Zoom Logo

Ethan Gutmann's Personal Meeting Room - Shared screen with speaker view
Just curious, with the NBM Probabilistic grids, do you see all of those being adjusted toward the official forecast using CDFs like we currently do with Probilistic Snow? How can we do this with limited computing iin the short term? Where will we go with cloud computing and how soon can we make the appropriate shift of our resources?
Perhaps better for Dale to type?
Dale Durran
Norwegian weather service used to provide a green/yellow/red indicator of forecast confidence. This seems to have disappeared. Any idea about the experience of other weather services with this basic indicator of ensemble confidence?
Have you verified the NBM mean against NDFD? If it beats NDFD wouldn’t that increase forecaster’s willingness to adopt?
Justin Minder
Question: With regards to the Reeves et al. (2016) SBC method for doing P-Type. Is the plan for the NBM to run this scheme in a deterministic sense on each ensemble member, or to also consider uncertainty in the SBC diagnostic tool? In the R2016 paper, they run sensitivity tests by perturbing things like assumed ice-nucleation temperature and degree of riming. These factors may contribute importantly to spread.
How well does the NBM handle extreme events? Especially in complex terrain. I have been doing verification of NBM in three locations in coastal CA for two years and find it does very well for the first 60 hrs routinely but not so well for extreme events. As you mentioned you need higher resolution models but these have very poor skill scores as they can put the heavy precip in the wrong basins. So for something like the Heidke they suffer. Are not extreme events really what the NWS is focusing on?
Joseph Bellier
With NBM doing most of the job, do you still expect forecasters to manually modify the values based on their expertise?
Robert Capella
Thanks for the presentation Matt