Science Wiki

Atmospheric water vapor is water in its gaseous phase in the atmosphere. It is a variable constituent that makes up about 0.25 percent of the atmosphere by mass on average (about 25 kg·m⁻² of precipitable water in the global mean), and from a few parts per million in the dry lower stratosphere to about 4 percent by volume in the warm, humid tropical boundary layer. It is the most important natural greenhouse gas, accounting for roughly half of the clear-sky greenhouse effect, and with clouds roughly three quarters of it. It is also the working fluid of the water cycle, and the release of latent heat when it condenses is a major source of energy for weather systems. Water vapor is studied in atmospheric physics, meteorology, climatology, and Earth system science.

Unlike the long-lived, well-mixed greenhouse gases, such as carbon dioxide, methane, and nitrous oxide, water vapor is a condensable gas with a residence time in the atmosphere of only about 8 to 10 days. Its concentration varies strongly in space and time and is determined mainly by temperature and by the supply of moisture from evaporation and transpiration, and by atmospheric transport and precipitation, and not by direct emissions.

Several variables describe the amount of water vapor:

• The vapor pressure e is the partial pressure of water vapor, and the saturation vapor pressure e_s(T) is the maximum partial pressure at which vapor can coexist with a flat surface of liquid water (or ice) at temperature T.
• The relative humidity is RH = e / e_s (expressed as a percentage). It depends on both the moisture content and the temperature, so that air that is cooled at a constant vapor pressure reaches saturation at the dew point.
• The specific humidity q is the mass of water vapor per unit mass of moist air, and the mixing ratio r is the mass of vapor per unit mass of dry air. They are related to e and to the total pressure p by q ≈ 0.622 e / p, where 0.622 is the ratio of the molar masses of water (18.02 g·mol⁻¹) and dry air (28.97 g·mol⁻¹).
• The precipitable water is the depth of liquid water that would result from condensing all the vapor in a column. It is about 25 mm globally on average, near 50 mm or more in the tropics, and a few millimeters or less in polar regions and deserts.

Saturation vapor pressure

For a condensable gas in equilibrium with its liquid, the saturation vapor pressure rises steeply with temperature according to the Clausius-Clapeyron relation:

where L_v is the latent heat of vaporization (about 2.5 × 10⁶ J·kg⁻¹ at 0 °C) and R_v is the specific gas constant of water vapor (461.5 J·kg⁻¹·K⁻¹). The fractional rate of change is therefore:

For T = 288 K, this gives 2.5 × 10⁶ / (461.5 × 288²) ≈ 0.065 K⁻¹, which is about 6 to 7 percent per kelvin. The percentage is smaller at higher temperatures (about 6 percent per K near 300 K) and larger at low temperatures (about 7.5 percent per K near 273 K). The statement of "approximately 7% per 1 °C" is a good approximation for typical near-surface conditions. An integrated form with a constant L_v is:

with e_s,0 = 6.11 hPa at T₀ = 273.15 K. For example, at 293 K (20 °C), this gives e_s ≈ 6.11 × exp[5,417 × (1/273.15 − 1/293)] ≈ 6.11 × exp(1.34) ≈ 23 hPa, which is close to the accepted value of 23.4 hPa. Air at 30 °C can hold about 1.8 times as much water vapor as air at 20 °C.

Not a limit on water vapor as such

The Clausius-Clapeyron relation gives the maximum water vapor pressure at equilibrium. It does not mean that the atmosphere is always saturated. The mean relative humidity of the troposphere is about 50 percent, and it varies from below 10 percent in the subsiding dry regions of the subtropics to near 100 percent in clouds. The climate models, and the observations, show that the relative humidity changes only slightly as the climate warms, so that the specific humidity increases at a rate near the Clausius-Clapeyron value. Water vapor in the atmosphere responds to temperature on timescales of days, because the excess condenses and falls as precipitation, and this is why it acts as a feedback and not as an independent driver of long-term climate change. Water vapor does not "unilaterally" start warming, but it also can be forced locally and for short periods, for example by evaporation from irrigation, and this has only a small effect on the global climate, because the extra vapor is removed within days.

The global mean precipitation does not increase at the same rate as water vapor with warming. It is limited by the energy balance of the atmosphere, as the latent heat released in precipitation must be balanced by radiative cooling, and it increases at about 1 to 3 percent per kelvin in models. As a result, the atmospheric residence time of water vapor increases with warming, and extreme precipitation, which depends on the moisture content of the air, increases at a rate closer to the Clausius-Clapeyron value in many regions.

Water vapor absorbs and emits infrared radiation in the vibration-rotation bands near 6.3 μm and in the pure rotation band at wavelengths longer than about 15 μm, and weakly in the continuum of the 8 to 12 μm window, where its absorption increases with the square of the vapor pressure. It also absorbs near-infrared sunlight (near 0.94, 1.1, 1.4, 1.9, and 2.7 μm), which accounts for most of the solar absorption in the troposphere.

Water vapor is the largest contributor to the natural greenhouse effect, of the order of half of the clear-sky effect, but it is not an independent forcing, because its concentration is controlled by the temperature. A feedback analysis defines the "water vapor feedback" as the change in the radiative flux caused by the change in water vapor with surface temperature. Its strength is about +1.8 W·m⁻²·K⁻¹ (with an uncertainty range of about ±0.2), the largest positive feedback in the climate system. Because the absorption of infrared radiation by water vapor depends approximately logarithmically on its concentration, the change in the flux is smaller than the increase in the concentration. The feedback increases the warming that is caused by a forcing by a factor that, together with other feedbacks, results in an equilibrium climate sensitivity of about 3 K per doubling of CO₂ (a likely range of 2.5 to 4 K), and it roughly doubles the warming that would result from the Planck response alone. In terms of feedback factors, if the sum of the feedbacks other than the Planck response is λ_f and the Planck response is λ_P (about −3.2 W·m⁻²·K⁻¹), then the temperature change for a forcing ΔF is:

For ΔF = 3.7 W·m⁻² (a doubling of CO₂) and λ_P = −3.2, the warming without feedbacks is 1.2 K. With a water vapor feedback of +1.8, the net parameter becomes −1.4 and the warming 2.6 K. The lapse rate feedback (about −0.5 W·m⁻²·K⁻¹) partly offsets the water vapor feedback in the upper troposphere, and the combined water vapor and lapse rate feedback is about +1.1 W·m⁻²·K⁻¹. Observations show an increase in water vapor in the upper troposphere and in the total column in recent decades, consistent with this feedback, from satellites, radiosondes, and surface stations.

Historically, the question of whether water vapor causes the greenhouse effect or amplifies it was settled in the work of Manabe and Wetherald in the 1960s, whose calculations with fixed relative humidity showed a large amplification of the response to CO₂. The runaway greenhouse is a related but different limit, in which the water vapor feedback does not saturate and the oceans evaporate, which is expected on Earth only if the solar flux is much greater, by the order of 10 percent or more, as it may be in about a billion years.

Evaporation, condensation, and energy transport

Evaporation of water from the oceans and from land and plants requires energy. The latent heat of vaporization, about 2.5 × 10⁶ J per kilogram, is taken from the surface, which cools it, and it is released in the atmosphere where the vapor condenses. The global mean flux is about 80 to 90 W·m⁻², as part of the energy budget of the surface. For example, an evaporation rate of 3 mm per day (3 kg·m⁻² per day) corresponds to a flux of 3 × 2.5 × 10⁶ / 86,400 ≈ 87 W·m⁻². The statement that evaporation stores "solar energy to break molecular bonds" is a loose description, because the energy is needed to overcome the intermolecular attraction (hydrogen bonds) between the molecules in the liquid, and it is not the energy of the bonds in the molecules, which remain unchanged.

Water vapor is transported by the winds from regions of net evaporation (the subtropical oceans) to regions of net precipitation (the tropics and mid-latitudes), and the latent heat is transported with it. The transport of moisture is concentrated in narrow filaments of high vapor transport, atmospheric rivers, at mid-latitudes. In the zonal mean, the poleward transport of latent heat and dry static energy together carries energy from the tropics to higher latitudes, and the Hadley cell and the mid-latitude eddies are the main agents. The moist static energy is the quantity that includes the contributions of sensible heat, geopotential, and latent heat:

Convection and storms

When moist air rises, it cools, reaches saturation, and the condensation releases latent heat, which makes the air more buoyant than a dry parcel. This is why the moist adiabatic lapse rate (about 4 to 7 K per km in the lower tropical troposphere) is less than the dry rate (9.8 K per km), and it is the source of the energy of deep convection and of thunderstorms. In tropical cyclones, the evaporation from warm ocean water supplies the energy, and the release of latent heat in the eyewall drives the circulation, so that they form over water at temperatures of about 26.5 °C or more, and weaken over land. Mid-latitude cyclones are driven mainly by horizontal temperature contrasts (baroclinic instability), and latent heat release contributes to their intensity, and so the statement that latent heat "generates" the low-pressure systems of mid-latitude cyclones is not accurate. The conversion of latent heat into the kinetic energy of the winds takes place through the buoyancy of the air, by the warming of the atmosphere, and not directly as a conversion of latent into "sensible" heat and then into motion.

The concentration of water vapor falls rapidly with height, because the temperature falls. Most of the vapor is in the lowest few kilometers of the troposphere, and the scale height is about 2 km. The temperature at the tropical tropopause (about 16 to 17 km) is about −75 to −80 °C, and it is lower than that in the mid-latitudes, and not −60 °C, so that the saturation mixing ratio there is very low. Air enters the stratosphere mainly through the tropical tropopause layer, where it is dried by freeze-drying, and the water vapor mixing ratio that results is about 3 to 4 parts per million by volume. This is the "cold trap". The stratospheric entry mixing ratio is therefore set by the lowest temperatures along the trajectory of the air, and it varies with the seasonal cycle, the quasi-biennial oscillation, and ENSO.

Water vapor in the stratosphere increases with height, because of the oxidation of methane, to about 5 to 6 parts per million by volume in the middle stratosphere. Each methane molecule that is oxidized yields about two water molecules (and some hydrogen), and the oxidation is mainly by the OH radical, atomic oxygen, and chlorine. The increase of methane in the atmosphere has increased this source of water. The radiative effect of stratospheric water vapor was estimated to be a forcing of about +0.07 W·m⁻² for the change in water from the oxidation of the additional methane, which is a modest quantity and is included within the radiative forcing of methane. It is therefore not a highly potent separate forcing. Stratospheric water vapor also takes part in the chemistry of ozone, because it is a source of the OH and HO₂ radicals, and it is a component of polar stratospheric clouds. Volcanic eruptions such as that of Hunga Tonga-Hunga Ha'apai in January 2022 injected an exceptional quantity of water vapor, of the order of 150 teragrams (about 10 percent of the stratospheric burden), into the stratosphere, and its radiative effect was a cooling of the stratosphere and slight surface effect, as studied in several analyses, and the vapor remained for years.

Water vapor is observed by radiosondes, by ground-based GPS networks (which measure the delay of the signal and give the total column), by microwave and infrared sounders on satellites (such as AIRS, IASI, and SSM/I), and by limb sounders for the upper troposphere and stratosphere. Reanalyses combine these data with models. The total column water vapor over the oceans has increased by about 1 to 2 percent per decade since the late 1980s, which is consistent with the Clausius-Clapeyron rate for the observed warming, and the increase is detected in the upper troposphere as well, though trends in the upper troposphere and over land are less certain because of changes in the instruments. Over land, observed relative humidity has fallen in recent decades since about 2000, because land warms faster than the ocean, and the moisture supply from the ocean has not kept up, which contributes to the increase in the vapor pressure deficit and in the risk of fire and drought. An increase in the vapor pressure deficit raises the evaporative demand of the atmosphere on plants and soils.

The mean residence time of a water molecule in the atmosphere is the burden divided by the flux of precipitation (or evaporation). The global burden is about 1.3 × 10¹⁶ kg, and the global mean precipitation is about 5 × 10¹⁴ m³ per year of liquid water (about 5 × 10¹⁷ kg per year), so:

This short residence time, together with the Clausius-Clapeyron control, is why water vapor responds to climate and does not accumulate.

• Water cycle
• Greenhouse gas
• Clausius-Clapeyron relation
• Humidity
• Latent heat
• Climate sensitivity
• Atmospheric river