There was no IOP today.  The forecasters spent the day doing training and archive case playback on the various systems.  This summary will contain our notes from the Gridded Warning archive case playback.

Comments and thoughts from our forecasters:

Knobology of WDSSII getting in the way of science – not comfortable with tool. Perhaps in 3-4 more days, we could be more ready. Consider a 2 week tour with a one-week overlap between tours.

Would be nice to issue hail warnings on Ref and MESH in one pane, and tornado warnings on Ref and Vel on another pane, another monitor, or another workstation.

There was some discussion about the use of low-probabilities for “pre-warnings”. Actually, that is part of the NSSL concept of gridded threat areas – to provide some downstream users who need lower-probability longer lead time information between the watch and warnings time and space scales.

Paul started by issuing threat areas in “swath mode” as is done today in the WFOs, but quickly realized this and adjusted his threat areas accordingly (current threat only).

Paul would like to issue different probs for different parts of the threat area.

Paul and Dave had a hard time with motion vector because it is not shown on the screen. With warngen, there is an arrow and past and future positions.

Most of these technology issues go away if this is fully integrated into D2D.

Definitely helps to have someone who knows the system sitting next to the forecaster.

Dave comments that perhaps the 2nd-week forecasters in the above scenario are the ones issuing the warning, while the first weekers in in training mode – perhaps not working together, so that the “veterans” can really concentrate on the science.

Paul and Dave are concerned about the precision of the prob values – what about every 10%?

Andy – how do we calibrate our probabilities? Perhaps we can integratee the verification into the live system?

Andy – Verification system should be included in the NGWT from the get-go – lesson learned from WRH experience with GFE (see white paper).

Greg Stumpf (EWP Weekly Coordinator, 28 April – 2 May)

Our first day of operations with our visiting forecasters was a moderate success. The orientation briefing went about 30 minutes long, so we didn’t start WDSSII training until 330pm. This will probably be the norm. Once that was complete, we begin training and briefing on the gridded warning experiment. That didn’t end until about 530pm when we morphed into me demonstrating the gridded warning software on the tail end of the event in NE North Carolina and SE Virginia. We did not have a chance to issue warnings on the major tornadoes that occurred today as they happened before we were ready, but there we did issue a few low-probability tornado warnings on the remaining mesocyclones before they went offshore. Our “IOP” lasted only about 90 minutes and then we had a dinner break. Afterwards, each forecaster was given hands-on training with the gridded warning software, and we ended the day with a 30 minutes discussion, captured nicely by Kristin on the live blog. I also made a few “live blog” entries during our “IOP” that I’ll need to clean up.

No pictures to show today since our “IOP” really wasn’t a true IOP (I operated the software most of the time), and we won’t be using these data in any post-analysis. But we may play back a little of the data at the debriefing on Tuesday.

Greg Stumpf (EWP Weekly Coordinator, 28 April – 2 May)

From 3-5pm, I trained Liz Quoetone, Jim LaDue, and Pail Schlatter on WDSSII. At the end of the training session, I tested a bug fix for the gridded warning software, and now, none of the warnings are automagically disappearing any more! Just a few more small issues, and the software should be ready to go.

Starting around 545pm, we began our gridded warning IOP centered on several storm which developed in NW KS. I spent the first hour explaining the software to Liz and Jim. Then, Jim took over the driver’s seat, followed by Liz, and then we wrapped up at 830pm. Several severe storms developed in the area, including one supercell that was a very large hail producer, and possibly a tornado producer. We issue gridded warnings for both hail and tornado on that storm.

During the event, we decided to try a new thing – live blogging. Jim and Liz put their thoughts on the blog as the event was occurring. I’ve decided that this would be a great idea for the gridded warning experiment. The cognizant scientist and/or one of the forecaster/evaluators should blog live during the IOP to share their thoughts in real time. We’ll try to put these live blog on a separate page in the EWP Blog.

Friday is the first of our end-of-week debriefings at 10am. We really don’t have a firm plan for how we are going to conduct these just yet, so we will probably spend some time discussing how we are going to conduct these weekly debriefings in the following 6 weeks.

Greg Stumpf (EWP Weekly Coordinator, 21-25 April)