Why is it colder on mountain plateaus?

[I like Serena]

The fact that it is colder on mountain plateaus has been bugging me.
(Or isn't it?)

As I understand it, air is heated by the ground and expands while rising up, lowering the temperature by (adiabatic) expansion.

But how does that work on a mountain plateau?
Shouldn't the temperature there be more or less the same as on sea level, since the ground there is heated directly by the sun?

 

[D H]

You are forgetting that the troposphere is in constant motion. Atmospheric mixing dominates over slower diffusive processes. The air mass that is over a mountain plateau today may well have been over sea level yesterday. Two things happen to that air mass as it is driven upwards: It loses thermal energy due to adiabatic expansion, and it loses absolute humidity due to the marked sensitivity of absolute humidity to temperature. Both of these effects tend make higher altitudes be cooler than lower altitudes.

The effect of loss of thermal energy on temperature is obvious; air tends to cool adiabatically as it rises. Regarding humidity: Water vapor is an incredibly powerful greenhouse gas. That humid air at low altitudes makes for a nice thermal blanket that keeps nighttime cooling at bay. Once the temperature hits dew point that is pretty much it for nighttime cooling in humid areas. That effect is pretty much gone in those mountain plateaus. Temperatures start dropping precipitously as soon as the sun sets and may keep dropping all night long. The dew point in those mountain plateaus can be very, very low.

 

[rcgldr]

The Texas High Plains area, about 3,000 to 4,100 foot elevation is apparently large enough that it's not significantly colder than the nearby Hill Country to the south east (elevation 500 to 2500 feet (the hills)), although it is dryer. There a fairly sudden increase in altitude at the south east edge of the High Plains area, and I'm not sure of the temperature in the immediate vicinty of the south east edge.

 

[I like Serena]

Thank you both for answering.

@D H: I understand that at night the temperature would drop faster due to lack of humidity and possibly clouds.
But doesn't that also mean that the temperature would rise faster during daylight?

@rcgldr: How is that on the Texas High Plains? Is the day-night temperature difference extreme?

Btw, I was triggered by the thread on the temperature that is lower in the Antarctic than on the North pole.
One of the arguments was that the Antarctic is elevated quite a bit.
It made me wonder if that was relevant...

 

[rcgldr]

How is that on the Texas High Plains? Is the day-night temperature difference extreme?

You have to realize that the Texas High Plains extends into the bodering states, New Mexico, Colorado, Kansas, Oklahoma, so it's a huge area. I'm not sure about day-night temperature swings for the high plains compared to other inland (not near an ocean) areas.

Btw, I was triggered by the thread on the temperature that is lower in the Antartic than on the North pole. One of the arguments was that the Antartic is elevated quite a bit.
It made me wonder if that was relevant...

White snow is also a factor (reflected light doesn't heat the air). From wiki:

Much of the sunlight that does reach the surface is reflected by the white snow. This lack of warmth from the sun, combined with the high altitude (about 2,800 metres (9,186 ft)), means that the South Pole has one of the coldest climates on Earth (though it is not quite the coldest; that record goes to the region in the vicinity of the Vostok Station, also in Antarctica, which lies at a higher elevation

http://en.wikipedia.org/wiki/South_Pole

 

[klimatos]

The fact that it is colder on mountain plateaus has been bugging me.
(Or isn't it?)

As I understand it, air is heated by the ground and expands while rising up, lowering the temperature by (adiabatic) expansion.

But how does that work on a mountain plateau?
Shouldn't the temperature there be more or less the same as on sea level, since the ground there is heated directly by the sun?

The atmosphere as a whole receives most of its heat from the surface of the Earth (69%) and only about a quarter of its heat (26%) from the sun. Some 4% comes from conduction from the planetary surface to the atmosphere; and the final 1% from hydrologic cycling. Obviously, then, the farther you get from the mean surface elevation of the earth, the cooler the atmosphere gets.

The rate at which the atmosphere cools with elevation is termed the normal atmospheric lapse rate. It is roughly 6.49°C per kilometer for a world-wide average over the course of many years.

Since the adiabatic lapse rate must work in both directions (just as much air must come down as goes up), net adiabatic cooling of the atmosphere is insignificant in any climatic (long-term) sense.

 

[klimatos]

1) You are forgetting that the troposphere is in constant motion. Atmospheric mixing dominates over slower diffusive processes. The air mass that is over a mountain plateau today may well have been over sea level yesterday. Two things happen to that air mass as it is driven upwards: It loses thermal energy due to adiabatic expansion, and it loses absolute humidity due to the marked sensitivity of absolute humidity to temperature. Both of these effects tend make higher altitudes be cooler than lower altitudes.
The effect of loss of thermal energy on temperature is obvious; air tends to cool adiabatically as it rises.

2)Regarding humidity: Water vapor is an incredibly powerful greenhouse gas. That humid air at low altitudes makes for a nice thermal blanket that keeps nighttime cooling at bay. Once the temperature hits dew point that is pretty much it for nighttime cooling in humid areas. That effect is pretty much gone in those mountain plateaus. Temperatures start dropping precipitously as soon as the sun sets and may keep dropping all night long. The dew point in those mountain plateaus can be very, very low.

1) The normal (non-adiabatic) lapse rate is sufficient to explain virtually all temperature drops with elevation. In point of fact, many high elevations experience descending air currents during the nighttime hours. This makes the higher slopes actually warmer than the valley bottoms. Climatological textbooks explain the decrease in temperature with increase in elevation as a consequence of simply being farther away from the primary source of heat--the Earth's surface. I don't recall ever seeing the adiabatic argument for long-term cooling.

2) Very powerful indeed. Along with both liquid and solid water it accounts for 69% of the radiant heat absorbed by the atmosphere. Carbon dioxide is next with 14%. (I'm just elaborating, I'm not getting into climate change. Honest!)

 

[I like Serena]

The atmosphere as a whole receives most of its heat from the surface of the Earth (69%) and only about a quarter of its heat (26%) from the sun. Some 4% comes from conduction from the planetary surface to the atmosphere; and the final 1% from hydrologic cycling. Obviously, then, the farther you get from the mean surface elevation of the earth, the cooler the atmosphere gets.

The rate at which the atmosphere cools with elevation is termed the normal atmospheric lapse rate. It is roughly 6.49°C per kilometer for a world-wide average over the course of many years.

Since the adiabatic lapse rate must work in both directions (just as much air must come down as goes up), net adiabatic cooling of the atmosphere is insignificant in any climatic (long-term) sense.

This is new to me. I thought *all* energy comes from the sun.

As for adiabatic cooling, I'm assuming it evens out into an equilibrium, where the temperature is highest on ground level, where the ground is actually heated, and gradually diminishing higher up.I found this picture on:
http://en.wikipedia.org/wiki/Earth's_energy_budget

It seems to show that all heat is from the sun, 19% directly absorbed, and 45% is absorbed by the atmosphere indirectly from the Earth after being heated by the sun.
How does this relate to your numbers?Furthermore I found this article on lapse rate:
http://en.wikipedia.org/wiki/Lapse_rate

It names 3 types: ELR, DALR, and MALR of lapse rates of temperature.
If I understand correctly the ELR indicates the actual average rate, which is in between the theoretical DALR and MALR, which are based on adiabatic expansion.
Am I misunderstanding?

 

[klimatos]

You have to realize that the Texas High Plains extends into the bodering states, New Mexico, Colorado, Kansas, Oklahoma, so it's a huge area. I'm not sure about day-night temperature swings for the high plains compared to other inland (not near an ocean) areas.

Amarillo (on the High Plains) has an average annual day-night temperature range of 26°F. Oklahoma City, which is roughly the same latitude and is some 2,500 feet lower, has an average annual day-night temperature range of only 21°F. Elevation does make a difference.

 

[D H]

Beat me to it! That is exactly the comparison I was about to make.

There's one thing you didn't say: Almost all of the difference between Amarillo and Oklahoma City can be attributed to daily lows. While Amarillo's average highs are a tiny bit lower than are Oklahoma City's for July through September, the average daily highs for the two cities are nearly identical for the rest of the year. OTOH, the daily lows are markedly lower in Amarillo compared to Oklahoma City throughout the year.

 

[klimatos]

1) This is new to me. I thought *all* energy comes from the sun.

2) As for adiabatic cooling, I'm assuming it evens out into an equilibrium, where the temperature is highest on ground level, where the ground is actually heated, and gradually diminishing higher up.


3) I found this picture on:
http://en.wikipedia.org/wiki/Earth's_energy_budget

4) It seems to show that all heat is from the sun, 19% directly absorbed, and 45% is absorbed by the atmosphere indirectly from the Earth after being heated by the sun.
How does this relate to your numbers?


5) Furthermore I found this article on lapse rate:
http://en.wikipedia.org/wiki/Lapse_rate

It names 3 types: ELR, DALR, and MALR of lapse rates of temperature.
If I understand correctly the ELR indicates the actual average rate, which is in between the theoretical DALR and MALR, which are based on adiabatic expansion.
Am I misunderstanding?

1) Virtually all of the Earth's heat comes ultimately from the Sun. However, the proximate source of most of the atmosphere's heat is the Earth's surface.

2) Adiabatic temperature changes are not a factor in the Earth's climate. And, we are talking climate here, not weather.

3) Yes, this is based on the seminal article by Kiehl and Trenbarth (1997) "Earth's Annual Global Mean Energy Budget". It has been modified a bit since publication, but is still the starting point for any modern understanding of the atmosphere's heat budget. We differ somewhat on handling the enthalpy of condensation, but agree on the important points. If you want to use their percentages instead of mine, The argument remains valid.

4) Keep in mind the difference between ultimate and proximate.

5) I'm referring to "ELR" the environmental lapse rate. The other two are adiabatic and irrelevant to the question of why climatic temperatures decrease with increased elevations.

 

[I like Serena]

Hmm, trying to *compute*...

So with high elevation (on a plateau) you have higher day-night-differences, but the same average temperature?
Or am I misunderstanding?

 

[klimatos]

Beat me to it! That is exactly the comparison I was about to make.

There's one thing you didn't say: Almost all of the difference between Amarillo and Oklahoma City can be attributed to daily lows. While Amarillo's average highs are a tiny bit lower than are Oklahoma City's for July through September, the average daily highs for the two cities are nearly identical for the rest of the year. OTOH, the daily lows are markedly lower in Amarillo compared to Oklahoma City throughout the year.

You have good sources, too! Mean daily highs for Amarillo/Oklahoma City = 70/71. Mean daily lows are 44/50. July highs and lows are 90/65 for Amarillo and 92/71 for Oklahoma City. January figures are 49/24 and 47/27 respectively.

I suspect the warmer winter daytime temperatures for Amarillo are due to lower (42% versus 51%) daytime cloud cover and consequent increased solar heating for Amarillo.

 

[klimatos]

Hmm, trying to *compute*...

So with high elevation (on a plateau) you have higher day-night-differences, but the same average temperature?
Or am I misunderstanding?

Misunderstanding. The mean (average) temperatures on the plateau will be lower than the mean temperature on the nearby lowlands by the environmental lapse rate -6.5°/km.

 

[I like Serena]

Misunderstanding. The mean (average) temperatures on the plateau will be lower than the mean temperature on the nearby lowlands by the environmental lapse rate -6.5°/km.

So where does the environmental lapse rate come from?
The wiki article doesn't say.

As for heat from the earth, how would altitude from the sea level have an effect?
The difference from the Earth's radius should be negligible.

 

[rcgldr]

The mean (average) temperatures on the plateau will be lower than the mean temperature on the nearby lowlands by the environmental lapse rate -6.5°/km.

In the case of Amarillo versus Oklahoma City mentioned in this thread, the temperature difference seems to be about 2° Farenheight, but the altitude difference is .75 km. Other possible causes of the temperature difference: average humidity and cloud cover, how reflective the ground cover is, prevailing winds, ...

 

[klimatos]

So where does the environmental lapse rate come from?
The wiki article doesn't say.

As for heat from the earth, how would altitude from the sea level have an effect?
The difference from the Earth's radius should be negligible.

The environmental lapse rate is a function of the totality of Earth-Sun relationships and atmospheric composition. There are entire shelves of books in libraries on the details of these relationships.

Radiant heat intensity is not a function of the percentage of some arbitrary distance (the Earth's radius) but of actual distances in meters following the inverse square law and a whole flock of factors involving longwave absorption by different components of the atmosphere.

 

[klimatos]

In the case of Amarillo versus Oklahoma City mentioned in this thread, the temperature difference seems to be about 2° Farenheight, but the altitude difference is .75 km. Other possible causes of the temperature difference: average humidity and cloud cover, how reflective the ground cover is, prevailing winds, ...

Yes. The ELR is a world-wide long-term average. It rarely applies to a particular place and time. Local climatic peculiarities are best dealt with by local climatologies. German scholars have done some excellent pioneer work in this field.

As a general observation, many of the misunderstandings displayed in these postings could be cleared up by an introductory course in Climatology or by anyone of the many good popularizations of the field.

 

[rcgldr]

I think the basic question is if an elevated area is large enough, say 1000 km², and has similar ground based heat radiation as a large lower elevation area, would the typical temperature difference between the elevated and lower areas still be about 6.5 ° (is this Farenheight or Celcius) per km of altitude?

 

[klimatos]

I think the basic question is if an elevated area is large enough, say 1000 km², and has similar ground based heat radiation as a large lower elevation area, would the typical temperature difference between the elevated and lower areas still be about 6.5 ° (is this Farenheight or Celcius) per km of altitude?

The mean environmental lapse rate (ELR) is 6.5°C per kilometer. As I mentioned elsewhere, it is a world-wide long-term average. It is rarely found at any specific place and time. Instead, what you get is something either less or more, depending upon local circumstances at that time and place. It's essentially the difference between climate and weather.

To answer your question, Yes. There would still be a difference due to elevation--although not necessarily exactly 6.5°C/km. The reason for this is that weather phenomena serve to mix the planetary atmosphere rather well. The farther you get away from the Earth's surface, the more uniform the temperature from place to place becomes. Shoe-top temperatures are far more diverse than temperatures at the standard height (1.5m) of an instrument shelter. And shelter temperatures are far more diverse than temperatures at 1000 meters.

Therefore, a 2000-m plateau is more likely to have a temperature in common with the atmosphere at 2000-m over a nearby sea-level location than it will have with that sea-level location itself. Atmospheric circulation is often more important in influencing temperature than the simple radiative heat balance.

Without global circulation, the tropics would be much hotter than they are, and the polar regions would be much colder than they are.