Open Science

At iHARP, we leverage open science practices to engage the broader scientific community in a scholarly journey to navigate multiple research artifacts (including models/algorithms, datasets, publications, presentations, and AI/ML benchmarks). By adhering to Findable, Accessible, Interoperable, Reusable (FAIR) principles, we envision furthering purposeful collaboration (thereby aligning with the HDR initiative of convergence) and reproducibility (to enhance outcomes), all geared towards increased scientific breakthroughs and discoveries.

Explore iHARP’s Repositories

Open Models Repository

Try out our accessible notebooks and algorithms on Github

Open Data Repository

Learn about our data sources and data preprocessing strategies on Zenodo

Open Polar Science Community

Learn about our network of polar science contributors on Ghub

Open Access

(Published Articles, Proceedings, Posters, Presentations)

Scholarworks@UMBC offers a starting point for accessing iHARP’s scholarly research publications. This is complementary to other partner institutions.

Preprints

Share your feedback to help us characterize the impact of our work at

ArXiv and EarthArXiv.

Open Science Spotlights

 

FAIR in Multi-disciplinary Spaces: Is the Data AI Ready, Shareable and Encourages Reproducibility?

Dr. Vandana Janeja (Director, iHARP), University of Maryland, Baltimore County

FAIR in ML, AI Readiness, & Reproducibility (FARR) Workshop, AGU Conference Center in Washington D.C. (October 9 – 10, 2024)

NSF HDR Machine Learning Challenge (A FAIR Perspective)

Dr. Philip Harris (Director, A3D3), Massachusetts Institute of Technology (HDR Ecosystem)

FAIR in ML, AI Readiness, & Reproducibility (FARR) Workshop, AGU Conference Center in Washington D.C. (October 9 – 10, 2024)

ML Reproducibility: Sources of Algorithmic, Implementation, and Observational Variability

Kevin Coakley is a Computational and Data Science Research Specialist at the San Diego Supercomputer Center and UC San Diego, focusing on AI reproducibility. Kevin holds a MAS in Architecture-based Enterprise Systems Engineering and Leadership from UC San Diego and is pursuing a PhD in Computer Science at the Norwegian University of Science and Technology. Kevin specializes in training and evaluating machine learning models for accuracy and reproducibility in applications like image recognition, time series prediction, and natural language processing.

Open Science Practices in Support of Arctic Research Applications

Dr. Anthony Arendt is a Senior Data Science Fellow and the Director of Community Engagement at the University of Washington’s eScience Institute. He has a background in Arctic and alpine glaciology, sea level change, and remote sensing. Currently, he leads the eScience Institute’s hackweek program for geospatial research applications. His team has hosted these participant-driven events during the past 8 years with the goal of fostering collaboration, training, and community software development.

Ghub: A science gateway for unifying ice sheet science and education

Dr. Sophie Goliber received her PhD at the University of Texas, Austin, focused on quantifying marine-terminating outlet glacier change in Greenland. She was partially supported as a NASA Earth System Science Fellow (NESSF). She also worked as a postdoc at the University of Buffalo with Dr. Sophie Nowicki on the GHub project.

Share or Perish: Open Science, Open Data and What it All Means to iHARP Researchers

Dr. Maria Esteva has a PhD in Information Science and is a Research Scientist at the University of Texas at Austin. Her research focuses on modeling large-scale data derived from diverse research methods and domains as interactive curation and publication pipelines and measuring the impact of data publications and reuse. Maria works at the Texas Advanced Computing Center where she is the lead data curator in DesignSafe (https://designsafe.org), an NSF-funded platform for management, analysis, and publication of natural hazards and social science data, Her work in the NSF GCR Community-Embedded Robotics: Understanding Sociotechnical Interactions with Long Term Autonomous Deployments involves building a convergent data model that allows transdisciplinary collaborative analysis of multiple datasets collected during the research project

Abstract: Often, scientific research outcomes from interdisciplinary efforts are not fully aligned with existing open science tools and platforms, as well as the participating communities associated with them. Even so, the constraints on these tools/platforms, such as data sizes, computing capabilities, support for DOI registration, utility cost, and long-term sustainability of platforms, to mention a few, further amplify the effective utilization of the tools/platforms in question. This research study entails conducting a comprehensive assessment of open science platforms through feature-based micro-mapping and macro-mapping processes to identify platforms that align with the needs for open knowledge sharing and collaboration for interdisciplinary research in polar science and artificial intelligence. More so, this study aims to define features associated with ‘openness’ at different stages in the research life cycle process, which will be measured against the adherence of open science tools/platforms to the – Findable, Accessible, Interoperable, and Reusable principles. The overall outcomes of this research project will serve as stepping stones for best practices in advancing open science efforts among interdisciplinary niche research communities.

Abstract: As part of the interdisciplinary work at the NSF HDR Institute for Harnessing Data and Model Revolution in the Polar Regions (iHARP), we have developed several research tools using AI for polar science challenges. However, to make these tools available to the scientific community and meet them where they are, we developed a preliminary open science pipeline on Ghub, which is a domain-specific access platform that facilitates developers to create, store, and share methods/models/tools particularly geared to the polar science community. In this study, we illustrate the stepping-stones aimed at enhancing accessibility and reproducibility within the broader scientific community, leveraging concepts which outline a workflow detailing the processes to facilitate shared data access and collaboration. This study builds upon the efforts to create awareness regarding the shared data use at iHARP involving the engagement of a broader community of expert stakeholders to develop streamlined process(es) for shared data use that will also provide support for complex datasets often birthed through multi-faceted research processes and teams.

The iHARP Data Catalog

Team: Lydia Fletcher & Julie Hammons

Start Date: Summer 2024

End Date: To Date

Description: This project focuses on establishing a collection of datasets relevant to polar science research conducted at iHARP.

An Evaluation Tool for Open Science Platforms

Team: Shamita Nandeeswaran, Josephine Namayanja, Lydia Fletcher & Vandana P. Janeja

Start Date: Fall 2024

End Date: Spring 2025

Description: This project focuses on multi-level framework to characterize open science platforms based on their features and functionalities such as computation, data access and storage, and facilitation of collaboration for researchers and broader communities of practice, thus highlighting each platform’s merits and gaps.

 

An open science guidance for iHARP is available here.