ESPIn: A cybertraining summer school for earth surface processes modelers

Abstract:

Modeling is fundamental to predictive Earth Surface Processes (ESP) sciences; however, Earth science education in the U.S. does not usually equip students with skills to use modern cyberinfrastructure. Although teaching students how to build, test, and apply models is essential for preparing current and future ESP scientists, we found from course catalog review that curricula of more disciplinary-focused departments do not commonly include this component. Additionally, scientific software development has radically changed over the last decade. Open source code is now the recognized way of making science more transparent and reproducible. Team-based code development is more common, and individual scientists can make quick gains by using modeling toolboxes, such as those developed through the Community Surface Dynamics Modeling System. The next generations of scientists need to be trained to be able to do all of this adeptly.

To address this need for the ESP community to teach modern programming to develop innovative models that predict how the Earth’s surface responds to environmental change and human influences, we developed the “Earth Surface Processes Summer Institute” (ESPIn). ESPIn is an immersive cyberinfrastructure training for graduate students and early career scientists that addresses best programming practices, such as open source software development, advanced use of version control systems (using Github), and introduces participants to a suite of libraries and tools for dealing with geoscience models. Tutorials rely heavily on existing cyberinfrastructure, the Landlab toolkit and Python Modeling Tool, delivered through Jupyter notebooks for demonstrations and code sharing. ESPIn dedicates several days to work collaboratively on authentic research projects with small teams with peer mentors. Participants develop their own coding projects, and make use of existing modeling toolboxes and frameworks, with solutions being presented on the final course day.

Although ESPIn was originally designed to be an in-person program, it was reorganized as an 6-day online course. We will report on what worked and what did not work during 2020. We will also discuss results of quantitative evaluations of learning efficacy, which are administered to all participants through a pre- and post-course surveys.