ESRL scientists help lead summer school in Earth system modeling
For three days this summer, 25 students sat at computers at the National Center for Atmospheric Research (NCAR), running some of the most cutting-edge Earth system models around, models usually only accessible to experienced scientists. The students—from Colorado State University’s Center for Multiscale Modeling of Atmospheric Processes (CMMAP) and other universities—listened to lectures and initiated model runs on a supercomputer.
One enthusiastic student wrote on the workshop evaluation, “I liked getting hands-on experience with a supercomputing facility (a first for me.) I also liked meeting people in the field.”
ESRL scientists from the Global Systems Division (GSD), and colleagues from Colorado State University, NCAR, NOAA’s National Centers for Environmental Prediction, and NASA led this first Summer School on Atmospheric Modeling (SSAM). The goal of the collaboration between those partners and NOAA’s Global Interoperability Program (at ESRL) was to build the professional skills of undergraduate and graduate students interested in global modeling for weather forecasting.
“This is the first time we’ve been proactive about getting a new modeling system out there for graduate students to test and use,” said GSD Director Steven Koch. “We’re hoping to pull a new generation of young scientists into the world of global Earth system modeling.”
This summer focused on ESRL’s new global Flow-Following Finite-Volume Icosahedral Model (FIM). Students were required to have senior-undergraduate training in atmospheric science with a focus on large-scale dynamics. Familiarity with numerical methods for modeling was recommended.
The three-day summer school presented an overview of FIM, with detail about the model’s innovative dynamical core and physical parameterizations. Classes included discussions on computational performance and hands-on experience with FIM. The students learned about hybrid vertical coordinates, geodesic numerics, FIM diagnostics, computational issues, and vertical discretization as they evaluated the accuracy of summer and winter forecasts.
Students reported that they appreciated training on research-and-development systems, especially when their lecturers were the very scientists and engineers who developed those systems. “I liked most the interaction with the scientists who built the model and the lectures concerning the creation/working of the model,” wrote one student. They also liked the open interactions among modelers, including lively discussions about multiple approaches to solving some of the problems they deal with.
Tom Henderson, a GSD software engineer supporting the effort, said that he and his colleagues put in many hours over weeks to prepare content, graphics, and slides for the lectures; to coordinate sessions; to test models; and address technical problems.
“It was a pleasure to see the genuine interest of all these NOAA scientists in educating the next generation of global modelers,” wrote David Randall of CMMAP in a letter of appreciation to ESRL Director Sandy MacDonald. Randall also thanked the behind-the-scenes staff from NCAR for “superb support” of SSAM. “Not only did SSAM benefit the students tremendously as documented by their many favorable comments, but presenters benefitted as well,” Randall said. “SSAM helped raise the profile of FIM, bringing well-deserved credit to ESRL and its Global Systems Division.”
Next summer SSAM will be hosted at NOAA’s Geophysical Fluid Dynamics Laboratory, and the focus will be on the AM3 atmospheric model. Organizers will incorporate lessons learned during this student-modeling workshop, such as performing more dry runs for hands-on sessions.
But overall, Koch said, this first SSAM was a successful attempt to initiate young talent into atmospheric science.
—by Annie Reiser