Local variability in snow concentrations of chlorinated persistent organic pollutants as a source of large uncertainty in interpreting spatial patterns at all scales

• Published in: Journal of Environmental Quality
• Main Authors: Pawlak, Filip, Koziol, Krystyna Anna, Kosek, Klaudia, Polkowska, Zaneta
• Other Authors: Narodowe Centrum Nauki, Norges Forskningsråd
• Format: Article in Journal/Newspaper
• Language: English
• Published: Wiley 2022
• Subjects: Arctic
• Online Access: http://dx.doi.org/10.1002/jeq2.20343; https://onlinelibrary.wiley.com/doi/pdf/10.1002/jeq2.20343; https://onlinelibrary.wiley.com/doi/full-xml/10.1002/jeq2.20343

Abstract Single point sampling, a widespread practice in snow studies in remote areas, due to logistical constraints, can present an unquantified error to the final study results. The low concentrations of studied chemicals, such as chlorinated persistent organic pollutants, contribute to the uncertainty. We conducted a field experiment in the Arctic to estimate the error stemming from differences in the composition of snow at short distances (1–3 m), including 13 single organochlorine pesticides and 6 polychlorinated biphenyls, thus providing the most detailed published dataset on the subject. We contrasted this variability with the uncertainty at larger spatial scales, both within one valley (regional scale, this study) and as described in the worldwide literature. The range of values for the coefficient of variation for local samples was 20–58% for single organochlorine pesticides (OCPs) and 33–54% for polychlorinated biphenyls (PCBs), and for regional samples it was 21–69% for OCPs and 65–93% for PCBs. We suggest that, to observe the actual changes in the concentration of selected compounds in snow, they should vary at the level of 40–60%, depending on the compound in question. The uncertainty margin remains much smaller than the current discrepancy between observation data and atmospheric deposition models considering snow, deeming field data on snow concentrations a useful ground‐truthing dataset. However, field observations on spatial differences at all scales need to be interpreted with caution, and the dataset provided here on the local sampling uncertainty helps define the margins of such interpretations.