Page 1 of 1

Accuracy of temp readings - Watson

Posted: Sun 31 Jan 2010 11:48 pm
by EvilV
I wonder if anyone has any comment on whether these Fine Offset units produce reliable temperature readings when sited properly?

I have taken a bit of care to get my temp sensor unit as far from buildings as I can within the bounds of my property, which is a typical British semi with a bit bigger garden than some. I moved the unit down out of the sunshine for the obvious reason, and I have it in a sheltered spot about 20 meters from any building. It is sited at about 5 feet above an earth surface and is always shaded from sun by a tall fence. The other side of the fence is an open field of about five acres - give or take.

The question of accuracy has come up because at night, my unit seems to be reading about 0.7 of a degree C lower than surrounding stations. I obviously can't vouch for the way these other stations are located. They could be right near houses and be getting warmed by that, but maybe the sensors are not that repeatable. I could easily put in an offset on the calibration screen, I suppose.

I'll check how it compares in day time temperatures. If it is down by the same amount, then I suppose an offset is the answer.

Re: Accuracy of temp readings - Watson

Posted: Mon 01 Feb 2010 2:15 am
by beteljuice
Why do you think it is an error ?
All locations are different - the only way you can check is to compare against a co-located thermometer. But even then, if there is a difference, how do you which one is the most correct 8-)

Re: Accuracy of temp readings - Watson

Posted: Mon 01 Feb 2010 8:02 am
by EvilV
Good point there Beteljuice. Before I read your post, I just did a search for a max / min thermometers. Then I just thought, 'What does it matter?' And yes - I'm familiar with the concept of 'frost pockets'. I should also compare the altitude of the sites and then take that into account. This house is on a bit of a hilltop, but from memory, I think there is a decline in temperature of one degree c per thousand feet.... It isn't that big a hill. I'm probably 200 feet above most of the stations I'm comparing with.

Cheers

Re: Accuracy of temp readings - Watson

Posted: Mon 01 Feb 2010 8:28 am
by steve
My VP2 and Fine Offset sensors are quite close together. The VP2 sensor (with its own radiation shield) is screwed to the outside of the Stevenson Screen, inside which is the FO sensor. They are both currently reading 2.2C.

Re: Accuracy of temp readings - Watson

Posted: Mon 01 Feb 2010 11:03 am
by EvilV
Thanks Steve.

I just did another series of checks since 10.30 when it has warmed up. There are a few localish stations that obviously have sunshine issues because they always heat up as soon as the sun comes out. Leaving these ones aside, my temp now compares very closely with the Newcastle Airport station, and the local Met Office station to the west of Newcastle. It is within 0.1 degree c of these at 0.9c. Since the 'problem' only seemed to be happening a couple to three degrees colder than that, I don't believe the thermal sensor is non-linear. I am concluding that my garden is just in a cold spot and collects really cold air. I had noticed that myself when sleeping in pretty much the same spot in my jungle hammock, before I took it on a cycling tour.

Thanks for your comments.