In Accuracy Of Weather Stations

[Habsish]

Are you certain that the data you collect and sometimes reported is correct?

I have an Oregon Wireless weather station BAA898HG with a couple of remote sensors.

It wasn't untill I put all 3 together I realised that one of the sensors was about 2 degrees C out and the humidity was about 10% out.

Contacted Oregon technical support who responded "Thank you for contacting Oregon technical support. for the unit to be 1 or 2c out is perfectly normal. there is going o be a fluctuation on the sensors as they are not manually calibrated and they are in fact 75% accurate."

Now that is something they do not say on the packet and strikes me as somewhat amazing when we all take readings down to 0.1 of a degree and state it as true.

Unfortunately I did not have a receipt to return it as it was a gift.

Suggest that all members who post should bear this problem in mind.

Regards

Graham

PS we all know that some people have faulty rulers as well that measure cm instead of mm. So it is not only electronics can be on the blink.

[suffolkboy_]

I think you get what you pay for alot of the time. Although 2C is way out! A traditional thermometer is better than that.

I still stand by Davis being the best. I have a Davis Vantage Pro 2....can't fault it, and you can send it back for calibration as often as you like, although it does cost.

[Tubs]

Habsish,

Really seems to be inexcusable to sell something such as a thermometer and accept that it could be only about 75% accurate. Should be a warning on the box and in their advertising.

Now have to check out my equipement using an ordinary thermometer.

Yes you pay for what you get but it should do the job reasonably.

What does anyone else think?

G

[johnholmes]

the data from nine was calibrated alongside manual thermometers which had in turn been calibrated alongside Met O equipment, calibrated at the National Physics Laboratory.

Having said that I've not checked them in the past couple of years.

[Polar Continental]

Hi Habsish

I have an Oregon Scientific WMR 928 Professional Automatic Weather Station (AWS) which is in full operation at my site and continually uploads live data to the Internet.

However the equipment I use for keeping records and reporting all meets Met Office specifications, all the appropriate instruments are housed in a Stevens Screen, including the electronic sensors belonging to the AWS.

The readings from the AWS have compared very well to the official MO Standard instruments since they were housed in the Stevenson’s screen. Plus or minus 0.2c.

However when the AWS sensors were on my north facing wall in permanent shade they often gave readings of 2c to 4c plus or minus in error.

Also, not sure if you are aware, but if your station is connected to your PC and you are running the appropriate software, (I use weather display) you can calibrate you sensors.

Hope that helps.

Paul

Edited December 2, 2008 by Polar Continental

[Tubs]

the data from mine was calibrated alongside manual thermometers which had in turn been calibrated alongside Met O equipment, calibrated at the National Physics Laboratory.

Having said that I've not checked them in the past couple of years.

Unfortunately no way of calibrating the Oregon ones other than trying to remember how much they are out.

[WhiteXmas]

That's ridiculous! It's not even as if it is advanced technology. Ok if it was 0.1-0.2 out then I would accept it but 2c is out of order. That could mean that your monthly average could end up a degree out of range! Unacceptable if you ask me. We pay money, no matter how much to get products which provide accuracy. If we wanted inaccuracy then we would just damn guess it ourselves and pay nothing!

Edited December 2, 2008 by WhiteXmas

[johnholmes]

Unfortunately no way of calibrating the Oregon ones other than trying to remember how much they are out.

there is

simply use the max and min from your sensor and check with that from a standard mercury max and min thermometer; placing both in about the same place.

How to check the mercury - easy'ish

for freezing point emerse in ice and check its on zero

for boiling point SLOWLY heat water to 100C

Edited December 2, 2008 by johnholmes

[Habsish]

Paul,

Sounds as if you have a nice bit of kit to play with.

Have to put it on my list to Santa who will probably ignor the request and give me socks (again)

Pitwood,

Do you have the same sort of set up as me?

Habsish

[Tubs]

there is

simply use the max and min from your sensor and check with that from a standard mercury max and min thermometer; placing both in about the same place.

How to check the mercury - easy'ish

for freezing point emerse in ice and check its on zero

for boiling point SLOWLY heat water to 100C

Thanks. What I meant was there was there no way to actually set the sensor up so it reads correctly.

Think the water when boiled or frozen needs to be pure otherwise any pollutant will give an incorrect reading. When boiling is done it should be at sea level. Mind is going back to my physics lessons that were too many years ago. Not that the accuracy is likely to become apparent with the equipment I would use.

[osmposm]

I have three brand new Technoline wireless remote (temp only) sensors + base unit. It's pretty cheap kit (£37 for everything from Weather Front), so I wasn't expecting the best results. but actually they're not too bad.

To assess how they performed I've temporarily set all three up next to each other not far from the house. The results are odd, but interesting. Most of the time #1 & #2 are within 0.2C of each other, but #3 is 0.3 to 0.6 below the other two. For example they are currently showing (+)5.3, 5.5 & 5.0C** respectively. Similarly last night's minimum showed as -0.3, -0.2 & -0.7C. Pretty consistent - though which, if any, is accurate is anyone's guess.

Sometimes, however, just to annoy me, they have a swap round. #2, particularly, will suddenly drop away below the normally lowest #3. The night before last it showed a minimum 0.6C below anything else. Which leaves me with what I was hoping for, a greatest trust in one of them, #1, which also happens to show something close to the mean of all three (most of the time). However.....who's to say if the saintly #1, for all its consistency, is anywhere near the true temp?!

And besides, being a temp extreme obsessive, I shall always be tempted to believe the lowest reading in winter, and the highest in summer...don't worry, the "data" will stay firmly in my own diary until such time as I buy some grown-up eqipment.

Ossie

**Now dropped to +5.2, 5.3 & 4.9C