Skip to main content
ARS Home » Southeast Area » Florence, South Carolina » Coastal Plain Soil, Water and Plant Conservation Research » Research » Publications at this Location » Publication #401514
Research Project: Innovative Manure Treatment Technologies and Enhanced Soil Health for Agricultural Systems of the Southeastern Coastal Plain

Location: Coastal Plain Soil, Water and Plant Conservation Research

Title: Machine learning to quantify and upscale manure nutrients budgets in agricultural watersheds of the Southeast Coastal Plain of the United States

Author
item Sohoulande, Clement

Submitted to: Meeting Abstract
Publication Type: Abstract Only
Publication Acceptance Date: 3/6/2023
Publication Date: 6/13/2023
Citation: Sohoulande Djebou, D.C. 2023. Machine learning to quantify and upscale manure nutrients budgets in agricultural watersheds of the Southeast Coastal Plain of the United States. EcoSummit 2023 : Building a sustainable and desirable future: Adapting to a changing land and sea-scape, Gold Coast, Australia, June 13-17 2023.

Interpretive Summary: .

Technical Abstract: Recent assessments of agricultural nutrients budgets across the United States (US) revealed an overall low manure nutrients recycling and critical contrasts between nutrients deficient regions and regions with excess nutrients. Whereas the outcomes of these assessments are useful at regional levels, they may not be adequate to accurately trace manure nutrients patterns at sub-basins and watersheds scales, because their estimates of nutrient budgets are essentially based on county-level pooled statistics of livestock and manure management practices. Particularly, in the Southeast Coastal Plain region of the US, where soils are sandy with a low nutrients holding capacity, a comprehensive understanding of the pathways of manure nutrients generated by concentrated animal feeding operations (CAFOs) is critical for the local ecosystems. Hence, the present study aims to quantify and trace manure nutrients using site-specific data retrieved from one-meter resolution aerial imageries. Precisely, Machine Learning (ML) procedures are being developed to geolocate CAFOs and identify features such as barns, waste lagoons, spray-fields. The study is on-going, and the retrieved features of animal waste management systems will be used as site-specific inputs to estimate manure nutrients budgets at the sub-basins level. The sub-basins-level estimates will be considered to scale-up manure nutrients from farms to the watershed scale. The results will complement the existing agricultural nutrients budget assessments in the US. The study will provide insights into manure nutrients pathways at the watershed scale. Hence, recommendations will be elaborated to enhance manure recycling and increase the footprint of manure nutrients in crop production.