Atmospheric pressure is the pressure exerted by the gases of an atmosphere on a surface. On Earth, it results microscopically from molecular motion and, on large scales, reflects the weight of the overlying air under gravity. It acts in every direction, not merely downward, and varies with altitude, weather, and the distribution of atmospheric mass.
Pressure is a fundamental variable in describing the atmosphere of Earth. Its spatial differences help drive winds; its vertical decrease provides a basis for measuring altitude; and its absolute value influences boiling, gas exchange, and the operation of machinery. The conventional standard atmosphere is exactly 101,325 pascals, but actual atmospheric pressure is neither constant nor everywhere equal to this reference value.
Physical meaning, units, and reference pressures
Pressure is the normal force exerted per unit area. For a uniform force acting perpendicular to a flat surface, it is expressed as . A pressure of one pascal means one newton of force per square metre. In a gas close to local thermodynamic equilibrium, pressure is approximately isotropic: a small surface experiences the same normal stress regardless of its orientation. This distinguishes pressure from a directed force such as the weight of a particular object.
At the molecular scale, gas pressure arises from the transfer of momentum as molecules strike surfaces or cross an imaginary boundary within the gas. For an ideal gas, the relation between pressure, number density, and absolute temperature is
where is the number of molecules per unit volume and is the Boltzmann constant. The same relationship can be written , where is mass density and is the specific gas constant. These relations explain why compression generally raises pressure and why heating a sealed, rigid container raises its internal pressure. In the open atmosphere, expansion and redistribution of air complicate the response.
The International System of Units uses the pascal, abbreviated Pa. Meteorologists commonly use the hectopascal, with . The older millibar is exactly equal to one hectopascal. Other widely encountered units are the kilopascal, bar, millimetre of mercury, and inch of mercury. One standard atmosphere is exactly , or , approximately and inches of mercury. These equivalences are unit relationships, not statements that the atmosphere always has that pressure.
Absolute pressure is measured relative to a vacuum. Gauge pressure is the difference between a system’s absolute pressure and the surrounding atmospheric pressure:
A tyre gauge reading of zero therefore does not indicate an absence of gas pressure; it indicates equality with the surroundings. Differential pressure is the difference between any two specified pressures, neither of which need be atmospheric.
Despite its substantial magnitude, ordinary atmospheric pressure does not crush objects that have nearly equal pressure inside and outside. A horizontal square metre exposed to standard atmospheric pressure experiences a force of about on one side, but the opposing pressure usually balances it. Large structural loads develop when a pressure difference is maintained, as in an evacuated vessel or a pressurized aircraft cabin.
Discovery and the development of pressure measurement
The recognition that air has weight emerged from investigations of pumps, water columns, and the possibility of a vacuum. Suction pumps could not lift water indefinitely: under ordinary near-sea-level conditions, their theoretical lifting limit is approximately 10.3 metres, and practical limits are lower. Earlier explanations often invoked nature’s supposed resistance to empty space. Experiments during the seventeenth century instead connected the height of a supported liquid column to atmospheric pressure.
In 1643, Evangelista Torricelli demonstrated the mercury-tube experiment that became the basis of the barometer. A glass tube closed at one end was filled with mercury and inverted into a mercury reservoir. Part of the mercury remained suspended in the tube, leaving a low-pressure space above it. Because mercury is much denser than water, the supported column was only about 76 centimetres high rather than approximately 10 metres. Its height represented a balance between the weight of the mercury column and the pressure exerted by the air on the reservoir.
Blaise Pascal helped establish that this supporting pressure decreases with height. At his request, his brother-in-law Florin Périer carried out measurements on the Puy de Dôme in central France in 1648. The mercury column was lower on the mountain than at its base. This result supported the interpretation that a smaller weight of overlying air produces lower pressure at higher elevations. It also linked a laboratory experiment directly to the physical structure of the atmosphere.
Otto von Guericke’s experiments in the 1650s made the mechanical effects of atmospheric pressure conspicuous. In the famous Magdeburg hemispheres demonstration, air was removed from a pair of fitted metal hemispheres. The external atmosphere pressed them together so strongly that separating them required a large force. The experiment did not reveal a special attractive force produced by a vacuum; it demonstrated the consequences of an imbalance between external and internal pressure.
During the nineteenth century, pressure measurement became central to organized meteorology. Lucien Vidi’s aneroid barometer, introduced in 1844, used the deformation of an evacuated metal capsule rather than a liquid column. This made portable instruments and recording barographs more practical. Telegraph networks subsequently allowed observations from distant stations to be assembled into near-simultaneous weather maps. Atmospheric pressure thus changed from an important experimental quantity into a routinely mapped field, helping establish the analysis of moving storms and large-scale weather systems.
Vertical structure and the decrease of pressure with altitude
The dominant large-scale vertical balance in the atmosphere is hydrostatic equilibrium. The downward gravitational force on a thin layer of air is approximately balanced by the difference between pressure acting on its lower and upper boundaries. With height increasing upward,
Pressure therefore decreases upward, most rapidly where the air is dense. Integrating this equation upward shows that surface pressure is approximately the weight per unit horizontal area of the entire air column above the surface. Dividing standard atmospheric pressure by standard gravity gives an overlying column mass of about .
Combining hydrostatic balance with the ideal-gas relation gives a useful expression for an isothermal atmosphere:
The quantity , called the pressure scale height, is the height over which pressure decreases by a factor of if temperature and gravity are constant. For dry air at , it is approximately 8.4 kilometres. The actual atmosphere is not isothermal, so a single exponential is only an approximation over substantial height ranges.
For a layer with varying temperature, the hypsometric equation relates its pressure boundaries to its thickness:
Here is the specific gas constant for dry air and is a suitable layer-mean virtual temperature. Virtual temperature accounts for the lower density of moist air compared with dry air at the same pressure and temperature. A warmer layer is thicker between two specified pressure levels than a colder layer. This relationship is essential for interpreting upper-air weather charts and the vertical structure of temperature systems.
A standard atmosphere provides an idealized reference profile rather than a weather forecast. In the widely used standard model, sea-level temperature is , and temperature initially decreases by . Pressure is approximately at 5 kilometres and at 10 kilometres. The surface is commonly near 5.5 kilometres, meaning that roughly half the atmospheric mass lies below it and half above it, with small qualifications from variation in gravity.
Pressure-based coordinates are consequently natural in atmospheric science. Rather than comparing conditions only at fixed geometric heights, meteorologists often compare them on surfaces such as 850, 500, or 250 hPa. These surfaces rise and fall as the atmospheric temperature and mass distributions change. Hydrostatic balance remains highly accurate for many large-scale applications, although vertical accelerations become important in strong convection and other small-scale motions.
Instruments, observing practice, and pressure reduction
A mercury barometer measures pressure through the hydrostatic balance of a liquid column:
The pressure above the mercury is small but not identically zero, and precision measurements require corrections for mercury temperature, local gravity, and instrument geometry. Mercury expands as it warms, changing its density, while gravity varies with latitude and elevation. Reading the meniscus and establishing the reservoir level also introduce practical sources of error. Mercury instruments were long important reference devices, but their size and the hazards associated with mercury have encouraged replacement.
Aneroid instruments contain one or more partially evacuated, flexible metal capsules. Changes in external pressure deform the capsules, and a mechanical linkage converts the deformation into a dial reading or a continuous trace. Electronic instruments commonly use capacitive, piezoresistive, or resonant sensing elements. A thin diaphragm responds to the pressure difference between atmospheric air and a reference cavity. These systems are compact and suitable for automated stations, aircraft, radiosondes, and mobile devices.
Instrument exposure matters as much as sensor design. Wind flowing around a building, mast, or pressure inlet can create local pressure departures from the undisturbed atmospheric value. Properly designed static-pressure ports reduce these effects. Temperature sensitivity, sensor drift, and the height of the instrument above the station reference level must also be considered. Calibration establishes the relationship between the instrument response and reference pressure, while measurement uncertainty describes the remaining limitations of the reported value.
Station pressure is the pressure measured at the observing station’s elevation. Sea-level pressure is a derived quantity intended to remove much of the direct influence of terrain, allowing horizontal weather patterns to be compared across different elevations. Its calculation estimates the pressure that would exist at sea level beneath the station, using assumptions about the temperature of an actual or hypothetical intervening air column. It is therefore not simply a raw measurement shifted by a universal constant.
This distinction is especially important over high terrain. A plateau station can have low station pressure even when it lies beneath a strong anticyclonic circulation. Conversely, sea-level reduction through a deep layer of hypothetical air can introduce substantial dependence on the assumed temperature. Meteorologists therefore use additional fields, including geopotential height on pressure surfaces, to interpret weather over elevated regions.
Upper-air observations extend pressure measurement beyond the surface. Radiosondes carried by balloons observe pressure, temperature, and humidity through much of the atmosphere; pressure may be measured directly or derived within the observing system. Satellite instruments generally constrain atmospheric structure indirectly through radiation and related measurements. In numerical weather prediction, these observations are combined with dynamical models through data assimilation, producing a three-dimensional estimate rather than a simple collection of isolated readings.
Pressure gradients, winds, and weather systems
Air accelerates in response to spatial pressure differences. The pressure-gradient acceleration is
It points toward lower pressure when evaluated at the same geometric level. On a weather map, closely spaced isobars—lines joining equal pressure—usually indicate a stronger horizontal pressure gradient. They do not alone determine the wind, because Earth’s rotation, surface drag, flow curvature, and acceleration also matter.
Away from the equator and above the strongly frictional surface layer, large-scale winds often approach geostrophic balance. In this state, the horizontal pressure-gradient force is balanced by the Coriolis effect, and wind flows approximately parallel to isobars or height contours rather than directly from high to low pressure. In the Northern Hemisphere, lower pressure lies to the left of the geostrophic wind; in the Southern Hemisphere it lies to the right. Where trajectories curve strongly, centrifugal effects contribute to the balance.
Near the surface, friction weakens the wind and changes this balance. Air commonly crosses isobars toward lower pressure, producing convergence into surface lows and divergence out of surface highs. The associated vertical motion helps explain familiar weather patterns. Convergence can support ascent, cooling, and cloud formation, whereas broad subsidence often warms and dries air. Nevertheless, low pressure does not guarantee rain, and high pressure does not guarantee clear skies: moisture supply, stability, terrain, and the vertical structure of the circulation remain decisive.
Extratropical cyclones develop in regions with strong horizontal temperature contrasts and are associated with organized weather fronts. Their growth involves interactions between upper-level disturbances, temperature gradients, and the redistribution of atmospheric mass. A surface low deepens when the net mass of its overlying air column decreases. Upper-level outflow can contribute to that decrease even while air converges into the system near the ground. This mass-budget view is more precise than describing low pressure simply as the result of warm air rising.
A tropical cyclone has a different energy source and thermal structure. Heat and moisture supplied by warm ocean water, together with the release of latent heat during condensation, support a warm-core circulation. Its intense pressure gradient sustains strong winds around the centre. Central pressure is an important intensity indicator, but wind speed also depends on storm size, latitude, and the surrounding pressure field. Typhoon Tip reached a central pressure of in October 1979, a well-established example of exceptionally low sea-level pressure in a tropical cyclone.
Pressure tendency—the change over a specified time—is therefore interpreted alongside pressure itself. A rapid fall may indicate an approaching or intensifying disturbance, but its significance depends on the storm’s track, the observing location, and the larger circulation.
Daily cycles, global circulation, and climate variability
Atmospheric pressure changes on timescales ranging from seconds to centuries. Short fluctuations can accompany turbulence, gravity waves, thunderstorms, and the passage of fronts. More regular daily variations arise from atmospheric tides, which are global oscillations driven mainly by periodic solar heating. Absorption of sunlight by ozone and water vapour, together with heating at the surface, excites atmospheric waves that redistribute pressure over large distances.
The semidiurnal component, with two cycles per solar day, is particularly prominent in tropical surface-pressure records. Its amplitude is often of the order of one hectopascal, although it varies geographically and seasonally. Under undisturbed tropical conditions, pressure commonly has maxima near 10 a.m. and 10 p.m. local solar time and minima near 4 a.m. and 4 p.m. This regular variation can be obscured by stronger weather-related changes at higher latitudes. Atmospheric tides should not be confused with ocean tides, even though both involve periodic responses to external forcing.
Long-term mean pressure patterns are linked to Earth’s unequal heating and rotation. The Intertropical Convergence Zone is associated with low-level convergence and frequent deep convection. Subtropical high-pressure regions are connected with descending branches of the large-scale circulation. Farther poleward, transient cyclones and anticyclones contribute to the mean distribution. These are broad statistical patterns, not stationary belts with identical conditions at every longitude. Continents, oceans, mountains, and seasonal changes substantially modify them.
Pressure gradients also change seasonally as land and ocean warm and cool at different rates. These changes contribute to monsoon circulations, although monsoons cannot be explained by a simple land–sea pressure contrast alone. Moist convection, large-scale atmospheric dynamics, topography, and ocean conditions all contribute. The Asian summer monsoon, for example, involves a regional reorganization of circulation and rainfall rather than merely a local breeze enlarged to continental scale.
Climate variability is often described through indices based partly on pressure contrasts. The North Atlantic Oscillation characterizes variability in the pressure difference between the subtropical North Atlantic and the region around Iceland; its phases are associated with changes in Atlantic winds and storm tracks. The Southern Oscillation involves tropical Pacific pressure variability, commonly represented by the contrast between Tahiti and Darwin, and forms the atmospheric component of the El Niño–Southern Oscillation.
These pressure patterns are indicators of coupled dynamics, not independent switches controlling climate. Interpreting their relationship with climate change requires separating internal variability, observational limitations, and externally forced changes. A trend at one station may reflect changes in circulation or local conditions without implying a comparable change in the total mass of Earth’s atmosphere.
Boiling, gas exchange, and biological effects
Atmospheric pressure influences the temperatures at which liquids boil. Boiling occurs when a liquid’s equilibrium vapour pressure becomes sufficient to sustain vapour bubbles against the surrounding pressure, with additional effects from surface tension and the depth of the liquid. For pure water at one standard atmosphere, the boiling temperature is approximately . At lower atmospheric pressure, its boiling point is lower; at greater pressure, it is higher.
At a pressure near , water boils at approximately . This explains why boiling-based cooking processes differ at high elevations. A pressure cooker has the opposite effect: maintaining a pressure above the surroundings allows water and steam to reach temperatures above the ordinary sea-level boiling point. These are consequences of phase equilibrium, not changes in the chemical identity of water. Evaporation, unlike boiling, can occur at a liquid surface well below the boiling temperature.
For a gas mixture, each constituent has a partial pressure. Under ideal-gas conditions, Dalton’s law gives
where is the constituent’s mole fraction. Dry air contains approximately 20.95 percent oxygen, giving an oxygen partial pressure near at standard atmospheric pressure. At an ambient pressure of , the same dry-air composition produces an oxygen partial pressure of only about .
The oxygen fraction of the lower atmosphere therefore remains nearly constant with altitude even while oxygen availability falls. Inhaled air is also warmed and humidified before reaching the gas-exchanging regions of the lungs, and alveolar gas composition is further affected by oxygen uptake and carbon-dioxide release. Gas exchange depends on partial-pressure gradients across biological membranes and on ventilation, circulation, and the properties of hemoglobin, not merely on the total volume of air inhaled.
Sufficiently reduced oxygen availability can cause hypoxia. High-altitude exposure produces physiological responses that include increased ventilation and, over longer periods, changes in red-cell production and oxygen transport. Pressure changes also affect enclosed gas spaces in the ears, sinuses, and other cavities. Rapid decompression can create injury through gas expansion and the formation of bubbles from dissolved gases, although the mechanisms and circumstances differ among forms of decompression exposure.
Atmospheric pressure has ecological consequences as well. Lower air density at altitude changes aerodynamic forces on flying organisms, while reduced partial pressures influence respiratory exchange. These effects occur alongside colder temperatures, altered humidity, and stronger radiation exposure, making it important to distinguish the direct effects of pressure from other environmental changes associated with elevation.
Aviation, engineering, and environmental applications
A pressure altimeter infers altitude from ambient static pressure using a reference atmospheric relationship. It does not measure geometric height directly. Its indication depends on both the pressure setting and the actual temperature structure of the atmosphere. If the atmosphere differs from the assumed profile, indicated altitude and true altitude differ. This is particularly important in cold conditions, when pressure surfaces are closer together than in a warmer reference atmosphere.
Aircraft operating at higher levels commonly use pressure altitude referenced to , expressed operationally through flight levels. A shared reference helps maintain vertical separation even when the relationship to sea-level height varies with weather. Near airports, a locally specified altimeter setting supports interpretation relative to the airfield and surrounding terrain. These distinct conventions reflect the difference between coordinating aircraft on common pressure surfaces and determining actual clearance above the ground.
Air density also controls important aspects of performance. Density altitude describes the altitude in a reference atmosphere corresponding to the actual air density. High temperature, low pressure, and increased humidity can all reduce density under otherwise comparable conditions. Lower density reduces aerodynamic forces at a given true airspeed and can reduce the output of engines that depend on intake air. Thus two airfields at the same elevation may present different performance conditions, and those conditions can change substantially during the day.
Pressure differences impose major engineering loads. A vessel evacuated to near vacuum at sea level may experience an inward pressure difference close to . Across a broad wall, that produces a large compressive load and a risk of buckling. A pressurized aircraft fuselage experiences the opposite loading: cabin pressure exceeds the outside pressure, producing repeated tensile loading as the aircraft climbs and descends. Design therefore depends on differential pressure, geometry, material strength, and fatigue, not simply on whether the internal pressure seems large in absolute terms.
Atmospheric pressure also limits suction-based pumping. A pump does not literally pull an unsupported water column upward through an arbitrary height; it lowers pressure at the top, allowing external pressure to support the column. The ideal sea-level limit for water is about 10.3 metres, but vapour pressure, flow losses, and cavitation reduce the practical limit. Deep wells consequently require arrangements such as submerged pumps rather than a surface suction pump alone.
Environmental measurements require similar care. Gas volumes, pollutant concentrations, and gas-flow rates depend on pressure and temperature. Reporting a volume without specifying its reference conditions can make comparisons misleading. Barometric fluctuations can also alter gas movement through soils and pressure in wells, producing groundwater-level responses that must be distinguished from changes caused by recharge or pumping.
Atmospheric pressure on other worlds
Atmospheric pressure is not unique to Earth, and comparisons across the Solar System show why pressure cannot be inferred from temperature or composition alone. Surface gravity, atmospheric mass, molecular properties, and the exchange of volatiles with a planet’s surface and interior all matter. For an approximately hydrostatic atmosphere,
where is atmospheric mass and is planetary surface area. The equation makes explicit that a given pressure can be supported by different atmospheric column masses under different gravitational accelerations.
Mars has a predominantly carbon-dioxide atmosphere with a characteristic surface pressure of roughly , although elevation and season produce large variations. This is less than one percent of standard terrestrial sea-level pressure. Seasonal condensation and sublimation of carbon dioxide at the polar caps exchange a substantial quantity of gas with the atmosphere, changing its total mass over the Martian year. Deep basins have higher pressure than elevated terrain because a larger atmospheric column lies above them.
The low Martian pressure strongly constrains liquid water. Water’s triple-point pressure is approximately , close to typical Martian surface pressures. Above that pressure, an appropriate temperature range can permit stable liquid water, but most present-day surface conditions still favour freezing, evaporation, or sublimation. Pressure alone does not determine stability: temperature, salinity, and exposure to the atmosphere also matter.
Venus presents the opposite extreme among the terrestrial planets. Its surface pressure is about , roughly 92 times Earth’s standard atmospheric pressure, and its atmosphere is predominantly carbon dioxide. The dense atmosphere and high surface temperature create engineering conditions very different from those on Mars. Saturn’s moon Titan, by contrast, has a mostly nitrogen atmosphere with a surface pressure around , approximately 1.45 standard atmospheres, despite its much lower surface gravity. Its very low temperature allows methane to participate in a cycle involving clouds, precipitation, and surface liquids.
For giant planets, “surface pressure” is usually not meaningful in the terrestrial sense because there is no accessible solid surface marking the bottom of the atmosphere. A specified pressure level, often one bar, is instead used as a reference for radius and atmospheric structure. In the extremely tenuous exospheres of bodies such as Earth’s Moon, collisions are so infrequent that an ordinary continuum description of gas pressure becomes less useful.
Planetary observations exploit pressure-dependent effects to reconstruct atmospheric structure. Spacecraft entry measurements, radio occultations, and spectroscopic observations constrain density, temperature, and composition. Pressure broadening of spectral lines and the bending of radio signals provide information about atmospheric layers, although interpretation requires models of molecular interactions, gravity, and the vertical temperature profile.