"It feels like it's 30° today"
I'm fascinated by how human perception compares to electromechanical sensors in terms of accuracy. "It feels like 23 minutes have passed since I put the cake in the oven" or "This note sounds like an F#, though it's slightly flat".
This question concerns temperature identification accuracy in air, that is to say, how accurately a person placed in a room can identify the temperature of the air in that room. I managed to dig up a few papers close to the topic, but not quite what I'm looking for, including the following:
- C. Stevens Kenneth K. Choo, J. (1998). Temperature sensitivity of the body surface over the life span. Somatosensory & Motor Research, 15(1), 13–28. doi:10.1080/08990229870925
- Kenshalo, D. R., Holmes, C. E., & Wood, P. B. (1968). Warm and cool thresholds as a function of rate of stimulus temperature change. Perception & Psychophysics, 3(2), 81–84. doi:10.3758/bf03212769
- Darian-Smith, I., & Johnson, K. O. (1977). Thermal sensibility and thermoreceptors. Journal of Investigative Dermatology, 69(1), 146–153. doi:10.1111/1523-1747.ep12497936
In fact, the word "air" really doesn't appear in any of the studies above, though the second features the word "hair" once and the word "chair" thrice. I also found another Stack Exchange question concerning sensitivity of thermoreceptors to change, but that's not it either.
The problem is all the studies that actually check human temperature sensitivity seem to use thermal pads on a small area of the skin, leading to wildly different results from tests in still air. They also tend to focus on how gradually humans can perceive temperature changes, rather than how accurately humans can sense temperatures themselves. My personal experience, which is irrelevant to the science but is my original source of curiosity surrounding this topic, is that I can usually guess the temperature within around 1˚C / 2˚F of sensor measurements as long as wind is low and humidity is fairly consistent from test to test. I can't seem to find studies testing this though.
Does anyone know of any research on this area, with a better value than my "n of one" estimate above?