Forums (Forecast Builder)

Back

RE: Confusion about ProbIcePresent guidelines

JM
Joseph Moore, modified 8 Years ago.

Confusion about ProbIcePresent guidelines

Padawan Posts: 88 Join Date: 11/1/13 Recent Posts

In the latest FB webinar, it was instructed that ProbIcePresent should never under any circumstance be deleted. This seems contradictory to the training and my PoWT experience over the years, in that ProbIcePresent was mainly to be used when a loss of ice aloft was expected to impact precipitation type - i.e. it was fine to delete. I know from the training and watching some of the webinars that ProbIcePresent can now be included in more scenarios due to an improved smart init strategy for ProbIcePresent, but the instruction to never delete the grid seemed new to me. Is this a change or was I just misunderstanding the previous training?

On a related but separate topic, the Top Down Init GUI ("Step 3") instructs the user "For Simple Rain or Snow scenarios, press OK to skip this step." However, pressing OK doesn't delete the grids if they have already been loaded (or grids from a previous shift exist), so it seems like the instructions are incorrect. (I previously submitted a suggestion to maybe make the GUI more clear about what was happening in this step.) The GUI also states the make-up of the top-down grids (20% NAM12, 20% GFS20, 60% RAP, then 50/50 NAM/GFS), but I think this is outdated now that we have the higher resolution GEM-REG/GEM-NH data being utilized.

Finally, is there a good way to keep up on documentation as changes are rolled out? I've tried to watch the available webinars to stay up to date on the changes, but is there a "living release notes" or changelog geared towards forecasters available?

AJ
Andy Just, modified 8 Years ago.

RE: Confusion about ProbIcePresent guidelines (Answer)

Padawan Posts: 99 Join Date: 6/2/15 Recent Posts

Joseph,

  Not removing the top-down grids is "new."  Back in December we added the top-down grids to the Previous database, meaning that SuperBlend would include a small percentage of what you already had in its next run. Additionally, we were receiving feedback on struggles to collaborate when offices were removing top-down grids. Certainly like removing ProbIcePresent for example limits freezing rain unless the warm nose begins to exceed 2.5C. Therefore, in the latest webinar we wanted to communicate the need to keep these top-down grids in the forecast.  

  You are correct that the top-down GUI needs an update to include the % mention of the Canadian, and I'll also update it for the "Simple Rain/Snow Scenarios."  We have a new tech order coming out soon for testing, so you should see the changes within the next month. There's been some delay due to a number of factors:

  • Waiting for the next National SmartInit update
  • Changeover from the Software Collaboration Portal to VLAB
  • Various meetings coming up for critical members of the CRGMAT.

  As far as keeping up with what's changing, definitely the webinars are the #1 spot.  We, the CRGMAT, are hoping that the webinars are being distributed out to the forecasters to watch.  We also have a development tracker which lists various changes since the October 4th start and what's coming "down the pipe."

JM
Joseph Moore, modified 8 Years ago.

RE: Confusion about ProbIcePresent guidelines

Padawan Posts: 88 Join Date: 11/1/13 Recent Posts

Thanks for the clarification Andy - I wasn't aware of all of those changes, and that makes a difference in how we create our forecasts!

In my opinion I think the team must be publishing these changes to procedure and reasoning behind it in a central location so that all forecasters are kept up-to-date on all of these changes. As someone who is enthusiastic about FB/PoWT and has followed the project closely, even I feel like there's just so much going on and changing constantly that it can be difficult to know how to handle different weather scenarios.  The webinars are good, but I'm not sure if all were even recorded (is there a comprehensive list?), and some of these changes now contradict things like the FAQ here on Vlab and some of the initial documentation. The change you mention about the Previous being used in the top-down grids in the December update isn't even listed on the Development Tracker spreadsheet. I understand and appreciate the fast development cycle, but I'm just afraid that if I as a proactive user who searches for the latest info am struggling to keep up with all the changes, I can't imagine how those new to the whole PoWT/FB concept must feel. This leads to collaboration issues, too, since not all forecasters are caught up on all the latest changes.

Could the CRGMAT/FB group please consider creating a central, up-to-date source for documentation that includes information on both code that changed, reasoning for the changes, and impacts to how we create our forecast based on those changes? I suggest a "Living Release Notes" Google Doc or something of that nature that can be updated with each new release on changes.