I'm closely comparing my InstroMet and Blake-Larsen sun recorders, especially through this period of long days of sunshine, both of which work in different ways.

The Thresholds for the Blakes-Larsen are in some way closely based on a Cambell-Stokes device, which burns a trace using a magnifying glass onto a peice of paper.

Currently I am finding that the Blakes-Larsen misses the first 30 to 45 minutes of sun at dawn and similarily at dusk. As I have never seen a Cambell-Stokes recorder in operation I was wondering on a clear sunny day how long after sun rise (or before sunset) is it before it starts burning a trace.

The location of my sun recording instruments means in the morning it sees the sun within about 10 minutes of sun rise.

An example of the differences from the last 4 days are as follows:

18th
BL 12.71 hours
Instro 13.42

19th
BL 12.53 hours
Many thanks
Intro 13.51

20th
BL 10.03 hours
Instro 11.57

21st
BL 10.29 hours
Insro 11.76

The 18th and 19th had pretty much unbroken sunshine, so are the best comparisons for this exersise.

Keith (Southend)

On Sunday, April 22, 2018 at 12:45:08 PM UTC+1, Keith Harris wrote:
I'm closely comparing my InstroMet and Blake-Larsen sun recorders, especially through this period of long days of sunshine, both of which work in different ways.

The Thresholds for the Blakes-Larsen are in some way closely based on a Cambell-Stokes device, which burns a trace using a magnifying glass onto a peice of paper.

Currently I am finding that the Blakes-Larsen misses the first 30 to 45 minutes of sun at dawn and similarily at dusk. As I have never seen a Cambell-Stokes recorder in operation I was wondering on a clear sunny day how long after sun rise (or before sunset) is it before it starts burning a trace.

The location of my sun recording instruments means in the morning it sees the sun within about 10 minutes of sun rise.

An example of the differences from the last 4 days are as follows:

18th
BL 12.71 hours
Instro 13.42

19th
BL 12.53 hours
Many thanks
Intro 13.51

20th
BL 10.03 hours
Instro 11.57

21st
BL 10.29 hours
Insro 11.76

The 18th and 19th had pretty much unbroken sunshine, so are the best comparisons for this exersise.

Keith (Southend)

I keep on eye on the BL recorder at nearby Bosullow Keith, and agree that it does appear to miss some early & late sunshine. As with you reflected in the last 3 days, http://www.sunrecorder.net/googlemap...serial=101 01

However, there was a parallel trial at Camborne where it was very close to the official recorder overall, so there are probably swings & roundabouts.

Last year it managed to record 221 hours in April, way short of it this year. Though we did well down here for sun in March, relative to the rest of the UK. (We did even better for rain!)

Cloudy so far today, but sun not far away now http://www.scillyman.co.uk/Lowertown_Cam.html

Graham
Penzance

On 22/04/2018 12:45, Keith Harris wrote:
I'm closely comparing my InstroMet and Blake-Larsen sun recorders, especially through this period of long days of sunshine, both of which work in different ways.

Keith, I am finding that the BL under-records by about 15%. It seems to
have difficulty registering sunshine if conditions are slightly hazy as
the thresholds appears to be set too high even though there are
well-defined shadows. I have contacted Ole about this but he seems to
think the recorder is working fine.

A classic example was the other day when there were thin contrails and
these were enough to cause the reading to drop below the threshold even
though the sun was still shining strongly with very distinct shadows.

My R&D recorder, like yours turns on about 40 minutes or so before the
BL at dawn. At dusk the BL turns off about 20 minutes before the R&D.
The only way the two recorders nearly record the same is when the sky is
exceptionally pristine with no haze at all, i.e., extremely deep blue
all the way down to the horizon - and that doesn't happen that often.

Here are my sunshine data for the last few days noting that the skies
were cloudless on the 18th, 19th & 20th:
R&D BL
18th 13.1 11.3
19th 13.2 11.6
20th 13.1 10.9
21st 7.4 5.4

Another issue that I have with the BL recorder is that it is totally
reliant on the program running all the time. There have been occasions
when the program crashes and doesn't restart. All the data during the
downtime does not get recorded. It really could do with a data logger
and until it does I don't see it as being a serious contender to the
present sunshine recorder line-up.

--
Nick Gardner
Otter Valley, Devon
20 m amsl
http://www.ottervalleyweather.me.uk

On Sunday, 22 April 2018 17:00:00 UTC+1, Nick Gardner wrote:
On 22/04/2018 12:45, Keith Harris wrote:
I'm closely comparing my InstroMet and Blake-Larsen sun recorders, especially through this period of long days of sunshine, both of which work in different ways.

Keith, I am finding that the BL under-records by about 15%. It seems to
have difficulty registering sunshine if conditions are slightly hazy as
the thresholds appears to be set too high even though there are
well-defined shadows. I have contacted Ole about this but he seems to
think the recorder is working fine.

A classic example was the other day when there were thin contrails and
these were enough to cause the reading to drop below the threshold even
though the sun was still shining strongly with very distinct shadows.

My R&D recorder, like yours turns on about 40 minutes or so before the
BL at dawn. At dusk the BL turns off about 20 minutes before the R&D.
The only way the two recorders nearly record the same is when the sky is
exceptionally pristine with no haze at all, i.e., extremely deep blue
all the way down to the horizon - and that doesn't happen that often.

Here are my sunshine data for the last few days noting that the skies
were cloudless on the 18th, 19th & 20th:
R&D BL
18th 13.1 11.3
19th 13.2 11.6
20th 13.1 10.9
21st 7.4 5.4

Another issue that I have with the BL recorder is that it is totally
reliant on the program running all the time. There have been occasions
when the program crashes and doesn't restart. All the data during the
downtime does not get recorded. It really could do with a data logger
and until it does I don't see it as being a serious contender to the
present sunshine recorder line-up.

--
Nick Gardner
Otter Valley, Devon
20 m amsl
http://www.ottervalleyweather.me.uk

Hi Nick,

I've come to the conclusion you can't compare the two systems like for like.. However, I do feel the BL is more accurate, but may conflict with what the human eye regards as full sunshine and is basing it's functionality on solar radiation, not shadow. I'm sure Ole won't mind me quoting his reply as I think it sums up our delemma:

snip
The human interpretation of what sun shine is to a certain extent is also quite subjective – how deep is the shadow casted before we deemed is as sun shine?

I have discussed the WMO definition of 120W/m2, which is very old and also to a great extent on the Cambell-Stoke performance.

When observing a very nice clear sun rise, our human mind surely says – yes – sun is shining, but the 120W/m2 is never reached just after sun rise but at something like 5-7° above the horizon (as far as I remember).

So the big issue is if sun shine is based on the light or on the radiation.
snip

I know what you mean with regards if the programme crashes, mine has been quite reliable lately, it does still record if the internet goes down, but of cours you won't see the 1/2 hourly updates online. The *new* SMS and email 'alarms' and daily & weekly summaries are good. Of course the only time you can do any maintanance with the software is out of sun day hours.

Some Months ago I also got the data logger for my R&D InstroMet sun recorder, but the data logged does not agree with the data on the display, it is always a lot less. Adrian was going to look at this using an old device, but as yet not had an answer nor a solution.

Eg:
Today Display 11:25, logger output 11:15.
18th Display 13:42, logger output 13:25.

Regards

Keith (Southend)

On 22/04/2018 18:30, Keith Harris wrote: I've come to the conclusion
you can't compare the two systems like for like. However, I do feel the
BL is more accurate,
Hmmmm, at the moment I am far from convinced that the BL recorder is
'more' accurate. I find it hard to believe that when the sun is shining
brightly and you can feel its heat, with its disk clearly visible and
there is a distinct, sharp edged shadow on the ground - that the sun is
NOT shining. The BL recorder can say that.

Also, and Ole agrees with me that the BL recorder can be fooled by
indirect light, i.e., lots of small cumulus clouds giving plenty of
reflected light can raise (as it does with my solar irradiation and UV
measurements) the level somewhat above the threshold and record sunlight
even though the sun itself is partly obscured and not casting a shadow.
I have noticed this on occasions and I contacted Ole to report it.

The Met Office might take some convincing as to the 'accuracy' of the BL
as I believe the KZ recorder uses the same method as the R&D, i.e., the
difference between sunlight and shadow.

I prefer to keep it simple, if there's a sharp-edged, distinct shadow
then then the sun is shining.

--
Nick Gardner
Otter Valley, Devon
20 m amsl
http://www.ottervalleyweather.me.uk

On Sunday, April 22, 2018 at 1:23:48 PM UTC+1, Graham Easterling wrote:
On Sunday, April 22, 2018 at 12:45:08 PM UTC+1, Keith Harris wrote:
I'm closely comparing my InstroMet and Blake-Larsen sun recorders, especially through this period of long days of sunshine, both of which work in different ways.

The Thresholds for the Blakes-Larsen are in some way closely based on a Cambell-Stokes device, which burns a trace using a magnifying glass onto a peice of paper.

Currently I am finding that the Blakes-Larsen misses the first 30 to 45 minutes of sun at dawn and similarily at dusk. As I have never seen a Cambell-Stokes recorder in operation I was wondering on a clear sunny day how long after sun rise (or before sunset) is it before it starts burning a trace.

The location of my sun recording instruments means in the morning it sees the sun within about 10 minutes of sun rise.

An example of the differences from the last 4 days are as follows:

18th
BL 12.71 hours
Instro 13.42

19th
BL 12.53 hours
Many thanks
Intro 13.51

20th
BL 10.03 hours
Instro 11.57

21st
BL 10.29 hours
Insro 11.76

The 18th and 19th had pretty much unbroken sunshine, so are the best comparisons for this exersise.

Keith (Southend)

I keep on eye on the BL recorder at nearby Bosullow Keith, and agree that it does appear to miss some early & late sunshine. As with you reflected in the last 3 days, http://www.sunrecorder.net/googlemap...serial=101 01

However, there was a parallel trial at Camborne where it was very close to the official recorder overall, so there are probably swings & roundabouts..

Last year it managed to record 221 hours in April, way short of it this year. Though we did well down here for sun in March, relative to the rest of the UK. (We did even better for rain!)

Cloudy so far today, but sun not far away now http://www.scillyman.co.uk/Lowertown_Cam.html

Graham
Penzance

I've got figures for the MetO site at Camborne & Bosullow, (on the moors 4 miles from Penzance) over the last 5 years. Over this period the Blake Larsen has recorded 2% more sunshine.

When Penzance had a sunshine recorder it typically recorded 10-15% more sunshine per annum than Camborne. Bosullow is certainly more cloudy than Penzance, but also sunnier than Camborne - which is far more affected by inland CU development & the sea breeze front. It is reasonable to assume that the Bosullow / Camborne difference is between 5% & 10%.

OK, there are a few assumptions, but all this suggests that the BLake Larsen is under-recording slightly (compared to Camborne) by perhaps 5%, but certainly a fair bit less than the 15% that Nick is experiencing.

Also, the Blake Larsen appears higher (relative to Camborne) in Spring & Summer. I'd put this down to more inland Cu development further up the County (which is certainly the case) but Nicks observation regarding reflection from white CU, is certainly a possible factor.

Graham
Penzance

"Nick Gardner" wrote in message
news
On 22/04/2018 18:30, Keith Harris wrote: The Met Office might take some
convincing as to the 'accuracy' of the BL as I believe the KZ recorder
uses the same method as the R&D, i.e., the difference between sunlight and
shadow.

I prefer to keep it simple, if there's a sharp-edged, distinct shadow then
then the sun is shining.

--
Nick Gardner
Otter Valley, Devon
20 m amsl
http://www.ottervalleyweather.me.uk


I don't think that the KZ rcorder uses the same method as the R&D, Nick. The
latter uses a shadow post to generate a difference in output from a ring of
photo-diodes. The KZ uses '3 photo-diodes with specially designed diffusers
the determine when a threshold of 120 W per sq m is exceeded'
See also:
measuringtheweather.com/wp-content/uploads/2012/06/An-overview-of-sunshine-recorders-July-2012.pdf
by Stephen Burt.


--
Bernard Burton

Satellite images and weather data for Wokingham at:
www.woksat.info/wwp.html



---
This email has been checked for viruses by Avast antivirus software.
https://www.avast.com/antivirus

On 23/04/2018 10:24, Nick Gardner wrote:

The BL uses a simple solar irradiation measurement coupled with a
quite clever software code to determine a sunshine/no sunshine
output.


Are you sure that's how BL works? I thought that the later design tries
to emulate the 120W/sqm direct irradiance measurement via some simple
but well thought-out optics. BICBW

I think the 'clever software' you're referring to is just the
traditional way used by a number of different software packages to try
and estimate sunshine hours from global irradiance by setting some
arbitrary threshold above which measured irradiance as a % of maximum
theoretical radiance is classified as bright sunshine. It's a little
intricate calculating the theoretical max for time and day of year,
lat/long, timezone etc and not forgetting the equation of time but it's
basically just coding some well-defined trig. IIRC the only real gotcha
is remembering to allow for which quadrant the ATAN result ends up in.

"Bernard Burton"
I don't think that the KZ rcorder uses the same method as the R&D, Nick. The
latter uses a shadow post to generate a difference in output from a ring of
photo-diodes. The KZ uses '3 photo-diodes with specially designed diffusers
the determine when a threshold of 120 W per sq m is exceeded'
See also:
measuringtheweather.com/wp-content/uploads/2012/06/An-overview-of-sunshine-recorders-July-2012.pdf
by Stephen Burt.
Bernard Burton

Bernard, I've just read the link to Stephen B's article and he
says that the KZ works by having at least one of the three
photodiodes in the shade giving a voltage differential. This is a
similar principle to the R&D.

Stephen also writes in the book version of Measuring the Weather
that the KZ and the R&D are likely to give close results as they
use similar principles to determine whether the sun is shining or
not.

The BL uses a simple solar irradiation measurement coupled with a
quite clever software code to determine a sunshine/no sunshine
output.

--
Otter Valley, Devon
20 m AMSL
http://www.ottervalleyweather.me.uk

On Monday, 23 April 2018 11:22:20 UTC+1, John Dann wrote:
On 23/04/2018 10:24, Nick Gardner wrote:

The BL uses a simple solar irradiation measurement coupled with a
quite clever software code to determine a sunshine/no sunshine
output.


Are you sure that's how BL works? I thought that the later design tries
to emulate the 120W/sqm direct irradiance measurement via some simple
but well thought-out optics. BICBW

I think the 'clever software' you're referring to is just the
traditional way used by a number of different software packages to try
and estimate sunshine hours from global irradiance by setting some
arbitrary threshold above which measured irradiance as a % of maximum
theoretical radiance is classified as bright sunshine. It's a little
intricate calculating the theoretical max for time and day of year,
lat/long, timezone etc and not forgetting the equation of time but it's
basically just coding some well-defined trig. IIRC the only real gotcha
is remembering to allow for which quadrant the ATAN result ends up in.

http://www.sunrecorder.net/googlemap...serial=1010 9

I think you are correct John in what you say, as I quoted earlier in the thread. It was interesting to see that this morning, cooler and clearer, not as hazy, that my BL started recording about 15 minutes after mathematical sunrise, so it gave me some confidence that the Threshold(s), that run on a sliding ruler, are ok. The last few days have probably been hazier than I thought. So it is true what Ole said, and the algarithm seems correct based on the WMO definition of 120W/m2, which is very old and also to a great extent on the Cambell-Stoke performance.

Again, you can't really compare systems like for like, but the do what they say on the tin. Althoug I get some quite large day to day variabilities between the BL and the InstroMet systems, I tend to find the monthly totals are never that far apart.

Keith (Southend)