Liz, Greg and I are tracking a pair of storms northeast of GLD. The lead storm is a supercell while the trailing storm is more obviously multicellular consisting of ordinary cells. We’re raising many questions about the probability warning guidance issue.
Questions include:
- Liz asks how does she maintain some kind of connection with her user? Greg counters that the warning meteorologist should be separated from the user. Afterall there are many types of users, each with their own thresholds for warning. There will still be a need for some kind of binary product that the lowest common dissemination method (e.g., NOAA wx radio) can still transmit. Liz is not comfortable with losing the connection with the user. Greg says someone will likely be filling this role.
- Can we set a threshold probability in the prob warning tool for which when passed, the tool flags the swath as an official binary warning?
- Can a forecaster set his/her own threshold warning probability?
- Liz wonders if we can save each forecaster’s warning thresholds so that we can see how each one thinks? My 50% is your 30%, for example. This process may shed light on forecaster differences (biases) in the assessment of probabilities.
- Can metadata be added to each threat area ID so that any input not easily translated to grids can be added? Greg suggested that it could be a blog for each threat area. We’re thinking that metadata could include special call-to-action statements, forecaster reasoning, adding context to the event and others.
- Liz wonders how we could translate any of this metadata to something that is consistent with text in current warnings.
Greg would like to have the individual threat area translucent overlays (H,W,T) to be separated out so that we can toggle each threat area type on and off.
Jim and Liz would like product legends to be simplified. Hard to get time and elevation cut/height to stand out.
Jim would like to be able to change from irregular polygon to ellipse in mid process.
Jim LaDue (EWP Weekly Coordinator-in-Training)