Integrating point sources to map anthropogenic atmospheric mercury emissions in China, 1978–2021

<p>Mercury emissions from human activities persist in the environment, posing risks to humans and ecosystem, and are regulated by the Minamata Convention. Understanding historical mercury emissions is critical for explaining their presence in the environment, and a long-term gridded emission inventory is essential for simulation and evaluation. While previous studies have improved the spatial resolution of emission inventories for recent years, few have combined long timescales with high spatial resolutions. Here, we compile a new comprehensive point source database by fusing multiple data sources and integrate it with the previous China Atmospheric Mercury Emission model to develop a long-term gridded emission inventory for China, covering 1978–2021, named P-CAME. By integrating point sources, P-CAME improves the accuracy of gridded emissions, reducing the normalized mean error by 108 % compared to an inventory without point sources for the most recent year of 2021. P-CAME highlights potential pollution hotspots, revealing that 20 % of cumulative emissions originate from just 0.3 % of the grids, primarily in the provinces of Gansu, Yunnan, and Hunan. These areas are dominated by non-ferrous metal smelting or mixed emissions from coal-fired industries and cement production. P-CAME also demonstrates consistency with observed Hg<span class="inline-formula">⁰</span> (gaseous elemental mercury) concentration trends over the past decade and shows potential to enhance the simulation of atmospheric mercury concentrations in urban areas, though its capacity is still limited by overall model performance. With improvements in terms of spatial distribution accuracy and reliable long-term trends, this updated inventory will provide valuable data support for global emissions modeling, facilitate assessments of mercury cycling and legacy impacts, and aid in the evaluation of the Minamata Convention. The dataset can be found at <a href="https://doi.org/10.6084/m9.figshare.26076907">https://doi.org/10.6084/m9.figshare.26076907</a> (Cui et al., 2024).</p>