Instructor:

• Alexander Michels, University of Illinois Urbana-Champaign
• Erick Li, University of Illinois Urbana-Champaign
• Yunfan Kang, University of Illinois Urbana-Champaign
• Furqan Baig, University of Illinois Urbana-Champaign

Time & Day: 9:00–10:30 am Central, Tuesday June 17, 2025   Room: 1a

This session offers an introductory walkthrough of the I-GUIDE Platform, showcasing its capabilities in enabling reproducible, data-intensive geospatial research. Attendees will explore platform components, user interface, example notebooks, and how it supports workflows across disciplines such as sustainability, hydrology, and environmental modeling.

Skills and Background: General interest in geospatial platforms. No technical background needed.

Requirements: Participants should bring their own laptops.

Instructor:

• Yunfan Kang, University of Illinois Urbana-Champaign

Time & Day: 10:50 am - 12:00 pm Central, Tuesday June 17, 2025   Room: 1b

This tutorial demonstrates how the I-GUIDE Platform integrates OpenSearch, Large Language Models (LLMs), and spatial search to create intuitive, multimodal search experiences for researchers and educators. Participants will gain hands-on experience deploying OpenSearch on JetStream2, implementing keyword search, semantic search, and spatial search capabilities, hosting the LLM framework Ollama, and building Retrieval-Augmented Generation (RAG) pipelines with AI agents—using the I-GUIDE Platform as a case study. Attendees will leave with actionable insights into leveraging CI resources for scalable, privacy-aware AI workflows and scalable knowledge discovery.

Skills and Background: Familiarity with basic Python programming, APIs, and geospatial data is recommended. No prior experience with LLMs or OpenSearch is required.

Requirements: Participants should bring their own laptops.

Instructor:

• Jibin Joseph, Purdue University
• Venkatesh Merwade, Purdue University

Time & Day: 9:00 am – 12:05 pm Central (with a break), Tuesday June 17, 2025   Room: 2

The hands-on tutorial aims to train participants in accessing and processing geospatial data for hydrologic applications using cyberinfrastructure, with a focus on FAIR (Findable, Accessible, Interoperable, and Reusable) data principles. The tutorial will include a hands-on interactive training session utilizing Python-based Jupyter Notebooks on CyberFaCES and I-GUIDE platforms. The training will focus on accessing high-resolution Digital Elevation Model (DEM) data from the USGS 3D Elevation Program (3DEP) hosted on Amazon Web Services (AWS) for any location within the CONUS, and on processing DEM data for hydrologic applications. Typically, hydrology researchers rely on GUI-based tools such as ArcGIS and QGIS for their geospatial operations. However, performing repetitive tasks on multiple areas of interest can be time-consuming and redundant. By adopting Python-based workflows, participants can streamline these operations and efficiently handle repetitive tasks, leading to increased productivity and minimized redundancy. The tutorial will follow a structured format to ensure an engaging learning experience for all participants. Specifically, the tutorial will have an introductory overview including a discussion of learning objectives and hands-on exercises covering various aspects of DEM processing and the execution of these operations on HPC platforms for scalability. The hands-on section will be followed by a question and answer session to allow participants to seek any clarification and address topics related to individual research areas. This collaborative hands-on tutorial will allow participants to interact with instructors, seek assistance, and share ideas with fellow learners. A computer machine with an internet browser is needed to complete this tutorial effectively.

Skills and Background: Basic Python and geospatial data familiarity recommended. Jupyter Notebook experience helpful but not required.

Requirements: Participants must bring their own laptops.

Instructor:

• Greg Buehler, Open Geospatial Consortium (OGC)

Time & Day: 9:00 am - 12:05 pm Central (with a break), Tuesday June 17, 2025   Room: 3

The OGC API Training Workshop offers an introduction to OGC APIs, which are transforming how geospatial data is shared, accessed, integrated, and analyzed. Built upon OGC’s legacy of open standards, these APIs leverage modern RESTful web architectures to make geospatial data accessible to anyone and easily integrated with other types of information. Participants will explore these APIs through hands-on exercises that support the discovery, access, visualization, and processing of geospatial content in line with FAIR data principles. This is a half-day, hands-on activity designed to provide participants with practical experience. Attendees should bring a laptop to fully engage in the interactive exercises and activities. Prior to the workshop, participants are requested to review the OGC API workshop website for foundational information, ensuring they are prepared to make the most of this learning opportunity.

Skills and Background: Ideal for GIS professionals and developers. Some knowledge of web APIs and geospatial data formats is helpful.

Requirements: Attendees should bring a laptop and review the OGC API workshop website beforehand.

Instructor:

• Lance Legel, Ecodash.ai

Time & Day: 9:00 am - 12:05 pm Central (with a break), Tuesday June 17, 2025   Room: 4

This workshop introduces participants to DeepEarth, a self-supervised framework that integrates heterogeneous geospatial and ecological datasets—such as satellite imagery, topography, climate records, and biodiversity observations—for spatiotemporal modeling of Earth’s ecosystems. By training on partially masked observations, DeepEarth infers missing information on species distributions and environmental variables, enabling simulation of ecosystem dynamics across historical, present, and projected future periods. Attendees will learn to build a multimodal pipeline that addresses incomplete sampling, observational bias, and complex ecological interactions, culminating in hands-on exercises with data pipelines, training routines, and interpretation of results. The workshop will also cover methods for extracting and applying pre-trained embeddings—latent representations of species and habitats—to downstream tasks like classification, presence/absence prediction, and scenario simulation. By the end, participants will possess a theoretical foundation and practical skills to design and implement self-supervised spatiotemporal models for advanced ecological research and geospatial intelligence.

Skills and Background: Some experience with machine learning or geospatial modeling is recommended. Basic Python knowledge helpful.

Requirements: Attendees should bring a laptop to fully engage in the interactive exercises and activities.

Instructor:

• Alexander Michels, University of Illinois Urbana-Champaign
• Erick Li, University of Illinois Urbana-Champaign
• Yunfan Kang, University of Illinois Urbana-Champaign
• Furqan Baig, University of Illinois Urbana-Champaign

Time & Day: 1:30–2:45 pm Central, Tuesday June 17, 2025   Room: 1b

This session offers an introductory walkthrough of the I-GUIDE Platform, showcasing its capabilities in enabling reproducible, data-intensive geospatial research. Attendees will explore platform components, user interface, example notebooks, and how it supports workflows across disciplines such as sustainability, hydrology, and environmental modeling.

Skills and Background: General interest in geospatial platforms. No technical background needed.

Requirements: Attendees should bring a laptop to fully engage in the interactive exercises and activities.

Instructor:

• Peter Darch, University of Illinois Urbana-Champaign
• Kyra Abrams, University of Illinois Urbana-Champaign

Time & Day: 3:05–5:15 pm Central, Tuesday June 17, 2025   Room: 1a

This interactive workshop introduces participants to the I-GUIDE Data Ethics Toolkit (I-GUIDE DET), designed to support responsible, reproducible, and policy-compliant research across the full geospatial data science lifecycle. Drawing on long-term research of I-GUIDE platform leaders, this suite of lightweight tools was developed specifically for the needs of the I-GUIDE platform. Participants will learn how to integrate the toolkit into their projects from the outset and update it iteratively throughout. Each component of the toolkit is designed for manual use or future automation and supports compliance with policies such as GDPR, the EU AI Act, and major funding agency requirements for data management and Open Science.

Skills and Background: No programming experience required. Ideal for researchers, project managers, and students interested in responsible data use and open science practices in geospatial research.

Requirements: Attendees should bring a laptop to fully engage in the interactive exercises and activities.

Instructor:

• Zhe Zhang, Texas A&M University
• Honggao Liu, Texas A&M University

Time & Day: 1:30–5:15 pm Central (with a break), Tuesday June 17, 2025   Room: 3

Disasters become prominent global issues that simultaneously pose a threat to multiple countries or regions around the globe. The dynamics of a multi-scale networked society are inextricably entangled with geographical and built environments, where human-environment interactions shape community resilience under different disaster events. Disaster management is gradually empowered by increasing geospatial big data awareness and growing computing capabilities to produce spatial vulnerability and situational understanding for supporting timely decisions. This tutorial sheds light on high-performance geocomputational educational concepts important in understanding human-environment interactions in disaster management.

Skills and Background: No prior experience required, but familiarity with geospatial analysis or emergency response is beneficial.

Requirements: Attendees should bring a laptop to fully engage in the interactive exercises and activities.

Instructor:

• Mengling Qiao, Columbia University
• Masahiko Haraguchi, Columbia University
• Upmanu Lall, Columbia University

Time & Day: 1:30–5:15 pm Central (with a break), Tuesday June 17, 2025   Room: 2

The hands-on tutorial aims to train participants in accessing, processing, and analyzing multi-year, borough-level urban crime and socio-environmental datasets using reproducible Python-based workflows, with an emphasis on FAIR (Findable, Accessible, Interoperable, Reusable) data principles. Using Python-based Jupyter Notebooks on cloud-hosted environments, attendees will learn how to access standardized, borough-level crime and predictor datasets; perform exploratory spatial and temporal analyses; build machine-learning models; and visualize feature impacts geospatially.

Skills and Background: Basic familiarity with Python (pandas, scikit-learn) and geospatial data (geopandas) recommended. Experience with Jupyter Notebooks helpful but not required.

Requirements: Participants must bring their own laptops with an Anaconda or equivalent Python environment installed (including pandas, geopandas, scikit-learn, and matplotlib). Cloud-based notebooks will also be available.

Instructor:

• Jeff Carpenter, Advanced Visualization Lab (AVL), NCSA, UIUC
• Matt Turk, AVL, NCSA, UIUC
• Stuart Levy, AVL, NCSA, UIUC
• Bradley Thompson, AVL, NCSA, UIUC

Time & Day: 1:30–5:15 pm Central (with a break), Tuesday June 17, 2025   Room: 4

Effective science communication is more important than ever. In this era of agendas, misinformation, and a seeming mistrust of facts, cinematic scientific visualization is uniquely positioned to engage and educate broad audiences by combining vetted science with production-quality graphics and storytelling. This workshop will explain the how and why of creating such visualizations. Cinematic scientific visualization differs from traditional scientific visualization because its primary goal is not just to analyze or verify a dataset. Instead, it aims to communicate complex scientific concepts to the general public, who may lack both knowledge and interest in the subject. These visualizations must be not only scientifically accurate but also engaging and visually appealing.

Skills and Background: No programming required. Recommended for researchers interested in communicating science visually.

Requirements: Attendees should bring a laptop to fully engage in the interactive exercises and activities.