Topic: Sensor response time

[AllenWood]

What sort of response time can I expect from a Vantage Pro 2 temperature probe ? This is in a well aspirated environment.
From what I can determine it is very sluggish.   

Thanks.

[AllenWood]

Ok. Thanks to a posting by LABOB I can see thermal inertia delay has been discussed and answer found. I concur that it appears that the Advantage Pro  console display  displays and "apparent" value equal to an approximated 16 minute running average.

[George Richardson]

"Thanks to a posting by LABOB I can see thermal inertia delay has been discussed and answer found."

Would you please post the location of this thread.

George

[jerryg]

I found it here http://www.wxforum.net/index.php?topic=18805.0

[jerryg]

I want to add my two cents in here, after reading about the console showing temp update every minute and that being some kind of average is not right. The Davis sensor has some lag time to it and how fast the temp is changing can cause the reading to look long between updates but i have a rm young shield and homemade sensor to fit it and it updates the temp on the console about every 10 to 12 seconds as per the specks.

[johnd]

The VP2 ISS temperature sensor updates every 10 secs and, to the best of my knowledge, there's no averaging or smoothing carried out either in the ISS circuitry or in the console. The 10-second updates are effectively available in the LOOP data from the logger.

Any apparent lag in temperature readings will be down to the aspiration rate of the fan in the FARS and to the intrinsic thermal inertia of the sensor element; I strongly suspect that the aspiration rate will be the dominant factor here and that by using an external DC supply (perhaps overvolted within reason if required) then the rate can probably be increased substantially. Removing the protective cover around the sensor element would also improve the sensor response I presume, but obviously at the risk of increasing sensor contamination if used long-term like this.

[txagwx]

I have 2 other data loggers in two different shields (RM Young, Davis), and I can tell you the response time is very similar in all 3.

[dendrite]

I'm all for faster response times, but keep in mind that the "official" airport stations in the US (ASOS, AWOS, etc) all use a running 5-min avg as the reported current temperature/dewpoint. IIRC, this was the solution when the stations were transitioning to the automated sensors from the old max/min thermometers since the old thermometers had a slower response time.

[CW2274]

I'm all for faster response times, but keep in mind that the "official" airport stations in the US (ASOS, AWOS, etc) all use a running 5-min avg as the reported current temperature/dewpoint. IIRC, this was the solution when the stations were transitioning to the automated sensors from the old max/min thermometers since the old thermometers had a slower response time.

I never researched this, but the 5 minute average completely makes sense, and I knew it had to be, with the observations I've seen with the ASOS's in my area. Our Davis's have no such function. What it measures, is what you get.

[dendrite]

I'm all for faster response times, but keep in mind that the "official" airport stations in the US (ASOS, AWOS, etc) all use a running 5-min avg as the reported current temperature/dewpoint. IIRC, this was the solution when the stations were transitioning to the automated sensors from the old max/min thermometers since the old thermometers had a slower response time.

I never researched this, but the 5 minute average completely makes sense, and I knew it had to be, with the observations I've seen with the ASOS's in my area. Our Davis's have no such function. What it measures, is what you get.

The raw 1-min data is available for the ASOS sites at the end of every month...

https://www.ncdc.noaa.gov/data-access/land-based-station-data/land-based-datasets/automated-surface-observing-system-asos

Often you will see the max/mins be 1-2F higher/lower than the official max/mins because of the averaging. For instance, the official record high of 108F at EWR on 7/22/2011, actually had a raw 1-min max of 109F.

You will notice just how "jumpy" the temps are without the averaging which shows how strong the response is with these high end stations. Some people like smooth temp/dew graphs, but in reality they're anything but smooth.

[CW2274]

I'm all for faster response times, but keep in mind that the "official" airport stations in the US (ASOS, AWOS, etc) all use a running 5-min avg as the reported current temperature/dewpoint. IIRC, this was the solution when the stations were transitioning to the automated sensors from the old max/min thermometers since the old thermometers had a slower response time.

I never researched this, but the 5 minute average completely makes sense, and I knew it had to be, with the observations I've seen with the ASOS's in my area. Our Davis's have no such function. What it measures, is what you get.

The raw 1-min data is available for the ASOS sites at the end of every month...

https://www.ncdc.noaa.gov/data-access/land-based-station-data/land-based-datasets/automated-surface-observing-system-asos

Often you will see the max/mins be 1-2F higher/lower than the official max/mins because of the averaging. For instance, the official record high of 108F at EWR on 7/22/2011, actually had a raw 1-min max of 109F.

You will notice just how "jumpy" the temps are without the averaging which shows how strong the response is with these high end stations. Some people like smooth temp/dew graphs, but in reality they're anything but smooth.

Excellent info! I'll be as bold to say if I were to put my 67CFM fars (and sht15) next to an ASOS and disabled the averaging, they'd probably pretty much mirror themselves.

[johnd]

I never researched this, but the 5 minute average completely makes sense, and I knew it had to be, with the observations I've seen with the ASOS's in my area. Our Davis's have no such function...

If you set the Temp to be 'Average Temp over Archive Period' (Set Temp & Hum Calibration) in WL and choose a 5min archive interval then you will indeed log 5 min mean temps.

[CW2274]

I never researched this, but the 5 minute average completely makes sense, and I knew it had to be, with the observations I've seen with the ASOS's in my area. Our Davis's have no such function...

If you set the Temp to be 'Average Temp over Archive Period' (Set Temp & Hum Calibration) in WL and choose a 5min archive interval then you will indeed log 5 min mean temps.

Yes, I'm aware of this but never bothered with it. Will it send out the average to the internet consumers, or is that average data only available to me? While I await your reply, time to experiment.....

[johnd]

As far as I know, the average will only be logged to the WL archive data and hence will be available to any process that consumes the archive data (but presumably not otherwise).

[CW2274]

As far as I know, the average will only be logged to the WL archive data and hence will be available to any process that consumes the archive data (but presumably not otherwise).

Thank you. The 5 min. average is indeed calculated on the data page in WL. However, I believe my experiment has already answered the question. It appears that the temp. the console reads at the exact moment of TX is only what is sent, no averaging. Bummer. I will continue to monitor to be sure.

[CW2274]

Experiment over. Only the exact temp. at TX is sent. The reason I don't care for that, is that there was a 5 minute span my temp. went between 110.4 and 106.5, a whopping (at least to me) 3.9 degree difference. The temp broadcast online at that time was 111 (don't ask me how), but the average was 108.2, quite the difference. I think an average would be a truer measurement of the actual temp. instead of the ebb and flow that now occurs.