Contributions of Great Salt Lake Playa‐ and Industrially Sourced Priority Pollutant Metals in Dust Contribute to Possible Health Hazards in the Communities of Northern Utah

Abstract

Communities and ecosystems of northern Utah, USA receive particulate pollution from anthropogenic activity and dust emissions from sources including the Great Salt Lake (“the Lake”) playa. In addition to affecting communities, anthropogenic pollution is delivered to the Lake's playa sediments, which are eroded during dust events. Yet, spatial variability in dust flux and composition and their risks to human health are poorly understood. We analyzed dust in 17 passive samplers proximal to the Lake during fall 2022 for dust flux, the dust fraction of particulate matter, ⁸⁷Sr/⁸⁶Sr, and elemental geochemistry. We evaluated spatial patterns of 11 priority pollutant metals and estimated the hypothetical non‐cancer dust and soil ingestion health hazard for six age cohorts. We observed the highest dust fluxes proximal to the Lake's playa. The highest concentrations of and greatest number of metals occurred in and south of Ogden, UT. Sites to the northeast of Farmington Bay had the highest fluxes. Metal concentrations and ⁸⁷Sr/⁸⁶Sr suggest that the dust composition near Bountiful represents contributions from anthropogenic sources, whereas the dust composition to the northeast of Farmington Bay reflects the Lake's playa emissions. Evaluations of potential health hazards from dust ingestion suggest that children between birth and 6 years are vulnerable at higher ingestion rates. Thallium, As, Pb, Co and Cr contributed most to the estimated hazard. Among these, As and sometimes Pb are likely derived from the Lake's playa emissions. Thus, suppression of dust emissions from the Lake's playa may decrease possible health risks for children in northern Utah.

Key Points

• Priority pollutant metals occur in dust in northern Utah with spatially variable composition and flux

• Priority pollutant metals in dust are attributed to both anthropogenic emissions and dust emissions from the Great Salt Lake playa

• Children under 6 may be vulnerable to health hazards from ingestion of metals via dust

1. Introduction

Dust pollution threatens human and ecosystem health (Ardon‐Dryer et al., 2020; Brahney et al., 2024; Goudie, 2014; Griffin & Kellogg, 2004; Jones & Fleck, 2020; Kim et al., 2015). The western slope of the Wasatch Range in northern Utah, USA regularly experiences dust events that impact air quality and deliver dust to mountain snowpack and downwind ecosystems (Hahnenberger & Nicoll, 2012, 2014; Lang et al., 2023; Munroe et al., 2025; Nicoll et al., 2020; Steenburgh et al., 2012). Dust events tend to occur during periods of sustained high winds in the spring and fall (Hahnenberger & Nicoll, 2012; Steenburgh et al., 2012), and the dust comes from playas and disturbed lands (e.g., burn scars, mining impacted lands, and land affected by other development). Source areas include the lakebeds of Sevier Lake and Tule Lakes, the Great Salt Lake Desert, and the Great Salt Lake playa (Carling et al., 2020; Goodman et al., 2019; Grineski et al., 2024; Hahnenberger & Nicoll, 2012, 2014; Hahnenberger & Perry, 2015; Lang et al., 2023; Mallia et al., 2017; Nicoll et al., 2020; Skiles et al., 2018). However, total dust loading may come from a combination of regional dust events and more frequent contributions of dust from within and near‐city sources like construction or gravel mining (Marcy et al., 2024; Munroe et al., 2025; Putman et al., 2022).

Among these sources, the Great Salt Lake playa exposed in Farmington and Bear River Bays are emerging dust sources of concern to state resource managers and the local population (Great Salt Lake Strike Team, 2023; Utah Division of Water Resources, 2024; Figure 1). Dust eroded from the Great Salt Lake playa has been shown to contain metals of concern (Jung et al., 2024; Lopez et al., 2024; Perry et al., 2019) and is more likely to cause respiratory issues than other dust sources (Attah et al., 2024; Cowley et al., 2025). Since the mid‐1980s, the exposed playa area, characterized by unconsolidated sediments, has increased to $\sim$2,400 ${\text{km}}^2$ from $\sim$500 ${\text{km}}^2$ (Radwin & Bowen, 2024) due to prolonged drought and growing demands on water supply (Wurtsbaugh et al., 2017). In fall of 2022, the Lake elevation was less than 1276.8 m (4189 ft) at U.S. Geological Survey station 10010000 (GREAT SALT LAKE AT SALTAIR BOAT HARBOR, UT) (U.S. Geological Survey, 2023), more than 2 m below the long term median elevation and 4.5 m below the record high elevation at the same site. This Lake elevation exposed an estimated 194 ${\text{km}}^2$ of dust emission hotspot areas (Great Salt Lake Strike Team, 2023).

Figure 1.

Overview map of the study region in northern Utah. Map shows dust collection and surface sediment sampling sites. Note that two sediment collection sites had multiple samples collected at the site. Twelve sediment samples were collected and 10 sediment sampling sites are present on the map. The map also shows the locations of Utah Department of Environmental Quality ${\text{PM}}_{10}$ monitors that were active in fall 2022 (https://air.utah.gov/network/Counties.htm) and permitted emitters that are discussed in the text (U.S. Environmental Protection Agency, 2012). Satellite imagery courtesy of the NASA Goddard Space Flight Center and U.S. Geological Survey (Earth Resources Observation and Science (EROS) Center, 2020) and TerraMetrics (2023). Imagery fetched from Google via cartopy.io.img_tiles(). Major roads fetched from Utah Geospatial Resources Center (2022).

Because particulate matter (PM) smaller than 10 $\mathrm{\mu }$m in diameter (${\text{PM}}_{10}$), is considered an inhalation health hazard at certain concentrations, ${\text{PM}}_{10}$ is monitored nationally. However, the monitoring network is sparse enough that many dust events go undetected (Ardon‐Dryer, Clifford, & Hand, 2023; Ardon‐Dryer, Gill, & Tong, 2023). Northern Utah is also characterized by network sparsity: ${\text{PM}}_{10}$ is monitored by a few ($\sim$5, with exact number dependent on the period of interest) active sensors, managed by Utah Department of Air Quality, that report data at hourly to weekly timescales (U.S. Environmental Protection Agency, 2023b). At the time of publication, no ${\text{PM}}_{10}$ monitors were operation in the three northernmost counties in Utah. Furthermore, filters collected by active ${\text{PM}}_{10}$ monitors are not consistently analyzed for constituents. Yet, contaminants including heavy metals and pathogens are associated with dust particles in both urban and rural areas (Brahney et al., 2021) and have been shown to be present in dust collected in communities and ecosystems downwind of Great Salt Lake (Goodman et al., 2019; Munroe et al., 2025; Putman et al., 2022; Reynolds et al., 2010, 2014). Further, the composition of Great Salt Lake playa dust specifically may exacerbate respiratory issues relative to dust from other playas or local sources like gravel pits or construction sites (Attah et al., 2024; Cowley et al., 2025).

Current and past anthropogenic activity affects the composition of particulate matter in the region in two ways. First, particulate matter is emitted from modern industrial activity within cities and commingles with local or regional dust during dust events (Munroe et al., 2025). Second, because the Great Salt Lake Basin is endorheic, any pollution that was emitted, deposited, or dumped in the Basin is likely to be transported to Great Salt Lake and deposited in the sediments through hydrologic processes (Jung et al., 2024; Lopez et al., 2024; Wurtsbaugh et al., 2020). Metals posing health concerns occur in higher concentrations in Farmington Bay (Jung et al., 2024; Lopez et al., 2024; Perry et al., 2019; Wurtsbaugh et al., 2020), the outlet for the Jordan River and recipient of discharge from wastewater treatment plants. The Jordan River runs through the Salt Lake Valley, which has a history of industrial practices (Putman et al., 2022) and documented metal contamination of soils in public parks (Lee et al., 2024). Wastewater discharge from treatment plants and industrial activity has been shown to contribute metals to Farmington Bay (Waddell et al., 2009). Dust emitted from Farmington Bay may carry these pollutants to downwind cities and ecosystems.

Because of the spatial distribution of anthropogenic point sources of particulate matter and the spatial distribution and concentration of metals in the Great Salt Lake playa, individual metals have distinct spatial footprints (Putman et al., 2022). This means that some neighborhoods receive ${\text{PM}}_{10}$ and dust with higher concentrations and/or total masses of specific metals compared to other neighborhoods. These contaminants tend to be somewhat or highly bioavailable and are present in multi‐metal mixtures that include U.S. Environmental Protection Agency (EPA) priority pollutant metals (U.S. Environmental Protection Agency, 2023a) such as As, Pb, Cd, Cu, and Zn (Goodman et al., 2019; Munroe et al., 2025; Putman et al., 2022).

Long‐term and event‐based exposure to these mixed metals fluxes from dust may have negative health impacts for individuals (Bollati et al., 2010). Studies of total exposure to metal mixtures in children, through combined inhalation, dermal exposure, and ingestion pathways, have found associations between metal loads in urine or hair and the presence of behavioral problems and/or adverse neurodevelopmental consequences (Rodríguez‐Barranco et al., 2013; Stein et al., 2022). Because about 60% of dust in homes comes from outdoor dust and local soil material tracked indoors (Hunt et al., 2006; Ibañez‐Del Rivero et al., 2023), in‐home exposure to metals delivered via dust transport can occur after the dust event. Concentrations of specific metals, however, are more likely to represent local soil conditions or indoor sources of metals (lead paint, smoking) than the composition of outdoor dust. However, the two may be linked depending on the intensity of the pollution and the magnitude of the dust flux (Van Pelt et al., 2020).

Because of the potential impact of exposure to mixed metals from particulate matter, distinguishing between directly emitted particulate pollution and dust emitted from the Great Salt Lake playa is important for guiding the management of this dust source. Geochemically, dust from playas tends to be distinct from other sources, but geochemistry alone may not be able to distinguish among contributions from different landscape sources (Mangum et al., 2024). In the Salt Lake Valley, radiogenic strontium isotope ratios (⁸⁷Sr/⁸⁶Sr) are a geochemical tracer with distinct values for the Great Salt Lake playa relative to other regional sources of dust (Carling et al., 2020; Munroe et al., 2019). In this region, dust with ⁸⁷Sr/⁸⁶Sr closer to $\sim$0.715 is attributed to emissions from the Great Salt Lake playa, whereas dust with a lower ⁸⁷Sr/⁸⁶Sr ($\sim$ 0.710 to $\sim$ 0.712) is attributed to emissions from other sources, like other regional playas (Carling et al., 2020) and within‐city land disturbances like construction or gravel mines (Marcy et al., 2024; Putman et al., 2022).

To identify the areas with the greatest potential human exposure to complex metals mixtures, we designed a synoptic dust deposition study to evaluate dust composition in unmonitored or undermonitored cities, towns, and ecosystems close to and downwind from the Great Salt Lake playa in northern Utah (Figure 1). Dust samples were analyzed for depositional flux, metals concentrations, and ⁸⁷Sr/⁸⁶Sr. These data were used to evaluate three objectives. The first objective was to evaluate the spatial variability in proportions and fluxes of EPA priority pollutant metals in dust in the region. We hypothesized that more priority pollutant metals would be present at higher concentrations in areas downwind of and home to industrial activity, and that total priority pollutant metal fluxes would increase proportionally with total dust loading. The second objective was to estimate the potential health hazard posed by incidental ingestion of metals from soil and dust for different community exposure sites. We hypothesized that areas with greater metals fluxes and high concentrations of more different types of metals would exhibit higher potential health risks from dust ingestion. The final objective was to infer source contributions of metals, including particulate matter contributions from the Great Salt Lake playa. We hypothesize that local industries (e.g., oil refining, copper mining, and mineral processing) and legacy land uses (e.g., Superfund sites) contribute specific metals to the dust. Other metals, such as As and some Pb, are hypothesized to come from Great Salt Lake playa emissions.

This study is the first detailed, synoptic study of dust composition in the under‐monitored communities adjacent to Farmington Bay and Bear River Bay. The study provides new information about the composition and origin of metals pollution in dust to communities and ecosystems in northern Utah. Results from this study can be used to identify different sources of metals pollution to communities most vulnerable to dust‐mediated metals exposure.

2. Methods

2.1. Field Collection and Sample Processing

2.1.1. Dust Samplers

We deployed 17 passive dust samplers across Davis, Weber, Box Elder, and Cache Counties between August 22 and 4 November 2022 (Table 1, Figure 1, refer to Blakowski et al. (2023) for exact deployment dates). Following Reheis et al. (1995) and described in Putman et al. (2022), the samplers consisted of circular cake pans filled with clear glass marbles and mounted on 1.5 m posts. The design described in Putman et al. (2022) was updated: silicone pans replaced metal pans to limit metals contamination. The flexible, silicone pans (surface area = 0.0353 ${\mathrm{m}}^2$) were nested inside aluminum pans for stability. Marbles were suspended 3–4 cm below the rim of the pan on rigid, plastic mesh and bird spikes were secured to the outer rim of each pan using plastic cable ties. All components of the collectors that came into contact with sample material were acid‐washed prior to deployment.

Table 1.

Site Code, Name, Location, Elevation, and Grouping

Site ID	Site name	Sampling media	Latitude	Longitude	Elevation (m ASL)	Group
LNP	Legacy Nature Preserve	Dust	40.85	−111.96	389	South
BJH*	Bountiful Jr. High	Dust	40.89	−111.88	403	South
HTR	Holbrook Trailhead	Dust	40.88	−111.84	497	South
DTC*	Davis Technical College	Dust	41.03	−111.92	413	South
SCH*	Syracuse City Hall	Dust	41.08	−112.06	399	Central
ONC*	Ogden Nature Center	Dust	41.25	−112.01	392	Central
PWS*	Pineview Water Systems	Dust	41.33	−111.81	430	Central
UST*	Utah State University Tremonton	Dust	41.72	−112.17	400	North
LIN*	Logan Island neighborhood	Dust	41.73	−111.83	422	North
UGF	Utah State University Greenville Farm	Dust, sediments	41.76	−111.81	429	North, soil
PPM	Perry playa monitoring	Dust	41.03	−112.12	388	GSL playa (Farmington)
GSP	Great Salt Lake Shorelands Preserve	Dust	41.05	−112.10	389	GSL playa (Farmington)
OBW	Ogden Bay Waterfowl Management Area	Dust	41.14	−112.18	388	GSL playa (Farmington)
HyL	Lakeside	Dust	41.22	−112.85	389	GSL playa (other)
HyS	Saline	Dust	41.25	−112.50	390	GSL playa (other)
BRM	Bear River Migratory Bird Refuge	Dust	41.47	−112.17	389	GSL playa (other)
SCW	Salt Creek Waterfowl Management Area	Dust	41.63	−112.27	393	GSL playa (other)
DAVIS BS		Sediments	40.83	−11191	415	Soil
DAVIS AG		Sediments	41.10	−112.05	401	Soil
WEBER AG		Sediments	41.19	−112.11	391	Soil
WEBER BS		Sediments	41.31	−111.93	440	Soil
BOXE AG		Sediments	41.72	−112.18	402	Soil
BOXE PIT		Sediments	41.77	−112.05	471	Soil
DAVIS PL		Sediments	41.13	−112.18	389	Playa
BRB PL WE		Sediments	41.34	−112.40	390	Playa
BRB PL CN		Sediments	41.51	−112.35	389	Playa

Note. Sites listed roughly west to east and south to north within groupings. All sites annotated with * are community exposure sites, which are situated either at a school, park, or within a neighborhood.

The spatial distribution of the samplers was designed to capture the west‐to‐east land use gradient, beginning at the eastern margin of the Great Salt Lake playa. Samplers were placed in areas with varying development density as well as east of the developed area (Figure 1). Sites were chosen to fill spatial gaps in past data collection by prioritizing sites near the shoreline and proximal to unmonitored developed areas. Sites were categorized by geographic region and named “South,” “Central,” “North,” “GSL playa (Farmington),” and “GSL playa (other)” (Table 1), where the acronym GSL refers to Great Salt Lake. All but one site (called PPM) represent dust deposition locations. Because it is situated a few kilometers onto the Great Salt Lake playa, the PPM site represents Great Salt Lake playa dust emissions.

At the end of the field campaign, we transported the dust samplers to the U.S. Geological Survey laboratory in Salt Lake City, UT in large, clean plastic bags. Marbles were rinsed from each dust trap into acid‐washed, polypropylene tubs using ultra‐pure deionized water (DI water). The mesh was scrubbed with a polypropylene brush, and the brush, mesh, and pan were rinsed into the tub, using approximately 1.5 L DI water per sample. To separate each dust sample into sand and dust fractions, we poured the bulk sample through a nylon 63 $\mathrm{\mu }$m sieve into separate, acid‐washed tubs. Large organic debris was removed from the coarse fractions using tweezers. The samples were then centrifuged at 3000 rpm, and the supernatant was decanted into acid‐washed high density polyethylene bottles. The sediments were retained in the centrifuge tube. Dust samples were dried in an oven for 48 hr at 40°C, and each sample was weighed. Finally, we acidified the supernatant with 2 mL 6M Omni Trace hydrochloric acid to lower the pH to <2 for preservation.

During sample processing, the sample from the LIN site (Table 1) was compromised. The beaker broke after washing the sampler and before sieving the sample into the <63 $\mathrm{\mu }$m fraction. Because more or all of the <63 $\mathrm{\mu }$m fraction was likely in suspension relative to the >63 $\mathrm{\mu }$m fraction, the loss of supernatant likely preferentially affects the mass measurements of the <63 $\mathrm{\mu }$m fraction. We assume that all of the <63 $\mathrm{\mu }$m fraction was in suspension during the spill, and thus the effect of the accident on the metals concentrations is minimal. However, if this assumption is incorrect, the loss of sample may also affect the concentrations of elements in the <63 $\mathrm{\mu }$m fraction, as finer particles typically host more metals (Gunawardana et al., 2014). In this case, the accident would decrease the concentrations of metals in the sample due to preferential loss of the smallest particles. Because of these considerations, this sample is not included in any analysis of mass ratios or flux, but has been retained for analyses that concern element concentrations.

2.1.2. Sediments

We collected 12 surface‐soil or sediment samples (hereafter referred to as sediments) at 10 sites in Davis, Weber, Box Elder, and Cache Counties in late fall of 2022 (Figure 1). These sediments were chosen to fill spatial and geologic gaps in possible dust sources, including samples of the Great Salt Lake playa from Bear River Bay. Soil was collected from current or recent agricultural fields that were being developed for residences to evaluate the potential that agricultural activity or new construction contributes dust. Finally, we sampled various soil formations from the Lake Bonneville paleoshoreline (Chen & Maloof, 2017) because there are numerous open pit‐gravel mines where these sediments are extracted (U.S. Geological Survey, 2005).

At each site, sediments were scooped into plastic sample bags using acid‐washed plastic trowels and a hub and spoke method (Gold et al., 2019; Hess et al., 2019; Putman et al., 2022). We identified a central sampling location and collected sediment from 6 to 10 locations within a 10 m radius of the central point using spatial compositing to ensure collection of a representative sample. In the laboratory, we disaggregated and ground the sediments using a mortar and pestle. Samples were composited using a stainless‐steel splitter until approximately 200 g of the bulk sample remained. We removed any coarse sand or pebbles by passing the dry sample through a 500 $\mathrm{\mu }$m sieve, and then isolated the <63 $\mathrm{\mu }$m fraction by wet sieving with DI water. These samples were dried for 48 hr at 40°C.

2.2. Air Quality and Meteorology

To evaluate the air quality during the period of record, we retrieved ${\text{PM}}_{10}$ monitoring data for the HV (Harrisville, 490571003) (U.S. Environmental Protection Agency, 2023e; U.S. Environmental Protection Agency, 2023d) and BV (Bountiful, 490110004) ${\text{PM}}_{10}$ monitors (U.S. Environmental Protection Agency, 2023d) (Figure 1). Hourly and daily 24 hr average measurements were available at HV daily and at BV once weekly. At the time of access, hourly data were available through 31 October 2022.

We obtained daily total precipitation and daily maximum and minimum temperature data as well as daily 5 s maximum wind speeds and 2 min average wind directions (National Oceanic and Atmospheric Adminstration (2023b), Table S1 in Supporting Information S1). Temperature and precipitation data were available at all sites, whereas wind speed and direction data were available at a subset of sites (Table S1 in Supporting Information S1). Temperature, precipitation and maximum 5 s wind speed data for the study region were averaged. The 2 min wind direction for the Ogden‐Hinckley Airport was used because it was the closest site to the HV sampler. Records for dust events occurring in the region were obtained from the National Centers for Environmental Information: Storm Events Database (National Oceanic and Atmospheric Administration, 2023a). However, because the Storm Events Database is incomplete for dust events (Ardon‐Dryer, Clifford, & Hand, 2023; Ardon‐Dryer, Gill, & Tong, 2023), we also obtained records for high wind events for the region from the same database (National Oceanic and Atmospheric Administration, 2023a).

2.3. Sample Analysis

2.3.1. Geochemical Analysis

Geochemical analysis of the <63 $\mathrm{\mu }$m dust and sediments was performed at the ICP‐MS Metals and Sr Isotope Facility of the Geology and Geophysics Department at the University of Utah. The samples were leached in 0.8 M ${\text{HNO}}_3$ at $\sim$ 22°C for 24 hr. This acid extraction, while stronger than other weaker alternatives used to evaluate the bioavailability of metals in environmental samples (e.g., ammonium acetate buffers), contributes to results that represent an upper end estimate of possible concentrations of bioavailable metals. One aliquot was taken from the supernatant to measure elemental concentrations via quadrupole ICP‐MS (Agilent 8900, Santa Clara, California, USA) using an external calibration method with an internal standard. A second aliquot was analyzed for ⁸⁷Sr/⁸⁶Sr via multi‐collector ICP‐MS (Neptune Plus, Thermo Finnigan, Bremen, Germany) after chromatographic purification of Sr (PrepFast, Elemental Scientific, Omaha, Nebraska, USA).

2.4. Data Quality Assurance

2.4.1. Method Blank for Method Contamination

One high‐purity quartz sample (grain size 125–250 $\mathrm{\mu }$m) was prepared using our sample preparation methods (aside from sieving) to test for contamination from sample processing. The blanks were leached and analyzed for their geochemical composition.

2.4.2. Analysis of Replicates for Homogeneity of Sample

We performed replicate processing and geochemical analyses on two samples (DAVIS AG and WEBER BS 1) to evaluate the reproducibility of these methods. Replicates were performed on the sediment samples because of the higher likelihood that subsamples of the substrate may differ from one another geochemically due to incomplete homogenization. Thus, performing replicate analysis on sediments provides a conservative estimate of the reproducibility expected from the homogenization process. The replicates were subjected to identical leach conditions and analyzed for their geochemical composition. To quantify the reproducibility of the methods, we calculated the mean absolute percent difference for each element from the two‐sample mean value, as well as the maximum and minimum percent differences from the mean.

2.4.3. Instrumental Analysis Quality Assurance

Total analytical uncertainty comprises three sources: blank correction, calibration interpolation, and instrument noise. All analyses using the ICP‐MS evaluated these sources of uncertainty by performing multiple analyses of an external standard during sample analysis.

For this project, the limit of determination was calculated as 3 times the standard deviation of the instrument background for the specific element. The instrumental limit of detection was converted to a measurement limit of detection by multiplication with the average total dilution factor for samples.

The standard SRM 1643f (Standard Reference Material called Trace Elements in Water, National Institute of Standards and Technology) was used to analyze instrument accuracy for elemental analysis. Not all elements have certified values for SRM 1643f. The standard was measured seven times throughout the ICP‐MS analysis run. The measured values were compared to the known values and mean.

Uncertainty in the ⁸⁷Sr/⁸⁶Sr measurements was analyzed using the standard SRM 987 (Strontium Carbonate Standard, National Institute of Standards and Technology). The standard was run 16 times, and the mean and standard deviation were reported. These values were compared to the known value by a Z‐score test.

2.4.4. Evaluation of Potential for Sampler Overflow

Passive dust deposition samplers also collect precipitation occurring during sampler deployment. If a substantial amount of water falls without a chance for evaporation, the samplers may overflow, causing mass loss.

To evaluate whether precipitation exceeded sampler capacity during sampling, we first calculated the volume capacity of the sampler by measuring the volume of water needed to fill the pan (with marbles and mesh installed) and relating it to the cross‐sectional area of the top of the pan to determine the depth (mm) of precipitation required to overflow the pan. Second, precipitation accumulation data for the period of deployment was retrieved from National Oceanic and Atmospheric Adminstration (2023b). Eleven sites in or near the study region had data available for the period of record (Table S1 in Supporting Information S1). Finally, we calculated the total rainfall per event (or cluster of events) in each sampler and compared it to our sampler's capacity. If more than a week occurred between events, it was that assumed all water in the sampler had evaporated given the low humidities and warm temperatures during sampler deployment.

2.5. Data Analysis Methods

2.5.1. Dust Fluxes and Size Ratios

Dust fluxes $(F)$ were calculated based on the masses $(M)$ of both total suspended particulates (TSP) and dust (<63 $\mathrm{\mu }$m), the area of the bundt pan opening ($A$ = 0.0353 ${\mathrm{m}}^2$) and the duration of deployment $(t)$ in days (Equation 1).

$$F=\frac{M}{(A\ast t)}$$ (1)

The dust mass flux was used to calculate the elemental mass flux $\left(F_x=C_x\ast F\right)$ where the elemental concentration is $C_x$ for the element $x$.

To evaluate the spatial variability in the ratios of size fractions present in the dust samples, we calculated the ratio of the dust mass (particulates <63 $\mathrm{\mu }$m in diameter) relative to the TSP mass. Values close to 1 indicate nearly all particulates collected were dust, whereas values close to 0 indicate that nearly all particulates collected were sand (>63 $\mathrm{\mu }$m in diameter). Size fraction influences the likelihood of uptake, with smaller particle sizes (i.e., the silts and clays that make up dust) being more susceptible to dissolution in stomach acid (Plumlee et al., 2006).

2.5.2. Enrichment Relative to Mean Upper Continental Crust and Great Salt Lake Playa Sediments

We evaluated the enrichment of elements in dust and sediment samples relative to the average composition of upper continental crust (UCC) and to average concentrations in samples from the Great Salt Lake playa. The comparison to UCC allows us to evaluate which elements may be present in greater abundance due to anthropogenic processes or local geology that may concentrate or dilute specific elements in the samples. However, average global crustal geochemistry does not adequately capture local variations in soil and sediment geochemistry, arising from both different parent materials and differences during soil development among regions (Aytop et al., 2023; Munroe et al., 2025). To address this concern, we also evaluated the enrichment of the samples relative to an estimate of the average composition of the Great Salt Lake playa. There are sufficient sediment samples from across the Great Salt Lake playa from this study (Figure 1, Blakowski et al., 2023) and a prior study (not shown on Figure 1, Blakowski et al., 2022) to reliably evaluate enrichments in the dust samples relative to an estimate of elemental concentrations in Great Salt Lake playa sediments. This calculation is used to evaluate which elements are likely to derive primarily or partially from the Great Salt Lake playa, and which may be contributed by particulate emissions from anthropogenic emissions.

Elemental enrichments $\left(E_x\right)$ were calculated as the observed concentration $(C)$ of the element $(x)$ in the material (dust or sediments) relative to the concentration of Al in the sample, divided by the reference concentration of that element ($R_x$, Equation 2) relative to the concentration of Al the data set chosen for enrichment evaluation.

$$E_x=\frac{\frac{C_x}{C_{Al}}}{\frac{R_x}{R_{Al}}}$$ (2)

Rudnick and Gao (2003) was used as the reference data set for the evaluation relative to UCC. The standard concentrations for Great Salt Lake playa values were derived from an arithmetic average of concentrations from sediment and playa emissions samplers from this study and data published in Putman et al. (2022). The samples and their metadata are reported in Table S2 in Supporting Information S1, and the Great Salt Lake playa elemental concentrations we computed from the samples are reported in Table S3 in Supporting Information S1.

2.5.3. Spatial Evaluation of Chemical Mixtures

The first objective of this study is to understand spatial variability in priority pollutant metals (As, Be, Cd, Cr, Cu, Pb, Ni, Sb, Se, Tl, and Zn) concentrations in dust in municipalities of northern Utah. To do so, we evaluated the composition (i.e., how many different priority pollutant metals are present at some minimum threshold?), intensity (i.e., are multiple priority pollutant metals present at levels that exceed some minimum threshold?), and flux (i.e., what mass of each element is delivered over a set period?) of priority pollutant metals in the samples. UCC average concentrations were used as a basis for scaling the priority pollutant concentrations. Using UCC to scale the elemental concentrations allows us to sort anthropogenically contributed or locally concentrated priority pollutant metals occurring at locally enriched levels from those that are present at background levels. However, use of UCC as the reference data set (a) will increase the likelihood that we observe metal enrichments in the data, relative to use of a local baseline and (b) does not contribute information about the potential health risks associated with the metals mixtures because enrichments relative to UCC (or a local baseline) may not always be cause for health concerns. The metals mixtures were evaluated in three ways. To evaluate composition, we calculated a normalized sum of ranked concentrations of priority pollutant metals. To estimate intensity, we calculated the sum of enrichments of priority pollutant metals relative to UCC. Finally, to evaluate flux, we calculated the total flux of priority pollutant metals.

For method (1), we assigned a rank (1 through the number of sites) at each site for each priority pollutant, with lower values corresponding to lower priority pollutant concentrations. The rank values for all priority pollutant metals were summed and normalized to the number of priority pollutant metals and sites measured. A rank sum value of 0 indicated that the sample had the lowest concentrations of all priority pollutant metals relative to the other samples in the collection. Conversely, a rank sum value of 1 indicated that the sample had the highest concentrations of all priority pollutant metals relative to the other samples in the collection. This method can detect, within the sample set, whether some sites tend to have higher concentrations of more priority pollutant metals relative to other sites, while addressing issues arising from the wide range of concentrations of different priority pollutant metals. This approach cannot determine whether those concentrations are enriched relative to UCC or any other baseline, or whether the concentrations pose potential health risks.

For method (2), the enrichment values relative to UCC (as calculated in Equation 2) were summed for all priority pollutant metals with values greater than 1. This method combines the effects of enrichment and composition of metals mixtures into a single metric. However, if one or a few priority pollutant metals are orders of magnitude enriched relative to UCC and the other enriched priority pollutant metals, this exercise will only track the results from that (or those) element(s) rather than the mixture as a whole. This was true of the element Zn, which was heavily enriched relative to UCC in all samples and all other priority pollutant metals. Thus, we calculated the sum of enrichments with and without Zn and present the results excluding Zn in the main text.

For method (3), we calculated the daily fluxes of priority pollutant metals $\left(F_x\right)$. The daily dust flux (Equation 1) for particulate matter <63 $\mathrm{\mu }$m $(F)$ was multiplied with the concentration of the metal $\left(C_x\right)$ in the sample (Equation 3).

$$F_x=F\ast C_x$$ (3)

The fluxes for all EPA priority pollutant metals were summed to produce the total average daily flux of priority pollutant metals at each sample site. The strength of this approach is that it allows us to evaluate the total loading of priority pollutant metals to different neighborhoods, where higher levels of total available material for exposure may elevate rates of human exposure to priority pollutant metals.

We assessed the results of the first three methods for elucidating metals mixtures together, using the spatial groupings introduced in Section 2.1.1. The results of the priority pollutant enrichment sum relative to UCC were compared to the total priority pollutant flux using graphical analysis.

2.5.4. Evaluation of Potential Health Risks Associated With Metals Mixtures

There are three primary pathways for metals exposure: inhalation, ingestion, and dermal exposure. These exposures can cause acute or chronic responses, and chronic responses may be cancer‐promoting or affect other systems. Preliminary studies suggest inflammatory effects and oxidative potential from inhalation of Great Salt Lake playa dust (Attah et al., 2024; Cowley et al., 2025). However, because we collected deposited dust, not temporally resolved concentrations of dust and constituents in the atmosphere, using the data to assess the health hazard posed by inhalation requires making numerous assumptions. This includes the concentration of dust in the atmosphere during dust events, rate of inhalation, and dust event duration. However, with this data set, which represents deposited dust, we can assess the non‐carcinogenic health hazard posed by dust ingestion. This is valuable because ingestion of deposited dust, including on food and from hand to mouth ingestion, may represent a persistent exposure pathway. This exposure pathway is expected to be critical to the health of small children who tend to consume more soil and dust relative to their size (Roberts et al., 2009; U.S. Environmental Protection Agency, 2011, 2017). Though they may be present, we did not evaluate carcinogenic health hazards from ingestion of priority pollutant metals due to a paucity of available thresholds for evaluation. For this part of the analysis, the ingestion evaluation was limited to community exposure sites, including at schools or in parks or neighborhoods (refer to Table 1 for sites).

To evaluate the non‐carcinogenic health hazard posed by soil and dust ingestion, we estimated the distribution of potential chronic hazard index ($HI$, Equation 4) using a Monte Carlo methodology. The approach followed the EPA method for estimating the combined effects of metals exposures (Teuschler & Hertzberg, 1995). The approach sums the hazard quotients ($HQ$, Equations 5 and 6) calculated for each element $(x)$ with an established reference dosage $(RfD)$ for a sample $(i)$. The elements we evaluated included most priority pollutant metals as well as additional elements with established $RfD$ values (Table 2). This approach assumes an additive effect from exposure to this mixture of metals, although synergistic or antagonist effects may also be present. Thus, the strength of the approach is that it allows for simultaneous evaluation of multiple metals of concern. The weakness of the approach is that it does not account for interactions among the components of the mixture, and results may depend on the total number of pollutants assessed.

$$H{I}_i=\sum _{x=0}^{n}H{Q}_{i,x}\text{…}H{Q}_{i,n}$$ (4)

Table 2.

Table of Reference Dose $(RfD)$ Values Used to Calculate $HQ$ for Priority Pollutant Metals and Other Metals

Element	RfD (mg ${\text{kg}}^{-1}$ ${\mathrm{d}}^{-1}$)	Value source
As*	0.0003	IRIS
B	0.2	IRIS
Ba	0.2	IRIS
Cd*	0.001	IRIS
Cr (IV)*	0.0009	IRIS
Mn	0.14	IRIS
Mo	0.005	IRIS
Ni*	0.02	IRIS
Sb*	0.0004	IRIS
Se*	0.005	IRIS
U	0.003	IRIS
Zn*	0.3	IRIS
Cu*	0.04	Taylor et al. (2023)
Pb*	0.0035	Gebeyehu and Bayissa (2020)
Tl*	0.000003	IRIS
V	0.007	Gebeyehu and Bayissa (2020)
Co	0.0003	Javed and Usmani (2016)
Fe	0.7	Javed and Usmani (2016)

Note. The source of each $RfD$ is included in the table, where IRIS refers to the Integrated Risk Information System Assessments (U.S. Environmental Protection Agency, 2024). All $RfD$s are for chronic oral exposure for a non‐cancer hazard assessment. All elements with * indicate EPA priority pollutant metals.

Element‐specific $HQ$ are the ratio of estimated average daily doses ($ADD$, mg ${\text{kg}}^{-1}$ ${\mathrm{d}}^{-1}$, Equation 6) to their reference dose ($RfD$, mg ${\text{kg}}^{-1}$ ${\mathrm{d}}^{-1}$, refer to Table 2 for values and data sources) (Agency for Toxic Substances and Disease Registry, 2018). Calculating the $ADD$ requires an elemental concentration ($C$, mg ${\text{kg}}^{-1}$) as well as estimates of ingestion rates ($IR$, mg ${\mathrm{d}}^{-1}$), exposure frequency ($EF$, d ${\mathrm{y}}^{-1}$), exposure duration ($ED$, y), body weight ($BW$, kg) and averaging time ($AT$, y).

$$H{Q}_{i,x}=\frac{AD{D}_{i,x}}{Rf{D}_x}$$ (5)

$$AD{D}_{i,x}=\frac{C_{i,x}\ast IR\ast EF\ast ED\ast 10^{-6}kgm{g}^{-1}}{BW\ast AT\ast 365d{y}^{-1}}$$ (6)

Many studies perform this calculation using single characteristic values of $IR$, $EF$, $ED$, $BW$ and $AT$. A deterministic approach yields a result applicable to the specific modeled scenario, which is typically thought to be broadly representative of a segment of the population (e.g., all adults, all children). However, body size and exposure frequency are critical variables in this assessment and vary across the population. To account for population variability in body size, uncertainty in the exposure frequency and duration and averaging time, we used a Monte Carlo approach to make estimations. That is, estimations were made using a set of 10,000 realizations of age‐weight‐exposure conditions for each of 5 scenarios: birth to 6 months, 6 months to 1 year, 1 year to 6 years, 6–12 years, 12–20 and 20 years and older.

The selected age scenarios correspond to the ages for which the EPA handbook provides estimates of median and ${95}^{th}$ percentile dust and soil ingestion (Table 5‐1, U.S. Environmental Protection Agency, 2017), which were used for the ingestion rate $(IR)$. The ingestion rates we used assume that any soil ingested will have the same composition as the dust analyzed in this study. To obtain the exposure duration and averaging time variables ($ED$ and $AT$), we first randomly selected ages from a uniform distribution constrained by the age boundaries of each scenario. For scenarios where individuals were 6 years old or younger, we assumed that exposure duration and averaging time were equal to the individual age, whereas for individuals older than 6, we assumed that exposure duration could be as low as 5 years, or as long as the individual's age, representing a lifetime of exposure. Averaging time was set to the individuals age. Age, and a randomly assigned sex, were used to estimate body weight $(BW)$, where the age‐weight distributions were sex specific distributions developed by the Centers for Disease Control and Prevention (Fryar et al., 2021). The exposure frequency $(EF)$ is not known and cannot be reliably measured. Exposure frequency may depend on dust event frequency (episodic exposure), or if deposited particulate matter is assumed to persist in the environment, be a function of the frequency of an individual interacting with dust‐bearing surfaces (e.g., hand‐to‐mouth transfer, consumption of dust‐bearing produce, chronic exposure). To account for this knowledge gap, we assigned this variable a uniform distribution, so as to minimize the effect of our assumptions on the results. The bounds of the distribution range from 45 to 350 d ${\mathrm{y}}^{-1}$. The low end of the distribution assumes 15 dust events per year for the area near Great Salt Lake based on 3 to 5 recorded dust storms per year (Steenburgh et al., 2012), with up to 10 additional locally important dust events not captured in the database (Ardon‐Dryer, Clifford, & Hand, 2023; Ardon‐Dryer, Gill, & Tong, 2023). This number assumes that exposures could occur on the day of a dust event and in the two following days. On the high end of the distribution, we assumed that ambient dust flux is an important contributor to total dust flux, and deposited dust persists in the environment, such that exposure could occur nearly daily. Although this may be less likely to occur for adults, it is a reasonable assumption for infants and small children, who spend more time on the ground and whose dust exposure largely occurs from hand‐to‐mouth or mouthing of food and objects that have been in contact with dusty or dirty surfaces. All distribution details and data sources are presented in Table S4 in Supporting Information S1.

These calculations yielded 10,000 estimates of $HI$ for each community exposure site, age scenario, and ingestion rate. The distribution of $HI$ is dependent on assumptions about the distributions of exposure frequency, exposure duration, body weight, and averaging time, rather than the actual distributions of these variables in the communities where samples were collected. Thus, the analysis does not represent an estimate of the actual population distribution of $HI$. Likewise, the estimates presented here are for dust and soil ingestion and thus cannot be considered a comprehensive hazard assessment because drinking water and food may also contribute to chronic loading of priority pollutant and other metals to individuals. Finally, the dust samples may not be representative of the actual compositions of dust and soil ingested by individuals, which may depend on factors such as local soil conditions, within home practices such as smoking, shoe wearing, and legacy pollutants (i.e., lead paint) in homes (Isley et al., 2022).

2.5.5. Assessing Sources of Priority Pollutant Metals

Inference about sources of priority pollutant metals to dust were made using a combination of known priority pollutant sources (refer to Table 2 in Putman et al., 2022), elements known to have high concentrations in the Great Salt Lake playa sediments relative to other regional sediments (including Li, Mg, Sr, Na, Ca, Goodman et al., 2019; Blakowski et al., 2022; Putman et al., 2022; Blakowski et al., 2023), and correlations among elements. Strong associations among elements indicated the elements derived from the same sources. Associations among elements were quantified using Pearson's correlations using the Python pandas corr() function (Pandas development team, 2020), which assumes a linear relation among elements. This approach assumes each element has a single particulate matter source and spatial variability in the element concentration arises from dilution of that source by other particulate matter sources. This assumption holds for some elements, and does not hold for others. Observed lack of linearity in associations among elements can be used to assess sources of priority pollutant metals.

3. Results

3.1. Air Quality and Meteorology During the Sampling Period

During the sampling period, no dust events were recorded by the Storm Events Database (National Oceanic and Atmospheric Administration, 2023a). However, multiple high‐wind events were recorded for northern Utah. The high‐wind events occurred on 15 October 2022, 22 October 2022 and November 1–2, 2022 (Figures 2a and 2b), and were characterized by windspeeds of 25–40 m ${\mathrm{s}}^{-1}$, depending on the location. These values are greater than the values of 8–14 m ${\mathrm{s}}^{-1}$ estimated to initiate saltation (Pelletier, 2006), and thus imply the likelihood of dust emission from the Great Salt Lake playa. During the sampling period, and concurrent with the dates of high‐wind, as well as other dates, the sampler HV (Harrisville, 490571003 U.S. Environmental Protection Agency (2023d)) recorded eight days with 24‐hr average ${\text{PM}}_{10}$ concentrations above the WHO 24‐hr threshold of 45 $\mathrm{\mu }$g ${\mathrm{m}}^{-3}$ (World Health Organization, 2021) and 29 days with a maximum hourly ${\text{PM}}_{10}$ concentration greater than 45 $\mathrm{\mu }$g ${\mathrm{m}}^{-3}$ (Figure 2a). However, the ${90}^{th}$ percentile of measurements was 42 $\mathrm{\mu }$g ${\mathrm{m}}^{-3}$, indicating that fewer than 10% of the measurements over the study period exceeded the WHO threshold. Maximum hourly ${\text{PM}}_{10}$ concentrations also exceeded the National Ambient Air Quality Standard (NAAQS, 40 CFR § 50.6, United States Federal Government, Environmental Protection Agency, Air Programs, 2025) 24 hr threshold of 150 $\mathrm{\mu }$g ${\mathrm{m}}^{-3}$ on 11 days, reaching levels as high as 874 $\mathrm{\mu }$g ${\mathrm{m}}^{-3}$. However, the 24 hr average ${\text{PM}}_{10}$ value did not exceed the NAAQS during the period of record assessed. This demonstrates the episodic and spatially heterogeneous nature of dust events.

Figure 2.

The meteorological and air quality conditions during the study period. (a) ${\text{PM}}_{10}$ (Utah Department of Environmental Quality, Air Quality Division, n.d.) and mean 5‐s wind speed data (National Oceanic and Atmospheric Adminstration, 2023b) from three airport sites that report wind data. (b) Daily average maximum temperature and precipitation amount from data from 11 stations in the study domain (National Oceanic and Atmospheric Adminstration, 2023b). Refer to Table S1 in Supporting Information S1 for weather station information. Wind direction data from the Ogden‐Hinckley Airport was categorized by whether the average ${\text{PM}}_{10}$ concentration on that day was less than (c) or greater than (d) 45 $\mathrm{\mu }$g ${\mathrm{m}}^{-3}$ the WHO 24 hr standard (World Health Organization, 2021). The wind bars, which indicate the number of days, were also binned by wind speed.

Consistent with the characterization of dust events for the region (Hahnenberger & Nicoll, 2012; Steenburgh et al., 2012), higher ${\text{PM}}_{10}$ concentrations tended to occur concurrently with wet or dry frontal passages, before rain events or during higher wind periods (Figures 2a and 2b). The majority of the wind recorded at the Ogden‐Hinckley Airport weather station came from the south (Figure 2c), and the majority of days with maximum hourly ${\text{PM}}_{10}$ > 45 $\mathrm{\mu }$g ${\mathrm{m}}^{-3}$ wind also came from the south (Figure 2d). However, elevated ${\text{PM}}_{10}$ also occurred on days when the wind came from other directions. Every day with a maximum 5‐s wind‐speed greater than 27 m ${\mathrm{s}}^{-1}$ had at least 1 hr with ${\text{PM}}_{10}$ concentrations >45 $\mathrm{\mu }$g ${\mathrm{m}}^{-3}$; however, not all days with a maximum 5‐s wind‐speed greater than 20 m ${\mathrm{s}}^{-1}$ were associated with elevated ${\text{PM}}_{10}$ concentrations. ${\text{PM}}_{10}$ concentrations tended to be higher during drier periods compared with wetter periods, consistent with documented controls on dust generation (Dickey et al., 2023; Merrill, 2023; Pelletier, 2006). Because the high spatial variability in ${\text{PM}}_{10}$ concentrations that characterize dust events and transport are not well captured by a single ${\text{PM}}_{10}$ monitor, we assume that other days in other parts of the study domain may also have been characterized by high ${\text{PM}}_{10}$ concentrations.

Precipitation occurred in the region during the deployment period. Data from the meteorological stations in the region indicated that between 40 and 116 mm of rain (mean: 74 mm and median: 73 mm) fell between 24 August 2022 and 5 November 2022 (National Oceanic and Atmospheric Adminstration (2023b); Figure 2b). The precipitation occurred in five discrete events, and no single event exceeded the sampler capacity (50 mm). About half of the precipitation fell in September and half in late October. This time between events was sufficient for the water collected in the sampler to evaporate, preventing overflow. When we retrieved the samplers, they contained varying amounts of water, likely from the November 1–3 precipitation event.

3.2. Dust Mass Flux and Particle Size

The highest total suspended particulate (TSP) flux occurred at the Farmington Bay playa site (PPM), with flux exceeding 1000 mg m⁻² day⁻¹. PPM is the only site that represents playa emissions as opposed to deposition. The median TSP flux for all sites was 97.7 mg m⁻² day⁻¹, an order of magnitude less than the maximum, indicating the much lower particulate fluxes at other sampling sites. For dust (grain sizes <63 $\mathrm{\mu }$m), the highest flux (976 mg m⁻² day⁻¹) also occurred at PPM (Figure S1 in Supporting Information S1). Similar to the patterns for TSP flux, the median dust flux was an order of magnitude lower (68.6 mg m⁻² day⁻¹). In general, largest TSP fluxes were found at playa group (Farmington or other) sites on the eastern edge of the Lake and lowest TSP fluxes were observed in and to the east of municipalities. The lowest TSP flux (44.6 mg m⁻² day⁻¹) and dust flux (28 mg m⁻² day⁻¹) occurred at the HyS site near Promontory Point, on the eastern shore of the north arm of Great Salt Lake.

Most sites exhibited high dust to TSP ratios (median = 0.88, Figure S2 in Supporting Information S1). The ratios of dust to TSP in deposited material were close to uniform among sites not on or near the playa. Among the sites, LNP had the highest dust to TSP ratio (0.9). A small subset of samplers indicated lower dust to TSP ratios (0.258–0.628). The sites with lower dust to TSP ratios include HyL on the west side of Great Salt Lake (ratio = 0.258), and HyS on the east shore of the north arm of Great Salt Lake (ratio = 0.628) and OBW north of Farmington Bay (ratio = 0.603).

3.3. ⁸⁷Sr/⁸⁶Sr

In this study, the highest, median, and lowest observed Sr isotope ratios of dust were 0.71601 (OBW), 0.71251 (SCW), and 0.71050 (UGF), respectively. Maximum, median, and minimum ⁸⁷Sr/⁸⁶Sr for the sediment samples were 0.71624 (DAVIS PL), 0.71273 (BRB PL CN), and 0.70954 (UGF), respectively. Great Salt Lake playa dust and sediment samples exhibited the highest ⁸⁷Sr/⁸⁶Sr (0.7126–0.7162). Dust samples collected further from, or in areas that are less frequently downwind of the Great Salt Lake playa, tended to have lower ⁸⁷Sr/⁸⁶Sr.

3.4. Elemental Patterns

Spatial patterns in element concentrations were assessed using Principal Component Analysis (PCA, Figure 3). The first principal component (PC) explained 51.2% of the variance in the dust data set (n = 17). In this axis of variability, separation was greatest between Great Salt Lake playa sediments (e.g., Mg, Ca, Li, Na, Sr, As, Se, Cd, U, Cs) and elements (Al, rare earth elements (REEs)) that are more abundant in the sediment from agricultural sites relative to those from the playa. Samples taken from on or near the playa, particularly in the GSL playa (Farmington) group had the lowest PC1 values (PPM, GSP, OBW). Samples in the North and South groups had the highest PC1 values (HTR, UST, BJH, UGF). The second PC explained 13.4% of the variance in the data set and was characterized by differences between samples from the South group (HTR, BJH, LNP) and samples from the north (UGF, LIN, SCW). The third PC explained 9.1% of the variance in the data set and reflected the variance arising from elevated Tl concentrations observed at OBW (not shown). Additional PCs represent other site‐ and element‐specific combinations, without broadly generalizable patterns. Elemental concentrations of priority pollutant metals exhibited complexity in their associations with one another. In the first two PCs, the EPA priority pollutant metals did not group closely together, with the exception of Cd, Se, and As.

Figure 3.

A principal component analysis of the geochemistry of dust samples, with dots colored by ⁸⁷Sr/⁸⁶Sr. The first principal component separates playa‐associated elements from elements associated with soils. The second principal component separates the data set by the geochemistry distinct to the south and north groups. Clusters of elements in the lower panel are exploded for readability. The first two PCs explain nearly 65% of the variance in the data set.

3.5. Element Enrichment Relative to UCC and Great Salt Lake Playa Sediments

All sites, including sediment collection sites, were enriched in all priority pollutant metals relative to UCC (Figures 4a and S3 in Supporting Information S1), with Pb, As, Se, Cu, Cd and Zn exhibiting the largest enrichments. GSL playa (Farmington) tended to be more enriched relative to UCC than most other groups, including the other playa sites. Among the non‐playa groups, the South group tended to be more enriched relative to UCC than the North group, though there was often overlap among the site enrichment distributions, and the rank of the group mean depended on the element. All groups of dust samples tended to be more enriched in priority pollutant metals, sometimes by orders of magnitude, relative to sediment samples (Figure S3 in Supporting Information S1).

Figure 4.

Enrichment of elements in dust samples, grouped by geography, relative to (a) average concentrations found in upper continental crust (UCC) and (b) an estimate of Great Salt Lake playa sediment composition. Values greater than 1 indicate that the element is more concentrated in a dust sample than would be expected in dust derived exclusively from weathering of UCC or delivery from Great Salt Lake sediments. Note that the enrichments relative to UCC are large enough that the y axis of (a) is on a log scale whereas the y axis of (b) is not. The order of elements in the two panels differs, reflecting the approximate order of elemental enrichment. EPA priority pollutant metals and other elements with reference doses ($RfD$s, Table 2) are indicated with shading.

Samples were also evaluated for their enrichment relative to the average elemental concentrations found in the Great Salt Lake playa sediments (Figure 4(b)). Assessing elemental enrichment relative to Great Salt Lake playa sediments for particulate matter deposited at sites further from the playa identifies elements that are likely to be Great Salt Lake playa‐derived and those deriving from other sources. For sites on or near the Great Salt Lake playa, this analysis highlights the spatial variability in Great Salt Lake playa sediment composition. The GSL playa (Farmington) samples were enriched in Sr, Cs, Ca, Ba, Cd, Mn, Tl, some REEs, Pb, Cu, and Zn relative to mean Great Salt Lake playa conditions, whereas the GSL playa (other) samples were enriched in REEs, Cu, and Zn relative to mean Great Salt Lake playa conditions. Among non‐playa group samples, most samples were enriched in REEs, Tl, Co, Pb, Cu, and Zn. Among these, the South group samples were the most enriched (both in number of elements and values of enrichments) relative to the Great Salt Lake playa, whereas the Central and North group samples were less frequently enriched, and enrichments were lower relative to the Great Salt Lake playa geochemical reference. Many elements that exhibited enrichments relative to UCC were depleted relative to the average concentrations in Great Salt Lake playa sediments.

3.6. Quantification of Metals Mixtures

Among all measured EPA priority pollutant metals (As, Be, Cd, Cr, Cu, Pb, Ni, Sb, Se, Tl, and Zn), all dust samples exhibited concentration enrichment relative to UCC. The spatial structure in the composition of the metals mixtures was assessed using ranked sum and sum of enrichments approaches. With the sum of enrichments approach, the HyS site exhibited the highest value (Figure S4, Table S5 in Supporting Information S1). This result arose because of the large Zn enrichment relative to UCC in the sample, relative to other sites. When Zn was excluded from the calculation (Figure 5a, Table S5 in Supporting Information S1), sites in the South group had the highest values, and sites in the North group had the lowest values. The normalized ranked sum approach exhibited similar spatial patterns to the enrichment sum approach (Figure S5, Table S5 in Supporting Information S1), confirming that the patterns in enrichment sums arose from concentrations of multiple metals. The South group sites tended to have the highest normalized rank sum scores (LNP, HTR), indicating that EPA priority pollutant metals exhibited higher enrichments relative to UCC. Sites in the North group had the lowest normalized rank sum scores (BRM, UGF), indicating that EPA priority pollutant metals exhibited lower enrichment relative to UCC relative to samples collected in other parts of the study area.

Figure 5.

The spatial structure of results of (a) sum of priority pollutant enrichments relative to upper continental crust (UCC) and (b) total priority pollutant fluxes. For panel (a), the sum of priority pollutant enrichments includes all priority pollutant metals measured except Zn. Sums can be high either due to a large enrichment in one metal, or to lower enrichments of a variety of metals, however the results from the normalized rank sum calculation (Figure S5, Table S5 in Supporting Information S1) indicate that these enrichments arise from combinations of multiple metals. For panel (b), where priority pollutant fluxes are higher, they tend to come from Farmington Bay (Figure S6 in Supporting Information S1). Satellite imagery courtesy of the NASA Goddard Space Flight Center and U.S. Geological Survey (Earth Resources Observation and Science (EROS) Center, 2020) and TerraMetrics (2023). Imagery fetched from Google via cartopy.io.img_tiles(). Major roads fetched from Utah Geospatial Resources Center (2022).

This spatial pattern shifted when dust flux was considered (Figure 5b, Table S5 in Supporting Information S1). The highest total priority pollutant fluxes tended to occur at sites on or near the playa that had high dust fluxes (, those northeast (roughly downwind) of Farmington Bay and the PPM site, including GSP and SCH), relative to other playa‐proximal samples. Sites with the lowest total fluxes of EPA priority pollutant metals were proximal to other parts of the playa (HyL, BRM) or situated in the foothills on the east side of the municipalities (HTR, SCW). The relationship between dust flux and priority pollutant flux highlights the potential for Farmington Bay to contribute priority pollutant–containing dust to community exposure sites, as well as the role of other particulate matter sources in controlling priority pollutant fluxes. For a transect of sites proximal to or likely to receive Farmington Bay dust, we found that dust flux was a significant predictor of total priority pollutant flux (p < 0.001) and explained 99% of the variance in the total priority pollutant flux (Figure S6 in Supporting Information S1). The prediction coefficient was 0.40, suggesting that for every 1 mg of Farmington Bay dust deposited at a site, about 0.4 $\mathrm{\mu }$g of priority pollutant metals would also be deposited. The intercept of 112 $\mathrm{\mu }$g ${\text{mg}}^{-1}$ suggests other sources of priority pollutants contribute to fluxes at community exposure sites. At other sites, total dust flux was not a predictor of priority pollutant flux. Instead, priority pollutant flux likely indicates anthropogenic contributions of priority pollutants unique to the site.

When the priority pollutant enrichment sum (excluding Zn) was compared to the total priority pollutant flux (Figure 6), high enrichment sums arose at both low and high dust and metals flux sites. Low metals flux sites with high enrichment sums were characteristic of the South group, whereas high flux and high enrichment sums were characteristic of sites composing the Central and GSL playa (Farmington) groups. Both groups exhibited similar priority pollutant compositions ($\sim$75% Zn, 10%–15% Cu, and 8%–10% Pb, (Figures S7a and S7b in Supporting Information S1). However, there were some differences among the priority pollutant composition of the two groups. The higher total priority pollutant flux data from the Central group tended to exhibit higher percentages of Pb, As, Cr, Zn, and Cd relative to the data from the South group (Figure S7c in Supporting Information S1). The lower priority pollutant flux data from the South group exhibited higher percentages of Cu and Ni relative to Central group (Figure S7c in Supporting Information S1).

Figure 6.

The association between priority pollutant enrichment sums and total priority pollutant fluxes reveals two modes of priority pollutant metal delivery to communities. The sites associated with high flux and high enrichment sums tend to be from either the Great Salt Lake (Farmington) or Central groups. These samples have higher ⁸⁷Sr/⁸⁶Sr, characteristic of Great Salt Lake playa sediments. The sites with low flux and high enrichment sums tend to have lower ⁸⁷Sr/⁸⁶Sr, characteristic of other local and regional dust sources, and most are part of the South Group. The North group and the Great Salt Lake (other) groups tend to be lower in both priority pollutant enrichment sums and flux.

3.7. Evaluation of Hazard Quotients and Indices

Across all community exposure sites (BJH, DTC, SCH, ONC, PWS, UST, LIN) and ingestion rates (${50}^{th}$ and ${95}^{th}$ percentiles), $HI$ tended to be lower and less variable for older age groups compared to younger age groups (Figure 7). For the 6 to 12, 12–20 years, and 20–80 years age scenarios, assuming a median ingestion rate, no realizations had $HI$ > 1. The same was true for the two oldest age groups when we used the ${95}^{th}$ percentile ingestion rate. Fewer than 4% of realizations exceeded the $HI$ threshold of 1 for the three youngest age scenarios using median dust and soil ingestion rates, with the greatest percentages associated with the 6 months to 1 year scenario. However, when the ${95}^{th}$ percentile dust and soil ingestion rates were used, between 31% and 76% of the realizations in the two youngest scenarios had $HI$ > 1, depending on the site and scenario. For the 1–6 year old scenario, as many as nearly half (8%–44%) of the realizations indicated $HI$ > 1, depending on the exposure site.

Figure 7.

$HI$ calculated for dust deposited at community exposure sites (BJH, DTC, SCH, ONC, PWS, UST, LIN). $HI$ of 1 is indicated by a horizontal gray line. For scenarios where $HI$ > 1, ingestion of dust and soil represents a health hazard. The data are displayed as letter‐value plots (Hofmann et al., 2017), where the central line is the data median, the innermost box contains 50% of the data, and the remaining boxes each contain 50% of the remaining data (and thus a diminishing proportion of the total data, i.e., 25%, 12.5%, 6.25%, etc.). The black circles indicate outliers. Results from the sites are grouped because the distributions of $HI$ among groups were not significantly different from one another. The distributions split by site are included in Figure S8 in Supporting Information S1.

No site was statistically distinguishable from another site in terms of the distribution of $HI$ calculated for a specific scenario and ingestion rate (Figure S8 in Supporting Information S1). However, among the community exposure sites and for both ingestion rates, the highest $HI$ values were associated with the BJH and PWS sites, and the lowest were associated with the UST and LIN sites. Spatial variability in $HI$ values tracks spatial patterns in the composition and flux of metrics of priority pollutant metals (Figure 5). Across all sites, the highest elemental contributors to the $HI$ (in descending order) are Tl, As, Pb, Co, Cr, Fe, Zn, Mn, and Cu (Figure 8, shown for the 0–6 months age group, though relative magnitudes of HQ are constant for all age groups). Although they were assessed, the elements V, Ba, Cd, Ni, Sb, U, B, Mo, and Se did not contribute notably to the non‐cancer $HI$.

Figure 8.

The distributions, shown as boxplots, of element‐specific $HQ$ calculated for median dust ingestion for the birth–6 months age scenario. Results from all community exposure sites are combined as they were not significantly different from one another. Whereas the magnitude of the $HQ$ values change with both ingestion rate and age scenario, the relative contributions of each element to $HI$ values do not change.

4. Discussion

Our study occurred during a period without dust storms or dust events. However, based on the assessment of meteorology and the dust fluxes delivered to samplers, dust transport occurred during the period of study. Dust and total particulate matter flux were largest for communities closest to the playa. That we collected particulate matter at all during the study period emphasizes the potentially important contribution of small‐to medium‐scale dust events with the majority of the particulate matter deriving from proximal sources. These events may be especially important for communities living close to possible dust sources. Because of their smaller spatial extent and duration, these smaller scale dust events are likely missed by the current network of ${\text{PM}}_{10}$ monitors (Figure 1, Ardon‐Dryer, Clifford, & Hand, 2023; Ardon‐Dryer, Gill, & Tong, 2023), leaving Great Salt Lake playa‐proximal communities unmonitored.

Principal component analysis informed the interpretation of spatial patterns in dust geochemistry. Similar to regional analyses (e.g., Mangum et al., 2024), the first principal component separated playa‐proximal sites from the sites located within and to the east of communities. The second PC followed south to north gradients in trace element composition. Based on multiple criteria, dust samples in the Central and South groups tended to exhibit more complex mixtures of priority pollutant metals than dust samples from the North group. Samples collected on and northeast of Farmington Bay (the GSL playa (Farmington) and Central groups) exhibited the highest total priority pollutant fluxes. These patterns in composition and flux were attributed to spatially variable contributions of two sources. The sources include (a) emissions from the polluted sediments of Farmington Bay (Jung et al., 2024; Lopez et al., 2024; Perry et al., 2019; Wurtsbaugh et al., 2020) and (b) direct (wet or dry) deposition of particulate matter bearing priority pollutant metals from anthropogenic activity (Goodman et al., 2019; Munroe et al., 2025; Putman et al., 2022). Due to the low concentrations of priority pollutant metals in non‐playa sediments, agricultural or former agricultural areas, and the Lake Bonneville paleoshoreline sediments are unlikely to be primary contributors of priority pollutant metals to dust (Figure S3 in Supporting Information S1), although these areas may contribute to total dust fluxes. Although other regional sources are known to contribute particulate matter to dust events (Carling et al., 2020; Munroe et al., 2025), regional sources were not considered to be important contributors to dust flux for these samples due to the absence of dust events during the study period.

4.1. Tracing Sources of Priority Pollutant Metals to Dust

We evaluated the sources of priority pollutant metals in dust using sediment geochemical information, spatial patterns, and previously established associations among metals. Specific priority pollutant metals were associated with either dust originating from Farmington Bay or particulate matter contributed by anthropogenic activities. Although other regional dust sources may have contributed to dust and metals fluxes, no dust events were recorded during this sampling period, so contributions of further sources are expected to be minimal.

4.1.1. Sites Downwind of Farmington Bay Received Polluted Great Salt Lake Playa Dust

The priority pollutant metals As, Cd, Se, Sb, and Pb are present in Great Salt Lake playa sediments at concentrations greater than UCC (Perry et al., 2019; Putman et al., 2022). Arsenic and Se are the elements most homogenously distributed around the Lake (Perry et al., 2019). High concentrations of Pb, Sb, and Cd co‐occur on the eastern shore of Farmington Bay. The spatial variability in Great Salt Lake playa geochemical composition is evident when comparing the GSL playa (Farmington) and the GSL playa (other) group to the Great Salt Lake geochemical reference values. The GSL playa (Farmington) group tended to exhibit higher concentrations of most priority pollutant metals than the GSL playa (other) group irrespective of the baseline for enrichment calculation (Figure 4). Likewise, both groups indicated enrichment or depletion of specific elements relative to the Great Salt Lake geochemical reference. In the absence of spatial variability, all Great Salt Lake derived sediments would exactly match the Great Salt Lake geochemical reference values.

In relatively high flux sites downwind of Farmington Bay (GSP, OBW, SCH and ONC, Figure 6) As, Cd, and Se were more enriched relative to UCC, but with no or little enrichment relative to Great Salt Lake playa sediments (Figure 4). Arsenic was well correlated with Sr (0.79) and ⁸⁷Sr/⁸⁶Sr (0.49). Consistent with Munroe et al. (2025), the observed element associations suggests that atmospheric loading of priority pollutant metals to the high flux sites downwind of Farmington Bay is largely controlled by dust transport from Farmington Bay. While the inferred source is likely for playa‐proximal GSP and OBW, Farmington Bay is also an inferred dust source for SCH and ONC. SCH and ONC are two community exposure sites. This finding implies that dust and priority pollutant metals are emitted from Farmington Bay and deposited to communities, even during periods without major dust events.

4.1.2. Sites in the Southern Part of the Study Domain Are Affected by Local Industrial Pollution

Metal types and amounts present in the South group samples likely arose from contributions of priority pollutant metals from industrial activity. Modern activities include copper mining and processing (Fitzpayne et al., 2018), oil refining, chroming, mineral production, and magnesium processing (U.S. Environmental Protection Agency, 2020; Womack et al., 2023). Industries in the region that have since been shuttered include trash incineration (Maffly, 2022) as well as lead‐arsenic and uranium processing (Hughes, 1990). Cleanup or maintenance of other legacy industrial activities remains ongoing, include activities associated with the Ogden Defense Depot Superfund site (Laplante, 2011; U.S. Environmental Protection Agency, 2023f) and the U.S. Magnesium Superfund site (U.S. Environmental Protection Agency, 2023g).

High La concentrations have previously been observed in dust collected in the area represented by the South group, and were inferred to arise from particulate matter produced by oil refining (Goodman et al., 2019; Putman et al., 2022). While La is not a priority pollutant, crude oil is known to contain heavy metals, including Cu, Cd, Ni and Pb (Ajeel et al., 2021; Van Pelt et al., 2020). These elements are all (except Ni) found at levels higher than expected from UCC in dust in this study in this area, consistent with prior dust sampling efforts (Munroe et al., 2025; Putman et al., 2022).

The high concentrations of Cu in the region likely arise from Cu mining activities. The primary source of Cu in the region is the Bingham Canyon Mine, which is mined by Kennecott Utah Copper LLC. Kennecott Utah Copper also operates a concentrator, smelter and refinery and tailings storage facility just north of Bingham Canyon (Figure 1). Copper is correlated with Mo (correlation coefficient = 0.94), and somewhat with Tl, when we account for the high Tl concentration at OBW (Figure S9 in Supporting Information S1, correlation coefficient = 0.51). This association is consistent with presence of Mo in Cu‐porphyry deposits in Bingham Canyon and elsewhere (Fitzpayne et al., 2018; Sinclair & Goodfellow, 2007), as well as the production of Mo associated with Cu mining (Nikonow & Rammlmair, 2022). The correlation is consistent with the minerals reported from Kennecott Utah Copper at the Bingham Canyon Mine (Krahulec, 2018), dust geochemical patterns in the region (Putman et al., 2022), and evaluations of heavy metals in the topsoils of parks downwind of the mine (Lee et al., 2024).

4.2. Non‐Source Specific Priority Pollutant Metals

A few elements, including Zn and Pb, did not exhibit clear spatial patterns but are nearly ubiquitously enriched relative to UCC. These important priority pollutant metals were reported to be enriched in other studies (Goodman et al., 2019; Munroe et al., 2025). We present interpretations of their possible sources.

4.2.1. Zinc Is Ubiquitous in the Study Region

Concentrations of Zn in dust samples were poorly correlated with most metals (Figure S9 in Supporting Information S1). The lack of association between Zn and other elements may be due to the presence of very high concentrations of Zn in dust at HyS, which is proximal to the Promontory Zinc Co. mine (Krahulec, 2018). Although the mine isn't presently active, Zn in dust may be sourced from proximal tailings piles or sediments associated with open‐pit mining.

When the sample from HyS was omitted from the correlation calculations, Zn's strongest positive correlations were with Ni and Mn (Pearson correlation coefficients of 0.58 and 0.56, respectively). Zinc was negatively correlated with elements associated with the playa (correlation coefficients between −0.25 and −0.39 for Mg, Ca, Sr, Na and Li). These correlations, in combination with the ubiquitous enrichment of Zn relative to the Great Salt Lake playa baseline (Figure 4), suggests that nearly all the Zn present in dust comes from non‐playa sources. Non‐mining or smelting sources of Zn, Ni, and Mn to the environment include metallurgy, vehicle emissions, and landfills (Putman et al., 2022; Table 2). Due to the ubiquitous high concentrations of Zn and higher concentrations near areas of high‐intensity development, the likely sources of Zn to the environment are fossil fuel combustion and (or) automobile brake or tire wear (O'Loughlin et al., 2023; Pavilonis et al., 2022).

4.2.2. Lead Comes From Both the Great Salt Lake Playa and Other Industrial Sources

Lead is an element of concern for human health and especially child development. In these data, Pb doesn't come from a single source. Scatterplots of Sr concentrations and ⁸⁷Sr/⁸⁶Sr against Pb (Figure S10 in Supporting Information S1) suggest at least two modes of delivery of the elements to dust, including transport from Farmington Bay (e.g., PPM, GSP, ONC, OBW), and a current or legacy anthropogenic source of Pb to dust from within the developed area (e.g., LNP, BJH, HTR). The inference that multiple sources contribute Pb to dust is supported by the modest enrichment of Pb in dust relative to average concentrations in the Great Salt Lake playa in some samples (Figure 4).

The highest concentration of Pb occurred at LNP, with BJH and HTR also exhibiting high concentrations. Putman et al. (2022) also observed high concentrations of Pb in the area near downtown Salt Lake City and the City of North Salt Lake. In the South group, most Pb is unlikely to have been derived from playa sediments (Figure S9 in Supporting Information S1). With the current data, we cannot pinpoint a specific Pb source considering that sources of Pb include mining and smelting, metallurgy, fossil fuel combustion, plumbing, battery manufacture, sewage sludge, landfills, waste incineration, pesticides, and paint industry (Putman et al., 2022; Table 2), as well as emissions from small planes (U.S. Environmental Protection Agency, 2023c). The sites where we observed the highest Pb concentrations are the closest to the Salt Lake City International Airport and the Skypark Airport. There are numerous previously active Pb‐producing mines in the region, including Lakeside, Promontory, Argenta, and Paradise (Krahulec, 2018). The Salt Lake Valley was also home to numerous smelters throughout the 19th and mid‐20th centuries who emitted Pb, As and other pollutants (Mitchell & Zajchowski, 2022). Finally, this area is among the older developed areas in the region and may still yield Pb from particulates associated with legacy fossil fuel combustion.

4.3. Implications for Human Health

Exposure to high concentrations of particulate matter negatively impacts human and ecosystem health (Ardon‐Dryer et al., 2020; Goudie, 2014; Griffin & Kellogg, 2004; Jones & Fleck, 2020; Kim et al., 2015). Metals in particulate matter enhance the risks associated with particulate matter exposure. For example, Great Salt Lake playa sediments have been shown in laboratory experiments to exhibit greater potential to produce a respiratory response than dust from other sources (Attah et al., 2024).

The spread in non‐cancer oral ingestion $HI$ results for scenarios with children younger than 6 suggests that interactions between high ingestion rates, smaller body size, and/or more frequent exposure may lead to $HI$ > 1 for metals through dust and soil ingestion. With the distributions of $BW$ and $EF$ that we modeled, $EF$ tends to be a better protective factor than $BW$ (Figures S11–S13 in Supporting Information S1). High enough $EF$ at any $BW$ and ${95}^{th}$ percentile $IR$ can lead to cases of $HI$ > 1, whereas for any $BW$, low enough $EF$ (<90 days), even with ${95}^{th}$ percentile $IR$ do not yield cases of $HI$ > 1. Our findings suggest that if $EF$ is episodic, like ingestion during a dust event or in the days after, $HI$ is unlikely to exceed 1 for individuals of any $BW$ and $IR$. However, if dust deposited from dust events persists in the environment and is regularly ingested (90 or more days a year, depending on the $BW$ and $IR$), certain individuals, particularly smaller individuals or those who ingest more dust and soil, may be at greater risk for having $HI$ > 1.

The finding that children under 6 are more vulnerable to possible $HI$ > 1 is consistent with another assessment of the vulnerability of small children and infants to dust‐mediated pollution (Roberts et al., 2009). Among the metals contributing most to the $HI$ were Tl, As, Pb, Co, Cu and Cr (Figure 8). Ingestion of Pb, As, Cd, Cr and Cu, especially where they occur together, has been demonstrated to correlate with the presence of behavioral issues and/or adverse neurodevelopmental consequences (Rodríguez‐Barranco et al., 2013; Stein et al., 2022).

Approaches for reducing children's dust and soil ingestion focus on in‐home and hygiene‐based interventions. Recommendations include using vacuum cleaners with a dust finder indicator and HEPA filtration (Yiin et al., 2008), high‐quality door mats, and HEPA air filters (Roberts et al., 2009). Other approaches include removing shoes while indoors, properly washing produce and toys, and regular hand‐washing, especially after outdoor play and before meals (U.S. Environmental Protection Agency, n.d.). Even if deposited dust persists in the environment, these practices may reduce exposure frequency and ingestion rates.

There may also be land and water management approaches for reducing likelihood of high $HI$ from dust and soil ingestion. In this data set, a large proportion of the As and Cd fluxes, and some proportion of the Pb load in specific areas, are likely to come from Great Salt Lake playa sediments. Thus, reductions in dust from the Great Salt Lake playa sources from dust mitigation (e.g., Owens Lake Scientific Advisory Panel et al., 2020) or increases in Lake levels (Great Salt Lake Strike Team, 2023) would reduce the probability, and likely frequency of exposure to these specific metals. Among the groups assessed, the effect is likely to be highest for the Central group. Conversely, increased dustiness, as could occur in a future with lower Lake levels (Great Salt Lake Strike Team, 2023; Grineski et al., 2024), could cause increased access to and direct ingestion of dust.

Our analysis assumes that the ingested dust and soil matches the composition of the deposited dust sampled (Figure 7). There are a couple reasons why this assumption may not be accurate. First, the synoptic design of this study is a major source of uncertainty; we could not assess temporal variability in dust fluxes or composition. The risk of realizing the hazard could be modulated by temporal variability in dust composition or dust flux, where individuals living in areas with more frequent or higher flux dust events may be more likely to have $HI$ > 1. Second, ingested dust and soil is usually a mixture of dust derived from local soils and deposited dust, and ingestion occurs both indoors and outdoors. A large proportion of indoor dust comes from outdoor dust and nearby soil tracked indoors on shoes (Hunt et al., 2006; Ibañez‐Del Rivero et al., 2023). Mixing deposited dust with soil may decrease the concentrations of priority pollutant metals in the ingested media. However, if delivery of particulate matter from pollution occurs nearly continuously to the surrounding area, the soil composition may resemble the dust composition (Van Pelt et al., 2020). This may be the case for some metals that are likely to be contributed by direct deposition of industrial particulate matter (e.g., Lee et al., 2024). Thus, the hazard index analysis may indicate a greater hazard than is likely to be realized.

Dust and soil ingestion constitutes one part of an individual's total $HI$. Ingestion may also occur through consumption of contaminated vegetables (Shahid et al., 2017), drinking water, and/or processed foods (e.g., dark chocolate, Hands et al., 2024). Further, the total $HI$ for an individual includes exposure to pollutants through inhalation and dermal exposure.

4.4. Influence of Great Salt Lake Water Levels on Priority Pollutant Metals in Dust

During the collection, period the Great Salt Lake elevation was at its lowest recorded level to date (less than 1276.8 m (4189 ft) at U.S. Geological Survey station 10010000 GREAT SALT LAKE AT SALTAIR BOAT HARBOR, UT, U.S. Geological Survey, 2023). Given the historic nature of the Lake level during the collection period and the few proximal ${\text{PM}}_{10}$ monitors, we cannot evaluate the representativeness of this data set relative to long‐term conditions. Because no dust events were recorded during the period of sampling, this assessment represents a conservative estimate of priority pollutant exposure via dust relative to the past. Future dust events occurring during similar or lower Lake levels and lower soil moistures may contribute greater dust fluxes to surrounding communities (Grineski et al., 2024; Ishizuka et al., 2008).

The spatial patterns in flux and the association among priority pollutant metals flux, enrichment sums, and priority pollutant metals composition suggest two sources of priority pollutant metals in dust. The first, associated with higher fluxes and higher enrichments of As, Se and Cd relative to UCC, is dust transport from Farmington Bay. During this period, dust from this source primarily affects Great Salt Lake playa‐proximal sites in the central part of the study area (GSP, OBW, SCH, ONC), although different wind conditions would mean other areas would be downwind (Grineski et al., 2024) and may reach the ecosystems of the Wasatch Range (Lang et al., 2023; Munroe et al., 2025). The second, associated with lower total fluxes and higher enrichments of Cu relative to UCC, is the deposition of particulate matter from industrial activity. This pathway affects sites in the southern part of the study area (LNP, BJH, HTR and DTC). Sites in the northern part of the study area (e.g., PWS, SCW, BRM, UST) and collections near the north arm of Great Salt Lake (e.g., HyL, HyS) may also receive Great Salt Lake playa dust. However, due to the spatial distribution of priority pollutant metals in Great Salt Lake playa sediments (different enrichments of playa groups in Figure 4 and Perry et al., 2019), the dust transported to these places from Great Salt Lake playa may contain lower concentrations of priority pollutant metals.

The total fluxes and concentrations of specific elements, like As, Pb, and Cd in dust would likely be lower if dust emission from the Great Salt Lake playa, and particularly Farmington Bay, were decreased or fully mitigated. This dust source area could be managed by increases in Lake levels or dust mitigation measures (Great Salt Lake Strike Team, 2023). Management of Lake levels or implementation of dust control measures would not completely eliminate ${\text{PM}}_{10}$ or occurrences of elevated ${\text{PM}}_{10}$ from the atmosphere, nor would these management practices prevent all priority pollutant delivery to communities. Thus, assessment of the efficacy of playa dust mitigation would require intensive spatio‐temporal ${\text{PM}}_{10}$ monitoring.

5. Conclusions

We collected dust across northern Utah during the fall of 2022. These samples were used to evaluate spatial patterns in priority pollutant metals in Great Salt Lake playa‐proximal communities that are not currently monitored for ${\text{PM}}_{10}$ nor have been previously studied. We observed spatial variability in total dust deposition and the composition, intensity, and flux of priority pollutant metals. More priority pollutant metals occurred in the samples collected at the southernmost sites. The greatest priority pollutant flux occurred at the sites in the central part of the study area. Fluxes and geochemistry of the dust deposited in these two regions suggested two mechanisms for delivery of priority pollutant metals in particulate matter to communities: contributions from industrial activity and in dust eroded from the Great Salt Lake playa. Ingesting metal‐laden dust and soil may constitute a health concern for small children with higher ingestion rates and chronic exposure, with smaller children more vulnerable than larger children in the same age cohort.

Our findings suggest the importance of the Great Salt Lake playa in contributing dust and certain priority pollutant metals to communities in northern Utah. Even in the absence of major dust events, playa sources still contributed to deposited dust. Small and medium‐scale dust events may not be captured by the current ${\text{PM}}_{10}$ monitoring network. Increases in Lake levels and/or implementation of dust control measures may decrease the fluxes of an important subset of priority pollutant metals (As, Cd, Se, and to a lesser degree Pb). However, industrial sources of priority pollutant metals in particulate matter remain a concern for particulate matter‐mediated metals exposure. Further evaluation of the variability and conditions controlling the relative contributions of these two main sources of particulate matter to communities requires continuous spatiotemporally intensive sampling.

Finally, ingestion of priority‐pollutant laden particulate matter is only one pathway for priority pollutant exposure, and only one component of the set of air and environmental quality issues faced by communities living along the Wasatch Front. Further investigation of the priority pollutant exposure during inhalation of ${\text{PM}}_{10}$ and ${\text{PM}}_{2.5}$ and from dermal exposure could help build a more comprehensive estimate of priority pollutant exposure through airborne particulate matter in the region.

Conflict of Interest

The authors declare no conflicts of interest relevant to this study.

Supporting information

Supporting Information S1

Putman, A. L. , Blakowski, M. , DiViesti, D. , Fernandez, D. , McDonnell, M. , Longley, P. , & Jones, D. K. (2025). Contributions of Great Salt Lake playa‐ and industrially sourced priority pollutant metals in dust contribute to possible health hazards in the communities of northern Utah. GeoHealth, 9, e2025GH001462. 10.1029/2025GH001462

Data Availability Statement

Data are publicly available on ScienceBase as a US Geological Survey data release (Blakowski et al., 2023) accessible at https://doi.org/10.5066/P9YF6Z4E.