Page 1 of 1
All time records - rainfall
Posted: Thu 19 Aug 2010 9:16 am
by serowe
Steve - I notice in the All Time Records there is an entry for All Time High Daily rainfall. The problem with this is that it asks for the Hour and Minute the event occurred - however, like the MOnthly Highest you can't realitically put a time ag ainst it.
Probably the same with the All Time Hourly rain - you can't realistically put minutes against that figure.
Re: All time records - rainfall
Posted: Thu 19 Aug 2010 9:53 am
by steve
serowe wrote:Steve - I notice in the All Time Records there is an entry for All Time High Daily rainfall. The problem with this is that it asks for the Hour and Minute the event occurred - however, like the MOnthly Highest you can't realitically put a time ag ainst it.
It's actually the time the last rain tip occurred on that day. Possibly interesting, probably unnecessary. The web tag for that just gives the date.
Probably the same with the All Time Hourly rain - you can't realistically put minutes against that figure.
The hourly rain isn't aligned to an hour on the clock so it's valid to have minutes in the timestamp.
Re: All time records - rainfall
Posted: Thu 19 Aug 2010 10:08 am
by serowe
Yes I noticed the webtag doesn't give - it was just that when I was updating the all time records I noticed it.
Hourly rain - is this something that is calculated then within Cumulus? If so, how do you calculate the hour? for example, if it rains from, say, 0410 to 0459, nothing for another 10 mins but then absoluely buckets down from 0509 to 0545 (or some other crazy figures) - how is the figure calculated? From the last rain recorded; from the highest amount etc? It's not something I could see easily being defined.
Re: All time records - rainfall
Posted: Thu 19 Aug 2010 10:21 am
by steve
serowe wrote:Hourly rain - is this something that is calculated then within Cumulus?
Yes - some weather stations supply the figure, others don't, so to keep things simple and consistent Cumulus always calculates it.
If so, how do you calculate the hour?
It's any 60-minute period. So hourly rain is the amount of rain that fell in the previous hour (or any given 60-minute period). I don't see how it can be anything else?
Re: All time records - rainfall
Posted: Thu 19 Aug 2010 10:38 am
by serowe
It's any 60-minute period. So hourly rain is the amount of rain that fell in the previous hour (or any given 60-minute period). I don't see how it can be anything else?
I understand that - but what 60 minute period si what I was trying to ask. From 00 to 59 minutes past the hour; 30 to 29 (ie a fixed period) or from minute nn to nn+60 whilst it is raining?
The example I gave before would be more like in a tropical area or summer thunderstorm area - heavy rain for 15-30 minutes, a clear period and then rain again.
Or is it a moving 60 minutes - whatever 60 min period gets the heaviest rain is what is recorded/labelled?
Re: All time records - rainfall
Posted: Thu 19 Aug 2010 10:59 am
by steve
serowe wrote:Or is it a moving 60 minutes
Yes. I can't really think of any other way of expressing it that I haven't used already. The hourly rain at any point in time is the amount of rain that fell in the previous 60 minutes. The all time record is the highest amount that fell in any 60-minute period, and the timestamp for the record is the end of the that 60-minute period.
Of course, if that rain actually fell over a period of only 30 minutes, say, then there will be more than one 60-minute period with that same record amount. As with any other record, Cumulus uses the first occurrence - this is simply a natural consequence of the way the code works, constantly checking whether the current data beats the existing record.
Re: All time records - rainfall
Posted: Thu 19 Aug 2010 11:12 am
by serowe
OK that's no problem - it was the concept of how you were obtaining the 60 minute period that I was trying to get clear in my mind (thus the comment about a 'moving 60 minutes')
I can work with this - I just needed a starting point so that I can process all of my old data (which doesn't give this 60 minute data) to come up with reasonably, reliable matching data from the original minute by minute recordings I have.