Author Topic: Making a Stevenson Screen (Read 23303 times)

einzack · « **on:** July 07, 2012, 05:15:22 PM »

A while ago I thought I'd have a go at making a poor-mans Stevenson's Screen.
Well, this morning I tried making one.
Currently my La Crosse 2355 sits in the permanent shade off the side of my wooden garage.
I know it's not ideal, but I think it works pretty good during the daytime, but reads too high overnight.

I found some instructions online (http://www.loganvillageweather.com/station/stevenson.html) which i based my first attempt off.
I bought a bunch of plastic 28cm pot plates and cut out the middle, drilled some holes in each of them and using some threaded bar, i made up a very simple and crude stevenson screen.
it took me a few hours to put together, and after putting it up sitting just over 2m off the ground and 1m off the side of the garage, plopped my ws2355 inside, and................it didn't work.

temps climbed rather quickly, with humidity dropping.
i left it in there for 40mins before deciding that it wasn't going to get any better, and then put the ws2355 back in it's usual spot.
temp then dropped back to where it was previously and stayed there.

So i'm wondering if anyone has any suggestions of how to improve my chances of saving anything from my hatchet job?
the colour of the pot plates were green. i'm going to buy some white spraypaint and give the thing a coat of paint.
other things I can think of that would have improved it were having bigger plates, with bigger space inside for the instrument.
but is there anything I could try using what I currently have and still get a decent result, or should I notch it down to experience and move on?

i've attached a picture of what my handywork looks like, and another showing what the temp did after i set it up - i moved it around 2.20pm, and put it back to it's usual spot at 3pm.
for reference, the NIWA site 500m away read 7.9°C at 3pm.

Derek · « **Reply #1 on:** July 07, 2012, 05:46:39 PM »

I made something similar for my old La Crosse temp/humidity sensor a few years back.

Anyway, that worked pretty well and in a comparison against my Davis it read about a degree or so higher when I tested them side by side.

So, to comment on your design, you've chosen green plates - white would have been your best choice (I found white bowls from the place you get a bargain) and I think you'll find that it will be better if you paint it white. I'd recommend a bit more spacing between layers to permit more air flow.
I had a top plate on mine (can't find the photos right now) which was actually a bigger diameter to provide shade to the stack below.
Will see if I can put some pics up for you shortly.

Tony · « **Reply #2 on:** July 07, 2012, 06:02:59 PM »

I've had this project booked marked on my "to do list" for a while now.

I was interested in your humidity reading drop off, I have long suspected my WS2355 is giving false high readings because of the lack of airflow around the sensor set up. How do your humidity readings compare with the NIWA site?

Tony

Derek · « **Reply #3 on:** July 07, 2012, 06:22:54 PM »

Here's a couple of pics of my one on the kitchen bench

It's made from dessert bowls with a dinner plate at the top to provide shade and an extra air space above the top bowl. The centres of the bowls are cut using a 100 mm diameter hole saw. Assembled using threaded rod with spacers and dome nuts.

One pic shows it sitting upright, the other what it looks like underneath (plus you get to see the design on the bowls

)

For this set up, the temp / humidity sensor was quite small profile and could be mounted on a wooden stake and this assembly fitted over the top.

Sent from my HTC Desire S using Tapatalk 2

einzack · « **Reply #4 on:** July 07, 2012, 06:46:17 PM »

wow derek, thats a mightily impressive contraption. must have weighed a tonne!

and tony, yes, my humidity readings are higher than the NIWA site. during the day approx 10% higher, but the same-ish overnight.
however, my intuition tells me (probably incorrectly) that if i was getting more airflow when using my stevenson screen then my temp readings would have been more better (as in lower). i took from the performance of my screen in terms of temp, that i was not getting a good airflow.

Derek · « **Reply #5 on:** July 07, 2012, 07:00:26 PM »

578 ±1 grams, to be reasonably exact....

If your screen (technically these are just radiation shields, not stevenson screens

) doesn't have a hole in its bottom pot plate, then I'd suggest you try it with one, and do your white spray painting. If it hasn't enough airflow and while it's still green, I think it's just likely to be doing the opposite of what you want... which is kind of what you found, isn't it?

einzack · « **Reply #6 on:** July 07, 2012, 07:10:31 PM »

mine has a hole in ever plate except the top one.
and as for having bigger gaps between the plates - i definitely think this would help, but i was concerned with making them too big and allowing the sun to get in at low angles - my plates don't have that great slow curve of your ones derek which would have worked a treat.

i think painting them white is a must. also, looking at the spaces between the plates and also to amount of space removed from inside the plate - i think mine was a bit small....
a few things to work on in the next few weeks.

Derek · « **Reply #7 on:** July 07, 2012, 07:22:43 PM »

Yeah, those plates had an ideal curve when I saw them, they were a couple of bucks each at The Warehouse (they are made from Melamine).
I think with the mods you're planning you'll get yours sorted - as an option, you're more than welcome to try mine, shouldn't cost much to post it down to you.... it's not like I'm using it....

TokWW · « **Reply #8 on:** July 13, 2012, 07:23:15 PM »

I just found this thread - so sorry for not posting earlier. I did something unconventional - I do have a La Crosse 2308 and i guess they all have much the same hygro unit. I just happened across a tall skinny plastic bucket in the Warehouse - most seem to spy cheap stuff in here...

All of a sudden, my thoughts saw it as an inverted bucket. I had seen a picture of another NZder DIY wx stn with a 1-Wire setup - more about that if you ask -

I layered some foil backed foam about 20mm thick in strips (I measured the diameters at top and bottom and made some triangle pieces with top cut off (trapezoids) of the right shape to glue onto the inside. I then cut a circle out of the top and got some (I hate the word "got") plumbing pieces to make a a pipe up and an elbow over and a pipe across and an elbow over and a pipe down which I could fit a PC power supply fan inside to pull air through all the time.

It works like a charm and has cured me of highs and flat temps - it is much more responsive than a Davis especially in calm conditions - night and heat of morning. I will try and find a photo of it all.

iomkiwi · « **Reply #9 on:** August 22, 2012, 09:34:33 PM »

Here are a couple of pics of my old FARS that I built for my LaCrosse.

The plates are from plant pot saucers - painted white.
The roof is actually two solar panels directly powering four PC fans - two small ones and two larger ones, when there was only a little sun the small fans would power up very quickly and as the solar intensity grew the two larger fans would also kick in.
The reflection off one of the fans can be seen in the second pic as the bright silver disc.
Fans need to be set so that they are 'pulling' air over the sensor so that there is no chance of them pushing heat from the fan itself onto the sensor.
As soon as I moved the sensor into the shield I found my graphs were a lot smoother - even though the sensor got moved from being fixed to a shaded south facing wall out into open space.

deepsouthweather · « **Reply #10 on:** August 22, 2012, 09:59:56 PM »

I too have experimented over the years with different idea's for making screens the best of them having the top significantly larger than the individual plates/slats below it (large enough to almost completely shade them) and this seems to take the direct solar effect out of the equation. Eventually i caved in and replaced my La Crosse station with a Davis VP2 which in my opinion is the best answer. I do have a daytime aspiration kit fitted which also seems to improve responsiveness and accuracy compared to not having it.

beteljuice · « **Reply #11 on:** August 22, 2012, 10:00:15 PM »

In accord with British MO requirements ....

OhauitiWeather · « **Reply #12 on:** August 22, 2012, 11:00:31 PM »

No one has yet answered the second part of the question I posted in the Shoutbox earlier. ". . . and why are they so deficient?"

"David commented in the Shout Box that it doesn't matter where you have located it, it's common knowledge that the radiation shields on those stations are inadequate." Why, I thought they measured Temperature not Radiation so can someone please explain? It seems to me that those that posted above had their stations located such that they were mainly in sheltered and sunny conditions, I notice for example that David's own station (As pictured on his website) is located on his roof with lots of other reflective stuff on or about the same pole. I am also left wondering how much these constructions affect the Min temperature, the insulation effect?

I firmly believe that the location of a Weather Station is critical, mine is purposely well out in the open, has plenty of ventilation and is high enough off the ground to not be shadowed by any buildings, trees or other obstructions etc. It is extremely rare for us not to have a breeze here, calm conditions are observed mostly at night or on cloudy/wet days so I fail to see why my station would not be adequately ventilated.

Based on the above posts from those who have built these shields all I can see that it did, apart from look butt ugly, was to flatten the graph, my guess is that it just averaged out the data creating a different and artificial modification to the data. If I put that kind of structure up my pole with the wind we get here it would be all accross the paddock after the first decent SW blow.

Trying to understand what all this is about as it does not seem that logical to me.

Oh and my High Temp from today of 22.6C is consistent with my manual temp gague that recorded a max today of 21.3C and is located in permanent shade behind the building, is less than 1 metre above ground level, and was fanned directly by todays predominantly SE/SW breeze. For comparative purposes the Max recorded Temp in August 2011 was 26.9C on 11/08/11 so 20C+ temps are quite achievable here in August.

Kind Regards.

beteljuice · « **Reply #13 on:** August 23, 2012, 02:35:28 AM »

Quote

Why, I thought they measured Temperature not Radiation

Don't you have radiators at home ?

The materials and design of any screen will absorb some radiation, this will materialize as heat which then may warm up the local environment either through direct transfer or convection.

Of course location is important (for comparison purposes), and strictly speaking the sensor should be 1.25m above short grass (or astro turf !) and either in an 'openly' ventilated sheild (white exterior, black interior) or within a concentric tube arrangement and fan assisted such that an air flow of > 3m/s is maintained.

As stated it is well known that the FO and clones sensor shield suffers in 'radiation' which may be direct or reflected. Reports suggest as much as 4 deg C difference on a 'normal' sunny day.

When the beteljuice had his station working, he accidentally copied the design of a *psychrometer and made one out stainless steel tube which proved to work incredibly well, and even when I tried (solar) fan aspiration it made typically less than a quarter degree C difference on a scorchingly hot day.

*Edit: The beteljuice originally erroniously said pyrometer.

David · « **Reply #14 on:** August 23, 2012, 07:36:23 AM »

Actually it is just the anemometer on the roof in the photos on my site, the rest is over a lawn down the back of our section

TokWW · « **Reply #15 on:** August 23, 2012, 07:54:58 AM »

My weather station - split up for better sitings of each component is here:

http://www.inmanavenue.com/sitephotos.html

Towards the bottom you can see the temperature and hygro unit and clicking on any photo will give a larger one.

My hygro has it's useless shield removed and I have about 0.5m/s air movement over it (2ft/s say) and feel this is a better breeze than on a hot summer's day or evening. My evening and dead calm responses are far more active than those without FARS - Fan Assisted Radiation Shield - and are better (more active) than those seen by a standard Davis VP2 unit and often 0.5deg lower on high temps than that device. Notably it was at least 2deg better than the original in a shaded upturned bucket without internal insulation which I used first at the same location.

einzack · « **Reply #16 on:** August 29, 2012, 08:44:34 PM »

Well i finally got round to setting up the radiation shield that derek very nicely sent me on Sunday.
it was a very nice and easy installation. just had to drill a couple holes in the top plate and it fitted very nicely on the contraption I had devised on my first attempt.

it's only been 3 full days so far, but it appears to be performing very well.
I had been tracking the performance of my station against the Reefton NIWA site over the last few months and these last 3 days have been much closer to the NIWA temps than I had previously been recording.
Average difference for daily maximums and mins are +0.3°C and +0.1°C respectively. this compares to +1.2°C and -0.3°C respectively for august up til the changeover.
I'll be keeping a close eye on it, but it's looking good.
humidity was another I was interested in, but I haven't had a chance to look at that yet.

again, many thanks to derek for sending me his handiwork.

News:

Author Topic: Making a Stevenson Screen (Read 23303 times)