Forums (Forecast Builder)

Back

Issue Adjusting Short-Term Temps

RSS (Opens New Window)(Opens New Window)
MS
Michael T Stavish, modified 2 Years ago.

Issue Adjusting Short-Term Temps

Youngling Posts: 12 Join Date: 9/20/12 Recent Posts

Need help with a silly issue I can't figure out - If I want to modify 1st pd MaxT for example, I click the grid in the grid editor, set the time range to current highlight of the MaxT grid, and make an adjustment. Then, in Analyze/Adjust, when I "Derive Hourly T and RH from MaxMin," the edits to the hourly T and RH grids don't start for several hours in the future - in this morning's case the hourly edits by FB started at 1 or 2Z, but the current time was around 17z.

 

Odd thing is when this problem was brought to my attention, I opened a GFE in practice mode and tried to replicate it. I did the same thing - clicked the MaxT grid, made a substantial edit, set the time in FB as the width of the MaxT grid, then in Analyze/Adjust ran Derive Hourly T and RH from MaxMin. This time the edits to hourlies started at the top of the next hour. This is the desired result!

 

Is there a reason the operational GFE is not editing the hourlies starting at the current+1 hour as in practice mode?

 

Thanks for your help!

Mike

 

 

MS
Michael T Stavish, modified 2 Years ago.

RE: Issue Adjusting Short-Term Temps

Youngling Posts: 12 Join Date: 9/20/12 Recent Posts

Well I tested more this afternoon and the hunch is that this may be time-dependent on when Derive Hourly T and RH from MaxMin is run. When running this afternoon, practice mode or not modifying MinT or MaxT within the 1st 12 hours and deriving hourlies resulted in hourlies being modified starting with the next hour from current.

 

I think most forecasters are aware of this varied behavior and try to work around it, but if this is normal behavior, what's the rationale?

MS
Michael T Stavish, modified 2 Years ago.

RE: Issue Adjusting Short-Term Temps

Youngling Posts: 12 Join Date: 9/20/12 Recent Posts

Per David Church...

 

This has been a bug that we have been chasing for some time now, and I think we just discovered the root issue earlier this week. We are currently testing a fix, but given the nature of how seemingly random the issue can be, we're just trying to ensure the fix is adequate before pushing it out. 

 

Basically what I think is happening now is that something is running in GFE (tool/procedure/formatter) before ForecastBuilder is launched that is changing the environmental time zone. So when ForecastBuilder launches a time variable is set in local time instead of UTC time, which, depending on your time zone, likely pushes the start time out by 7 or 8 hours in WR. 

 

Until we get this fix pushed out, restarting the GFE session and running ForecastBuilder resets everything and gets it working again. This has been the somewhat painful workaround I've advised our forecasters of in the meantime. This also explains the Practice Mode behavior you see, since you're starting a fresh session likely without running whatever is causing the issue first. 

 

I haven't seen this happen here since testing the fix, so I'm encouraged that this should be resolved soon.