Geospatial Data Science
We excel at leveraging the power of machine learning (ML) and artificial intelligence (AI) to extract meaningful insights from geospatial data and imagery. Our team of data scientists collaborates closely with subject matter experts to develop and deploy models tailored to specific client needs. We ensure scalable and efficient operations to handle large datasets and complex analyses.
Our Expertise
100 + Projects
-
AI and Machine Learning
-
Sample Design
-
Training Data Collection
-
Large-scale Data Processing
Forest Inventory Modeling and Post-Fire Change Detection using Lidar Data
RedCastle Resources (RCR) partnered with the USFS to produce forest inventory models across the full extent of North Kaibab Ranger District (655,000 acres) bordering Grand Canyon National Park. Using a combination of Bayesian variable selection and linear regression modeling, we produced eight forest inventory models at a 20-meter spatial resolution. In addition, we worked with pre- and post-fire lidar data to detect post-fire changes in vegetation cover and height. We also produced a burn severity map based on canopy cover differencing and qualitatively compared it to existing fire severity products, including Monitoring Trends in Burn Severity and Rapid Assessment of Vegetation Condition after Wildfire data.
Continental Landscape Change Monitoring
Understanding where, when, and what factors drive change across landscapes is essential for making informed land management decisions. To assist land managers and scientists, we helped develop the Landcover Change Monitoring System (LCMS) for the U.S. Forest Service through a collaboration between remote sensing experts from the USDA Forest Service, USGS, NASA, and several universities. LCMS is a remote sensing-based system that tracks landscape change across the United States from 1985 to the most recent full growing season. It provides the most accurate map of landscape change by utilizing advanced time series change detection techniques, Landsat and Sentinel data, cloud-based computing power, and big data analysis.
View LCMS
Vegetation mapping
Mapping vegetation types is crucial for effective environmental management and ecological conservation. We have decades of experience creating robust vegetation maps covering extensive areas. We have conducted over 20 mid- scale vegetation mapping projects covering diverse landscapes. Our in- depth approach leverages a combination of remote sensing expertise, ecological background and advanced modeling techniques providing a comprehensive view of the landscape. This enables vegetation mapping that supports informed decision-making in forest and ecological management.