Abstract

Environmental scientists spend a lot of time turning relatively simple scientific questions into complicated code, and much of that code is devoted to finding data, managing formats, and implementing familiar analytical operations rather than expressing the science itself. CubeDynamics is an open-source Python package that encapsulates these tasks in a grammar for environmental data analysis: environmental observations are source-qualified nouns, analytical operations are verbs, and their ordered composition provides the syntax of an executable scientific question. This allows scientists to build analyses quickly using concise, readable expressions while retaining access to the data, transformations, assumptions, and provenance underneath them. In this talk, I will use CubeDynamics to move from simple questions to analyses that combine multiple environmental datasets, showing both how the grammar speeds up analytical work and how it makes the resulting reasoning easier to inspect and communicate. I will close by exploring how this approach could support much larger environmental syntheses and AI-assisted analysis, where we increasingly need analytical systems that are fast for computers to execute but still easy for scientists to read and interrogate.

Speaker Bio

Ty Tuff is a Data Scientist with the Environmental Data Science Innovation & Impact Lab (ESIIL) and Earth Lab, in the Cooperative Institute for Research in Environmental Sciences (CIRES) at the University of Colorado Boulder. He is trained as a theoretical ecologist, with a particular interest in how relative motion shapes large, spatially expansive, and complex natural systems. Today, his work combines that perspective with environmental data science and cloud computing to tackle questions that require bringing together large and diverse environmental datasets. At ESIIL, he also serves as the Working Group Lead and Technical Lead, helping interdisciplinary teams turn ambitious environmental questions into collaborative, computationally tractable research. Much of his current work focuses on developing new tools and approaches for environmental synthesis and using advances in cloud computing and AI to help scientists better understand complex environmental systems and inform decision making.