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

Abstract

Abstract As the demand for livestock and poultry supply chain continues to rise, managing the ever-increasing amount of livestock manure has become a significant challenge. In this study, we employ two models of neural networks, namely the multi-layer perceptron (MLP) and radial basis function (RBF) models, to accurately forecast the production of livestock and poultry manure from 2020 to 2030. The aim is to aid decision-making processes in reducing greenhouse gas (GHG) emissions caused by manure storage. Our results reveal that the RBF model outperforms the MLP model in terms of accuracy and reliability. According to our predictions, the provinces of Iran are estimated to produce 10782.4 and 6469.44 Mm3.year− 1 of biogas and biomethane, respectively, from livestock and poultry manure in 2030. This is equivalent to 4.03% and 4.98% of Iran's annual gas and electricity consumption in 2030. Our findings also show that the manure management system will produce 14 million tons of carbon dioxide in 2030, equivalent to 16.71% of GHG emissions in the agricultural sector. Our scenario analysis indicates that using biomethane produced from biogas instead of natural gas in 2030 is the most effective action to reduce GHG emissions in the energy sector compared to the current trend of manure management. Our study highlights the potential of neural network models in accurately forecasting livestock manure production and in developing strategies for reducing GHG emissions.