Cooling with altitude
A Socratic walk-through of cooling with altitude — reasoned out one step at a time, not lectured.
The question we started with
THE QUESTION #Why does the air get colder as you climb a mountain toward the sun?
Climb three thousand metres and you are measurably closer to the sun, the light is fiercer, sunburn arrives faster — and there is snow on the ground. This is only a paradox if you assume the air is warmed by the sunlight falling on it. Suppose it is not. What would have to be true instead, and what would follow?
Reasoning it through
REASONING #First, dispose of the distance. The sun is about 150 million kilometres away, so five kilometres of altitude changes the separation by roughly one part in thirty million. Proximity is not a term in this.
So why is the mountaintop brighter and the burn faster? Because there is less atmosphere above you to scatter and absorb the light, particularly the ultraviolet — a clue rather than a contradiction. The air is doing something to the sunlight on its way down, and the summit has less of it overhead.
So what does clear air do with sunlight? Mostly it lets it pass. Nitrogen and oxygen barely interact with visible light — which is why the sky is transparent and you can see the sun at all. Some is absorbed on the way through, by ozone in the ultraviolet and water vapour in the near infrared, but the bulk reaches the ground and is absorbed there. The surface then radiates at much longer wavelengths, to which the air is far from transparent, warms the air touching it, and evaporates water into it.
So the atmosphere is heated from below, at its base — a fluid warmed from the bottom, in which the further from the heat source the cooler you should expect to be. That already reverses the intuition, but it does not say how fast temperature falls.
Take a parcel of air near the warm ground. Being warmer it is less dense, therefore buoyant, and it rises. As it climbs the pressure around it drops — pressure at any height is just the weight of the air above — and the parcel expands to match. Expanding means pushing the surrounding air outwards, and pushing takes energy. Where does that energy come from?
Here is the crucial step. Air conducts heat poorly and the parcel is large, so during the ascent it exchanges almost no heat with its surroundings. The work of expanding must therefore be paid out of the parcel's own internal energy — which in a gas is molecular motion, the thing temperature measures. The parcel cools without giving its heat to anything. Nothing was lost; it was spent.
That is an adiabatic process, and the rate follows: an unsaturated parcel cools by about 9.8 degrees Celsius per kilometre of ascent. The number contains no reference to the sun, the season, or the mountain. It is a property of air expanding.
One complication matters. As a rising parcel cools it eventually saturates and its water vapour begins to condense — releasing the latent heat that evaporation absorbed at the surface. That heat is delivered inside the parcel, partly offsetting the cooling, so a saturated parcel cools more slowly: around 5 degrees per kilometre in warm moist air, approaching the dry value in cold air with little vapour to give up.
What a climber experiences is neither rate. It is the environmental lapse rate — the temperature of the still air at successive heights — averaging about 6.5 degrees per kilometre in the lower atmosphere. It sits between the other two because the real atmosphere is a compromise: radiation works to steepen the profile, convection stirs whenever it is steep enough for parcels to keep rising, and moisture moderates it. That average hides much local variation, and on a still clear night cold air pools in valleys and the profile can invert entirely.
The analogy
THE ANALOGY #Work a bicycle pump hard and the barrel gets hot in your hand — you are compressing air, doing work on it, and that work becomes molecular motion. Now run it backwards: open a valve and the escaping gas is noticeably cold, because it does the work of pushing the outside air aside and pays for it out of its own store of motion. A parcel rising through the atmosphere is on the second half of that cycle continuously, expanding a little more with every metre it gains.
the pump's heating comes from an outside agent, your arms, while a rising parcel is lifted by its own buoyancy through a pressure gradient already there. And gas rushing from a valve also cools by mechanisms specific to a nozzle and mixes at once with the room, whereas the atmospheric parcel expands slowly and coherently — which is what makes the clean adiabatic accounting apply.
Clarifying the model
THE MODEL #Two ideas are easily run together, and both are needed. The first is where the heat comes from: the surface, not the air. The second is why a gradient forms and holds: rising air expands and cools adiabatically, so vertical stirring builds a profile that decreases with height.
A common misreading is that the air "loses" its heat as it rises. It loses it to nothing — that is what adiabatic means. The energy is converted into the work of expansion, and if the parcel sinks again it is compressed and warms back by the same amount. That reversibility is why air descending a mountain's lee slope arrives hot and dry, having shed its moisture on the way up.
The honest simplification is the environmental lapse rate: 6.5 degrees per kilometre is a global average through the troposphere, not a law, and the real profile depends on humidity, cloud, season, time of day and terrain. Nor does the decrease continue forever. Above the tropopause — roughly 11 kilometres in mid-latitudes — the stratosphere gets warmer with height, because ozone there absorbs ultraviolet directly. That is the confirming case: where air is heated by sunlight rather than from below, altitude does bring warmth.
A picture of it
THE PICTURE #How to readFollow the arrows down the page as one parcel's history. The first message is the step people assume warms the air, and it does not — the light passes through and is absorbed at the ground. The dashed self-messages are the parcel's own temperature changing with no heat crossing its boundary, first at the dry rate, then at the slower saturated rate once condensation begins. The return arrow to the ground shows the process is reversible, and the closing note is what a climber actually measures.
What became clearer
WHAT CLEARED #The mountaintop is cold not despite the sun but because of how the atmosphere is heated — at its base, by a surface absorbing the light the air let through. Add that rising air expands and must pay for the expansion out of its own warmth, and falling temperature with height is the expected result rather than a puzzle. The stratosphere, warming upward where sunlight is absorbed directly, is the exception that confirms it.
Where to go next
ONWARD #- Why the foehn or chinook wind arrives hot and dry on a range's lee side, using the same two lapse rates.
- How radiative-convective balance sets the depth of the troposphere.
Key terms
TERMS #| Term | What it means |
|---|---|
| Adiabatic | a change in which no heat crosses the parcel's boundary, so temperature changes only through work done or received. |
| Dry adiabatic lapse rate | the cooling of an unsaturated rising parcel, about 9.8 degrees Celsius per kilometre. |
| Saturated adiabatic lapse rate | the slower cooling where vapour is condensing, around 5 degrees per kilometre in warm moist air. |
| Environmental lapse rate | the measured decrease with height in still air, averaging about 6.5 degrees per kilometre. |
| Latent heat | energy absorbed in evaporation and released on condensation, with no temperature change at the moment of transfer. |
Every term the collection defines is gathered in the glossary.