RE: time periods in forecast builder

Forums (Forecast Builder)

RSS (Opens New Window)

Ruth Ebert, modified 9 Years ago.

time periods in forecast builder

Youngling Posts: 8 Join Date: 10/29/16 Recent Posts

I'm new to forecast builders (in WR) but am a little confused by the naming convention of the time periods in step one of the forecast builder. Among the list is tomorrow and day 2. However aren't these the same period? When I run forecast builder with day 2 included in the time selection it populates grids in day 3 in my GFE. This offset also occurs with all the subsequent days (day 3 is day 4, day 4 is day 5, etc). Has anyone else had a problem with this? It is kind of confusing.

Ruth Ann Ebert

WFO Glasgow MT

Andy Just, modified 9 Years ago.

RE: time periods in forecast builder

Youngling Posts: 89 Join Date: 6/2/15 Recent Posts

Ruth,

You are correct that there can be confusion with the ForecastBuilder time periods and those you might see in GFE. This is in my development agenda to have everyone standardize their time periods in GFE. During the initial ForecastBuilder development phase last winter, I had built ForecastBuilder to not use the time periods in GFE. There are a couple of reasons:

Time periods can differ from office to office
Time periods may not adjust to Standard Time / Daylight time switches (like they did here)

So a quick rundown of how the time periods are assigned. Today = Day 0, Tonight/Tomorrow = Day 1, Tomorrow Night / Day after Tomorrow = Day 2, Day 3, Day 4, etc.

Attached are some before and after pictures when modifying time periods here at ARX today to match those of ForecastBuilder. They're not perfectly in sync in terms of where the image was taken, but if you focus on say Nov 9 and 10th, you'll see the differences.

Before:

After:

Chris Geelhart, modified 8 Years ago.

RE: time periods in forecast builder

Youngling Posts: 8 Join Date: 6/22/13 Recent Posts

Hi Andy,

To be honest, I think that particular change would cause even more confusion. Forecasters are already used to the definition of "Day 1", "Day 2", etc. from other products such as SPC ounlooks. The existing periods in most peoples' GFE displays likely also conform to this terminology, like in your first time bar example. It also doesn't make much sense that "Day 3" is only 12 hours long, but "Day 4" is 24 hours long.

Chris G. @ ILX

Joseph Moore, modified 8 Years ago.

RE: time periods in forecast builder

Youngling Posts: 72 Join Date: 11/1/13 Recent Posts

I strongly agree with Chris - and in fact, "Day 1" is already defined by the NDFD directive as "Today/Tonight." http://www.nws.noaa.gov/directives/sym/pd01002001curr.pdf - see appendix B, page B-2.

So, I think it would be inconsistent with current practice and directly in conflict with the NDFD directive to define "Today/Tonight" as Day 0. Please consider using the official NWS definition of Day 1.

-Joe

On Wed, Nov 9, 2016 at 1:06 AM, VLab Notifications <VLab.Notifications@noaa.gov> wrote:

Hi Andy,

To be honest, I think that particular change would cause even more confusion. Forecasters are already used to the definition of "Day 1", "Day 2", etc. from other products such as SPC ounlooks. The existing periods in most peoples' GFE displays likely also conform to this terminology, like in your first time bar example. It also doesn't make much sense that "Day 3" is only 12 hours long, but "Day 4" is 24 hours long.

Chris G. @ ILX

--
Virtual Lab Message Boards https://vlab.noaa.gov/web/cr-soo/forecast-builder-forum/-/message_boards/view_message/1792180 VLab.Notifications@noaa.gov http://vlab.noaa.gov

Joseph J. Moore

Meteorologist | Social Media Program Leader | Open Source GIS Evangelist

NOAA/National Weather Service Duluth, MN

Andy Just, modified 8 Years ago.

RE: time periods in forecast builder

Youngling Posts: 89 Join Date: 6/2/15 Recent Posts

All,

Great points made here. One of the conflicting issues that was present with GFE time periods was Day 8, defined after 12Z. Time shifting to make Day 7 to encompass 00Z to 00Z, which matches the NDFD directive under UTC hour, translates backwards to adjusting Day 6, 5, 4, etc.

I do like the recommendation to matching the NDFD directive entirely. Unfortunately I don't think the NDFD directive matches that of SPC. SPC goes 12 to 12Z, the NDFD directive based on UTC hour goes 00Z to 00Z.

So I do think there needs to be an initiative to standardize time periods. I'm going to bring this up more with the CRGMAT with the information presented in this thread.

Joseph Moore, modified 8 Years ago.

RE: time periods in forecast builder

Youngling Posts: 72 Join Date: 11/1/13 Recent Posts

Hi Andy,

I was looking at the Diurnal Day definition, not UTC Day, in the table on page B-2. We usually label the days/nights in GFE by local time/diurnal day, not based on UTC day.

Andy Just, modified 8 Years ago.

RE: time periods in forecast builder

Youngling Posts: 89 Join Date: 6/2/15 Recent Posts

Joseph,

I see that as well and brought that part too the attention of the CRGMAT. In the end I think I may end up adjusting ForecastBuilder - but also ensuring that everyone's GFE time periods match that of the NDFD directive for Diurnal Day. To be perfectly frank I wish the directive didn't have both the Diurnal Day and UTC day in there - go with one or the other (even though I know how the UTC day is being referenced here).