Atmospheric wet deposition of trace metal elements: Monitoring and modelling

Abstract

Trace elements (TEs), a group of atmospheric pollutants, have attracted considerable attention from scientists and government administrators worldwide. The wet deposition fluxes of nineteen trace elements (NTE) were monitored at Wanqingsha, a coastal site in the Pearl River Delta, for three consecutive years (2016.9–2019.8). Significant seasonal differences in NTE between wet and dry seasons were observed. The fluxes of crustal elements (Ca, Na, Al, Mg, K, Fe, Zn and Ba) were significantly higher than those of anthropogenic elements, accounting for over 99 % of the total annual wet deposition of 19 elements. Analysis of PM2.5 and rainfall samples reveals that both the fraction of each TE in the PM2.5 (CQ) and the Apparent Scavengance Ratio for TE (ASR, defined as the concentration ratio in rain and PM2.5) follow lognormal distributions. The logCQ variation for each element is relatively small but shows substantial differences, with means ranging from −5.48 to −2.03, while the logASRs for all elements show similar means (varying from 5.86 to 7.64) and an extremely wide range of variation. The influences of meteorological factors on CQ and ASR were also investigated. A simple box model framework was constructed to reasonably simplify the TE removal process by precipitation. The corresponding regression analysis showed significant correlations between NTE and the precipitation rate, PM2.5 concentration, ASR, and CQ, with R2 ranging from 0.711 to 0.970. By substituting the effects of environmental factors on ASR and CQ into the above relationship, temporal variations in NTE can be predicted. The reliability of the model was demonstrated by comparing model simulations with observations over three years. For most elements, the models can predict the temporal variation of NTE quite accurately, and even for the worst predictions, such as Al, Mg, K, Co and Cd, where predictions exceed observations by only an order of magnitude.