Adiabatic Lapse Rate Calculator

An introduction to adiabatic lapse rates and parcel theory

A lapse rate is the rate at which temperature falls with height, written Γ = −∂T/∂z, so that a positive lapse rate means the air gets colder as you go up. Meteorology uses three of them and they answer three different questions. The dry adiabatic lapse rate Γ_d describes how fast an unsaturated air parcel cools purely because it expands as it rises into lower pressure, exchanging no heat with its surroundings. The saturated adiabatic lapse rate Γ_s describes the same parcel once it is at saturation, when condensation releases latent heat and partly offsets the expansion cooling. The environmental lapse rate Γ_e is not a parcel property at all: it is the temperature profile the surrounding atmosphere actually has right now, which you measure with a radiosonde or aircraft.

This calculator computes all three for the layer you specify and then does the thing the three rates exist for: it tells you whether that layer is absolutely stable, conditionally unstable or absolutely unstable. That classification is the reason parcel theory is taught. It is what separates a day with flat stratus and trapped pollution from a day that grows thunderstorms, and it depends entirely on where the environmental profile sits relative to the two adiabats.

The important departure from most quick lapse-rate tools is that Γ_s here is computed, not assumed. There is no single moist adiabatic lapse rate. It ranges from roughly 3.6 K/km in hot, humid air near the surface to more than 9 K/km in the cold upper troposphere, where the air holds so little vapour that latent heating becomes negligible and the moist adiabat converges on the dry one. Substituting a fixed 6.5 K/km — which is really the standard environmental rate, not a moist adiabat — is a modelling error that can flip a stability verdict from stable to unstable.

How to use this stability and lapse rate tool

Every field is a physical quantity, not a preference, so entering real observations gives a real answer. Take the temperature and elevation from a station report, the environmental lapse rate from a sounding if you have one, and leave the pressure blank unless you have a measured value.

• Air temperature at the starting level (°C): the temperature of the parcel and of the environment where the lift begins — a valley station, an airfield, or the surface below a developing cumulus.
• Starting elevation (m above sea level): used as the base of the layer and, when you leave the pressure field blank, to derive the ambient pressure from the standard atmosphere. Pressure matters because the saturation mixing ratio depends on it.
• Altitude change (m): the vertical displacement. Positive for ascent, negative for descent. The parcel is lifted or lowered through this depth.
• Station pressure (hPa, optional): supply a measured pressure when you have one. A real 1035 hPa anticyclone or a 985 hPa cyclone shifts Γ_s by a few hundredths of a kelvin per kilometre; leaving it blank uses the standard-atmosphere value for your elevation instead.
• Environmental lapse rate (°C/km, optional): the observed cooling rate of the surrounding air. Leave it blank to use the International Standard Atmosphere value of exactly 6.5 K/km. Enter a negative number to describe an inversion, where temperature rises with height.

Press Calculate and the panel returns Γ_d, Γ_s at the starting level, the layer-mean Γ_s, the parcel temperature at the top of the layer along each adiabat, the environmental temperature there, the parcel-minus-environment buoyancy difference, and the stability class. The level-by-level breakdown shows how Γ_s drifts as the parcel cools and depressurises, and Download CSV exports that profile for a spreadsheet or a plot.

The dry adiabatic lapse rate formula

Start from the first law for an adiabatic process, c_p dT = α dp, and substitute hydrostatic balance dp = −ρg dz with α = 1/ρ. The density cancels and the lapse rate collapses to a pure constant:

Notice what is not in that expression: no temperature, no pressure, no humidity. Γ_d is fixed by gravity and by the specific heat of the gas, which is why it is the same 9.77 K/km at the surface of the Sahara and at the tropopause. The familiar textbook figure of 9.8 K/km is this same number rounded; the calculator carries the unrounded value so that the level-by-level arithmetic stays self-consistent. Applying it over a finite layer is exact, because the rate does not change:

The saturated adiabatic lapse rate formula

For a saturated parcel the latent heat released by condensing water vapour has to be added to the energy budget. Carrying that term through the derivation gives the standard expression used throughout atmospheric science:

Here L_v is the latent heat of vaporisation, r_s the saturation mixing ratio, R_d = 287.04 J·kg⁻¹·K⁻¹ the specific gas constant of dry air, T the absolute temperature and ε = R_d/R_v = 0.622 the ratio of the dry-air and water-vapour gas constants. Both correction terms are proportional to r_s, so the whole departure from the dry adiabat is controlled by how much vapour the air can actually hold. The leading factor g/c_pd is exactly Γ_d, so the fraction is the ratio Γ_s/Γ_d — always less than one, and tending to one as r_s → 0.

The saturation mixing ratio follows from the saturation vapour pressure e_s and the ambient pressure p:

with t in degrees Celsius. That vapour-pressure fit is Bolton's 1980 formula, accurate to better than 0.1 % between −35 °C and +35 °C. The temperature dependence of the latent heat is carried as L_v = 2.501 × 10⁶ − 2370 t J·kg⁻¹, which is why Γ_s keeps changing even along a single ascent.

Because Γ_s depends on the parcel's own temperature and pressure, it cannot simply be multiplied by a layer depth the way Γ_d can. The calculator therefore integrates the saturated ascent numerically with a second-order Runge–Kutta step, updating the pressure from hydrostatic balance using the parcel's virtual temperature at every step, and reports both the starting-level value and the layer mean. Over a 3 km lift the difference between doing this properly and freezing Γ_s at its surface value can exceed 2 °C.

Environmental lapse rate and the stability criteria

The International Standard Atmosphere fixes the tropospheric environmental lapse rate at exactly 6.5 K/km from mean sea level to 11 km geopotential, with a sea-level temperature of 288.15 K and a pressure of 1013.25 hPa. That is a defined constant of the standard, adopted for engineering and altimetry, not a measured global average — real soundings depart from it by several kelvin per kilometre every day. When you leave the environmental field blank the calculator uses this defined value and, when you also leave pressure blank, the matching standard barometric relation for the troposphere:

Stability then follows from a single comparison. A parcel displaced upward is buoyant, and therefore keeps going, whenever it stays warmer than the air around it — which happens whenever the environment cools with height faster than the parcel does:

The equalities are the neutral cases: Γ_e = Γ_d is a dry-neutral layer, typical of a well-mixed afternoon boundary layer, and Γ_e = Γ_s is a saturated-neutral layer, typical of the interior of a deep cloud mass. A negative Γ_e is an inversion and is very strongly stable.

How the three lapse rates compare across conditions

The table below is the whole argument for computing Γ_s rather than assuming it. Every saturated value is produced by the expression above at the stated temperature and pressure; Γ_d and the ISA environmental rate are shown alongside for reference, and both are constants. Read across a row to see the pressure dependence, and down a column to see the far stronger temperature dependence.

Three things follow immediately. First, Γ_s spans a factor of three across the troposphere, so no single number can stand in for it. Second, at −40 °C the saturated rate is within 3 % of the dry rate, which is why the moist and dry adiabats on a skew-T diagram converge near the tropopause. Third — and this is the trap the ISA value sets — Γ_s happens to pass through about 6.5 K/km near 0 °C at low levels, which is very likely how the standard atmosphere's environmental rate came to be mislabelled as "the moist lapse rate" in so many places. It coincides at one point on a curve; everywhere else it is wrong.

A worked example: lifting a saturated Gulf Coast parcel

A late-afternoon coastal station reports 28 °C at sea level with the air effectively saturated in the cloud layer, and the 00 UTC sounding gives an environmental lapse rate of 7.5 K/km through the lowest 2 km. Lift a parcel 2000 m.

1. Ambient pressure. With the pressure field blank, the standard relation at h = 0 m returns p = 1013.25 hPa.
2. Vapour content. Bolton's formula gives e_s(28 °C) = 37.81 hPa, so r_s = 0.622 × 37.81 / (1013.25 − 37.81) = 24.11 g/kg. This is very moist air, and the latent-heat terms will dominate.
3. Dry adiabat. Γ_d = 9.80665 / 1004 = 9.7676 K/km, so an unsaturated parcel arrives at 28 − 9.7676 × 2 = 8.46 °C.
4. Saturated adiabat at the surface. With L_v = 2.4346 × 10⁶ J/kg and T = 301.15 K, the numerator is 1.679 and the denominator 4.401, giving Γ_s = 9.7676 × 0.3815 = 3.73 K/km — barely more than a third of the dry rate.
5. Integrate the ascent. As the parcel cools and depressurises, Γ_s rises to 3.93 K/km at the top of the layer. The integrated result is 20.36 °C at 2000 m, a layer-mean rate of 3.82 K/km. Freezing Γ_s at its surface value would have given 20.55 °C — a small error here, but one that grows with depth.
6. Environment. 28 − 7.5 × 2 = 13.0 °C at 2000 m.
7. Stability. 3.73 < 7.5 < 9.77, so the layer is conditionally unstable.

The buoyancy numbers make that verdict concrete. The saturated parcel arrives 20.36 − 13.0 = +7.36 K warmer than its surroundings and accelerates upward; the unsaturated parcel arrives 8.46 − 13.0 = −4.54 K colder and sinks back. Nothing rises spontaneously, but anything forced to saturation — by a sea-breeze front, by terrain, by an outflow boundary — takes off. That is exactly the setup for isolated afternoon thunderstorms along a humid coastline.

A second example: foehn warming on descent

Reverse the sign to see the other classic case. Air at 2 °C at 2500 m crosses a ridge and sinks 1800 m into a lee valley, so you enter −1800 m. Having already precipitated out its moisture on the windward side, it descends unsaturated and warms at Γ_d: 2 + 9.7676 × 1.8 = 19.6 °C on arrival. A 17.6 K warming with no heat added to the air at all is the entire mechanism behind chinook, foehn and Santa Ana winds, and it is why the calculator reports the dry result for descent even when you are thinking about moist air.

Reading the stability verdict in the field

The stability class is a statement about what the atmosphere will do if something disturbs it, and each class has a recognisable signature in the sky.

• Absolutely stable (Γ_e < Γ_s): vertical motion is suppressed whether or not the air is saturated. Expect flat-based stratiform cloud, smooth flight conditions, fog that lingers, and pollutants that accumulate near the surface. Nocturnal and subsidence inversions sit deep inside this class.
• Saturated neutral (Γ_e ≈ Γ_s): a displaced saturated parcel neither returns nor accelerates. This is the state a deep, actively overturning cloud mass relaxes toward, which is why sounding traces inside large convective systems tend to lie along a moist adiabat.
• Conditionally unstable (Γ_s < Γ_e < Γ_d): the most common regime in the warm-season troposphere, and the most consequential. Nothing happens spontaneously, but any mechanism that lifts a parcel to saturation releases the instability. Fronts, terrain, sea breezes and thunderstorm outflow all do this. The word conditional means conditional on reaching saturation.
• Dry neutral (Γ_e ≈ Γ_d): the signature of a well-mixed convective boundary layer on a sunny afternoon, and the reason mixed-layer depth grows so quickly once surface heating begins.
• Absolutely unstable (Γ_e > Γ_d): even unsaturated parcels accelerate upward. Superadiabatic layers are real but almost always shallow and short-lived, confined to the first few metres over strongly heated ground, because convection destroys the very gradient that creates it. A deep absolutely unstable layer in your input usually means a data error.

Alongside the class, check the sign of the buoyancy figures. A parcel that arrives warmer than the environment is positively buoyant and will keep rising under its own power; one that arrives colder will sink back. Comparing the dry and saturated buoyancies at the same level is the quickest way to see whether saturation is what unlocks the convection.

Assumptions and limitations of parcel theory

Parcel theory is a deliberate idealisation and every one of its assumptions is violated in the real atmosphere to some degree. These are the limitations that matter most when you compare an output here with an observation.

• No entrainment or mixing. A real thermal continuously entrains drier environmental air across its edges, which dilutes its buoyancy and evaporates condensate. Entrainment is the single largest reason observed updraught temperatures fall short of the undilute parcel value, often by several kelvin in the mid-troposphere.
• Reversible saturated ascent. The formula used here retains condensed water in the parcel. A pseudoadiabatic ascent, in which condensate falls out immediately, gives a slightly different rate; the two differ by well under 0.1 K/km at low levels but the gap widens aloft.
• Water only, no ice. Saturation vapour pressure is taken with respect to liquid water at all temperatures, and the latent heat of fusion released when droplets freeze is not included. Above the freezing level in a real cloud that fusion heating makes the true ascent slightly warmer than this calculation.
• Vapour-pressure fit range. Bolton's expression is quoted as accurate to 0.1 % between −35 °C and +35 °C. Outside that window the vapour amounts are tiny and the effect on Γ_s is correspondingly small, but the fit itself is being extrapolated.
• A single straight environmental profile. Real soundings are layered, with inversions, moist layers and dry intrusions stacked on top of one another. A single Γ_e over a deep layer averages all of that away, and can hide a capping inversion that would prevent convection entirely.
• Standard-atmosphere pressure when none is supplied. Leaving the pressure field blank assumes ISA conditions, which can be off by 30 hPa or more in a strong high or low. Supply a measured station pressure whenever precision matters.
• No hydrometeor loading, radiation or surface effects. The weight of suspended condensate reduces buoyancy; radiative cooling of cloud tops alters the profile; and near the ground, insolation, snow cover and terrain shading routinely dominate over adiabatic reasoning.

Treat the output as the undilute, idealised bound. It is the right first calculation and the right sanity check, but it is not a substitute for a plotted sounding when a decision depends on the answer.

Frequently asked questions about lapse rates

Sources and further reading

Every constant and equation on this page is taken from the following primary and institutional references rather than from secondary summaries.

• American Meteorological Society, Glossary of Meteorology — definitions of the dry-adiabatic, saturation-adiabatic, pseudoadiabatic and environmental lapse rates, and the form of the saturated lapse rate expression used above. glossary.ametsoc.org
• NOAA, NASA and USAF, U.S. Standard Atmosphere, 1976 (NOAA-S/T 76-1562) — the defined sea-level values T₀ = 288.15 K and p₀ = 1013.25 hPa and the molecular-scale temperature gradient of −6.5 K/km for the 0–11 km geopotential layer, together with the barometric relation reproduced above. ntrs.nasa.gov
• ISO 2533:1975, Standard Atmosphere — the International Standard Atmosphere, which adopts the same defined 6.5 K/km tropospheric temperature gradient.
• Bureau International des Poids et Mesures, The International System of Units (SI Brochure) — the standard acceleration of free fall, g_n = 9.806 65 m·s⁻², used for Γ_d.
• D. Bolton (1980), "The Computation of Equivalent Potential Temperature," Monthly Weather Review 108, 1046–1053 (American Meteorological Society) — the saturation vapour pressure fit e_s = 6.112 exp[17.67t/(t+243.5)] and the linear temperature dependence of the latent heat of vaporisation.
• J. M. Wallace and P. V. Hobbs, Atmospheric Science: An Introductory Survey, 2nd ed. (Academic Press, 2006), Chapter 3 — the derivations of Γ_d = g/c_p and of the saturated adiabatic lapse rate, the values c_pd = 1004 J·kg⁻¹·K⁻¹, R_d = 287.04 J·kg⁻¹·K⁻¹ and R_v = 461.5 J·kg⁻¹·K⁻¹, and the stability criteria used for the verdict.
• World Meteorological Organization, Guide to Instruments and Methods of Observation (WMO-No. 8) — observational practice for the radiosonde temperature profiles from which real environmental lapse rates are derived.

The saturated lapse rate figures in the comparison table were generated with the same code the calculator runs and agree with the standard published tables of Γ_s to within 0.05 K/km.