What factors contribute to increases in planetary temperature?

As a retired physicist who has given considerable thought to the question of planetary temperature, I can offer an answer that is scientifically rigorous and, in some sense, complete. Physics tells us there are only a few fundamental factors that drive global temperatures.

It is helpful to define what I will call the “energy-balancing temperature”, 𝑇ebal. This is the average global temperature that the planet’s surface would need to have in order for the rate at which energy leaves the planet by escaping into space to exactly balance the rate at which energy arrives into the planet’s atmosphere, oceans, and land.

The concept of the energy-balancing temperature is useful because whenever the energy-balancing temperature is higher than the actual average global surface temperature, 𝑇, there is an excess of energy arriving, which leads to energy accumulating and temperatures rising. In contrast, whenever the energy-balancing temperature is lower than the actual temperature, there is an excess of energy leaving, which leads to energy draining away and temperatures falling.

Thus, there is a strong tendency for the actual temperature to move towards the energy-balancing temperature.

It helps to divide the overall question into sub-questions:

What factors contribute to increases in the energy-balancing temperature, $T_\mathrm{ebal}$ ?
What factors affect the movement of the actual temperature, $T$ , towards the energy-balancing temperature, $T_\mathrm{ebal}$ ?

A. Factors affecting the energy-balancing temperature

The energy-balancing temperature is affected by three factors that determine the rate at which energy arrives:

The intensity of sunlight reaching Earth. This varies over the course of each year (by about 7%), as Earth’s distance from the Sun varies. (Earth is closest to the Sun in January.) The output of the Sun also varies (by about 0.1%) over an 11-year solar cycle. There are additional variations that occur over longer time scales.
The planetary albedo, which is a term for the fraction of sunlight reflected back out to space.
Non-solar heating, or heat from sources other than the Sun, such as geo-thermal heat and energy released by human burning of fossil fuels.

While all three of these factors have effects in principle, in practice, albedo (reflectivity) is the only heating factor changing enough to play a role in Earth’s current climate change.

The intensity of sunlight reaching Earth has not changed enough in recent decades to account for global temperature changes. On average, incoming sunlight hasn’t changed significantly.
Non-solar heating is 1,700 times smaller than the rate at which Earth absorbs heat from the Sun. There is not enough non-solar heating to significantly affect the planet’s temperature.
Earth’s albedo changes when sea ice coverage, atmospheric dust (aerosols), or cloud patterns change. All of these have changed in recent decades.

The energy-balancing temperature is affected by three factors that determine the rate at which energy leaves the planet, for a given average global temperature, $T$ :

The emissivity of Earth’s surface. Each material that makes up Earth’s surface is characterized by an “emissivity” value which indicates how efficiently the material radiates thermal energy. (If you look at two objects of the same temperature with an infrared scope, the object with a higher emissivity would seem brighter.)
The degree of variation in temperature between different places and different times. For a given average temperature, Earth’s surface as a whole would emit more thermal radiation overall if temperature varies than it would if the temperature was uniform and consistent. The level of temperature variation that exists on Earth increases the surface’s emission rate by about 1.8 percent.
The greenhouse effect, which is the name given to the atmosphere’s tendency to reduce the amount of thermal radiation that reaches space relative to what leaves the surface. On Earth, about 40% less thermal radiation reaches space than what leaves the surface. That 40% reduction is the greenhouse effect.

While all three of these factors have effects in principle, the greenhouse effect is the only energy-loss-rate factor capable of changing enough to play a role in Earth’s current climate change.

B. Factors affecting the greenhouse effect

The greenhouse effect is basically an “optical” phenomenon, like phenomena associated with the propagation of light. A scientist can use the laws of physics (e.g., Schwarzschild’s equation for radiation transfer) to compute how much thermal radiation will be emitted to space, provided the composition and temperature profile of the atmosphere are known.

The greenhouse effect is affected by these factors:

The concentration of what I’ll call “persistent” greenhouse gases, such as carbon dioxide and methane, which tend to remain in the atmosphere for long periods of time.
The humidity, i.e., concentration of water vapor, as a function of altitude. Water vapor is what I call a “responsive” greenhouse gas, insofar as its atmospheric concentration responds to temperature changes. Warmer temperatures tend to lead to more water vapor being present in the atmosphere. This is a “feedback process” in which any warming that occurs leads to more water vapor in the atmosphere, which leads to further warming. (This process increases warming, but does not lead to an unlimited run-away warming process.)
Cloud coverage. Clouds are currently responsible for about 16% of the greenhouse effect. High clouds contribute more to the greenhouse effect than do low clouds.
The lapse rate, or rate of decrease in temperature with increasing altitude. The typical lapse rate on Earth is about 6.5℃/km, but this varies by place and time. The greenhouse effect would be zero if the lapse rate was zero. Variations in the lapse rate cause the greenhouse effect to vary.

All four of those factors are considered by climate models.

Humans have had a direct effect of increasing the concentrations of persistent greenhouse gases.
Humidity and lapse rate changes are “climate feedbacks”, i.e., factors that change in response to temperature changes, leading to additional temperature changes. Both measured data trends and climate models indicate that, in combination, humidity and lapse rate changes tend to amplify any global temperature change that occurs.
Cloud changes also are considered to be “feedbacks” which could in principle amplify or mitigate temperature changes. In practice, over the past few decades, cloud cover has changed in ways that have decreased both albedo and the greenhouse effect. These two effects have largely canceled out one another, so that changing cloud cover has not had any significant net effect on planetary temperature (over the period 2000-2023).

C. Factors affecting shift towards the energy-balancing temperature

How quickly the actual temperature shifts towards the energy-balancing temperature is governed by:

The heat capacity and heat transport ability of the materials whose temperature is changing. The materials that make up land surfaces have moderate heat capacities and relatively low thermal conductivity. So, temperatures on land are able to change quickly to shift towards the energy-balancing temperature. However, water has a high heat capacity, and water’s ability to circulate give it a high heat transport ability. The result is that the oceans take decades to centuries (or longer, for the deep ocean) to shift temperature towards the energy-balancing temperature. Since heat can flow between land and ocean regions, the oceans also moderate temperature changes on land, to some degree.

Those factors affect the rate of warming, but do not alter the tendency of temperature to shift towards the energy-balancing temperature.

In the short-term, temperatures can be significantly affected by:

Changes in heat exchange between the ocean surface and deeper waters. This seems to be what produces the periodic El Niño and La Niña phenomena, which affect temperatures in the Pacific Ocean and on adjacent land masses.

D. Temperature fluctuations

While factors like the concentration of carbon dioxide in the atmosphere may change in a smoothly increasing fashion, this does not mean we should expect the actual temperature to change in an equally smooth fashion. The actual global temperature fluctuates, while tending to increase on average to reflect increasing greenhouse gas concentrations.

These natural fluctuations in global temperature occur because there are natural fluctuations in things like:

ocean currents (which affect humidity/lapse rate/clouds/greenhouse effect, sea ice/albedo, and ocean heat exchange between surface waters and deeper waters);
clouds (which affect both albedo and the greenhouse effect);
sea ice extent (which affects albedo)

These phenomena give rise to fluctuations in the energy-balancing temperature (when albedo or the greenhouse effect fluctuates), as well as fluctuations of temperature relative to the energy-balancing temperature.

These sort of natural fluctuations mean that global temperature should not be expected to rigidly track the rising concentration of atmospheric carbon dioxide. However, both the energy-balancing temperature and the actual temperature tend to rise as the concentration of persistent greenhouse gases (like carbon dioxide) increases.

E. Conclusion

The list of the factors which directly affect the energy-balancing temperature is scientifically rigorous and complete.

Many other factors can influence the energy-balancing temperature indirectly, but only by influencing one of the primary factors, such as planetary albedo or the greenhouse effect.

Changes in heat transfer within the oceans can cause short-term fluctuations in temperature, but have only an indirect influence on the long-term energy-balancing temperature.

While there are a number of factors that can affect planetary temperature, there is no need to guess about which factors are currently leading to Earth’s warming. One can measure the various quantities involved, to see which are responsible for the increase in Earth’s energy-balance temperature. Doing such an analysis of data from the period 2000–2023 indicates:

The largest contribution to increases in the global energy-balancing temperature has arisen from an increase in the non-cloud component of the greenhouse effect, i.e., the effect which we would expect to increase due to the increase in atmospheric carbon dioxide.
The second-largest contribution to increasing temperatures has come from a decrease in Earth’s non-cloud albedo. This change seems to be a result of decreases in sea ice extent and snow coverage, presumably as a result of increasing temperatures. So, this change is a “feedback” which has amplified the effect of warming from other causes.
Although cloud coverage changed significantly over this period, cloud effects on albedo and the greenhouse effect largely cancelled out, so that cloud changes had a slight net cooling effect overall.

APPENDIX. Math

If math is a turn-off for you, please ignore this appendix. However, I’ll offer a few key equations to support those who enjoy understanding how things work quantitatively.

The global energy-balancing temperature of a planet is given by

$T_\mathrm{ebal} = \sqrt[4]{\frac{H}{\epsilon_\mathrm{eff}\,\sigma}}$

where the rate of heat arriving, $H$ , is given by

$$H = (1-a)\,S + J$

and the planetary effective emissivity, $\epsilon_\mathrm{eff}$ , is given by

$\epsilon_\mathrm{eff} = (1-g)\,(1+v)\,\epsilon_s\,$

In these equations, $S\approx 340$ W/m² is the intensity of sunlight before it enters the atmosphere, averaged over the surface of the globe; $a \approx 0.29$ is the planetary albedo (fraction of sunlight reflected back to space); $J \approx 0.14$ W/m² is the rate of heat arriving from geothermal energy and human energy production; and $\sigma = 5.67\times10^{-8}$ W m⁻² K⁻⁴ is the Stefan-Boltzmann constant.

The planetary effective emissivity, $\epsilon_\mathrm{eff}\approx 0.61\times\epsilon_s$ , is a measure of how efficiently the planet emits thermal radiation into space, for a given surface temperature, $T$ . It is composed of three terms:

The surface emissivity, $\epsilon_s\approx$ 0.94, is a weighted global average of the emissivity of the materials that make up Earth’s surface.
The temperature variation factor, $v \approx$ 0.018, accounts for fact that, for a given average temperature, emissions will be larger if there are variations in temperature, e.g., the variation between tropical and polar temperatures. This factor is defined by $1+v = \left< \underline T^4 \right>/ \left< \underline T \right>^4$ where angle brackets indicate averaging. This factor makes Earth about 1℃ colder than it would be if Earth’s temperature was uniform. (The most important temperature variations are associated with differences between tropical and polar temperatures, though differences between January and July, and between day and night, also play a minor role.)
The normalized greenhouse effect, $g\approx 0.40$ , is the fraction by which thermal radiation reaching space is less than the amount of thermal radiation leaving the surface. The flux of thermal radiation reaching space is about 60% as large as the flux that leaves Earth’s surface.

Note that the temperature, $T$ , which shifts towards the energy-balancing temperature is technically the surface “skin” temperature, i.e, the temperature of a thin layer of material (maybe 20 microns thick) at the interface between the surface and the air. The “skin” temperature is likely to be warmer than the more commonly reported “near air surface temperature”, though the skin temperature and near-surface air temperature are strongly correlated.

Earth’s Energy Imbalance, $\mathrm{EEI} \approx 1$ W/m2, is the rate at which Earth’s net energy is increasing. This is directly tied to the difference between the energy-balancing temperature and the current temperature:

$\mathrm{EEI} = \epsilon_\mathrm{eff}\,\sigma\,\left( T^4_\mathrm{ebal} - T^4 \right)$

There is a local energy-balancing temperature, ${\underline T}_\mathrm{ebal}$ , specific to a particular place and time. The local temperature, $\underline T$ , will tend to shift towards ${\underline T}_\mathrm{ebal}$ . When computing the local energy-balancing temperature, one needs to take into account advection, $\underline L$ , i.e., local heat delivery due to oceanic or atmospheric circulation. Local heating is given by

${{\underline H}} = (1-{{\underline a}})\,{{\underline S}} + {{\underline J}} + {{\underline L}}\,$

The local effective emissivity is given by

${\underline\epsilon}_\mathrm{eff} = (1-{\underline g})\,{\underline\epsilon}\,$ .

The equations I’ve offered in this appendix are completely rigorous, without any approximations or assumptions—beyond the assumption that energy is conserved, as must always be the case. The rest is simply a direct mathematical consequence of the definitions of the various quantities.

Note: this essay originally appeared as an answer on Quora.