So I was in this place called Chamundi hills, Mysore.

I was riding back at around 7pm down hill and noticed this strange behavior. About every 50-100 meters the temperature fluctuated.

It was extremely cold in some areas and in the next 100 meters I could feel the warmth. The whole cycle repeated all the way (about 8-10 kilometers).

Is there any scientific explanation about this?

  • $\begingroup$ Are you saying simply that the temperature changes, which we might expect due to altitude, or that it fluctuates or alternates between hotter and cooler? $\endgroup$ Jan 30, 2023 at 23:38
  • $\begingroup$ A couple things you could check next time (or now [Also Google satellite imagery])... how was the elevation changing... what was the ground cover nearby (cement, trees, water, fields?). Maybe you could write us a followup answer from the further results? :-) $\endgroup$ Jan 31, 2023 at 0:07
  • 2
    $\begingroup$ @RobbieGoodwin I have been in places where the temperature fluctuated quite noticably just climbing up a hill of 20 meters or so... $\endgroup$
    – Michael
    Jan 31, 2023 at 16:39
  • 1
    $\begingroup$ @Michael Yes… party why we need a better description. Perhaps useful is that scied.ucar.edu/learning-zone/atmosphere/… says the standard (average) lapse rate means as you climb a mountain, you can expect the air temperature to decrease by 6.5 degrees C for every 1000 meters. Prolly unless your terrain is rather regular, the road moving in and out from the main mass of the mountain would combine with that lapse rate to create temperature change, if not accompanying winds. $\endgroup$ Jan 31, 2023 at 16:47
  • 1
    $\begingroup$ @RobbieGoodwin I think what I experienced matches the accepted answer as I have experienced a temperature increase just going up 20 meters. Conversely, it's possible for it to feel like it dropped a lot more if there is wind on a ridge that the hill shields you from until you reach it. $\endgroup$
    – Michael
    Jan 31, 2023 at 16:50

4 Answers 4


To extend @Poutnik's invisible river analogy, the cold air can also 'pool' in small dips.

On clear nights, temperature fluctuations are quite a frequent occurrence due to radiative cooling. If the air cools sufficiently to reach the dew point, then cloud can act as a 'tracer' for the air pockets.

An example of this are fog patches that you potentially pass through in certain conditions. The air in the fog patches is a little colder (or higher absolute humidity, for example over a lake) than the surrounding fog-free areas. Even if the air wasn't cold enough to form cloud, the same temperature variation can occur invisibly.

Here's an example from Wikimedia user Simo Räsänen: Radiation fog over flat fields and pond in morning

In the mountains, the invisible rivers of cold, descending air that @Poutnik mentions can sometimes be seen by tracer clouds flowing down hillsides. Colder air is denser than the surrounding air and has less friction so gravity pulls it down. The areas with cloud will be colder than the adjacent cloud-free zones. This effect can lead to strong down-slope katabatic winds. In Antarctica this effect is responsible for some of the strongest winds in the world.

Here's a good example from Wikimedia user Andrew J. Kurbiko: Radiation fog tracing katabatic winds off a plateau

To summarise, the key to the effect you noticed is almost certainly radiative cooling and the movement of colder denser air. Radiative cooling occurs due to any object above absolute zero giving off radiation. On a clear night this radiation passes through the Earth's atmosphere and is lost into space. This cools down the object (land) that the radiation came from. The land, in turn, cools down the air above it. On a cloudy night, the radiation is reflected by the clouds and reabsorbed by the land keeping the temperature constant.

  • $\begingroup$ (Not sure if you can say the land cools down the air above it in the last paragraph with reliability... the low level atmosphere is primarily heated by conduction from the Earth... I'd say the atmosphere is radiatively cooling just as the land, but unlike the daytime has very low amounts of heat transferred in from the ground? I think many/most nights, the ground temperature is warmer than the atmosphere just above) $\endgroup$ Jan 31, 2023 at 0:05
  • $\begingroup$ (I should finish that comment fully. I think many/most nights, the ground temperature is warmer than the atmosphere just above... so the ground can't be cooling the atmosphere) $\endgroup$ Feb 1, 2023 at 2:33
  • $\begingroup$ Look up 'Radiation Fog'. The UK Met Office's definition is: "Radiation fog usually occurs in the winter, aided by clear skies and calm conditions. The cooling of land overnight by thermal radiation cools the air close to the surface. This reduces the ability of the air to hold moisture, allowing condensation and fog to occur." Many meteorological textbooks refer to the same process. The same process can occur without reaching the dew point. metoffice.gov.uk/weather/learn-about/weather/types-of-weather/… $\endgroup$ Feb 1, 2023 at 9:05
  • $\begingroup$ I didn't say it doesn't happen, just that I'd think many/most nights it doesn't :-) Indeed, in winter/early spring when the ground has had ample time to cool (early winter, snow/ice often won't stick because the ground is warmer than the moisture and often air temp in such events). $\endgroup$ Feb 2, 2023 at 6:19
  • $\begingroup$ It's certainly just my thinking out loud... as frost/dew likewise are some proof that the ground can indeed cool the air. I guess SHC has a lot to do with it too, so surfaces like metal and cement will cool quicker than the air just as they heat faster. But for other surfaces like grass and soil, I picture that significant atmosphere-to-ground conduction will mainly only be once the air cools to the dewpoint (and thus stops cooling quickly due to condensation heat release). I could be wrong. $\endgroup$ Feb 2, 2023 at 6:31

I assume local downstream flows of cold air near surface, that got cool down by surface radiation on the hill slopes. You can imagine it as invisible rivers of cold, descending air.


Yes, there could be several reasons for the temperature fluctuation you experienced. Some possible explanations are:

Altitude change: As you were riding downhill, you may have experienced changes in altitude, which can affect temperature.

Microclimates: Different areas of the hill may have their own microclimates, with different temperatures and humidity levels, which could explain the fluctuations you experienced.

Air currents: The movement of air currents can also cause temperature fluctuations, as warm air rises and cold air sinks.

Shadows: Shadows cast by buildings, trees, or other obstacles can also affect temperature, as they can block the sun's warmth.

Localized heating: Localized heating sources, such as factories or power plants, can also affect temperature in a specific area.

It's possible that a combination of these factors contributed to the temperature fluctuations you experienced.

  • $\begingroup$ Furthermore, in early morning in hilly terrain (as the OP states) some places are likely to have been warming in the sun for a while others nearby may still be as in the shadow and as cold as they had been in the night. $\endgroup$
    – Pere
    Feb 11, 2023 at 16:41

Something I have observed more than once: Katabatic and anabatic winds tend to follow drainage patterns--there's more air flow when you're lined up with the terrain. Even when you don't feel the wind it can change the air temperature.


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.