160 Overview of National Blend of Models Version 3.1. Part II: Bulk Verification and High Impact Examples

Monday, 8 January 2018
Exhibit Hall 3 (ACC) (Austin, Texas)
Jeffrey P. Craven, NOAA/NWS, Silver Spring, MD; and T. L. Huntemann, D. W. Plumb Jr., and D. E. Rudack

The National Blend of Models (NBM) version 3.0 (NBM V3.0) was placed into operations in July 2017 at the National Centers for Environmental Prediction (NCEP). This gridded guidance has five domains including the CONUS, Alaska, Hawaii, Puerto Rico, and a large Oceanic Domain that covers much of the Pacific and Atlantic Oceans. The NBM V3.0 currently runs hourly (24 times per day) on the Weather and Climate Operational Supercomputing System (WCOSS), harnessing over a dozen different inputs including a variety of high resolution models, global models, and statistically post-processed guidance.

There are a number of significant enhancements and capabilities that will be described in the presentation. The NBM V3.0 (with exception of the Oceanic Domain) is issued hourly using a post-processing run cycle rather than traditional model cycle notation, essentially running each hour with the latest available suite of deterministic, ensemble, and statistically post-processed guidance. The number of model inputs roughly tripled to 12-15 for the four CONUS and OCONUS sectors, and the number of ensemble members for Oceanic winds doubled to over 40 inputs.

In July of 2018, the next upgrade will be NBM V3.1. Part 1 describes the capabilities and upgrades of NBM V3.1, which will focus on adding service program gaps within the aviation, marine, and fire weather areas. Part 2 is designed to provide bulk verification statistics on a variety of elements to date in 2017, and to document any improvements/degradations between NBM V3.1 (parallel) and the operational NBM V3.0. Several high-impact weather examples are presented to show the utility of the NBM during heavy precipitation events, winter weather, strong wind events, and severe weather.

While providing the greatest possible lead time on high impact events is critical, some decision makers require a level of precision on the timing of weather events of around one hour. The provision of hourly time steps through 36 hours was designed to provide more information to assist forecasters in the Enhanced Short Term Forecasts in the first 1-2 days. This presentation is intended to demonstrate that one hour precision on the timing of weather events is skillful.

- Indicates paper has been withdrawn from meeting
- Indicates an Award Winner