presstrendval webtag definition

[freddie]

I have been doing some work on my website involving the presstrendval webtag and have been puzzled by the values I have been seeing. I found a definition is the Wiki that matches the values I have been seeing:

The average rate of pressure change over the last three hours, calculated from (<#press> - <#RecentPressure h=3>) / 3

I was wondering why the value is made into an hourly average. The three-hour pressure change time period is significant in meteorological theory - it is the typical response time between a change in pressure gradient at a location and the resulting re-establishment of geostrophic balance at that location.

If the 3-hour value is used, it means that comparisons between PWS stations and NMS "professional" stations would be easier.

[mcrossley]

I don't know the history behind the tag - maybe our historian Paul does? - but MX quotes the rate of change (hPa/hr), you are saying the meteos just use the absolute change over 3 hours?

[freddie]

you are saying the meteos just use the absolute change over 3 hours?

Correct. The 3-hour change can be compared between locations (typically on a chart whereby the values can be contoured) and tied directly to the geostrophic equation, which enables reference to all sorts of fluid dynamics theory. The only region of the world where this value is relatively unimportant is the tropics where the diurnal component of pressure variation is an order of magnitude larger than the dynamic component (away from hurricane activity!) - so here a 24-hour difference is used.

[mcrossley]

It looks like it has always been defined like that in both Cumulus 1 and CMX.

Of course converting back to absolute is just a matter of multiplying by 3, but maybe we should have a dedicated tag for that.

[freddie]

I thought that may have been the case. If there was a dedicated tag then I would definitely make use of it.

[HansR]

I saw version b3130 has the tag #PressChangeLast3Hours. But the change rate is also in realtime.txt
So what will be shown in realtime.txt? shouldn't that be an option then?