Nuclear Science Seminar Speaker Schedule Sign Up

Current Schedule

Speaker = Prof. Zach Meisel of Air Force Institute of Technology

Host = Hendrik Schatz schatz@frib.msu.edu

When = 2025-02-05 Wednesday 15:30 (note the new time!)

Where = FRIB 1300

Please feel free to sign up below, but note that some slots may need to be re-assigned to accommodate those with a specific need to meet the speaker. If this happens, the Admin or Speaker Host will reach out to you via email.

Title = As simple as possible but no simpler: Modeling for cosmic and terrestrial applications

Abstract

Modeling is an essential component of understanding the physical world. However, the closer one looks, the more details there are to model. Selection of the modeling approach comes down to the principle “as simple as possible, but no simpler”. In this presentation I will show the application of this principle to problems in nuclear astrophysics, homeland security, and radiological emergencies.

Two-Sentence Advertisement: This seminar will attempt to tie together some commonalities I've experienced between modeling in industry and academia. One goal is to show FRIB students how their training is preparing them for a diverse set of career paths.

Relevant background with commentary from Prof. Meisel:

• A thought-provoking discussion on modeling phenomena, what we're trying to capture with a model, and whether or not there's one best approach is Chapter 1 of McElreath's Statistical Rethinking, titled The Golem of Prague. It has my favorite quote summarizing modeling, where the model is the “golem” and accurate results are “Prague”: “Make no mistake: You will wreck Prague eventually. But if you are a good golem engineer, at least you'll notice the destruction.”
• The first third of the seminar will discuss advantages and disadvantages of varying levels of modeling complexity, using X-ray bursts as an example. Hendrik's 2008 classic Physics Today article nicely summarizes the main physics questions.
• The second third will provide an example where the space to model is huge and so the simplest possible models are required. The example itself is with regards to homeland security, so I'll have to use analogies (like the game Clue). There's a nice article that captures the same general idea when performing physics-based models of the natural disaster risk phase-space.
• The final third will provide an example where the models in use are demonstrably too-simple and need higher fidelity. This is the example of modeling radiation health effects for animals. There's a nice article demonstrating the problem for fish.
• Extension of the nuclear mass surface for neutron-rich isotopes of argon through iron (MSU PhD 2015)