248-7 Mapping Livestock Infrastructure Using Object-Based Image Analysis

Session: Expanding Geology’s Horizons: Geoinformatics, Open Science, and Open Data

Presenting Author:

Authors:

Abstract:

As demand for meat and dairy continues to rise, the livestock industry in the United States has increasingly turned to animal feeding operations (AFOs), facilities where large numbers of animals are confined in centralized areas to optimize production. While efficient, these operations can lead to significant environmental and public health concerns due to concentrated waste emissions. At the same time, they represent potential sites for resource recovery, particularly manure reuse. Despite their significance, there is currently no nationwide, standardized dataset detailing the locations, animal populations, or manure management practices of AFOs, creating a major gap in monitoring and policy development.

This study seeks to close that gap by enhancing an existing automated detection method using machine learning and object-based image analysis. Building on previously identified AFO hotspots, we introduce a structural feature extraction technique to confirm and characterize facility locations. Key metrics such as footprint area, shape complexity, and boundary length are used to estimate permitted animal counts, which in turn inform calculations of potential manure output. These spatial indicators show strong promise for predicting facility-scale impacts.

Ultimately, this approach not only improves our ability to monitor AFOs at regional and national scales but also lays the groundwork for adapting this methodology to global livestock systems, enabling better-informed strategies for environmental mitigation and sustainable agricultural management.

Geological Society of America Abstracts with Program. Vol. 57, No. 6, 2025

doi: 10.1130/abs/2025AM-9220

© Copyright 2025 The Geological Society of America (GSA), all rights reserved.