New rain rate calculation

[steve]

When did you start Cumulus running?

Steve

[Samuel]

I looked at wrong value on the webpage it was "22,5 mm/hr at 09:36 on 26 november 2008"

[Samuel]

I started Cumulus this morning and it has run all day, but when I found this strange value I restarted Cumulus but it has same value when it started again.
Now I change in records to the old record, and today.ini to a more normal value.

[steve]

OK, thanks. There's clearly some other problem that I haven't spotted yet.

I think I will start again. I think I'll go for something much simpler, for now, which is: Take the rainfall in the last 5 minutes, multiply that by 12, and call that the current rate (i.e. per hour). So, if 2 mm fell in the last 5 minutes, that's a rate of 24 mm/hr.This is like the old 'rain in last hour' method, but finer grained. It also corresponds more closely to the calculation using the history data (it's exactly the same if you have your station set to log every 5 minutes).

Steve

[dc1500]

Fine with me, it's not exactly a figure that's widely used in standard weather record keeping.
Dave

[steve]

Fine with me, it's not exactly a figure that's widely used in standard weather record keeping.
Dave

That's pretty much what I was thinking, Dave. The station doesn't bother to calculate it, Easyweather doesn't, there don't appear to be any official guidelines on calculating rain rate with a TBR, so the way I see it, anything that gives a reasonable figure is probably good enough.

Thanks,
Steve

[steve]

OK, I think I have something. I'll upload it tomorrow.

Steve

[Super-T]

By my thinking, "Rain Rate" means the rate at which the rain is falling at the time. If no rain then no rain rate. That would mean that you would have to decide if there was any rain currently and work the rain rate from there.
I can see all sorts of difficulties there but logically, you can't have a rain rate if there is no rain at the time. It's like saying that a car is going at 50kph and that is it's "Rate". When it stops then it's rate is 0.
Just my thoughts

[steve]

Can't argue with any of that. Now, if only our weather stations had sensors which told us whether it was currently raining, and better still, at what rate.

Steve

[Super-T]

Lets see.....the way I understand it is that the rain gauge works by a tilting bucket arrangement? I must have a look inside sometime.
If the bucket tilts then it is raining. When it hasn't tilted for a set time then it has stopped raining, or evaporation is faster than the rainfall? The time between tilts gives the rate and can be measured over a set time. If it rains for more than an hour then it would be fifo. I guess this is something like you are already doing? I don't know enough about how long it takes to tilt, or the size of the bucket, to guess how long you should be looking for a no tilt or indeed the next tilt.
Different models with different parameters must be a nightmare to figure out.
My logic is probably all wrong.

[steve]

My logic is probably all wrong.

No, you have it exactly right. This is the algorithm the Davis stations use internally, and the algorithm I tried to implement initially. It's a nightmare because of the asynchronous nature of Cumulus reading the data every 10 seconds (or whatever), and the station updating every 48 seconds, independently. The code to work out the time between tips is horrendous.

Yes, it's do-able - I've been designing and writing software professionally for nearly 30 years, and I've solved much harder things than this in my time. But it's so tedious having to write the code, give it to you to test, wait for it to rain, discover something I hadn't thought of, repeat... And all the time I'm thinking, is it really worth the effort for rain rate? Wouldn't something not quite so fancy do instead? So that's what I'm going for

We can revisit rainfall rate in Cumulus 2. It's probably much easier in C#

Steve

[Super-T]

Rain rate doesn't really matter to me either....if I stand outside and get wet then it's raining Just trying to analyse the problem.
My programming is fairly limited and I would never class myself as a programmer, more a problem solver. Most programmes have been on robotics and PLC's but I should really get my teeth into something a bit more modern.
Maybe I should make a hardware mod for the wh-1081

[steve]

Maybe I should make a hardware mod for the wh-1081

What would be nice would be if someone could write some new firmware for it

Steve

[TrOjAn]

I have a rain detector outside attached to my home automation system, its there so if the wife has washing on the line and it rains then we are told over the house speakers its raining. Saved my neck a few times

IF? I went to the xAP side of things then MAYBE it could be used to let cumulus know its actually raining.. would get complicated though, thats for sure.

TrOjAn

[steve]

Excellent

I assume it works by the rain causing conductivity? And you can adjust the sensitivity?

Steve