WXforum.net
Weather Station Hardware => Ambient Weather and Ecowitt and other Fine Offset clones => Topic started by: Topcat on April 03, 2020, 07:09:36 AM
-
Hi
Can anyone assist on calibrating the UV / solar readings. Mine seem extremly high currently showing 8/824 and its partly cloudy. I checked some nearby stations that seemto be a lot lower
Its been up and running for nearly 2 weeks now.
current readings can be found at https://theforsters.me.uk
Thanks
-
I would suggest that you contact Lucy (support@ecowitt.com)
Whilst you can apply an offset in Calibration page for the solar radiation setting your station would suggest something is very wrong.
Check that you have not accidentally applied an offset (it should be 1.0)
At this time of year in the UK you should not really see more than 6-7 max.
You may however see a spike if you have water droplets on the sensor as they act as magnifying lens. Not much you can do about those other than keep the sensor clean.
-
Hi Mandrake
Thanks for the response. Checked offset and its set at 1.0. I will email Lucy to see if they will ship a replacement. The rest is all working fine.
I might reposition unit to see if it makes a differnce, im not sure but sunlight might be refecting towards it from a neighbours green house but then it would not be all day due to sun position, could that be an issue? unit is at a height of about 6 metres currently until i mount on the roof.
Regards
N
-
Its possible that reflected Sun might be an issue.
I have been recording a max of about 7 around the middle of the day in the current fine weather we are having in the south.
-
However, I have observed that all Fine Offset devices have very high UV values compared to nearby Davis stations. For example, my station and others in my area (Fine offset clones too) gave UV values of 6-7 in February. Davis stations in my area on those same dates never exceeded a UV value of 3.
My station (February):
(https://i.gyazo.com/b87f573e9e55d8a6fa89692561ec9540.png)
Nearby Davis station (February)
(https://i.gyazo.com/48420884b0f8d8fd56b4ec165491301c.png)
In March the same. My station values 9-10 many days, Davis station maximum values of 5
I don't understand what this difference is, maybe it will be the sensor, but it is not something that bothers me :grin:
-
It looks like the Fine Offset UV sensors are of poor quality. At Awekas.at, almost all Fine Offset stations have UV sensors blocked.
I have introduced UV calibration = 0.55 at my stations.
Currently UV = 7 in the UK is a bit too much. How much will your station show in the summer?
-
My ws80 Ecowitt is ok compared to Davis Vantage pro, same exposition
last two hours, no offset
[ You are not allowed to view attachments ]
-
My ws80 Ecowitt is ok compared to Davis Vantage pro, same exposition
last two hours, no offset
[ You are not allowed to view attachments ]
Here we write about the UV sensor. You give a comparison of a solar sensor. These are two different things.
-
My ws80 Ecowitt is ok compared to Davis Vantage pro, same exposition
last two hours, no offset
[ You are not allowed to view attachments ]
Here we write about the UV sensor. You give a comparison of a solar sensor. These are two different things.
Sorry ;)
-
My ws80 Ecowitt is ok compared to Davis Vantage pro, same exposition
last two hours, no offset
[ You are not allowed to view attachments ]
Here we write about the UV sensor. You give a comparison of a solar sensor. These are two different things.
Sorry ;)
Mauro, write what Your UV readings look like.
-
So to eliminate any glass refections unit now moved and mounted at approx 8 meters above ground level. sensors are clean no signs of any moisture.
So just after midday I see UV of 8 and solar of 877 w/m2 so i should be burnt to a crisp :grin:
I guess i can try to adjust offset to get a more realistic figure. I did some research and as Mandrake says 7 max for the UK in the middle of summer and that would be a rare reading average is 5-6 in the hieght of summer in the south.
Does the Solar reading look more realistic.
Regards
N
-
On page:
http://www.meteovista.com/Europe/United-Kingdom/Sunpower-London/4163337
there are today's maximum UV values for Great Britain.
-
My ws80 Ecowitt is ok compared to Davis Vantage pro, same exposition
last two hours, no offset
[ You are not allowed to view attachments ]
Here we write about the UV sensor. You give a comparison of a solar sensor. These are two different things.
Sorry ;)
Mauro, write what Your UV readings look like.
I have to admit that I never gave importance to this parameter, but only because I don't care
having said that, I believe that my ws80 overestimates this parameter, today often has touched the value of 7
Mauro
-
My ws80 Ecowitt is ok compared to Davis Vantage pro, same exposition
last two hours, no offset
[ You are not allowed to view attachments ]
Here we write about the UV sensor. You give a comparison of a solar sensor. These are two different things.
Sorry ;)
Mauro, write what Your UV readings look like.
I have to admit that I never gave importance to this parameter, but only because I don't care
having said that, I believe that my ws80 overestimates this parameter, today often has touched the value of 7
Mauro
By page:
http://www.meteovista.com/Europe/Italy/Sunpower-Italiani/4864373
in Italy, UV should be between 3 and 5, and around 3 to 4 in your area.
Could Fine Fine Offset UV sensors be of poor quality?
-
My ws80 Ecowitt is ok compared to Davis Vantage pro, same exposition
last two hours, no offset
[ You are not allowed to view attachments ]
Here we write about the UV sensor. You give a comparison of a solar sensor. These are two different things.
Sorry ;)
Mauro, write what Your UV readings look like.
I have to admit that I never gave importance to this parameter, but only because I don't care
having said that, I believe that my ws80 overestimates this parameter, today often has touched the value of 7
Mauro
By page:
http://www.meteovista.com/Europe/Italy/Sunpower-Italiani/4864373
in Italy, UV should be between 3 and 5, and around 3 to 4 in your area.
Could Fine Fine Offset UV sensors be of poor quality?
In my opinion, every entry-level weather station use low cost and low-quality sensor.
for some parameter, like temperature, humidity, wind, rain, barometric pressure, there's no problem, now it's possible to have very good performance with very low cost sensors, but there are some parameters, like UV, that need something other, more professional, more accurate etc
Mauro
-
If a sensor is already installed, it should show real values. Overstating by 200% or 300% is a misunderstanding.
-
If a sensor is already installed, it should show real values. Overstating by 200% or 300% is a misunderstanding.
in fact I always thought that a weather station should have all the sensors separate
I will only choose the sensors based on the parameters that interest me, in fact my main stations are like this, these are toy stations and must be taken for what they are, for certain things they can also go well, for others not
Mauro
-
well I admit I tend to use these as indicators as far as solar UV is concerned
For reference my WH65 (Triwing sensor) recorded 480 w/m3 (UVI 4)
My WS80 which is 2 feet away recorded 667 w/m3 (UVI 7)
This is for a cloudless sunny day in the south of the UK today
I am more inclined to believe the WH65 though in most aspects the WS80 is the superior instrument for everything else
I have not tried calibrating the UV sensor as it seems to be more of an indicator than anything else.
I must admit I have done any research in the sensor accuracy or discussed this with Lucy.
-
For Fine Offset stations the UV and Solar Radiation all comes from the very same sensor. It is a Lux sensor. This then gets converted to both UV and Solar Radiation based on some predetermined formulas.
That said we do have the capability to make calibration adjustments as Krojan has done.
Davis VP2+ stations have separate UV and Solar Radiation sensors that are of much higher quality.
-
I have emailed Lucy so will see what they say.
In the mean time how would you go about calculating what offset to apply to make reading more realistic. I note in the manual it says to recalibrate Solar/UV every 2-3 months but no instructions on how you would do this.
Regards N
-
On my Osprey sensor the reported UV Index is double to what it is on the nearby Davis Vue Pro station, so I applied offset of 0.5 to my GW1000 and the readings coresponding with Davis station and to the predicted UVI for my location. I only noticed that after reading this post.
-
We are comparing the UV result in a lot of Fine Offset stations in our network and had found that an offset of 0.65/0.7 can give a result with good accuracy compared to official maps and Davis UV sensors
this comparison work will continue next days thank to the good weather conditions
Mauro
-
I don't want people to take anyone else's recommendation of what is a proper UV or Solar Radiation calibration adjustment number. The reason being that the calibration required is going to be very different for people living in different parts of the world. These stations use a Lux sensor and they try to approximate Solar Radiation and UV based on some loose definition of how much radiation there is in visible light. This is going to differ if you live in Italy versus the UK and it will differ if you live in different parts of the US.
Therefore please do not go by anyone else's calibration recommendation... Unless you live where they live AND you really trust them.
The simple explanation is that these stations don't have a real Solar Radiation and UV sensor. They have a Lux sensor which is a different animal and FO is trying to deliver Solar Radiation and UV by making some broad assumptions on Lux.
-
I don't want people to take anyone else's recommendation of what is a proper UV or Solar Radiation calibration adjustment number. The reason being that the calibration required is going to be very different for people living in different parts of the world. These stations use a Lux sensor and they try to approximate Solar Radiation and UV based on some loose definition of how much radiation there is in visible light. This is going to differ if you live in Italy versus the UK and it will differ if you live in different parts of the US.
Therefore please do not go by anyone else's calibration recommendation... Unless you live where they live AND you really trust them.
The simple explanation is that these stations don't have a real Solar Radiation and UV sensor. They have a Lux sensor which is a different animal and FO is trying to deliver Solar Radiation and UV by making some broad assumptions on Lux.
we are not doing this, we are telling our users
"hey, the UV data on the map of your station are not realistic compared to the official station that you can find at the following link ...... and to those of the official maps that you find at the following link ......
try to apply an offset, starting from this example value until you have a data as close to the official one "
it is very different
Mauro
-
Mauro63,
I agree that you are not telling people to use your calibration unless they are near you. I was just trying to make it clear to others that may stumble on posted calibration offsets and blindly accept them.
-
Mauro63,
I agree that you are not telling people to use your calibration unless they are near you. I was just trying to make it clear to others that may stumble on posted calibration offsets and blindly accept them.
Thanks Galfert,
now the first step in our network is to disable UV sensors of Ecowitt/Fine Offset/Froggit and other weather station of the same type
we are inviting our users to test their personal offset comparing data with the closest professional weather station and with official maps
the second step will be a very accurate check of the results and, where will be possible, to rehabilitate the sensor in our maps
for my part, Ecowitt is updated about this problem and I'm at their disposal to find a technical solution
in my opinion, there's no technical solution, so, due to the importance of this parameter, could be better to specify with great emphasis on the equipment manuals that UV data is approximative and must not be used for human health safety
Mauro
-
Seems to me that this is a stand out opportunity for Ecowitt/Fine Offset to develop a stand alone genuine UV instrument to complement the growing sensors they have.
I imagine it must be possible and might be cost effective now rather than deriving from lux values.
I'll suggest it that this is added to products to investigate if its not already on their roadmap
-
Mandrake,
Great idea for FO/Ecowitt to come out with real Solar Radiation and UV sensors! Let us know what they say. All they have to do is come up with a standalone upgrade Radiation and UV sensor unit that overrides the current data, just as the standalone rain and anemometer override the all-in-one array data.
I would be willing to pay for that upgrade. They could then incorporate it into a newer complete kit. HP7557 anyone??? (I just made that model number up because it looked cool....7 is 2 more than 5.....so there could be HP7556, HP7557, and HP7558....to replace HP2551, HP2552, and HP2553).
-
Mandrake,
Great idea for FO/Ecowitt to come out with real Solar Radiation and UV sensors! Let us know what they say. All they have to do is come up with a standalone upgrade Radiation and UV sensor unit that overrides the current data, just as the standalone rain and anemometer override the all-in-one array data.
I would be willing to pay for that upgrade. They could then incorporate it into a newer complete kit. HP7557 anyone??? (I just made that model number up because it looked cool....7 is 2 more than 5.....so there could be HP7556, HP7557, and HP7558....to replace HP2551, HP2552, and HP2553).
I totally agree, and I would be willing to pay also
this could be the final step to have a very powerful weather station at right price ;)
Mauro
-
Yes, that seems like a very large correction factor to have to apply. - Rich K.
-
Mandrake,
Great idea for FO/Ecowitt to come out with real Solar Radiation and UV sensors! Let us know what they say. All they have to do is come up with a standalone upgrade Radiation and UV sensor unit that overrides the current data, just as the standalone rain and anemometer override the all-in-one array data.
I would be willing to pay for that upgrade. They could then incorporate it into a newer complete kit. HP7557 anyone??? (I just made that model number up because it looked cool....7 is 2 more than 5.....so there could be HP7556, HP7557, and HP7558....to replace HP2551, HP2552, and HP2553).
I totally agree, and I would be willing to pay also
this could be the final step to have a very powerful weather station at right price ;)
Mauro
Well the idea has been well received and Ecowitt are looking into the feasibility.
I will advise here when I get any feedback!
-
Today I raised the offset to UV = 0.75.
Now it looks correct in Poland.
I would also buy a separate solar / UV sensor.
In the HP2551 manual I found such information as in the attachment:
[ You are not allowed to view attachments ]
In the WS80 manual, it says that UV accuracy is ± 2.
[ You are not allowed to view attachments ]
-
I have been experimenting with the offset for my station and have currently settled with 0.57. This gives a more realistic indication of UV reading. Having looked around nearby stations on various PWS sites its difficult to find anything to reference against that may have a more accurate reading. I did find a couple of websites I tried to reference against https://uk-air.defra.gov.uk/data/uv-index-graphs and https://uv.willyweather.co.uk/se/oxfordshire/oxford.html but dont know how the data is provided or how accurate it is.
It would be easier if output values stepped in .1 increments as opposed to just a full jump from 1 to 2 etc ie see 1.4, then 1.5 as it increases, maybe if ecowitt build a seperate sensor we could get this feature.
Another question is how accurate is the Solar radiation reading in W/m2.
Thanks for all the input on this thread
Regards
Nick
-
The solar radiation value again is a derived value from the light sensor (Lux) which as you indicate has a 15% +- tolerance.
-
Sorry to resurrect this thread but I am finding similar inaccuracy with my WH2910c - point taken that the UV is merely calculated instead of measured but it's so far off that it seems pointless to include.
Here is a comparison of the official data for Melbourne Australia (arpansa.gov.au) with my readings, on two random days (figures rounded):
[ You are not allowed to view attachments ]
[ You are not allowed to view attachments ]
Ignoring obvious cloud cover, there is a big discrepancy with the Ecowitt reading much higher. I don't think a correction factor will quite do it - at times it is 2/3 and at other times double.
It is a pretty important measure, at least in Australia where the UV can reach 10-12 due to the Ozone hole :sad:
I'd definitely buy a separate sensor if it was available.
-
Sorry to resurrect this thread but I am finding similar inaccuracy with my WH2910c - point taken that the UV is merely calculated instead of measured but it's so far off that it seems pointless to include.
I'm not sure what sensor is in the WS68, but if it's the same as the WS80 (which I would expect) the calibration strategy is a bit troublesome due to the sensor design. The handy thing is you (in Australia) can rent a UV meter from ARPANSA for comparison purposes for just $55.
For those wanting to actually go and calibrate their WS68 or WS80 sensor, the following may be good considerations when you do it:
- If you want to calibrate it for lux/fc readings (most accurate way), then calibrate the solar gain against a photometric CIE-1931 sensor. More accessible is a handheld lux sensor. Most commercial ones have decent accuracy while the cheap ebay ones can be 15-20% off (more useful for relative measurements than calibration). I would say this would be the 'proper' calibration method provided that the measurement values used are lux/fc.
- If you want to calibrate for irradiance, then calibrate against a pyranometer by adjusting the solar gain as well as the w/m2 conversion factor. The tough part is a pyranometer has a completely different response than a lux sensor so you will be at the mercy of the sensor chip algorithm.
- If you want to calibrate for UV, the manufacturer recommend method for UV sensors (not lux sensors) is to calibrate against a UVB meter with the sun in a cloudless sky at an angle >60 degrees but less than 90 degrees. Ecowitt only gives adjustment for UV gain, so I expect if you are in an extreme areas there will be some challenges to get a good calibration. It looks like the Ecowitt UV index is calculated by simply taking irradiance (w/m2) and dividing that by 100, floored to be an integer. Then you can look to adjust the UV gain to find a suitable setpoint.
There is a scientific place that rents such instruments in my city, but I have no idea how much they are charging for the daily rate. If it's a reasonable price, in the spring I may try to get rent a logger and see how the value compare and then calibrate accordingly for my locale.