Variation and Spatial Distribution of Emissions from Livestock Manure Management in Iran: An Evaluation and Location Analysis

Abstract

Rising livestock and poultry production necessitates sustainable manure management practices to curb greenhouse gas (GHG) emissions. This study employs two artificial neural networks, Multi-Layer Perceptron (MLP) and Radial Basis Function (RBF), to forecast manure production in Iranian provinces (2020–2030). The RBF model demonstrated superior accuracy compared to a Multi-Layer Perceptron model. Our forecasts predict the significant potential for biogas and biomethane production from manure by 2030, estimated at 10,782.4 and 6469.44 Mm³.year-1 respectively. This translates to replacing 4.03% and 4.98% of Iran's projected annual gas and electricity consumption in 2030. While this offers a renewable energy source, conventional manure management practices are projected to increase agricultural methane emissions. Our analysis highlights that utilizing biomethane from biogas represents the most effective strategy for reducing GHG emissions in the energy sector. The study projects that by 2030, manure management will still produce 14 million tons of carbon dioxide, equivalent to 16.71% of the agricultural sector's GHG emissions. Scenario analysis indicates that adopting biomethane as a natural gas substitute offers the most significant reduction in energy sector emissions compared to current practices. These findings underscore the importance of effective manure management for climate change mitigation. Furthermore, they highlight the need for long-term pollution reduction policies informed by accurate livestock growth forecasts. This study also contributes by demonstrating the potential of artificial neural network models for accurate manure production forecasting and developing GHG reduction strategies.