Shown below is the temperature readings I am getting from 3 different sources for the same location!
The 1st 2 are community devices, the 3rd is the built-in device. They are @bangali's ApiXU Weather Driver V2.0.0, @Cobra's Custom WU Driver V2.5.0 and the built-in Weather Underground respectively.
What gives the huge discrepancies for temperature? Anybody got some insight into this?
There is an old saying...... "A man with one thermometer knows what temperature it is, a man with two thermometers does not know what the temperature is"
Seriously my Wunderground site is about 3 miles away from me and the temperature is between 2.5 and 3 °C higher. I would think this is normal. However I would think the 36 °C (I'm assuming your in °C) is in the sun or has some other issue.
Do you have an old fashion liquid thermometer (used to be mercury)? I find them to always be in the "ball park" and useful for a sanity check.
One of my drivers uses either apixu or wu and I find wu most accurate.
Probably because there is possibly a station nearer to me.
Now I just use my own pws and (after some adjustment) I know it’s correct.
Actually I would be interested to use a station id in the wu settings instead of a location.
Both in mine and the built-in driver
That way you would guarantee that they are using the same data
Interesting...
I just did the comparison
Temp is 0.5 degrees C lower with the built-in driver (could just be rounding)
But... humidity is way off (about 5%)
I know the station is correct because it’s mine (realised I’m still pushing data to wu. But will probably continue until they cut off access)
It might be worth checking the Wunderground location temperature in you web browser to see if it is the same as on you display. Just a simple sanity test.
If different look at other sites in your area, just to get a feel of the site to site variation.
The truth is, you are looking at three uncalibrated thermometers, or one calibrated, and two uncalibrated thermometers. Three sources at the same location would be three weather stations right next to one another. I'm sure that's not what you're looking at. Remember, weather is local, so if you don't have your own calibrated station, you need to accept an average that seems correct based on a large enough sample size.
This is a very good point @JohnRob makes. If you don't have the expensive equipment and a controlled environment of calibrate your own thermometer (and how many would really) then looking at temperatures from various weather stations in close range of each other on WU is your best bet to find an average that you can live with and follow that one.