SoilTwin is being developed by the Center for eXascale Spatial Data Analytics and Computing at Colorado State University. The project is led by Shrideep Pallickara, with Sangmi Lee Pallickara serving as co-lead.
Housed in the Department of Computer Science, SoilTwin brings together expertise in geospatial data systems, high-performance computing, artificial intelligence, visualization, and environmental data analytics. The platform is being built with the contributions of graduate and undergraduate students at Colorado State University, whose work spans data ingestion, scalable storage, query processing, machine learning, and interactive web-based visualization.
SoilTwin is supported by grants from the National Institute of Food and Agriculture (NIFA) and the National Science Foundation (NSF). These investments support our broader goal: to make soil variation visible, queryable, and usable at decision-relevant scales. By combining foundational soil data, satellite and in-situ observations, physics-guided AI models, and scalable geospatial infrastructure, SoilTwin aims to help researchers, educators, students, and decision-makers better understand the living landscape beneath our feet.
Interpretability
Physics-informed AI
Guiding AI/ML models with the physics that governs soil and water interactions.
Data-driven
Big Data for the Sciences
Turning voluminous, longitudinal environmental datasets into usable scientific insight
Interactivity
High-performance Visualizations
Making hidden soil processes visible, explorable, and understandable
Situational awareness
Geospatial Intelligence
Connecting place, time, and process across the agricultural landscape.
Insights
From Data to Understanding
Transforming complex soil data into explanations people can use
Insights
Use-informed
Engagement, demos, and user studies with potential users, students, growers, and stakeholders
See soil as a living system. SoilTwin brings together soils, terrain, weather, water, land cover, sensors, satellites, and models into one coherent view.
Ask better questions across layers. Explore relationships across space and time, from “what is happening here?” to “why is it happening here, now?”
Move from maps to insight. GPU-accelerated visualization, physics-guided AI, and scalable infrastructure make complex soil processes visible, queryable, and usable.