WXSIM users?

[skyewright]

How many WXSIM users do we have here now? Quite a few I think!

I've recently joined the club, and now have forecasts uploading 3 times a day (but not yet added to the menu, so you need to know that it is at
http://www.isleofskyeweather.org.uk/24elgol/wxsimforecast.php)

Amazing piece of software...

[Budgie]

I also joined quite recently, just prior to the v12 update. 
It's been on the site for a few weeks now and seems quite accurate, well it got today spot-on from about 4 days ago: chucking it down!! 

There seems quite a difference in the rain forecast between yours & mine from Sunday night onwards though: http://www.hebrides-photos.com/weather/forecast.php
It could be down to the times that I'm running the forecasts, 04:40, 12:40 & 20:40. I run them after half past so the webcam movies have already been produced as they also use a fare bit of CPU memory. 

[carseman]

I think that that news of an impending v12 release prompted a lot of us to take the plunge 
I had iniital problems with my ISP's firewall blockng active ftp to nws.noaa.gov and after sorting that, the buoy data went down for a few days. All is well now and the auto forecasts appear to be fairly accurate.
My forecast is here: http://www.carseweather.co.uk/wxfore.php

Chris

[skyewright]

I also joined quite recently, just prior to the v12 update. 

Me too.

There seems quite a difference in the rain forecast between yours & mine from Sunday night onwards though: http://www.hebrides-photos.com/weather/forecast.php
It could be down to the times that I'm running the forecasts, 04:40, 12:40 & 20:40.

09:40, 13:40, 20:40 here.

The difference may be down to Tom's site specific customisations.

I've noticed on SCWN that we often seem to get more rain than you. I just took a quick look at your year to date too, that's less than half ours - we passed the 1m mark sometime in July.

[skyewright]

All is well now and the auto forecasts appear to be fairly accurate.

Yes. Mine had Thursday down as likely to be a nice day (which it was) well before the MO (as well as getting the weekend right, as Martin mentioned).
I'm very impressed. 

[Budgie]

I've noticed on SCWN that we often seem to get more rain than you. I just took a quick look at your year to date too, that's less than half ours - we passed the 1m mark sometime in July.

I really need to move my rain gauge as it does get shelter from the West round to NNW so it tends to read on the low side. Problem is, it's waayyyyy down the "to do" list at the moment. 

[munrobaggins]

I have been running WxSim for a couple of years now.  Initially, it didn't have any of the auto features and each forecast had to be done manually.  Now that the auto side of things has been implemented so successfully by Tom and the efforts of WD forum members, it can generally be left to "do its own thing" for weeks on end.  Obviously it is also possible to tweak the fcsts on a daily basis, but I now leave mine well alone, as the fcsts, surprisingly, are generally better than the beeb/MO anyway.  Sometimes, though the wxsim fcst has mysterious abberations, but these usually disappear during the next cycle of downloaded data. 

Mine is all setup for scheduled fcsts at 0700, 1300 and 1900.  WxSimate uses 17 mins past the hr downloads and WxSim does the consequent fcsts at 35 mins past.  My web fcsts auto upload (using system scheduler) at the 50 plus minute mark...I tend to do a lot of deconfliction on my FTP uploads etc so as not to overload my poor old low spec Wx PC.

Getting a suitable webpage fcst and graphic was a long process and it must have taken the best part of a year before the PHP gurus etc managed to sort it and I could understand/get it to work...hey ho..  Even now there are many methods of doing this including the good old WD method of display that Tokkiwi still appears to use.

I'm sure that you'll all have plenty of fun with WxSim and spot even more desirable improvements...I would like the output graph to show the rain as a blocked in green column rather than the line that makes it to the final graph product...strangely, this blocked in method is that which is used on the WxSim fcst graph at initial showing. 

[skyewright]

...I would like the output graph to show the rain as a blocked in green column rather than the line that makes it to the final graph product...

You may have noticed that on the Weather-Watch WXSIM board I queried the possibility of 'total precipitation' in 'Comparison to Actuals'. Tom thought it sounded like an interesting idea, so maybe it may appear someday...

Maybe rain is of greater interest here in Scotland than elsewhere.  LOL

[skyewright]

There seems quite a difference in the rain forecast between yours & mine from Sunday night onwards though: http://www.hebrides-photos.com/weather/forecast.php

As well as 'we probably get more rain than you (due to topography?)', the apparent rainy bias might also have been influenced by the fact that I just discovered that my plaintext-parser-data.txt was an old copy (the copy that came with the rest of the php files, rather than the  latest one specifically from the page about using the WXSIM php forecast).

I discovered that because it seems odd that a 10% chance of rain was being described as "Rain likely"! Hunting for the text "likely" led me to the answer...

Updating that file (and plaintext-parser.php, but the difference there was trivial), has changed "Rain likely" to "Chance rain". The newer plaintext-parser-data.txt also has an extra line in it and suddenly the forecast icons on my page show a lot less of the 'cars in rain on a (slight incongruous around here) multi-lane highway'. 

Maybe there was an 'off by one' problem without that extra line?

[Budgie]

That could have been it, the temp & wind are much closer now.