Cumulus Support

Hi all

In the channel islands the pressure curve for the last few days has been just a bit excessive, starting from 999.3hPa on the 23rd @ 21:18 and rising to 1033.49 on the 25th @11:18 and falling today to 982.63 @ 09:00 and still falling. The following screen shots are attached for info. It will be interesting to see just how far it goes today.

We made it over the 45, with the maximum drop being 45.4 between 0330 Saturday and 0330 today, now levelled off. 31.0 of that was over 13 (1110 to 2310) though

Mapantz wrote: ↑Sun 27 Dec 2020 1:29 am That's a good record.

I got around 35hPa for my daily stat. It's a shame that the pressure started to fall later in the day, otherwise it would have been a greater drop.

Worth watching tomorrow night and in to Monday as well. This low pressure deepens again across Western Scotland and then starts moving Southeast. You may get close to 960hPa around 3am on Monday, Ray.

Pressure bottomed out at 04:44 with 972.7hPa. Exactly 24 hours earlier it was 1019.7. So that's 47hPa in 24 hours which is exceptional but not known if it is a record. I suppose if others can also measure their lowest and the reading 24 hours earlier we might get a useful comparison.

My December record is 964.3 back in Dec 2013. If your suggestion is right Mapantz my December record may go and conceivably my all-time of 963.1 back in Feb 2014.

RayProudfoot wrote: ↑Sun 27 Dec 2020 9:43 am Pressure bottomed out at 04:44 with 972.7hPa. Exactly 24 hours earlier it was 1019.7. So that's 47hPa in 24 hours

That's not the right definition to use - you need the 'biggest fall over any 24 hours', not the 'amount of fall in the 24h before lowest' (and let us put to one side that you don't yet know that the lowest point has been reached)

If you look at a pressure graph, mine is roughly sigmoid and by the lowest point the pressure is no longer falling. You should select the 24h period with the "most fall", ie including the most downward slope and the least of any 'gradual start' or 'gradual end' to the fall.

Good news: this will give a larger value and bring you closer to any record!

@Cortmalaw, your post comes across as pretty patronising. How exactly am I meant to determine that when all I have is the Mk 1 eyeball and looking at my stats on WU?

I haven't a clue what 'sigmoid' means. I'm not a professional meteorologist, just a keen amateur. And where is your station based? No location shown against your profile or signature.

How would you define the record...
a) The greatest change within any sliding 24 hour period window, (ie max value in that window minus min value in that window) or
b) The greatest change over a sliding 24 hour period (ie, fixed start/end times 24 hours apart).

I think the former?

Mark, I agree that (a) makes most scientific sense (the weather doesn't have a built in clock, it has no concept of calendar days, and pressure drop could easily be part-way-through across midnight).

Ray, I'd no intention of being patronising, and re-reading I think my post is helpful (as is shown by my final comment "Good news: this will give a larger value and bring you closer to any record!"). I'm amateur too (thanks for pointing out need to update my profile - got omitted when I needed to rejoin the forum following a gremlin).
Sigmoid just means S-shaped. You should be able to identify the 24h period with the most drop from the plots in WU, or from the graphs in the Cumulus local web interface, which even make it easy to read off the values (numbers appear when you mouse-over).

mcrossley wrote: ↑Sun 27 Dec 2020 1:42 pm How would you define the record...
a) The greatest change within any sliding 24 hour period window, or
b) The greatest change over a sliding 24 hour period (ie, fixed start/end times 24 hours apart).

I think the former?

I agree Mark, a sliding 24 hour window which allows the user to define the point at which pressure levels out and then examine the preceeding 24 hour period.

I'm attaching a Select-A-Graph screenshot from Cumulus 1. The end time was a couple of minutes after pressure levelled out and the start time is 24 hours earlier. I've rounded times on the graph by a few minutes for clarity. I fail to see how the change in pressure could be greater than this 24 hour period.
@Cortmalaw, sorry for coming down hard on you. I prefer to scan numeric tables rather than a graph but you're welcome to view my info on WU. My station ID is ISTOCKPO4. But I still think the 24 hour period shown in the graph above reflects the greatest change of pressure.

In the last 12 hours pressure here has only dropped around 4hPa.

RayProudfoot wrote: ↑Sun 27 Dec 2020 2:54 pm But I still think the 24 hour period shown in the graph above reflects the greatest change of pressure.

Ah - in your case, by chance, it does not make any difference. But that will rarely be the case.

My graph is the same shape, but spread out over nearly 3 days from peak to lowest, so choosing the 24h before lowest gives only part of the steepest downslope. Largest drop is between about 2200 on 25th and 2200 on 26th (rather than 1200 on 26th to 1200 on 27th)

Cortmalaw wrote: ↑Sun 27 Dec 2020 3:06 pm so choosing the 24h before lowest gives only part of the steepest downslope.

I get what you are saying, But Ray was looking for the large fall in 24 hrs (if I got his point well) and you are doing it in 3 days (72 hrs).

You have roughly a 70 hPa drop in 72 hrs which is 1 hPa per hr. Ray gets to > 2 hPa/hr. Your drop of 70 may be high because it is a continuous downfall (mostly pressure drops with some intermediate rises in between). So if there would be a webtag for the record fall in 24 hrs it is just a matter of measuring a sliding window. Sigmoid or sinusoidal or whatever curvature: just measure the begin and the end of the 24 hr window (one of Marks definitions) and you know the difference.

Like Hans I'm curious why Cortmalaw has shown a three day graph when we're all talking about the greatest pressure change in 24 hours.

Looking at his graph the greatest change appears to be between 06:00 on the 26th to 06:00 on the 27th. 1015 down to 973 = 42hPa.

Sorry if I've confused you.
I displayed a 3-day window precisely so that you could imagine or overlay or slide any 24-hr window of your choosing, so as to find the 24-hr period with the largest drop (which is what Ray did from my graph).

Compared with Ray's comment "greatest change appears to be between 06:00 on the 26th to 06:00 on the 27th. 1015 down to 973 = 42hPa", I see a slightly larger drop between 2200 on 25th and 2200 on 26th, which is what I described previously. This is the 24-hr period selected to cover the steepest part of the curve, rather than Ray's own graph (which by chance covered the whole of the descent within 24-hr) , and which he described as being the 24-hr before current lowest.

I don't think we're actually disagreeing here, just that we don't have a set of nomenclature that we use uniformly.

Cortmalaw wrote: ↑Mon 28 Dec 2020 12:32 am I don't think we're actually disagreeing here, just that we don't have a set of nomenclature that we use uniformly.

As non-native English speaker I know everything about non-uniform nomenclature.
But it's new to me that the same can be observed between Scots and English

Ahem... sorry for this.

Btw: I already observed years ago that any communication which is only written electronically (email, forum, twitter etc...) is sensitive to misunderstanding and diverge. This sentence in itself may even contribute to that so my advise would be not to respond. And for what it's worth: the same effect does not seem to occur when comms is written on paper and send by snail-mail.

Well, that broke my daily pressure change record...

And my top 15 greatest pressure changes. The problem using the Min and Max readings it doesn't capture the greatest change over a 24 hour period. I'm sure Steve must have thought about this but maybe too complex for the number of times it would be of interest. Microsoft Access still very useful when interrogating a txt file loaded into a table.