WXforum.net

Weather Station Hardware => What Weather Station Should I Buy? => Topic started by: DesertBlue on December 19, 2021, 03:46:48 PM

Title: Any web reviewers who actually test the stations? / Reference sensors
Post by: DesertBlue on December 19, 2021, 03:46:48 PM
Hi all – I found a couple of weather station review websites, but it looks like they're doing the same thing that I've noticed for all products. They're not actually testing the stations, and are just posting Amazon affiliate links after some sort of "review" write-up. In most cases, they don't actually possess the product – i.e. they've never been in the same room with the product. They're just sort of winging a review based on manufacturer marketing material, Amazon product listing info, and maybe the shopper reviews on Amazon (which may or may not be real, much less competent), and collecting all the affiliate commissions.

One of the things I'm most interested in is accuracy, mostly temperature but the other variables too. A website called Weather Station Advisor seems to actually possess the stations they reviewed, at least in some cases. But their accuracy review was just this sentence, with a lot of repetition and reformulation: "I was pleasantly surprised to see this model hold up so well against the Vantage Vue, however, expect a little more variability in terms of accuracy."

There were no numbers, no data, nor any kind of supplement or appendix with the data. Is it typical to not report accuracy data? It's weird to report accuracy with mere words, especially completely ambiguous language like "hold up well".

Accuracy is interesting because of the reference issue – how do we know that a station is accurate, or the degree of its accuracy/inaccuracy? We need some sort of reference sensors that we know are accurate to some standard. I was going to ask about that issue separately – how are people testing accuracy? What do they use as reference sensors? The reviewer above implied that the Vantage Vue is an acceptable benchmark. Is that the best approach? Is it known to be super accurate?

I'm used to having calibration weights for high precision weight scales, and gauge blocks and rules for length and straightness references. Reference objects for things like mass and length are pretty straightforward. But needing reference sensors is a new challenge for me, and I reckon it will be expensive to get high-accuracy reference gear. Are there simple, focused sensors for that purpose, without being a redundant full-fledged weather station like the Vantage Vue?

Thanks.
Title: Re: Any web reviewers who actually test the stations? / Reference sensors
Post by: johnd on December 19, 2021, 05:09:13 PM
The only solid reviews of Davis stations I'm aware of are by Stephen Burt:

http://measuringtheweather.com/wp-content/uploads/2012/06/Stephen-Burt-Davis-Instruments-Vantage-Vue-review-Sept-2013-c-Stephen-Burt.pdf

http://measuringtheweather.com/wp-content/uploads/2012/06/Davis-Vantage-Pro2-AWS-review-2009-c-Stephen-Burt.pdf

A bit old now, and the VP2 specs have improved since that review was done, but might be a starting point.
Title: Re: Any web reviewers who actually test the stations? / Reference sensors
Post by: mauro63 on December 20, 2021, 04:52:29 AM
Hi all – I found a couple of weather station review websites, but it looks like they're doing the same thing that I've noticed for all products. They're not actually testing the stations, and are just posting Amazon affiliate links after some sort of "review" write-up. In most cases, they don't actually possess the product – i.e. they've never been in the same room with the product. They're just sort of winging a review based on manufacturer marketing material, Amazon product listing info, and maybe the shopper reviews on Amazon (which may or may not be real, much less competent), and collecting all the affiliate commissions.

One of the things I'm most interested in is accuracy, mostly temperature but the other variables too. A website called Weather Station Advisor seems to actually possess the stations they reviewed, at least in some cases. But their accuracy review was just this sentence, with a lot of repetition and reformulation: "I was pleasantly surprised to see this model hold up so well against the Vantage Vue, however, expect a little more variability in terms of accuracy."

There were no numbers, no data, nor any kind of supplement or appendix with the data. Is it typical to not report accuracy data? It's weird to report accuracy with mere words, especially completely ambiguous language like "hold up well".

Accuracy is interesting because of the reference issue – how do we know that a station is accurate, or the degree of its accuracy/inaccuracy? We need some sort of reference sensors that we know are accurate to some standard. I was going to ask about that issue separately – how are people testing accuracy? What do they use as reference sensors? The reviewer above implied that the Vantage Vue is an acceptable benchmark. Is that the best approach? Is it known to be super accurate?

I'm used to having calibration weights for high precision weight scales, and gauge blocks and rules for length and straightness references. Reference objects for things like mass and length are pretty straightforward. But needing reference sensors is a new challenge for me, and I reckon it will be expensive to get high-accuracy reference gear. Are there simple, focused sensors for that purpose, without being a redundant full-fledged weather station like the Vantage Vue?

Thanks.

Hello,

very interesting speech and that I have already had the opportunity to face many times, I deal with amateur meteorological networks, specifically their quality, and I have a purely "metrological" approach.
As you well know, in metrology the characterization of a sensor, whatever it measures, is a necessary practice to determine accuracy, precision, and in general the characteristics of the sensor, and, as far as temperature sensors are concerned, it is a procedure that is carried out to direct comparison in climatic chambers with primary sensors in turn characterized to the fixed points.

Unfortunately, when we talk about a meteorological station, or in any case a temperature sensor, in the atmosphere we find ourselves having a data that is the result of a myriad of thermal transfers, not always desired, which makes it impossible to determine the exact measurand that interests us, that is, the air temperature.

Personally, as regards our tests as an association, we use some test stations, mine is one of them, where we use reference sensors, in my case with certification, inserted in high-level solar shading, based on these tests we try to obtain a global measurement uncertainty (sensor + solar radiation shield) as accurate as possible but, although this is done with care and criteria, and in the most varied climatic conditions, the data will always be approximate and not verifiable.

M.