Amount Awarded: $20,000
We propose to host a summer school to teach modern methods for the modeling and analysis of glaciology data, combining statistical and machine learning techniques with physical, first-principles methods. Students will learn computational skills for open, collaborative and reproducible science, using a cloud-hosted platform tuned to the needs of this scientific community. Remote sensing has revolutionized the way we study glaciers, allowing regional to global scale analyses from vast datasets, but traditional exploratory analysis methods are insufficient to analyze the complex interactions present in glaciers at this scale. Machine Learning can help bridge this gap in understanding. However, most glaciologists are not formally trained in these statistical methods. We propose to organize a series of annual summer schools led collaboratively between glaciologists, physicists, and statisticians at UiB, UiO, and UC Berkeley, that will train the next generation of glaciologists in the usage of ML methods for physical analysis. These events will foster interdisciplinary collaboration among the scientists and PhD students in UiB, UiO, UC Berkeley, and beyond.