Development of COMET web-based professional development training on numerical weather prediction for NWS field forecasters: Stephen Jascourt, COMET Postdoctoral Fellow (National Centers for Environmental Prediction)

"Development of COMET web-based professional development training on numerical weather prediction for NWS field forecasters": Stephen Jascourt, COMET Postdoctoral Fellow (National Centers for Environmental Prediction)

Final Report

As Numerical Weather Prediction (NWP) modeling systems become increasingly sophisticated and accurate, forecaster reliance upon them is increasing. However, changes in model physical parameterizations, data assimilation, resolution improvement, and local model use are producing more intricate and complex error characteristics for forecasters to grapple with. Also, forecasters are faced with model output depicting increasingly realistic-looking structures, which may verify well under some conditions and poorly under other conditions.

Training programs are needed to provide forecasters with an understanding of how NWP modeling systems work and the capabilities and limitations of the current operational model suite, updating this information as the models are changed. This training has the goal of improving forecasters' ability to wisely utilize NWP models in the variety of particular weather scenarios they face each day. COMET is serving this training need through development of an internet-based Professional Development Series (PDS) containing Professional Competency Units (PCU) on operationally significant aspects of how NWP models work in general ("PCU1"), operationally significant details of specific models ("PCU2"), and application of this knowledge in assessing model initial conditions and using model output ("PCU3/4"). Further details can be found at http://meted.ucar.edu/nwp/.

The NWP PDS project is a collaborative effort involving COMET meteorologists, COMET instructional, graphics, and computer staff, help and input from NCEP, and so on. Instead of reports and journal articles, PCU1 and PCU2 products are the Web training modules, which can be found on the PDS matrix page http://meted.ucar.edu/nwp/pcu2/index.htm. Note that the left column has links to PCU1 modules while the table entries form PCU2. In addition, a teletraining session has been developed based on the PCU1 material, with further details at http://www.cira.colostate.edu/ramm/visit/nwptop10.html and a path through the PCU1 material has been arranged into an online course at http://meted.ucar.edu/nwp/course/. PCU3/4 consist of online case studies pointing out specific points of how certain aspects of the models affected a particular forecast. The first of these can be found at http://meted.ucar.edu/nwp/pcu3/cases/301200/.

My role in the NWP PDS team has included

Serving as a subject matter expert.
Scoping out module content and determining content priorities and organization for use by operational forecasters within the limited training windows they have available.
Writing and reviewing text.
Preparing and reviewing conceptual graphics and data plots.
Obtaining information/details about NCEP models and transferring that information in a presentation understandable and relevant to a typical operational forecaster.

Accomplishments

Completed PCU1. I was primary author for the modules on data assimilation, convective parameterization, and "Why Parameterize", and contributed substantially to the team efforts on many of the others.
Completed the PCU2 portions on convective parameterization for Eta and AVN/MRF and Eta Introduction. Updated the PCU2 "Model Structure and Dynamics" Eta model sections (previously prepared by Dr. Robert Rozumalski) when the Eta model was changed.
Completed preparations for, including generation of some new graphics, written compilation of "talking points", and creation of the Visitview lesson, and began delivery of Visitview teletraining sessions. This work was performed in collaboration with COMET meteorologist Dr. Bill Bua with much Visitview help from the Visit Team. Information on Visitview teletraining is at http://www.cira.colostate.edu/ramm/visit/visithome.asp and details for this particular lesson are at http://www.cira.colostate.edu/ramm/visit/nwptop10.html.
Completed, with Bill Bua, the first PCU3/4 case study, online at http://meted.ucar.edu/nwp/pcu3/cases/301200/.
The COMET PDS team accomplishments have also been described in several AMS conference preprints (lead author Rich Cianflone) and presented at several AMS conferences and NWS workshops by Rich Cianflone and Wendy Abshire.
Participate in NCEP "synergy" meetings, including collaborating with HCP SOO Pete Manousos and COMET meteorologist Bill Bua on the agenda, facilitation, and meeting notes. These meetings promote communication between all the co-located NCEP centers (EMC, HPC, MPC, CPC, NCO) so that model development, computer implementation, and operational model needs and use may be coordinated and potential problems identified and addressed. Participating in this has provided an excellent big-picture perspective on the integrated picture of NWP model development and implementation and allowed me to keep abreast of upcoming changes and issues of concern.
Attend EMC mesoscale branch meetings, providing more detailed information on Eta model issues and projects under development for future model implementations
In the course of examining model code and model output, I have identified a few minor model bugs and inconsistencies and performed some new verification/assessment statistics for QPF from the convective parameterizations in the Eta and AVN/MRF.
Answered field questions about the models.

Ongoing and Future Work (as COMET Project Scientist)

Continue teletraining delivery (collaboration with Bill Bua). Fourteen sessions were scheduled for March 2001 and nine for April 2001, with an average of roughly five forecast offices and four people at each office participating.
Complete PCU2, in particular, the data assimilation parts for both Eta and AVN/MRF.
Update all PCU2 to reflect changes in the operational Eta model (major changes anticipated twice per year).
Develop PCU3/4 case studies, including
- Many short, near real-time vignettes identifying one or few key points of a very recent forecast.
- A few more in-depth cases.
Ensuring that the case studies hold together coherently and meet learning objectives for PCU3/4.
Cases should include variety of forecast scenarios across the country - different seasonal and geographic influences.
Exchange information between NCEP and the field through talks at NWS workshops (including the Western Region SOO Workshop May 15-17, 2001 in Reno).
Exchange information between NCEP and the field through an email list similar to the RUC forum.
Continue to participate in NCEP synergy meetings and EMC mesoscale branch meetings.
Consideration of further NWP training needs and methods (after PCU2 completed and PCU3/4 cases developed): where do we go from here?
Possibly (hopefully) becoming involved in Eta model experiments (including experimenting on workstation Eta running on my PC). Also, developing forecaster-friendly Web pages for the Eta experimental and parallel runs, promoting more field involvement in evaluating possible model changes and anticipating their forecast impact.