The Atmospheric Radiation Measurement (ARM) Data Center is responsible for the timely collection, archival, and curation of science data products. These products are freely available through an online data repository. Metadata creation is paramount for scientific users to find and access over seven petabytes of atmospheric science data. The hierarchical metadata structure allows users to search for information at both broad and narrow levels. This project aims to leverage 30 years’ worth of manually created metadata to enable machine predictions of broad-term classifications from narrow-term descriptions. These classification predictions would assist metadata coordinators with their term selections. This paper discusses the cleaning and preprocessing of the training data, the pipeline developed to determine the best model for this task, and the creation of an API metadata classifier for ARM measurement metadata. Our results show that the Linear Support Vector Classification (LinearSVC) algorithm, along with the Term Frequency – Inverse Document Frequency (TF-IDF) vectorizer, is well-suited for our multi-class classification task. Lengthier input training data led to better results, and artificial balancing was unnecessary for this particular use case. This predictive classifier enhances efficiency in metadata creation, as well as supports greater consistency and accuracy in metadata tagging.

We introduce the Atmospheric Radiation Measurement (ARM) Data Workbench, a web-based interface that combines access to atmospheric datasets and computational resources. It is developed to facilitate a interactive data exploration and data analysis for the research community. The ARM Data Workbench offers a comprehensive solution that combines data locality, user-friendly access, and collaborative capabilities within a single platform. To empower researchers with a familiar analysis and compute environment, the data workbench integrates a JupyterHub which is a popular platform for scientific computing. In this paper, we discuss the features of ARM Data Workbench through which users can access the ARM hosted scientific datasets, order them and get them delivered. We discuss different types of user accounts we provide in order to have access to ARM data and the compute platform to perform analysis. We explain the process of navigating to ARM’s Data Discovery interface to access datasets and JupyterHub platform for scientific users with an example case study. The platform targets new researchers, graduate students and established researchers to help them in their research in a collaborative environment.