Nitrous Oxide Removal from Cattle Manure Emissions Using Inorganic and Carbon-Based Adsorbents: A Statistical and Machine Learning Approach

Abstract

Nitrous oxide (N₂O) emissions from agriculture and livestock significantly contribute to climate change. Consequently, reducing emissions from livestock waste is crucial. This study aimed to assess the removal of N₂O emitted from cattle manure using various adsorbent materials. Nine adsorbents—bentonite, vermiculite, zeolite, pumice, perlite, colemanite, activated carbon, graphite, and boron nitride—were evaluated over 24 days in a total of 10 laboratory-scale reactors, including a control reactor. Each reactor was loaded with 1,000 g of cattle manure; gas samples were measured daily, and the data were analyzed using the Shapiro-Wilk, Levene, Kruskal-Wallis, discriminant analysis, time-series clustering, and random forest methods. The Kruskal-Wallis test indicated no significant difference between the reactors and the control (p=0.416). Time-series clustering identified temporal differences only for pumice and graphite. The random forest analysis predicted a decreasing trend for activated carbon and increasing trends for pumice and perlite. The findings suggest that, over a 24-day monitoring period, the adsorbent types did not exhibit a statistically significant advantage in N₂O removal, highlighting the need for longer-term and multiparameter monitoring.