Exploring Earth and Planets with Large Databases

By learning scientific data analysis, undergraduate students can develop problem solving skills. In data analysis courses, large research databases can provide a new opportunity for active learning for students. At Arizona State University, we have developed undergraduate-level data analysis courses in Earth and planetary sciences. The main learning outcome of the courses is: students can build strategies and then write data analysis codes for solving complex scientific problems about chemical and physical processes of Earth and planets. To solve a variety of questions, students research background information, explore the large databases (such as PetDB, USGS earthquake, and exoplanet databases), develop codes for analyzing data in Jupyter, review the strength and weakness of their solutions, and communicate their results with other students. Some problems require students to combine a variety of data, such as earthquakes, volcano eruption records, and petrology. In order to support these courses, we have developed a library of learning modules in Jupyter notebook. The notebooks include Python codes for exploring databases, statistical analysis, visualization, and modeling. We found that code examples and manuals on search APIs provided by some databases are very useful for exploration of databases without being limited by web interfaces. Open-access scientific publications with data analysis codes included in open-source formats, such as Jupyter notebook, make immediate use of research results possible in such data analysis courses. Therefore, coordinated efforts among researchers, educators, and database developers can innovate undergraduate education in Earth and Planetary sciences. The collaboration would be also important considering the recent demands for new modalities of learning, such as remote, online, and hybrid forms of remote and in-person learning.