Cumulus Support

Hi Jim,

yes as Steve says, my understanding is that exceptional is outside the parameters. I guess we could have used 'error' or similar. If we actually have it appearing at all I guess I'd investigate at the time, but we felt it will be unlikely.

Re the different forms of showers or showery, we used those as per the original wording, so to be honest, I don't know enough to quantify and therefore change them. Beteljuice may be able to help on that also. Our aim was purely to remove some of the old fashioned wording like 'mending' because it wasn't meaningful to us.

JimDantin wrote:"Showery" is noted in some places, "showers likely" in others -- is that the same thing? Could "Showery" be changed to "Occasional showers" or "Scattered showers", which would indicate less chance of precip than "showers likely"?

'Occasional Showers' and 'Scattered Showers' however, do not indicate the same things and are also completely different to 'showers' and 'showers likely'.

This mixup of terminology happens because of the prepondewrance of private and government (read unofficial and official) forecasting bodies. One is there to provide a service (the official services) the other to make money (the private or company body).

As an aside .. from a link in another thread, ... the google translation of a Davis weather condition:
Tasty temperature, occasionally a shower ... I like that

Exceptional weather is a PREFIX ... I keep typing this .... PREFIX to condition 1 or 26

If conditions fall OR rise beyond the 'normal' set guide, then you are nearing the Sun or building an ark !
If you leave the string blank "", then the output is the same as condition 1 or 26 (whichever is applicable)

The forecast, which has no satellite or other external input has no concept of scattered rain (a very localised condition). Simply no rain, chance of rain, some rain, rain ...

Off Topic I know......but could somebody post the original "strings.ini" text file on here? I've been doing some editing and have lost the original wording.
Thanks...

The original should be in your Cumulus installation folder, called samplestrings.ini

steve wrote:The original should be in your Cumulus installation folder, called samplestrings.ini

Yes, but I made a mistake and overwritten that file. Therefore I could use a copy of someones....

bnwrx wrote:Yes, but I made a mistake and overwritten that file. Therefore I could use a copy of someones....

Ah, right. Attached is the latest one from 1.9.2.

Here is my first attempt at "modernizing" the forecast strings. My intent is to present the forecast in a readable, non-technical way that gives the user a general idea of what to expect.

Comments and suggestions would be appreciated, especially with regards to the model's indication of the probability/frequency of something happening. Did I get the "mostly", "frequent", etc. modifiers correct?

A funny note -- the program has generated a forecast19 message this morning. There have actually been light showers and thunder rolling through for the past few hours here. Nice correlation! The NWS, however, is simply forecasting "Partly sunny"!

Please be nice!

[Forecast]
Forecast=Forecast
notavailable=Not available
exceptional=Conditions are outside of the forecast model's boundaries
forecast1=Dry, mostly clear, with little change expected
forecast2=Dry and clear
forecast3=Improving, becoming dry and clear
forecast4=Mostly clear, but with increasing clouds and periods of showers
forecast5=Mostly clear, with some showers
forecast6=Isolated showers becoming mostly clear
forecast7=Mostly clear, but with some showers early
forecast8=Mostly clear, but with occasional showers later
forecast9=Occasional showers early, improving later
forecast10=Occasional showers mixed with clear periods, with improving conditions expected
forecast11=Mostly clear, but with occasional showers likely
forecast12=Occasional showers or storms probable, clearing later
forecast13=Occasional showers or storms, probably clearing later
forecast14=Occasional showers mixed with clear periods
forecast15=Occasional showers, becoming less settled
forecast16=Varying conditions with some precipitation expected
forecast17=Occasional showers or storms possible, mixed with clear and dry periods
forecast18=Occasional showers or storms possible, changing to periods of rain later
forecast19=Occasional showers and storms possible, some periods of rain
forecast20=Frequent showers and storms possible
forecast21=Occasional showers or storms possible, conditions may worsen
forecast22=Rain at times, very unsettled conditions
forecast23=Frequent periods of rain
forecast24=Rain, with storms possible
forecast25=Frequent storms, may improve
forecast26=Frequent storms, significant precipitation expected

Thanks Steve..

exceptional=Conditions are outside of the forecast model's boundaries

Yeeeees - but the forecast (1 or 26) is still valid, It is more exceptional than you would expect for your location (range settings) eg, a tornado is approching fast or perhaps drought conditions.

beteljuice wrote:Yeeeees - but the forecast (1 or 26) is still valid, It is more exceptional than you would expect for your location

Oops, that may be my fault - I may have given Jim a misleading reply

OKaaaaay -- now I'm really confused! Since "exceptional" seems to be a multi-meaning term, please suggest something different. Could it be "Extreme conditions expected" or "Warning - extreme weather possible"?

Hi Jim,

Congratulations on updating your forecast text, it’s obviously a local based interpretation as storms are probably quite frequent in your region. You may also consider whether you will use precipitation (17, 26) or rain as your standard.

I have attached a table which may serve to clarify or confuse (beteljuice can confirm or counter my interpretation) your revamp of the forecast definitions.

It is colour coded to show if the forecast is associated with a rising, falling or steady pressure (or a combination). What I also noticed was that the “steady” only forecasts are rarely seen (or for very short periods of time). To view the historical trend of forecasts I included the Forecast number in my tweet message so I could see how it changed etc.
Conclusion : Within the cumulus.ini you can add a parameter FCPressureThreshold, the default I believe is 0.1 which means provided the pressure is within -0.1 to +0.1 the forecast will use the appropriate “Steady” forecast (I’m sure Steve will correct me if I’ve got this wrong). Based upon my observations this tolerance is quite rarely met for any reasonable length of time. I played around with mine and decided (for now) that 0.2 gives me a better fit and means I get the “steady” text forecast for a reasonable and appropriate time period, certainly something to consider/play with.

I hope this makes sense but don’t be afraid to ask (or shoot down).

JimDantin wrote:OKaaaaay -- now I'm really confused! Since "exceptional" seems to be a multi-meaning term, please suggest something different. Could it be "Extreme conditions expected" or "Warning - extreme weather possible"?

Either of those would be true, but remember that will appear in front of forecast 1 or forecast 26 and may become a bit 'lengthy' for your page format.


exceptional=Exceptionally
forecast1=Dry, mostly clear, with little change expected
....
forecast26=Frequent storms, significant precipitation expected

Would give:
Exceptionally Dry, mostly clear, with little change expected
OR
Exceptionally Frequent storms, significant precipitation expected

Does that make better sense ?

nking wrote:You may also consider whether you will use precipitation (17, 26) or rain as your standard.

I think the original uses "rain" rather than "precipitation", and so did Cumulus. But then came the complaints from people where the temperature was -20 and it was forecasting rain...