Cumulus Support

I am just a beginner at this and I am noticing some odd things with rainfall.

I use Cumulus 1.8.7 build 819 from a WH1091 weather station with USB connection. I use the Fine Weather mode on Cumulus.

First: I added a YTD rainfall from the weather bureau a ferw days after I started using the programme. That was OK but it showed 9mm rain for this month afterwards. The YTD total was correct at ~400mm.

Second: Since then we had our first rain 0.3mm yesterday which transferred accurately from the WH1091. Today, according to the WH1091 we got 0.6mm rain in a short period of time, but the Cumulus shows today's rainfall as only 0.3mm. I waited for it to update the Cumulus data but it still hasn't done so a few hours later.

Any suggestions for a newbie?

My web display can be seen at http://www.kynetoncfa/greg/

Hi; I have a similar unit to yours, a WH1080.

I've had it about 3 weeks now and been trying to calibrate it for rainfall.

You will need to make alterations on the Configuration/calibrate section of Cumulus.
The station remote units seem to read quite low compared to a rain gauge.
If you start at say Rainfall 1.3 to 1.7 as a starting point you may get somewhere near to calibration.

The type of device used in the rainfall sensor is not that reliable or accurate. It doesn't pick up small amounts of rain that well.
One thing you must do it stop insects; mainly spiders getting inside the rain unit.
I made a tube from dressmakers net which covers the unit and is tightly fitted round the support arm.
A heavy metal nut placed in the top will keep the net in a concave fashion and help the rain collection.

It's best to try and calibrate the unit with a rain gauge; a tin can will do if you haven't got a proper one. But you need 10mm or more to be able to measure it with a mm rule and set the calibration on Culumbus; the more rain the better the calibration.

You will also have to calibrate most of the other sensors as well, especially the Humidity; set that to 99% in heavy rain with the offset and/or multiplier. Mine is set to -1.000 offset, 1.10 multiplier). The temperature can be set as well if you have some very accurate thermometers.

Hope that helps a little.

Mike.

P.S. Your link to your site doesn't work; can get to the main site though.

Something I meant to reply to in your original thread !

One thing you must do it stop insects; mainly spiders getting inside the rain unit.
I made a tube from dressmakers net..... But you need 10mm or more to be able to measure it with a mm rule and set the calibration on Culumbus; the more rain the better the calibration.

All units (with the posible exception of Ultrasonic kit) have a certain amount of hysterisis to overcome before 'collection' can start.

i.e. the droplets have to overcome their viscosity in order to join / flow into the collector.
So there will always be a shortfall / delay on 'start up'.

By introducing a mesh, you have increased both the initial amount that will stick to the mesh and it's reluctance to collect into the bucket, the finer the mesh the worse it gets, also there will be a greater propensity to 'splatter', thus losing some rain - triple whammy !

Thanks for your comments. I am sure I need to calibrate, but the immediate issue is that that WH1090 iteself is recognising 0.6mm of rain, but only uploading 0.3mm to the PC (Cumulus programme). I am thinking it may be related to teh start & finish of the respective 24 hour periods. I'll see what it looks like after 0900h tomorrow when the new day starts in Cumulus.

Greg

PS I think I put an extra space in the URL before greg - should be www.kynetoncfa.com/greg/

vetgmc wrote:First: I added a YTD rainfall from the weather bureau a ferw days after I started using the programme. That was OK but it showed 9mm rain for this month afterwards. The YTD total was correct at ~400mm.

The year-to-date figure is just added to the yearly total. I don't know how it could work any other way?

Second: Since then we had our first rain 0.3mm yesterday which transferred accurately from the WH1091. Today, according to the WH1091 we got 0.6mm rain in a short period of time, but the Cumulus shows today's rainfall as only 0.3mm. I waited for it to update the Cumulus data but it still hasn't done so a few hours later.

Could you attach your Oct09log.txt file, please?

vetgmc wrote:I am thinking it may be related to teh start & finish of the respective 24 hour periods. I'll see what it looks like after 0900h tomorrow when the new day starts in Cumulus

Ah, yes. If you're using a 9am-9am day in Cumulus, you can't compare the daily totals with the midnight-midnight figures from the console.

I tried to look at the graphs on your web site to see when the rain had actually fallen, but the graphs don't work. You don't seem to have created the images folder?

beteljuice wrote: By introducing a mesh, you have increased both the initial amount that will stick to the mesh and it's reluctance to collect into the bucket, the finer the mesh the worse it gets, also there will be a greater propensity to 'splatter', thus losing some rain - triple whammy !

Yes, I guess it's a 'no win' situation. However with no mesh how do you keep the wild life out?
You don't really want to go climbing on the roof ever few days to clean the collecting box out.
I've used a relatively course mesh which shouldn't stop too much water, but tulle (very fine) would be a problem.
As you say, small amounts of rain will always be a problem to measure but large amounts should be possible to calibrate to an acceptable degree. We've got a lot forecast for tomorrow so it will be a good day to calibrate here.

Mike.

... go climbing on the roof ever few days to clean the collecting box out.

It shouldn't be that high !

Plus it does need a surprising amount of maintence - wiping out dust / algae etc. from both the chute and the spoons. I always give mine a squirt of WD40 (sic.). It seems to reduce those build-ups and make the chute more water 'repellent'

beteljuice wrote:

... go climbing on the roof ever few days to clean the collecting box out.

It shouldn't be that high !

Plus it does need a surprising amount of maintenance - wiping out dust / algae etc. from both the chute and the spoons. I always give mine a squirt of WD40 (sic.). It seems to reduce those build-ups and make the chute more water 'repellent'

It's the only place that's convenient to put it where the rain won't get sheltered. Good idea the WD40. I'll try that next time I'm up there.

Mike.

Been doing some tests today with the Rain sensor.
It read approximately 40% high today during the windy morning.
It normally reads low in periods of low or no wind.
I made one quick mod this afternoon; wired a piece of plastic packing material under the box to stop the wind getting into the lower vents. I'll see if that makes any difference next time it rains. It does wobble a little in the wind but not much, so if that affects it I may have to re-site it as well; not easy.
Picture below shows pad underneath as well as the net sock over the whole unit.
Other picture shows where it is situated on the tower.
Mike.

.. wired a piece of plastic packing material under the box to stop the wind getting into the lower vents.

You mean you've blocked the drain holes !

You really must get your rain bucket and thermo/hygro unit lower down, else you will have to use some formulae (which are only approx.) to make your readings in any way comparable with 'published' figures given for 'standard' sensor heights.

eg. If you didn't have the rocking / updraft rain collector problem, then because of its exposure to wind at a greater elevation, It should be accurate when calm, but would drastically under report when there was a wind.

beteljuice wrote:

.. wired a piece of plastic packing material under the box to stop the wind getting into the lower vents.

You mean you've blocked the drain holes !

You really must get your rain bucket and thermo/hygro unit lower down, else you will have to use some formulae (which are only approx.) to make your readings in any way comparable with 'published' figures given for 'standard' sensor heights.

eg. If you didn't have the rocking / updraft rain collector problem, then because of its exposure to wind at a greater elevation, It should be accurate when calm, but would drastically under report when there was a wind.

I haven't blocked the drain holes; there's a small gap for the water to get out.

The unit is approximately 30ft up so the Pressure will be approximately 1mb less up there than the ground reading.

The thermometer is not far out but the hygrometer readings are difficult to calculate accurately anyway and I have got it fairly well calibrated with ground level.

I think the re-site job will have to wait till after the Winter when I'll get some extension leads and sort it properly.

My main need was wind direction and speed and that seems to be working fairly well although I haven't been able to calibrate the wind speed against anything else yet.

Mike.

Great picture. Our weather bureau recommends that rain guages be no more than 300mm from the gtround surface because the stronger winds up high cause erratic rain readings. Hand recommendation, gien the need for regular cleaning! I think there is a lot of connecting cable - it should be possible to site it a fair way from the hygro/thermometer & anemometer. I haven't sited my sensors permanently yet - so you're way ahead of me! Well done.

Greg

sort of unrelated, but this is the closest thread I can find, and in particular hopefully some WH1091 users can help.

I have noticed lately when I disconnect my weather station from my computer (WH1091), then reconnect it, cumulus seems to miss the period of data. I restart cumulus, and it says "loading data", but the graphs seems to have gaps in them like so:



that flat period is when the USB was disconnected, and it has resumed tracking now i've got it reconnected.

Do others encounter this gap? I dont want to keep the unit connected to the computer the whole day (which i did today)

Mine seems to catch up OK. I can't see any reason to unplug it even if you aren't running cumulus at the time. It might be worth changing the graphs by unticking the 3-D option & seeing what they look like then. Another idea might be to check how often the data are being logged.

Greg