Passive air temperature measurement

Being able to compare two different systems, I’ve found that there are significant errors in passive temperature readings of a few


passive air temperature.docx (182 KB)

Hi

Is your sensor the same as the attached image, if so looking at the top where the solar sensor is housed .It looks like it has a clear plastic cover over the top of the sensor. This would cause temperature spikes , it would heat up the the air inside the housing. There’s a small stand off tube if this is Hollow then this also has the possibility of holding heat. For a test I would black out with aluminium tape the solar sensor cover just leaving a small opening where the solar sensor can be exposed, the cover looks like a glossy plastic perspex with somehow space it could be heated up and also may be cool at night

Batteries inside the housing then having the solar shield over the top, I would have thought this would not make that much of a difference

If you’ve rehoused the senses in your own solar screen would need to see an image to make any comments as from different designs they can have a big effect on how accurate the senses maybe
Hope this makes sense

Mick


images.jpg

It has always surprised me that the small, 4"/105mm diameter, Vaisala DTR502/503 non-aspirated shield performs well enough to be the standard in many official installations :?

My $0.02. One of the issues with low end passive shields is poor airflow and windspeed can certainly affect that. I would be surprised if the correction factor holds for a wide range of wind conditions.

My first station was a Honeywell one which had unshielded temperature sensor. I built a wooden house and put it in. No issues with direct/indirect sun. Due to some reasons, I’ve bought a similar priced WH3081 (basically similar to 3083) and put it to the same rod. Surprisingly, temperature read from 3081 was significantly higher than from the Honeywell. After few days, I’ve intentionally broken the housings and put a 3081 temperature sensor (including a shield) into the wooden house and a solar on top (outside). It worked OK since.
Due to several (connection) issues with a 3081, I’ve added a Davis VP2 to same location. Davis reading matches the Honeywell’s down to 0.X degree C. Out of curiosity, I have a 3081 outside again. Right now (this moment), the readings are:

Honeywell (in the wooden house): 17.6 degC / 71%
WH3081 (outside): 22.3 degC / 49%
Davis VP2 (standard): 17.0 degC / 74%

Solar is about 670 W/m2 from Davis and 59.3klux from WH3081

Months ago I contacted the seller that sold me the WH3081 claiming the shield was not efficient. Their response was the unit was sold in thousands with no one complaining about the issue.

Many (if not all) official AWS use these here in the UK including the roadside stations.

Quote from: niko on 17-06-2017, 17:33:55

It has always surprised me that the small, 4"/105mm diameter, Vaisala DTR502/503 non-aspirated shield performs well enough to be the standard in many official installations :?

Mick, thanks.

Yes, the unit is as you illustrate. The top bit houses only the solar and UV sensors, as well as a micro solar panel ‘to charge the niMH cells’ (bad terminology). It is in no way connected to the temperature/humidity measurement which is housed in the screen below. I haven’t looked at it in detail but it seems to work OK, closely compared to the Davis solar (I don’t have the Davis UV), with differences up to

As I said, in the article, the wind speed had little effect on both systems. This rather surprised me.

Thanks for your input.
Typical!

As supplied, Aercus cover their backsides. I quote from their documentation:

Temperature sensors need to be shaded (often times under the eaves of a house is a good place) whereas rain and wind sensors need to be unobstructed.
If you are located in the Northern Hemisphere, you can site your temperature sensors on north-facing walls slightly below the eaves. This will provide shelter from the rain while still having the sensor low enough that its readings are not affected by heat emanating from the roof.
[quote]If you are unable to mount your temperature sensor in the shade then you are likely to record temperatures higher than what you see from your local weather service who measure temperature in the shade. This is because the solar shields that most common stations ship with are ok but don

Mick, thanks.

Yes, the unit is as you illustrate. The top bit houses only the solar and UV sensors, as well as a micro solar panel ‘to charge the niMH cells’ (bad terminology). It is in no way connected to the temperature/humidity measurement which is housed in the screen below. I haven’t looked at it in detail but it seems to work OK, closely compared to the Davis solar (I don’t have the Davis UV), with differences up to

Agreed, that isn’t very logical. It suggests that adding fan aspiration would not change the result :?

In Reply #8 you are quoting me as having written “Months ago I contacted the seller that sold me the WH3081 claiming the shield was not efficient. Their response was the unit was sold in thousands with no one complaining about the issue.” - That wasn’t me…

My apologies for misquoting you.

I agree re forced aspiration but there may be differences: the vertical fan aspiration may involve higher air volumes/speeds round the sensor than the horizontal wind in the chicane of the screen ???

True, I should have been more clear. Since it’s clearly not designed to be converted to a true forced air vertical aspiration I was thinking more of simply blowing air on the shield with a fan.

I’m not sure what range of windspeeds you had during the test but I would tend to think that the WH chicane is very effective at blocking airflow.

FWIW here’s the original Davis shield comparison test document. Wind isn’t addressed other than by specifically testing on low wind days.

I’ve just started on my ‘Plan B’, mentioned above. I had difficulty separating the WH3083 parts so that the solar/UV/wind/rain sensors remained in the open, while the temperature/humidity/wireless were moved to ‘where the sun don’t shine’ (no, it wasn’t painful!). When I have some definitive results, I’ll explain it in detail and how to do it. Preliminary results show that the Davis was reading ~2