Pieter Tans, Ed Dlugokencky, and Ben Miller
NOAA Global Monitoring Laboratory, R/GMD, 325 Broadway, Boulder, CO 80305-3328
Updated October 2020
We present a direct physical measure of the direct climate heating influence of greenhouse gas enhancements, showing how they have increased dramatically since the onset of the industrial revolution. The conclusion that humans are nearly 100% responsible is inescapable. The contributions of different greenhouse gases have evolved over time. We present tangible examples of the excess heat being retained in the Earth system, easily large enough to force global and regional climate change. The direct heating influence of greenhouse gases is well understood. It is explained by the same science that gave us lasers, fluorescent lights, LEDs, cell phones and more. It is thus not surprising that we experience the impact of our emissions on the climate. This heating influence has been reported by NOAA through a range of national and international assessments.
There are two requirements for a gas to be called a greenhouse gas (GHG): it must absorb infrared radiation (sometimes called heat radiation), and it must have a long residence time in the atmosphere. There must be enough time for emissions and removals of the gas to mix globally, causing its concentration to be fairly uniform. Climate projections have large model uncertainties due to the complexity of the climate system. However, the uncertainties in GHG measurements themselves are extremely small. Furthermore, the extra heat retention in the atmosphere due to the increased abundance of each gas can be calculated fairly accurately when other factors are kept the same. We call this the direct heat retention or direct heating influence. We present here a measure that summarizes the increase of the direct heating influence since 1800 (also known as “climate forcing”) supplied from these gases. We also present it as a fraction of the Earth’s weather and climate engine, namely the total solar energy absorbed by Earth and its atmosphere. Because it is based on the observed amounts of GHGs in the atmosphere and on well-known physics of how gases absorb and emit radiation, this measure has little uncertainty.
The Intergovernmental Panel on Climate Change (IPCC) defines climate forcing as “An externally imposed perturbation” of the radiative energy budget of Earth’s climate system [Ramaswamy, 2001]. “Externally” means external to the natural climate system. Specifically, that includes the emissions of greenhouse gases, changes in Earth’s albedo (reflected portion of sun light) caused directly by human activities such as the replacement of (darker) forests with (more reflective) crop lands and urban areas, human emissions of aerosols and aerosol precursors, irrigation, etc. It does not include changes in solar radiation, clouds, precipitation, snow and ice cover, etc. The latter are determined by the climate system itself, and they are a part of the response of the climate system to human intervention. The largest components of climate forcing are the changes in the global atmospheric burdens of carbon dioxide (CO2), methane (CH4), nitrous oxide (N2O), and a number of industrial gases containing bromine, chlorine, or fluorine. The first three are natural greenhouse gases that have been in the atmosphere for at least hundreds of millions of years, but have very sharply increased recently. The other greenhouse gases are, or have been, made by the chemical industry, and were not present in the atmosphere before the mid-20th century. We call them collectively “industrial gases”.
Although water vapor absorbs and emits infrared radiation, it is not included as a climate forcing agent. Water evaporates from the oceans and land surface (and falls back as rain or snow) in amounts at least 40,000 times larger than direct emissions of vapor due to our combustion of fuels. It is not an “externally imposed” influence on climate, but its abundance is internally generated by the climate system itself, primarily by the temperature of the surface. Furthermore, Lacis and coworkers showed in climate modeling studies [Lacis et al., 2010] that infrared absorption by CO2, CH4, and N2O is essential to sustain temperatures needed for water vapor to remain in the air in high enough amounts. With their model starting from pre-industrial steady state conditions, when the GHGs were removed from the model, Earth turned into an ice ball in three decades, with most of the water vapor frozen out of the atmosphere and the planetary albedo increased. Other spatially heterogeneous and short-lived climate forcing agents, such as aerosols and tropospheric ozone, have uncertain global magnitudes and are also not included here.
Direct climate forcing can be calculated by global climate models by changing only the GHG abundances, while leaving the rest of the modeled climate system such as clouds, surface albedo, temperature, precipitation, etc. unchanged, except for some small immediate temperature changes in the stratosphere [Ramaswamy et al., 2001]. This creates an imbalance in the Earth’s heat budget, less infrared energy leaving Earth, while the incoming solar energy remains the same. The model calculates the imbalance “on the side”, not letting it influence the evolution of its climate. The most important consequence of our definition is that climate forcing by GHGs is well-known and caused by human activities alone. We do not want to entrain the uncertainties of the climate system and uncertain predictions of its response to human influence into our definition of climate forcing.
The NOAA monitoring program provides high-precision measurements of the global abundance and distribution of GHGs that are used to calculate changes in radiative climate forcing.
Air samples are collected through the NOAA/GML global air sampling network, including a cooperative program for the carbon gases which provides samples from 40-50 global background air sites, including measurements at 5 degree latitude intervals from ship routes (see Figure 1). Beginning in 1980 global averages are calculated for CO2 based on a subset of sampling sites in the marine boundary layer (“MBL”) that give access to well-mixed clean air with minimal influence from local vegetation and from nearby human activities. Weekly data are used from the MBL sites to create a smoothed north-south latitude profile [Masarie and Tans, 1995] from which a global surface average is calculated (See: gml.noaa.gov/ccgg/trends/global.html). We started using the MBL for CH4 in 1983, and for N2O in 2000. Before those dates the air samples came from atmospheric sampling networks with fewer sites and from snow and ice.
For CO2, CH4, and N2O we used the ice and firn data from Law Dome in coastal Antarctica, which have the highest temporal resolution during the last two millennia [Etheridge et al., 1998; MacFarling Meure et al. 2006], and merged them with the MBL data to produce continuous records for global CO2, CH4, and N2O from the year 1 CE through 2019, taking into account time-dependent offsets between the concentrations measured over Antarctica and the global surface mean as observed in the MBL data.
The atmospheric histories of the industrial gases were derived by one of several means depending on the availability of data. For CFC-12, -11, -113 and CCl4, we used the GML data (e.g., https://gml.noaa.gov/hats/combined/CFC12.html), while augmenting the earlier period with abundance estimates taken from Walker et al. , which provides estimates back to zero abundance. For many of the remaining gases (e.g., HCFC-22), the data records are more sparse. Here we rely on current GML measurements (approximately 1994 to present) for global coverage. To extend this time series further back in time, we made use of analyses of the Cape Grim Air Archive [CGAA, Langenfelds et al., 1996] to create a Southern Hemisphere observation record as far back as 1978. To extend this Southern Hemisphere (SH) record back to the pre-industrial atmospheric mole fractions (typically zero) before 1800 CE, we back-extrapolate from the oldest years of the CGAA with an exponential fit to zero. Since Northern Hemisphere (NH) observations of many of these gases in the earliest period are lacking, an estimate of the NH record is then created by using the more recent observed interhemispheric difference (IHD) in the oldest GML flask-air data. Because most of the emissions take place in the NH, trace gas abundances are lower in the SH (which includes Cape Grim at ~40°S). So the CGAA record needs to be scaled up a bit to give a global estimate. We assume that the IHD itself scales with the emissions, and therefore with the changing year to year abundances in the CGAA. The resulting NH and SH records are then averaged to produce a global annual average history.
The atmospheric abundance of CO2 has increased by an average of 1.78 ppm per year over the past four decades (1980-2019). The CO2 increase is accelerating — while it averaged about 1.6 ppm per year in the 1980s and 1.5 ppm per year in the 1990s, the growth rate increased to 2.4 ppm per year during the last decade (2010-2019). The annual CO2 increase from 1 Jan 2019 to 1 Jan 2020 is 2.6 ppm (For more information, see https://gml.noaa.gov/ccgg/trends/gl_gr.html).
The growth rate of methane slowed down from 1983 until 1999, consistent with its concentration approaching steady-state, assuming no trend in CH4 chemical lifetime. Superimposed on this decline is significant inter-annual variability in growth rates [Dlugokencky et al., 1998, 2003]. From 1999 to 2006, the atmospheric CH4 burden was nearly constant, but since 2007, globally averaged CH4 has been increasing again. Causes for the increase during 2007-2008 include warm temperatures in the Arctic in 2007 and increased precipitation in the tropics during 2007 and 2008 [Dlugokencky et al., 2009]. Isotopic measurements argue for continued increasing microbial emissions after 2008 (e.g., likely from wetlands or agriculture) [Schaefer et al., 2016; Nisbet et al., 2019; Schwietzke et al. 2016]. Since 2014, the global within-year increase (1 Jan to 1 Jan) in methane has become even larger, averaging 9.3 ± 2.2 ppb yr-1 through 2019 compared to an average annual increase of 5.7 ± 1.1 ppb yr-1 in the preceding seven years 2007-2013 (https://gml.noaa.gov/ccgg/trends_ch4/).
The atmospheric burden of nitrous oxide continues to increase steadily over time, with an average rate of 0.97 ± 0.17 ppb yr-1 over the past decade 2010-2019, faster than in previous decades. (gml.noaa.gov/ccgg/trends_n2o/)
Radiative forcing from the sum total of observed changes of the industrial gases continues to increase. The abundances of gases included in the original Montreal Protocol on Substances that Deplete the Ozone Layer have declined, but other gases, some of them included in later amendments to the Montreal Protocol, are still increasing. The increased radiative forcing from the sum of the latter group more than compensates for the declines of the original group, so that climate forcing from all industrial gases continues to go up.
The IPCC’s choice of 1750 as the start of the industrial era, and as the reference year for gas abundances, was unfortunate because it led to an anomalously low value for pre-industrial CO2 during the “Little Ice Age” from about 1600 to 1800 (Figure 2). Instead, we pick 1800 CE as the start of the industrial era, and for CO2 we take its average during 1-1800 CE of 279.35 ppm as the pre-industrial value. Both CH4 and N2O tended to be slightly lower during the years 1-1000 than during 1000-1800 (Figure 1), so that we take their averages over the latter period as our estimates for their respective pre-industrial values (716.6 ppb for CH4, 269.4 ppb for N2O). The use of these new pre-industrial values guarantees that there is no net long-term climate forcing between 1 and 1800 CE, or between 1000 and 1800 CE, consistent with our definition.
We have used the IPCC [Ramaswamy et al., 2001] recommended expressions, with our new pre-industrial values, to convert from the global abundance changes of CO2, CH4, and N2O to instantaneous radiative forcing (see Table 1), and we used the recommended radiative efficiency factors (Watt/m2 per ppb) in the IPCC 5th Assessment [IPCC (2014), Ch. 8, Table 8.A.1] for the industrial gases. The empirical expressions are derived from atmospheric radiative transfer models and generally have an uncertainty of about 10% (95% confidence interval). The relative uncertainties in the global average abundances of the long-lived greenhouse gases are much smaller (<0.1% for CO2, and CH4 <0.5%).
|Trace Gas||Radiative Forcing, ΔF (Wm-2) (simplified expressions)|
|CO2||ΔF = 3.35 [g(C) − g(Co)], with g(C) = ln(1+1.2 C + 0.005 C2 + 1.4x10-6 C3 )|
|CH4||ΔF = 0.036 (M½ − Mo½) − [ f(M,No) − f(Mo,No) ]|
|N2O||ΔF = 0.12 (N½ − No½) − [ f(Mo,N) − f(Mo,No) ], with
f(M,N) = 0.47 ln(1 + 2.01x10-5 (MN)0.75 + 5.31x10-15 M (MN)1.52)
|Industrial gas||ΔF = αX (X-Xo)|
|Notes: “ln” is the natural logarithm function (base “e”), C is the CO2 global measured abundance (in ppm), M and N are the same for CH4 and N2O respectively (in ppb), and X is the symbol for any of the industrial gases (in ppb). The subscript “o” denotes the pre-industrial global abundance, Co = 279.35 ppm, Mo = 716.6 ppb, No = 269.4 ppb. Xo = 0 except for CF4 in which case Xo = 0.0347 ppb. The radiative efficiency factor αX (Watt/m2 per ppb) is different for each gas.|
Figure 3 portrays how much our emissions of greenhouse gases have changed Earth’s heat budget. We could express it as a fraction of total solar radiation absorbed by the atmosphere and Earth’s surface, which is the energy that runs our entire weather and climate system, driving the winds, ocean currents, evaporating water from the surface, fueling hurricanes and tornadoes, etc. In 1996 the fraction passed 1.00%, in 2019 it was 1.37%. The solar constant is taken as 1360.8 W m-2 [Kopp and Lean, 2011]. This is the energy intercepted by Earth’s disk (creating the earth’s shadow). The full area of Earth’s surface is four times larger than a one-sided disk, so that intercepted solar radiation is 1360.8/4 = 340.2 W m-2 averaged over the entire surface of Earth. Of that, 30.6% is reflected back to space (“Bond albedo”, D. Williams, 2019), leaving (1-0.306) times 340.2 = 236.1 W m-2 as the total absorbed solar radiation.
In the section below, “How much energy are we talking about”, we will give some examples of what the globally averaged heat retention of 3.22 Watt/m2, the value for 2019 that corresponds to 1.37% of absorbed solar radiation, can do.
|Enhancements of natural gases (W m-2)||Industrial gases||TOTAL|
|Year||CO2||CH4||N2O||sum of CO2 CH4 N2O||since mid-20th century||W m-2|
Climate forcing by CO2 during the years 1800-1900 was dominated by land use change, such as deforestation as well as the conversion of grass lands to crop lands, accompanied by the loss of organic carbon in soils. These processes were aided by increasing mechanization. Meanwhile, industrial production fueled by burning coal and oil continued to increase, and overtook land use as the main cause of CO2 emissions in the late 19th century, based on a mass balance calculation of emissions and observed changes in the atmosphere and oceans [Tans, 2009]. That same calculation showed that from the mid-20th century on, terrestrial ecosystems (including agriculture) became a net “sink” of carbon (removing CO2 from the atmosphere) when globally averaged, despite ongoing deforestation and fires that convert wood to CO2. This could be partially due to fertilization of plant growth by higher CO2 and by deposition of reactive nitrogen, which includes fertilizer use. Today, emissions resulting from manufacturing, transportation, and heating/cooling have grown so large that the net global terrestrial sink of ~1 billion metric ton C per year (GtonC/yr) [Friedlingstein et al., 2019] is dwarfed by emissions from fossil fuel burning, at ~10 GtonC/yr. Note that 1 GtonC equals 3.67 Gton CO2. The latter includes the mass of the oxygens, and is often used in emissions statistics.
Table 3 shows that, although CO2 is responsible for 66% of the climate forcing by all greenhouse gases, its rate of increase during the last five years accounts for 82% of the total increase in forcing. From well-known chemistry of the carbonate system in the oceans we can estimate that, when the atmosphere and oceans are again in chemical equilibrium (after about 1000 years), ~83% of the excess CO2 resides in oceans and ~17% in the atmosphere. In the natural system, very slow calcium carbonate dissolution (which includes coral reefs) increases the alkalinity of the oceans allowing them to ingest the remaining 17% from the atmosphere, but that is expected to take between 3000 and 7000 years [Archer et al., 2009]. This is how long future generations will likely have to deal with the enhanced atmospheric CO2 unless ways will be found to pull the excess CO2 back out of the atmosphere and out of the oceans. Note that for this purely chemical estimate we assume that other factors, such as ocean circulation and ocean biology do not change, which is unlikely when the climate changes. If we would merely pull excess CO2 out of the atmosphere, the oceans would emit enough of its excess carbon back into the atmosphere to re-establish chemical equilibrium. The same could happen with the terrestrial biosphere, if the main reason for their current net uptake is fertilization by high CO2.
After the pause in the CH4 growth rate from 1999 to 2006, climate forcing by CH4 has increased by 0.034 W/m2 in 13 years. That is about equal to one year’s worth of the increase of climate forcing by CO2. There have been some proposals to remove CH4 from the atmosphere by (yet to be developed) industrial processes, in order to decrease its climate forcing. That would be foolish. It is much simpler and less energy intensive to decrease its emissions, and let the atmosphere remove the CH4 by photochemical processes, ultimately powered by solar energy, with a time scale of about 9 years.
Nitrous oxide is responsible for 39% of the current climate forcing that is due to methane, but the increase of forcing by N2O over the last five years is almost equal to that of methane. Since fertilizer production for agriculture is the largest single cause of the N2O increase it seems likely that N2O’s contribution to climate forcing will continue to accelerate.
|2019||5-yr avg. 2015-2019 rate of increase|
|All industrial gases||0.373||0.116||0.00182||0.041|
The 5-year average rate of increase of forcing by all industrial gases comprises only 4.1% of the increase of forcing by all gases because the abundances of several important gases controlled by the original Montreal Protocol decreased.
Let’s look into the contribution of individual industrial gases in more detail. First we split them up into four groups: 1. Those that were controlled (scheduled to be phased out) in the original Montreal Protocol (“MP gases”) which came into force in 1987, six chlorofluorocarbons (CFCs), three halons (H-1211 etc.), carbon tetrachloride (CCl4), and methyl chloroform (CH3CCl3), 2. Hydrochlorofluorocarbons (HCFCs) intended to replace the MP gases, 3. Hydrofluorocarbons (HFCs) intended to replace the HCFCs when the latter also became controlled because the HCFCs still had an impact, albeit smaller, on stratospheric ozone destruction. Subsequently, since the Kigali amendments to the Montreal protocol went into force in early 2019, the HFCs also became controlled substances because some of them have long lifetimes and therefore contribute significantly to global heating on a per-molecule basis, and 4. Other greenhouse gases (called “non-MP”), often with extremely long chemical lifetimes, that are not covered by the Montreal Protocol and its amendments. On a per-molecule basis their influence on global heating is enormous.
Figure 4 shows a summary of the industrial gases. Climate forcing by the MP gases is declining, but the other three groups are all increasing. Between 1970 and 1990 climate forcing by all industrial gases increased nearly linearly. If we extrapolate that linearly to 2019, their total climate forcing would have been larger by 0.30 Watt m-2 in 2019 than what actually happened. This “avoided global heating” is a success of the Montreal Protocol process, in addition of the protection of the stratospheric ozone layer. However, forcing by all industrial gases is still increasing, and perhaps even accelerating, with the increase in the 5 years from 2014 to 2019 at 9.10 milliWatt m-2, while the preceding 5 years from 2009 to 2014 saw an overall increase of 8.3 milliWatt m-2. The rate of increase of the HCFCs is slowing down, during the last 5 years forcing increased by 3.9 milliWatt m-2, while during the preceding 5 years it increased by 7.6 milliWatt m-2. The other two groups, the HFCs and non-MP gases are still accelerating.
Back to the original MP gases, the decrease of CH3CCl3 was rapid (Figure 5) when its production stopped because of its relatively short lifetime of 5 years. The rate of decrease for most CFCs was slower than the rate one would expect for zero emissions, because of stored “banks” in insulation materials, equipment, etc., but very recently the rate of decrease of CFC-11 slowed down significantly further because of new production of the chemical [Montzka et al., 2018]. CCl4 is still being produced and emitted in large quantities, as its rate of decrease is much slower than would be expected from its chemical lifetime.
When we multiply the 2019 average heating intensity of 3.224 Watt m-2 by the surface area of the Earth we have 1644 TeraWatt (TW). For comparison, a large electrical power plant produces 1 GigaWatt (GW) of electrical power. One TW equals the output of one thousand of such 1 GW power plants. So the heat retention by greenhouse gases in 2019 equals the electrical output of 1.64 million large power plants. Global electricity production from all power plants in 2019 was 3.15 TW (extrapolated from 2018, BP Statistical Review 2019).
Let’s also compare 1644 TW with all the direct heat produced from all energy uses, the production of electricity including nuclear, transportation, heating/cooling of buildings, industrial processes, biofuels, waste. That total is 18.1 TW in 2019 [Ritchie, 2014]. Therefore, the excess heat retention by greenhouse gases in 2019 was 91 times larger than all the direct heat produced by humanity.
On a personal level, when we burn one gallon of gasoline, and if we would collect the CO2 coming out of tail pipe, it would fill a balloon with a diameter of about 7 feet. That CO2 will be capturing infrared heat in the atmosphere day in day out, for millennia to come. In the first 100 years the total amount of heat captured by that CO2 in the atmosphere (accounting for the portion that leaves the atmosphere by transferring into the oceans) is 120 times larger than the heat released by the original burning of the one gallon. After 500 years it is 350 times larger.
Now we will take a look at what 1644 TW could do in the climate system. If all of that energy were (hypothetically) directed into the Greenland ice cap, in one year it would heat up the ice, and then melt, 5.0% of the Greenland ice cap, which would raise global sea level by 36 cm, or 14”. How much ice is contained in the ice cap? The volume of Greenland’s ice cap could cover the U.S. states of Georgia, Tennessee, Missouri, Iowa, South and North Dakota, and all states to the north and east of that “line” with a 1 km (or 5/8 of one mile) thick layer of ice. When all of that ice melts it would raise global sea level by 7.4 m, or 24 ft.
Alternatively, the energy could go toward heating the upper layers of the oceans. In one year the upper 100 m of all oceans would warm by 0.35 degree centigrade (°C , or 0.64 °F). If all of the energy could be aimed exclusively at the Great Lakes in North America (their water volume is ~22,600 km3), they would completely evaporate in 14 months.
In the above examples, all of the heat was applied to just one purpose, either to melt ice, or to warm sea water, etc. To be sure, we have observed heating of sea water, loss of glaciers and ice sheets, sea level rise, and melting of sea ice plus warming of permafrost soils in the Arctic. So the excess heat gets spread out over multiple things. Furthermore, the Earth’s land surface has warmed up from the 1960s and 1970s by 1.45 °C (2.6 °F), and the ocean surface by 0.70 °C (1.3 °F). A warmer surface emits more radiation, so that at the wavelengths where greenhouse gases do not absorb (the so-called “window regions”) a portion of the excess heat is sent to space and is thus not available to heat the surface. In addition, there are additional climate forcings that are directly influenced by human activities. Greatly increased loading of the atmosphere with small particles (aerosols), clearly caused by human activities, is providing a counterforce. The aerosols scatter light directly, some of it to space, and, on average, enhance the brightness of clouds somewhat. The latter is a fast response of the climate system. It is currently still difficult to quantify the magnitude of these effects. Replacing forests by crop lands increases the reflectivity (albedo) of the surface, so that less sun light is absorbed. At the same time decreasing ice and snow cover (a climate response) leaves the surface darker so that more sun light is absorbed. As already mentioned, there is the known feedback from water vapor; a warmer surface evaporates more, and water vapor absorbs infrared radiation, thus enhancing the effect of all the long-lived greenhouse gases while also injecting more energy (latent heat) in the atmosphere from the condensation of water vapor.
Earth transitioned from the last ice age to the warm interglacial period (the Holocene) we are currently in over a period of 6000 years, between 17500 yr BP (years before present) and 11500 yr BP, the start of the Holocene. The increase in GHG forcing was 2.5 W/m2 and global average temperature increased by 2.5 °C (4.0 °F) between these dates, with GHG forcing leading temperature [Shakun et al., 2012]. What will today’s 3.2 W/m2 do? A striking difference between today’s increase and the last de-glacial transition is the speed, which averaged 0.41 W/m2 over the last decade, whereas it averaged 0.0042 W/m2 per decade during the deglacial. Seen from a geological perspective, the current increase is an explosion.
The above comparisons demonstrate that our collective activities have a large impact on the Earth’s climate, both heating and cooling. Indeed, the global average heating since 1970 has been very rapid compared to the temperature changes we have observed in ice cores over the last 800,000 years. All cumulative CO2 emissions from fossil fuel burning from 1800 until 1970 were only 25% of what they are today. In other words, 75% of all emissions took place during one human lifetime. There is no question that humanity is collectively responsible for the recent changes, but not in the sense that we have control over how the climate system responds to our interventions. We do not understand the climate system and ecosystems well enough. In recent years we have seen extreme rainfall events, as well as extreme droughts, massively destructive fires, temperatures over 100 °F for several weeks in a large region of northern Siberia. The frequency and severity of these events were not predicted as little as a decade ago. It does not bode well for how “manageable” global heating might be. Fortunately it is well-known how we can replace the emissions of CO2 and other gases with renewable energy, and how to improve energy conservation and efficiency. Will we fulfill our responsibility toward future generations?