Vantage Pro2 Sun Sensor

[onepk]

Hello, I've had a Vantage Pro2 system for a couple of months now and I'm curious about some data from the sun sensor. On partially cloudy days it often (actually it always at some point) shows values greater than are achieved at the same time on cloudless days (see attached plot, 1 min interval data from 26 & 28 May*). Is this caused by reflection from clouds being "brighter" as seen by the sensor than the patch of sky they fill or something else, instrumental or otherwise?

Also does anyone know of a source of quantitative info on the sun sensor (spectral response etc.)? I've asked Davis but not had any reply so far.

Thanks for any ideas.

* the decline at around 12:20 is an artificially induced shadow.

[onepk]

Just an update and to add at least one reply in case anyone is still looking!

I got a 'reply' from support@Davis only to say it must be caused by a reflection or reflections from something. Haven't time to list the reasons why this seems unlikely to say the least.

The folk at Prodata Weather Systems, from whom the kit was purchased, at least took time to listen and to look at the data in detail, but had no suggestions apart from possible reflections or connection/signal transmission problems. Even on that basis they did concede that it was strange the phenomenon only occurred on cloudy days.

They also pointed out that day-to-day comparisons may not be legitimate because of atmospheric clarity changes - a cloudless sky does not necessarily mean a very transparent sky . That is true, but I have examples where internally the day's irradiance envelope is quite obvious (from lengthy spells of unobstructed sun) and can easily be interpolated, yet between episodes of cloudiness during that same day, the irradiance still goes well above that day's envelope, let alone above the envelope of adjacent cloudless days.

I can't help but feel I'm missing something obvious here - still nobody any ideas?

Can anyone out there with a Davis sun sensor at least confirm that they do NOT ever see the same effect?

Thanks

[reef]

Sorry, just seen this! Welcome to the forum

Its sounds unlikely it would be a reflection as no surface would reflect 100% of light and certainly not as much as direct sunlight itself. That said, if the unit is picking up reflections from multiple sources on a partly cloudy day, in theory a higher solar radiation level could be recorded than direct rays on a clear day - similar to requiring sunglasses in winter when the solar radiation is much weaker I guess. It sounds a bit far fetched though.

Unfortunately I dont have a solar sensor on my VP2 due to issues locating it (why they dont allow it to be attached in a wireless anemometer port when its there in the unit I dont know!). As always with these issues (whether solar/anemometer/ISS) I would check first for water ingress, check the seal and hairdryer the unit on a light setting if any moisture is found and possibly re-grease the connectors.

[Evening thunder]

Hi, I actually wrote this post last night but forgot to post it so have recovered it from the forums 'autosave' feature. It mentions some things that Prodata Weather Systems say, but I am also not really sure of the reason either so my post might not be that useful. Like reef I don't have a solar or UV sensor either, mainly due to cost but also partly due to not being sure if I could easily locate it in a good environment. I guess one thing you could do is find online websites from people's Davis stations that publish graphs from solar sensors and see if they show a similar thing on partly cloudy days.

Anyway, what I wrote yesterday is:

'That's interesting, I was going to say that maybe the sunny day was more hazy/had a less clear atmosphere from dust/pollution from the continent, and the day with some clouds had a clear air mass so there was stronger solar radiation reaching the ground, or maybe there was less ozone in the upper atmosphere above the UK (not sure how much that varies or effects solar radiation in the UK though).

But then I noticed the clear spell from roughly 1pm-3pm showed very similar to the sunny day. Your theory of some clouds reflecting some sunlight to the ground could be correct, and I have heard of some thin high clouds also having the effect of intensifying the solar radiation that reaches the ground somehow (not sure how much or if that's completely true though)

But then it could be something instrumental like the sensor 'overcompensating' when the sun comes out from behind a cloud, if that makes sense.

Just some Ideas I could think of as I can't really think of much else at the moment, hopefully there might be someone out there with more info than me. '

[onepk]

Thanks for the comments reef and Stormmad26.

I note from the trouble-shooting section of the Vantage console manual that they suggest that unusually "high" solar sensor readings could be caused by the presence of high, thin cirrus. As a general meteorological question, any idea of the physical mechanism behind such a phenomenon?

One supplementary question I had, specific to the Vantage, concerns the time-stamp in the data records. Each stored data value is supposed to be the average readout over the requested interval (1 min.... 30 mins). The answer may be buried in the documentation - apologies if so - but does anyone know whether, for example, a data record time-stamped 08:01 with a 1 min archive interval will be the averaged data from

08:00:00 --> 08:01:00

08:00:30 --> 08:01:30 or

08:01:00 --> 08:02:00

Comparing the data displayed on the console with the archived data, I have my doubts that it's any of them! Also what is the sensor readout frequency (ie readings per time interval not electromagnetic!) - does it vary from sensor to sensor? Just wondering what it means when the manual says the console displays the "current" value.

Cheers

[Evening thunder]

On my station the data archive interval is set to every 5 minutes, and I think that on my station at least much of the data is the value at that specific moment, i.e the temp shown at 14:50 would be the actual temp at that specific time rather than an average of some kind.

I can tell this when the temp is rising or dropping, as the current value is nearly always the highest/lowest in the 5 minutes prior, rather than some mid-point. Though I can't remember if I changed any settings (or if it's possible to) 4+ years ago. For example when the temp dropped due to a shower this afternoon it showed:

Time.....Temp Out.....Hi Temp.....Low Temp

14:45 17.2 17.3 17.2

14:50 16.7 17.2 16.7

14:55 16.3 16.7 16.3

'Hi temp' and 'Low temp' are the high/low values recorded in the 5 minutes since the last archive time up to that current one.

With some variables like wind speed then I think it is the average for the time interval, as mine tends to differer slightly to the 10 minute average displayed by the console and Weatherlink, with slightly higher peaks, suggesting it's a 5 minute average instead. Values like rain rate are the highest recoded in the time frame as well.

Sensor read out frequency does vary from sensor to sensor, I know the anemometer is something like '2.5-3 seconds', and I think the temp might be every 10 seconds with the humidity perhaps less frequent, though I don't know without checking.

I don't really know what it would be like for solar sensors of course as I don't have one, but I hope that helps a bit if my post made sense!

Edit: Do'h, just realised you might have been asking about the archived data for the solar sensor specifically, but if not hopefully that helps.

[onepk]

Thanks again for your thoughts Stormmad26. What the data logger does in terms of averaging (rather than the timing of the interval) is given in this extract from the WeatherLink help file.

"What we call "current solar radiation" is technically known as Global Solar Radiation, a measure of the intensity of the sun’s radiation reaching a horizontal surface. This irradiance includes both the direct component from the sun and the reflected component from the rest of the sky. The solar radiation reading gives a measure of the amount of solar radiation hitting the solar radiation sensor at any given time, expressed in Watts per square meter (W/m2). The value logged by WeatherLink is the average solar radiation measured over the archive interval."

It seems to be different for temperature in the sense that averaging is optional:-

"The Vantage Vue/Vantage Pro/Vantage Pro2 either samples the temperature at the end of the archive period and writes it to the data logger, or samples the temperature throughout the archive period and calculates an average value at the end of the period.

Note: Temperature averaging is available as an option only for Vantage Vue, Vantage Pro consoles with firmware is dated 7/18/2002 or later. "

On the latest Pro2 software, the temperature averaging option is settable under "Set temperature and humidity calibration" and the temperature data to be averaged is sampled every 5 secs for any archive interval <= 10 mins.

This still leaves open the question of when an archive time interval starts. I've asked Davis to advise on this - hopefully they're not on a 4-day weekend!

Any ideas on where I might find a pundit to answer the "high cirrus" question? I could try the Met Office I suppose, they usually (ie are presumably legally obligated to) reply to all dumb questions!

Cheers

Onepk

[onepk]

Just a quick update in case anyone is still awake!

Adrian Hudson (http://www.willandweather.org.uk/index.php) was kind enough to confirm that he sees the same effect with his sun sensor and said he thought it was a well-known effect caused by the sun's illumination of clouds. I am attempting to track down any academic study which explains or models the physics of the effect in detail (replies awaited from Davis support and the Met Office).

In the meantime I have gained access to some pyranometer data kept by the BADC through which I hope at least to be able to characterize the effect so that others don't have to go through the "is it me?, is it my equipment? is it the site?" process I did. I might even be able, with evidence, to inform "support" at Davis that reflections are not the only possible answer or cause!

Cheers

Onepk

PS. it appears the Davis sun sensor MIGHT indeed archive average values, but only those taken on the minute time stamp. For a one minute archive interval, you still get the average.... but only of one reading!