Fig. 2: Spatiotemporal variations in global atmospheric particulate matter (PM) isotopic fingerprints.
From: Blockchain-based isotopic big data-driven tracing of global PM sources and interventions
a, c, e, g Temporal trends of fM-14C for EC (a) and OC (c) δ15N for NH4+ (e) and NO3− (g) were analyzed using formal bootstrap testing procedures. The abbreviations EC, OC, NH4+, and NO3- represent elemental carbon, organic carbon, ammonium, and nitrate, respectively. The different-colored lines in (a, c, e) and g represent liner fits for different periods, with the corresponding shading indicating their 95% confidence intervals. b, d, f, h Moving subset window analysis of the temporal change trends of fM-14C-EC (b), fM-14C-OC (d) δ15N-NH4+ (f) and δ15N-NO3− (h) over 5-year periods. The analysis step was 1 year. The bars represent isotopic change rates (k) every 5 years. Details on the amount of isotopic data in each plot are provided in the Source Data file. ***P < 0.001, **P < 0.01, *P < 0.05. Statistical robustness was verified using the Theil–Sen estimator (Supplementary Figs. 6 & 7). The color scale of the bars correspond to temporal trends, without specific indication. i Global distribution of atmospheric particulate matter isotopic observations and pollution levels. The global atmospheric particulate matter pollution levels are represented by the annual average concentrations of atmospheric fine particulate matter (PM2.5) for 2022. Isotopic observations are primarily concentrated in Asia, Europe, and the Americas. The map was generated with Natural Earth, which provides free vector and raster map data available at naturalearthdata.com. j The t-distributed stochastic neighbor embedding (t-SNE) map shows global atmospheric particulate matter isotopic affinities within the IDGAR. Most isotopic data form distinct clusters rather than random distributions.
