The global climate can be nearly balanced at the top of the atmosphere while individual latitudes, landscapes, and hours of the day have large energy gains or losses. Keeping the boundary clear—top of atmosphere or surface—is the key to reading every budget correctly.

Solar energy is reflected, transmitted, scattered, or absorbed

The top of the atmosphere receives about $340\ \mathrm{W\,m^{-2}}$ when sunlight is averaged over Earth’s full sphere. Roughly 30% is reflected to space by clouds, the atmosphere, aerosols, and the surface. The remaining 70% is absorbed by the atmosphere and surface.

Exact numbers differ slightly among datasets, but a useful rounded shortwave budget is:

The entries sum to 100%. Reflection returns energy to space; atmospheric absorption heats gases and clouds; surface absorption supplies energy that can later become longwave radiation, sensible heat, latent heat, chemical energy, or stored heat.

Albedo controls absorbed shortwave radiation

Albedo $\alpha$ is the reflected fraction of incoming shortwave radiation:

$$\alpha=\frac{SW^\uparrow}{SW^\downarrow}.$$

For an opaque surface with negligible transmission,

$$SW_{\mathrm{net}}=SW^\downarrow-SW^\uparrow =(1-\alpha)SW^\downarrow.$$

Suppose radiometers above a hay field measure $SW^\downarrow=800\ \mathrm{W\,m^{-2}}$ and $SW^\uparrow=160\ \mathrm{W\,m^{-2}}$. Then

$$\alpha=\frac{160}{800}=0.20,$$

and the field absorbs

$$800-160=640\ \mathrm{W\,m^{-2}}.$$

Fresh snow with $\alpha\approx0.85$ under the same sunlight absorbs only

$$(1-0.85)(800)=120\ \mathrm{W\,m^{-2}}.$$

Albedo varies with wavelength, solar angle, moisture, roughness, vegetation, snow age, and contamination. For semitransparent media such as water, ice, snow, and vegetation canopies, the fuller relation is reflectance + absorptance + transmittance = 1.

The atmosphere has bands and windows

The atmosphere is not simply transparent to shortwave and opaque to longwave. Its behaviour depends on wavelength.

• Ozone absorbs much ultraviolet radiation.
• Water vapour absorbs broad infrared regions.
• Carbon dioxide absorbs strongly in bands including the region near $15\ \mu\mathrm{m}$.
• Methane and nitrous oxide absorb in additional bands.
• Visible wavelengths pass through the clear atmosphere relatively well.
• A thermal-infrared atmospheric window, roughly $8$–$13\ \mu\mathrm{m}$, allows some surface radiation to escape more directly to space.

The window is not completely open: water vapour and ozone absorb within parts of it, and clouds absorb and emit broadly across thermal infrared wavelengths.

The natural greenhouse effect

Earth’s surface absorbs solar radiation and emits thermal infrared radiation. Greenhouse gases and clouds absorb some of that infrared and emit infrared both upward and downward. Downward atmospheric radiation adds energy to the surface; upward emission eventually carries energy toward space.

Earth’s absorbed-solar and outgoing-longwave balance corresponds to an effective radiating temperature of about $255\ \mathrm{K}$, or $-18\,^{\circ}\mathrm{C}$. In a simplified no-greenhouse estimate with the same planetary albedo, the surface would also approach that value. The observed global mean surface temperature is closer to $288\ \mathrm{K}$, or $15\,^{\circ}\mathrm{C}$. The difference is not caused by radiation being permanently trapped. The atmosphere changes the height and temperature from which much of Earth’s infrared energy escapes. Because the troposphere is generally colder aloft, raising the effective emission level reduces outgoing radiation until the surface–troposphere system warms enough to restore balance.

Adding greenhouse gases creates a radiative forcing: an imposed change in the planetary energy budget relative to a reference state. Radiative forcing is not the same as the instantaneous net radiation measured over one wetland or parking lot.

Top-of-atmosphere budget

At the top of the atmosphere,

$$Q^*_{\mathrm{TOA}} =SW^\downarrow-SW^\uparrow-LW^\uparrow.$$

There is effectively no climatologically important $LW^\downarrow$ arriving from space. In an older rounded equilibrium budget,

$$342=107+235\ \mathrm{W\,m^{-2}},$$

where $342\ \mathrm{W\,m^{-2}}$ is incoming solar radiation, $107\ \mathrm{W\,m^{-2}}$ is reflected shortwave, and $235\ \mathrm{W\,m^{-2}}$ is outgoing longwave.

Present-day Earth has a small positive imbalance: absorbed solar energy exceeds outgoing longwave energy. That residual is tiny beside the gross fluxes of hundreds of watts per square metre, but it accumulates across the planet and through time. Most of the added energy enters the ocean; smaller shares warm land and atmosphere or melt ice.

Global balance does not imply balance at every latitude. Low latitudes absorb more radiation than they emit, while high latitudes emit more than they absorb. Atmospheric circulation, storms, water vapour, and ocean currents carry energy poleward, preventing an ever-hotter equator and ever-colder poles.

Surface radiation budget

At the surface, both shortwave and longwave have upward and downward terms:

$$Q^*=SW^\downarrow-SW^\uparrow+LW^\downarrow-LW^\uparrow.$$

Here:

• $SW^\downarrow$ is sunlight reaching the surface;
• $SW^\uparrow$ is reflected sunlight;
• $LW^\downarrow$ is infrared emitted downward by the atmosphere and clouds;
• $LW^\uparrow$ is infrared emitted upward by the surface.

This equation uses positive magnitudes and explicitly subtracts upward fluxes. Some graphs instead plot upward fluxes as negative. In that convention the signed terms are added, so an upward term must not be subtracted a second time.

Wetland example

At noon, suppose

$$SW^\downarrow=800,\quad SW^\uparrow=80,\quad LW^\downarrow=350,\quad LW^\uparrow=450$$

in $\mathrm{W\,m^{-2}}$. Then

$$Q^*=800-80+350-450=620\ \mathrm{W\,m^{-2}},$$

and the albedo is $80/800=0.10$.

At midnight, shortwave terms are zero. If $LW^\downarrow=320$ and $LW^\uparrow=380\ \mathrm{W\,m^{-2}}$, then

$$Q^*=320-380=-60\ \mathrm{W\,m^{-2}}.$$

The negative value means net radiative cooling. If the surface cools to the dew point or frost point, condensation or deposition can occur. Clouds often keep nights warmer because they increase $LW^\downarrow$, but wind, humidity, advection, and cloud temperature also matter.

Net radiation must go somewhere

A simplified surface energy balance is

$$Q^*=Q_H+Q_E+Q_G,$$

or equivalently $R_n=H+LE+G$.

• Sensible heat $Q_H$ changes the temperature of air or another material. At the surface, molecular conduction acts across a very thin layer and turbulence carries most of the heat into the atmosphere.
• Latent heat flux $Q_E=LE$ is the turbulent energy transfer carried by water-vapour exchange. Evaporation and sublimation absorb energy at the surface; condensation and deposition release it when the vapour changes phase elsewhere.
• Ground heat $Q_G$ conducts energy into or out of soil, rock, water, snow, or built surfaces.

Here $Q_H$ and $Q_E$ are positive upward from the surface, while $Q_G$ is positive downward into the substrate. These are signed fluxes, unlike the positive radiation magnitudes that were explicitly added or subtracted above. Sensible heat can be negative: warmer air over a colder upwelling ocean can transfer heat downward to the surface. Sign conventions differ among disciplines, so every budget must state its choice.

A complete local budget may also need storage in vegetation, water, buildings, snow, or the near-surface air; horizontal advection; melt or freezing energy within the surface; and anthropogenic heat. Melting and freezing are latent-heat processes, but they belong in a separate phase-change or storage term unless the defined $Q_E$ explicitly includes them. Measurement errors also make real field budgets fail to close perfectly.

Phase changes move large amounts of energy

Water absorbs latent heat during melting, evaporation, and sublimation. It releases latent heat during freezing, condensation, and deposition. Representative values are

• fusion: about $0.334\ \mathrm{MJ\,kg^{-1}}$;
• vaporization: about $2.45\ \mathrm{MJ\,kg^{-1}}$ near ordinary environmental temperatures;
• sublimation: about $2.8\ \mathrm{MJ\,kg^{-1}}$.

Evaporation cools the surface, but it does not destroy energy. Water vapour carries that latent energy until condensation releases it elsewhere, often higher in the atmosphere.

Water availability controls energy partition

Two surfaces can receive similar net radiation but develop different temperatures because they divide that energy differently.

A wet, vegetated landscape can spend much of $Q^*$ on evapotranspiration, so $Q_E$ is large and $Q_H$ is smaller. A dry desert has little water available, so $Q_E$ is small and more energy becomes sensible and ground heat. The dry surface and nearby air warm strongly.

The Bowen ratio summarizes the split:

$$B=\frac{Q_H}{Q_E}.$$

• A low Bowen ratio is typical of wet surfaces where latent heat dominates.
• A high Bowen ratio is typical of dry surfaces where sensible heat dominates.

This is why a watered park can feel cooler than a nearby parking lot. The park uses more available energy for evapotranspiration; the pavement directs more into surface heating, sensible heat, and storage. Vegetation, stomatal resistance, humidity, wind, soil texture, rooting depth, and season also affect the split.

Reading spatial and daily patterns

The usual broad patterns follow from available energy and water:

• Net radiation generally decreases from tropics to poles.
• Latent heat flux is large over warm oceans and moist, vegetated tropical regions.
• Sensible heat flux is large over hot, dry continental regions.
• Oceans usually direct more energy into evaporation than deserts do.
• A midlatitude daily budget is commonly negative at night, turns positive after sunrise, peaks around midday, and becomes negative again near sunset.

The afternoon air-temperature maximum occurs later than the radiation maximum because the surface continues gaining more energy than it loses for several hours after solar noon. Temperature peaks only when total gains and losses become equal.