A related aside - when we instrument research aeroplanes, it is very difficult to get OAT even faintly accurate. The effects of moisture inside clouds and the variable radiation effects depending upon the angle of the sun (or re-radiation by the earth) relative to the aeroplane introduce numerous error factors that are really hard to design out.
Given all of that, you can only possibly regard a normal aircraft OAT gauge that cost a few tens of pounds, as a vague guess at the temperature.
Clearly, if in doubt, assume it's a bit worse than the gauge says.
G
The effects of moisture inside clouds and the variable radiation effects depending upon the angle of the sun (or re-radiation by the earth) relative to the aeroplane introduce numerous error factors that are really hard to design out.
Interesting Genghis. I've been plagued by inaccurate OAT readings and haven't really come up with a good solution. I have the Aspen glass cockpit which contains a Remote Sensor Unit on the top of the tailcone of my Cessna. This unit contains among other sensors also a digital OAT which is used for the Air Data Computer to calculate TAS. I've seen it indicate too high and I thought it is because of its exposure to direct sunlight.
Aspen added the option to have an external OAT probe which I installed on the aircraft's belly. The result was even worse as it's heated up by exhaust even though it's at least 3 meters behind the exhaust pipe.
When I talked to people about the OAT in direct sunlight and the alleged design mistake by Aspen, several told me that is no problem because heating through sun radiation should be negligible as the unit is in the airflow. I have no idea whether that is true.
The best reading I get so far is a probe in the pipe of the cabin vents. This is used for the engine monitor to calculate %BHP.
Aspen have had many many problems with their rooftop sensor unit. Reading about the countless tales of hassle in the US pilot community, some informal surveys suggest the "issue rate" is closer to 100% of installations than any lower figure. They have been trying everything including fiddling the errors by software hacks.
several told me that is no problem because heating through sun radiation should be negligible as the unit is in the airflow.
I think that is true, but it's harder to deal with the heat conduction from the cockpit, into the roof material. If the airflow is supposed to be say -10C and the cockpit is +25C, you have a large thermal gradient across the roof material so a lot of heat will be conducted into the sensor housing. And if the sensor was separated from the roof with an air gap (which is actually very hard to do because there has to be "something" holding it) you still have the problem of the roof (which in this example might be sitting at say +5C) radiating at the probe.
I have an "OAT probe" on my house, sticking out of the wall by about 10cm, and there is no way to make it read right due to the radiation from the wall which is warmer than the air, due to conduction from inside the house. If the OAT is say -5C, the probe reads about -3C.
It's easy enough to get an OAT probe to read accurately on the ground, against a reference probe strapped to it. You can buy a PT100 handheld thermometer which is accurate to 0.1C, for about £300.
In flight, one has to make assumptions about what one is measuring because different shapes will have different amounts of aerodynamic heating (or cooling, if the pressure has dropped). IMHO, a simple cylindrical probe of the type commonly used, mounted into the underside wing skin, gives a pretty good readout, hardly affected by sunlight. In flight, I have found mine to be a very reliable indicator of the actual air temperature on the wing leading edge, as evidenced by ice accretion.
One thing I noticed very early on in my flight training was the casual attitude to OAT measurement. We had the crappy "dial" gauges, screwed through the windscreen, and a lot of them were simply stuck at -30C or whatever. Yet it is quite easy to get this right - at least to within say 1C.
For me pitot heat on on the runway, off after landing.
For IFR flights always. I had almost every system fail on me at least once over the years, but never the pitot heater(s). Apart from the fact that on some Piper singles and light twins the heaters are so weak that the pitot tube and static port ice-up even in light icing conditions. One more reason to stay outside ice with these aircraft...
One thing I recall from PPL training was that none of the pitot tubes ever got more than slightly warm.
Then I got the TB20, on which you nearly burn yourself in the time it takes to turn it on, walk to the opposite wing to check the wingtip lights, walk to the wing with the pitot on it, check the wingtip lights on that, and walk to the pitot tube. That is some tens of seconds, and it is impossible to hold the pitot tube after that.
In retrospect I can't say whether the C152/PA28 pitots actually worked or whether they just got a bit warm from one holding them in one's hand.
Are static ports heated? Mine aren't. I've never seen heated static ports in GA. Maybe if one buys the full TKS, when one gets a heated stall warner...
I use pitot heat anytime the OAT is below +5C. One should not get icing at say +3C but I recall seeing "slightly odd IAS readings" which went away when the pitot was turned on, so now I don't take a chance on it.
One thing I recall from PPL training was that none of the pitot tubes ever got more than slightly warm.
Whenever I'm doing ground avionics tests and there is a burn smell, I have to run to the pitot tube and remove the cover. I saw someone having one of these over his pitot tube, roasted chicken
Are static ports heated? Mine aren't. I've never seen heated static ports in GA. Maybe if one buys the full TKS, when one gets a heated stall warner...
Never seen that either, not even with a fully TKS'ed plane. Normally the solution to that is alternate static which draws from the cabin.
Well I have a heated stall warner. Never really heard of static icing up given lack of airflow through it.
Fuel tank vents can ice up.
With the full TKS system you get these covers
They are cheap little fibreglass bits but the firm charges about €250 for a pair 
Some TB20s had them fitted anyway - not sure if it was factory or the original dealer.
A heated pitot would also be easy to install.
That huge plastic chicken is very funny, but do they sell matching ones to stick into the static vents
Are static ports heated? Mine aren't. I've never seen heated static ports in GA.
Almost every Piper has one, as pitot tube and static port are part of the same probe. And every twin turboprop and bizjet (they are GA planes as well...) has heated static ports. Amber warning lights and flashing "Master Caution" will come on, should any of the heaters fail (or left forgotten to turn on).
Almost every Piper has one, as pitot tube and static port
Mine doesnt.
