2026 Big Sky GeoCon - From Data to Design

Esri’s GeoAI capabilities fuse advanced machine learning, deep learning, and spatial analysis to automate feature extraction and accelerate insight generation across imagery, video, point clouds, text, and other geospatial data sources. By integrating pretrained models, configurable geoprocessing tools, and extensible Python APIs, ArcGIS empowers users to rapidly transform raw sensor data into actionable information. These capabilities now extend directly into field workflows through Survey123’s computer vision integration, image classification and feature detection to validate field observations, streamline QA/QC, and enrich survey data collection with automated intelligence.

Building on advancements in large language models, Esri is enhancing GeoAI with natural-language-driven analysis and intelligent assistants that help users discover data, generate workflows, and interact with GIS more intuitively. LLM-powered text AI complements traditional feature extraction by enabling semantic understanding of unstructured content while Survey123’s vision-based automation strengthens the connection between field capture and enterprise analysis. Together, these innovations make spatial AI more accessible, scalable, and efficient by unlocking new levels of productivity for analysts, field crews, and decision-makers across industries.

This workshop provides a practical, end-to-end overview of using unmanned aerial systems (UAS) with Esri Site Scan to capture, process, and deliver high-value geospatial data. Participants will learn how drone-collected imagery and reality capture products integrate directly into the Esri ecosystem to support mapping, analysis, and decision-making workflows.
The session will cover core concepts of drone operations, data collection best practices, and how Site Scan streamlines flight planning, data management, processing, and collaboration. Emphasis will be placed on turning raw aerial data into actionable GIS deliverables such as orthomosaics, digital surface models, point clouds, and 3D content that can be shared across teams.
Attendees will gain insight into real-world use cases including surveying, construction monitoring, asset inspection, public safety, and infrastructure management. The workshop will also highlight accuracy considerations, operational efficiencies, and how organizations can scale drone programs while maintaining consistency and data quality.
By the end of the workshop, participants will understand how to effectively connect drone workflows with Esri Site Scan to move from flight to finished GIS products faster, more efficiently, and with greater confidence.

A large rainfall event triggered multiple landslides across Bridger Bowl Ski Area in southwest Montana on July 29, 2025. The Geospatial Core Facility at Montana State University conducted a UAV-based LiDAR survey over a landslide path along the northern boundary of Bridger Bowl and generated a 1-m resolution post-event digital elevation model (DEM). A pre-event 1-m DEM was obtained from the U.S. Geological Survey. Raster differencing between the pre- and post-event DEMs was used to estimate a volumetric displacement of 16,000 m3. These results demonstrate the value of UAV LiDAR for post-event landslide assessment and terrain management in alpine environments.

Extreme heat and rapid urbanization are converging challenges for the Arabian Peninsula, yet their fine-grained interactions remain poorly understood. We present a high-resolution assessment of land-cover land-use change, population growth, and land surface temperature (LST) from 2000 to 2020. We found that newly urbanized areas converted from desert exhibited significantly lower warming (+1.78°C) than existing urban areas (+2.39°C) and unchanged desert (+2.97°C). However, these newly urbanized areas maintained higher mean LST than existing urban areas by 2020 (42.68°C versus 41.29°C), creating a counter intuitive thermal where 13.1 million new residents live in places that experienced relative cooling yet higher absolute LST exposure. We furthermore identified distinct population-thermal pathways: small Gulf states achieved population growth with minimal warming through densification, while larger countries showed sprawl-dominated patterns with varied thermal outcomes. Our findings demonstrate that desert cities experience fundamentally different thermal dynamics than temperate regions and require a revised adaptation framework accounting for urban cooling potential and extreme baseline temperatures.

Many of the major themes within Environmental Education rely upon remote sensing and orbital data collection. This unit was developed to introduce students in grades 11-12 to the science behind the maps and graphics they see in class. It ranges from a review of the physics of the electromagnetic spectrum to data manipulation, ground-truthing, and ArcGIS workflows.
The unit begins with a physical modeling of active vs. passive sensing and spectral signatures, teaching students to recognize that every land cover—from healthy forest to impervious urban surfaces—possesses a unique "spectral fingerprint." Students use resources from NASA and the USGS to develop an understanding of how global change can be detected remotely.
As the unit progresses, students access primary data via USGS EarthExplorer, perform change-detection analysis of their local area using Landsat time-series data, and calculate the Normalized Burn Ratio (NBR) to model ecological disturbance. An infrared sensor and an air quality sensor are used to collect data around the school campus. This field data is crowdsourced via ArcGIS Survey123 and visualized alongside satellite imagery in ArcGIS Living Atlas in an attempt to demonstrate "ground-truthing." The unit culminates in an ArcGIS workflow that helps students build a comprehensive ArcGIS StoryMap portfolio detailing their learning progress throughout the unit.