to represent the warming generated by that specific gas relative to CO2. For instance, the IPCC uses a GWP100 value of 28 for methane, based on the premise that releasing one kilogram of methane will result in 28 times the warming impact of one kilogram of CO2 over a century.
To grasp why the conversion factor of 28 faces criticism, one must recognize that different greenhouse gases persist in the atmosphere for varying durations. Unlike CO2, methane is a short-lived greenhouse gas. It exerts a powerful warming effect in the short term but dissipates rapidly. This contrasts with CO2, which can remain in the atmosphere for many centuries. Consequently, methane has a high impact on warming in the short term but a lower impact over the long run. This leads to frequent confusion regarding how we should quantify the climate impacts of methane.
Researchers are therefore developing new metrics and methodologies intended to provide a more accurate representation of the warming potential of different gases.
Michelle Cain, Myles Allen, and their colleagues at the University of Oxford’s Martin School lead a research programme on climate pollutants that addresses this challenge. Dr. Michelle Cain, a lead researcher in this field, discusses the difficulties of GHG metrics and the role of a new GWP application that accounts for methane’s shorter lifespan (known as GWP*) in an article in Carbon Brief here.
Because methane has a shorter lifetime, the standard CO2-equivalence does not accurately reflect how it influences global temperatures. Therefore, the CO2eq footprints of foods that produce high levels of methane—primarily beef and lamb—do not inherently capture their short-term or long-term temperature impacts.
The question remains: Do these measurement challenges affect the carbon footprint of various foods? Are the large differences solely due to methane?
In the visualization, I compare the global average footprint of different food products, both with and without the inclusion of methane emissions.
As in my original post, this data is by Joseph Poore and Thomas Nemecek (2018), published in the journal Science. The study examines the environmental impacts of foods across more than 38,000 commercial farms.
This chart compares emissions in kilograms of CO2eq produced per kilogram of food product.
The red bars represent the greenhouse emissions that would remain if methane were removed entirely; the grey bar indicates the emissions originating from methane. The combined red and grey bars represent total emissions, including methane.
For example, the global mean emissions for one kilogram of beef from non-dairy herds is 100 kilograms of CO2eq. Methane accounts for 49% of that total. If we exclude methane, the remaining footprint is 51 kgCO2eq, as shown in red.
As we can see, methane emissions are significant for beef and lamb. This is because cattle and sheep are ruminants, meaning they produce large amounts of methane during digestion. If methane were removed, their emissions would decrease by approximately half. This is also significant for dairy production and, to a lesser extent, for farmed fish and shrimp.
This is not the case for plant-based foods, with the exception of rice. Paddy rice is typically grown in flooded fields, where microbes in the waterlogged soil produce methane.
This indicates that beef, lamb, and dairy products are particularly sensitive to how methane is treated in greenhouse gas metrics. Few would suggest eliminating methane entirely, but as noted, there is an ongoing debate regarding how to weight these emissions—specifically, whether the grey bar should be larger or smaller in these comparisons.
So, is it accurate to say that red meat and dairy only have large carbon footprints because of methane? As the red bars demonstrate, that is not the case.
Although the scale of the differences changes, the ranking of the food products remains the same.
The disparities remain substantial. The average footprint of beef, excluding methane, is 36 kilograms of CO2eq per kilogram. This is still nearly four times the average footprint of chicken, or 10 to 100 times the footprint of most plant-based foods.
Where do the non-methane emissions from cattle and lamb originate? For most producers, the key emissions come from agriculture, the land required for animal feed, pasture management (including liming, fertilizing, and irrigation), and emissions from slaughter waste.
What about producers who do not raise livestock on converted land? Do they have a lower footprint? In our related article, I examine the distribution of GHG emissions for each product in detail, from the lowest to the highest emitters. When excluding methane, the absolute lowest beef producer in this global dataset of 38,000 farms across 119 countries had a footprint of 6 kilograms of CO2eq per kilogram. In this instance, emissions resulted from nitrous oxide from manure, machinery and equipment, the transport of cattle to slaughter, slaughter processes, and food waste, which can be high for fresh meat. While 6 kilograms of CO2eq (excluding methane) is significantly lower than the beef average, it is still several times higher than most plant-based foods.
Is it potentially misleading to compare foods based on mass? After all, one kilogram of beef does not provide the same nutritional value as one kilogram of tofu.
In the other visualization, I present these comparisons as the carbon footprint per 100 grams of protein. Once again, methane emissions are shown in grey, while emissions excluding methane are shown in blue.
The results are similar: even if methane were excluded entirely, the footprint of lamb or beef from dairy herds is five times higher than tofu, ten times higher than beans, and more than twenty times higher than peas for an equivalent amount of protein.
The weight assigned to methane influences the magnitude of the differences in carbon footprints between food products. However, it does not alter the primary conclusion: meat and dairy products remain at the top of the list, and the differences between food types remain significant.
We would like to thank Dr. Joseph Poore for providing the underlying data for this analysis, and Dr. Michelle Cain for her feedback on earlier drafts of this article.
The standard metric used to quantify GHG emissions is “carbon dioxide-equivalents.” This is the metric adopted by the United Nations Framework Convention on Climate Change (UNFCCC), used in official GHG reporting and target-setting by countries and institutions, and is the most widely adopted metric in scientific literature. As some researchers have noted, the lack of life-cycle assessment data disaggregated by gas can lead to the loss of vital information that could assist in developing more effective climate mitigation strategies.
What are carbon dioxide-equivalents? Carbon dioxide (CO2) is the most significant greenhouse gas, but it is not the only one; gases such as methane and nitrous oxide also drive global warming. Carbon dioxide-equivalents (CO2eq) attempt to aggregate the warming impacts of all greenhouse gases to provide a single measure of total emissions. Two factors complicate this: the gases have different warming “strengths,” and they persist in the atmosphere for different lengths of time.
To convert non-CO2 gases into their carbon dioxide-equivalents, we multiply their mass (e.g., kilograms of methane emitted) by their “global warming potential” (GWP). GWP measures the warming impact of a gas relative to CO2; it essentially measures the “strength” of the greenhouse gas averaged over a chosen time horizon. The standard approach is to evaluate GWP over a 100-year timescale (GWP100). GWP100 is the accounting metric adopted by the IPCC in inventory guidelines, although their Fifth Assessment report (AR5) did not explicitly mandate its use. Chapter 8 of that report described both GWP and Global Temperature-change Potential (GTP) as examples of metrics that are useful depending on the specific question being asked.
The GWP100 value for methane from AR5 is 28 (or 34 if climate feedback processes are included). This means that emitting one kilogram of methane creates 28 times the warming of one kilogram of CO2 averaged over the next 100 years. However, this does not account for the fact that methane is a short-lived greenhouse gas. It exerts a very strong warming impact when first emitted, but this impact diminishes over subsequent decades, whereas an equivalent amount of CO2 could persist for centuries.
Using the GWP100 metric can therefore misrepresent the impact of short-lived gases like methane in both directions. It underestimates short-term warming: the warming impact of methane when it is first emitted and in the following years is much higher than the “28” value assigned by GWP100. Consequently, some argue for using a value that represents global warming potential over 20 years (GWP20), as it provides a better indication of short-term warming. The IPCC reports a GWP20 value of 84 for methane (86 if feedbacks are included). Others argue that GWP100 overestimates the long-term impacts of methane, as the methane emitted today will not be present a century from now. These differences are reflected in the significant changes in GTP over different time horizons. The GTP100 value for methane is 4, whereas the GTP20 value is 67.
This makes it challenging to reconcile these warming impacts into a single metric. Our choice of metric can influence how we prioritize GHG reduction strategies: should we first target strong but short-lived gases like methane? This might slow warming in the short term—a reasonable argument if we are concerned about approaching temperature-induced tipping points. Or should we instead focus on persistent CO2 emissions, which will be the primary driver of long-term temperature impacts?
Some researchers have developed new methods aimed at providing a closer representation of the actual temperature response to different gases. Myles Allen, Michelle Cain, and their colleagues at the University of Oxford’s Martin School lead a research programme on climate pollutants that addresses this challenge directly.
They have proposed a new way to represent short-lived greenhouse gas emissions—GWP*—which aims to be more representative of the warming response. Dr. Michelle Cain, a lead researcher in this area, discusses the challenges of GHG metrics and the role of the new GWP* metric in an article in Carbon Brief here.
GWP* is used to calculate CO2-warming-equivalent emissions, which reflects that (a) increasing methane emissions would immediately increase global temperature, (b) rapidly decreasing methane emissions would immediately reduce global temperature, and (c) a gradual decline in methane emissions would stabilize the global temperature attributed to methane. Scenarios (b) and (c) differ significantly from CO2, as rapidly or gradually decreasing CO2 emissions only slows the rate of temperature increase rather than reducing it.
This is explored further in an Oxford Martin School briefing note, found here, and in a recent publication by researchers John Lynch, Michelle Cain, Raymond Pierrehumbert, and Myles Allen (2020).
