DOI: 10.1029/2025jd045536 ISSN: 2169-897X

Near‐Equal Agricultural and Non‐Agricultural Contributions to Atmospheric Ammonia in the Erhai Lake Basin, China: Evidence From δ 15 N Isotope‐Based Source Apportionment

Jiahui Kang, Yuepeng Pan, Sijie Feng, Chenxi Yao, Xiaoyun Du, Qikun Shen, Bowen Tang, Mathew R. Heal, Xuejun Liu, Wen Xu

Abstract

Atmospheric ammonia (NH 3 ) deposition poses a threat to aquatic ecosystems by contributing to eutrophication, particularly in vulnerable alpine lakes. While emission inventories prioritize agricultural sources, the role of non‐agricultural emissions remains poorly quantified in many regions. We present a high‐resolution source apportionment of atmospheric NH 3 in the Erhai Lake Basin in China, integrating a dense monitoring network (45 sites across six land‐use types) with δ 15 N isotope signatures and Bayesian mixing modeling. A localized δ 15 N‐NH 3 source fingerprint library was established from seven specific sources and grouped into five end‐member categories: agricultural (fertilizer and livestock) and non‐agricultural (waste, transportation, and biomass burning). The results reveal strong spatiotemporal variability with NH 3 concentrations peaking in farmland (13.58 ± 6.92 μg m −3 ) and a wide δ 15 N range (−26.84‰ to +3.86‰) that enabled clear source discrimination. The model quantified near‐equal contributions from agricultural (approximately 51%) and non‐agricultural (approximately 49%) sources at the basin scale. Fertilizer was the largest individual source (31 ± 4%), followed by livestock (20 ± 4%), waste (19 ± 4%), transportation (16 ± 2%), and biomass burning (14 ± 2%). The overall source structure remained robust across different isotopic fractionation scenarios. This study indicates that non‐agricultural emissions are a co‐dominant source of atmospheric NH 3 , suggesting that conventional inventories may underestimate their contributions and underscoring the need for integrated, cross‐sectoral mitigation strategies in sensitive lake basins.

More from our Archive