model release times.

[timeless]

lm to understand that they are released every 4 hours? and pages like the storm forecast/will it thunder pages are every 6 hours...

 

was just wondering if there was any possibility of some sort of timer basically telling us when the next run will be released.

 

on a completely unrelated subject the storm forecast map seems pretty confusing, the will it thunder/storm forecasts sometimes give higher numbers than the storm forecast map seems to suggest.. a good example would be that currently its suggesting a 42% chance (earlier it was 52%) yet when looking at the storm forecast map at most it suggests 20%, that said saturday is just as strange for suggesting 32% for the afternoon and if lm reading it right there seems even less possibility there.

[Summer Sun]

All times can be found here - https://forum.netweather.tv/topic/72242-model-output-update-times/

 

[Paul]

The models update every 6 hours, and the charts/forecasts follow that schedule. The storm forecasts on the site use our NMM models, whereas the storm risk maps you're referring to are from the GFS, so they will be different (the NMM is higher resolution).

[timeless]

thanks for the clarification, however it does still confuse me why two different variables are used.. to be more specific the storm forecast model seems to use a somewhat similar way to gather storm potential into percentages, and since you use a similar system to work out possibility elsewhere it kinda throws me off that you wouldnt use the higher res model for the storm forecast..

 

that said l was kind of confused a few days ago at the high possibility given to possible storms despite there being no convective cloud being forecast (ld always assumed you need convective cloud for storms to happen).

[Paul]

We do use the higher res nmm for the storm forecast (the model doesn't create separate convective cloud in the first 48 hours, it's bundled together within non-convective cloud), the storm risk maps use the GFS which is lower res

[timeless]

so l guess given the above information, l kinda wonder how valid the GFS are. dont get me wrong its been interesting learning a little. but the difference between the GFS and NMM going by how wildly the numbers differ seems to make me wonder the point of the GFS given that from what lve seen over the past few weeks that there have been times where potential has been high on the areas you have used NMM where as the GFS has always showed low probability.. its really added to my confusion (which l must admit is somewhat clearer now l know two different models are used, but the fluctuations between the models still weirds me out considering the huge differences in potential some days).

[Paul]

The NMM will pick out more local detail, the GFS is a broader brush, but is still a useful tool. 

[timeless]

just seen the update notice on the GFS page, are you starting to use NMM models on the new updated version? reason l ask is the significant differences in whats forecast.

 

in fact lve just compared the results for friday and there are pretty huge variations in the storm forecast for friday between the GFS and the GFS-P viewer on both the cape and storm forecast model, for example on the GFS model page it doesnt show my area covered where as the GFS-P does have my area and a chunk of the west midlands covered, even the cape models seem very different.

[Paul]

The GFS-P is a new version of the GFS model run at NOAA so will have quite different output at times to the current GFS.

[timeless]

but the storm forecasts that the will it thunder page uses l assume is still NMM and l guess the GFS-P its kinda like NMM but gives a bit more resolution compared to the normal GFS.