Air Quality, Atmosphere & Health (2024) 17:2445–2459

https://doi.org/10.1007/s11869-024-01581-8

Spatial variability of dust concentration and deposition around an

industrial port in South Africa emphasises the complexity of sources
and transport
Heleen C. Vos1,2 · Kaukurauee I. Kangueehi1 · René Toesie3 · Frank D. Eckardt4 · Grant Ravenscroft5 · Susanne Fietz1

Received: 31 December 2023 / Accepted: 30 April 2024 / Published online: 14 May 2024
© The Author(s) 2024

Abstract
The port and industrial zone of Saldanha Bay in South Africa accommodates activities related to the transport, processing,
and production of commodities such as iron ore, manganese ore, and steel. The visible emission of dust from this area
raised concerns for public health and to address this, the municipality has monitored the fine particulate matter (PM2.5)
concentration and dust deposition since 2015. Here, this monitoring data served to assess spatial and temporal changes
and to evaluate the potential contribution of industrial and meteorological processes to these changes. We observed high
temporal variability in both PM2.5 concentration and dust deposition, and high spatial variation in dust depositions. Dust
originated from local sources such as industry and traffic, but industrial activities could not explain the observed spatial
variability, and concentration and deposition fluxes did not significantly increase over the years despite the extension of
industrial activities. Meteorological factors such as rain, wind speed, wind direction, as well as topography exerted an
important influence, but could also only partially explain the observed variability in both dust concentration and deposi-
tion. Furthermore, the PM2.5 concentration and dust deposition are not significantly correlated, which highlights the chal-
lenges in appropriate dust monitoring. It follows that such monitoring efforts, though meeting national standards, require
improvement to assess risks accurately. Our study illustrates that in areas with such high complexity of industrial activities,
the high variability of dust load and deposition must be considered to evaluate implications for public and environmental
health, adherence to guidelines, and mitigation strategies.

Keywords Air pollution · Dust deposition · PM2.5 concentration · Public health · Mining

Introduction

Dust emission, transport, and deposition play several criti-

cal roles, affecting land degradation, climate, the environ-
ment and public health. For example, the removal of fine
particles can deplete emitting surfaces in nutrients, clay,
Heleen C. Vos and silt (Bridges and Oldeman 1999). Furthermore, through
[email protected] cloud condensation and direct radiative forcing, suspended
particles can modulate regional (Marcella and Eltahir 2014;
1
Department of Earth Sciences, University of Stellenbosch, Pan et al. 2018) and global climate (Mahowald et al. 2014;
Stellenbosch 7600, South Africa
Samset et al. 2018). The depositional of dust can lead to
2
BIOGRIP Water and Soil Node, University of Stellenbosch, an enrichment in nutrients and fines and therefore play cru-
Stellenbosch 7600, South Africa
cial roles in ocean fertilization (Grantz et al. 2003; Dansie
3
Municipality of Saldanha Bay, Vredenburg et al. 2017). Any toxic particles present in dust can also
7380, South Africa
negatively affect the environment (Farmer 1993; Paytan
4
Department of Environmental & Geographical Science, et al. 2009). Lastly, dust can strongly affect public health,
University of Cape Town, Cape Town 7700, South Africa
which is a phenomenon that has attracted attention from the
5
Argos Scientific Pty Ltd.,, Cape Town 7405, South Africa

13
Content courtesy of Springer Nature, terms of use apply. Rights reserved.
2446 Air Quality, Atmosphere & Health (2024) 17:2445–2459

health sector (Li et al. 2019; Pandey et al. 2021). The World and crude iron are processed and transported, along with
Health Organisation (WHO) estimates that in 2019, 4.2 mil- phosphate from nearby phosphate mines (AEC 2022). The
lion people died prematurely due to outdoor air pollution, Saldanha Bay Municipality furthermore includes densely
whereby 89% of these deaths occurred in middle- and low- populated residential areas and key national protected areas,
income countries (WHO 2021). The exposure to air pollu- making it a region with a high diversity of land use.
tion and the number of deaths from outdoor air pollution has The production and export of materials generally
decreased in most South American, North American, and expanded since early 2000 (AEC 2022), despite the closure
European countries during the last 30 years, but increased of one of the steel plants in February 2020 and the fact that
in Asian and African countries (Ritchie and Roser 2019). In the general economy has been affected by the COVID-19
response to this global threat, studies have focussed on the lockdown in 2020 and 2021. Especially the export of man-
precise effects of dust on human health and the characteris- ganese ore and lead increased rapidly over the last eight
tics of this effect from different dust sources. years. Future development includes increasing the storage
The impact of dust on human health depends on both the and handling capacity of iron ore, expanding the rail and
concentration and the physical and chemical characteristics port, increasing the industrial capacity, developing biofuel
of the dust particles (Entwistle et al. 2019; Al-Swadi et al. and gas production facilities, and a general expansion of the
2022), which are determined by their emitting source (Csav- industrial activities.
ina et al. 2012; Goudie 2013; Bauer et al. 2019; Querol et The high quantity of mining material transported, stored,
al. 2019). Dust sources can be divided into natural dust and processed in this area resulted in the visible emission
sources, such as dunes, ephemeral rivers and lakes, and of dust. The dust emitted from the iron ore can be observed
shrublands, and anthropogenic dust sources, which mainly in the area as a “red dust plume” (AEC 2022). Further-
consist of agricultural areas, mining areas, fires, and indus- more, the rail and road transport of mining materials has
trial processes (Koch et al. 2007; Ginoux et al. 2012; Xia been acknowledged as a possible source of dust (DEA&DP
et al. 2022). Generally, anthropogenic dust has a relatively 2019; Saldanha Bay Municipality 2023). The dust in this
high negative impact on public health and the environment region has raised concerns about its impact on public health.
(Csavina et al. 2012; Bauer et al. 2019; Xia et al. 2022) since It should furthermore be considered that the smaller, less
dust from these sources often consists of finer particles and visible particles from transport and smelting activities could
contains contaminants such as heavy metals, pesticides, also be harmful. Lastly, the dust could impact the environ-
and pathogens (Landrigan and Baker 1981; McCartor and ment, either by depositing nutrients or toxic particles in the
Becker 2010; Csavina et al. 2011; Salawu-Rotimi et al. surrounding areas (AEC 2022).
2021). To support health risk assessments in the residential
Csavina et al. (2012) estimated that from the variety areas surrounding the port, the Saldanha Bay Municipality
of anthropogenic sources, activities related to mining and installed dust monitoring equipment in 2015. This equip-
the processing of these mining materials have the highest ment consists of a fine particulate matter (PM2.5) monitor-
potential risk to human and environmental health, and many ing station and several dust buckets and aims at addressing
studies have addressed this negative impact (Zheng et al. whether the National Ambient Air Quality Standards
2010; Csavina et al. 2011; Al-Swadi et al. 2022). Hereby, (NAAQS) have been crossed in these areas. The data that
significant amounts of dust can not only be emitted by quar- was produced from this monitoring has not been further
ries, but also by secondary mining processes, such as crush- analysed on any temporal and spatial patterns and any
ing, grinding, separation, smelting, and transport (Csavina anthropogenic influences on these patterns. This knowledge
et al. 2011; Kristensen et al. 2015). Such diversity and spa- would be of importance for understanding the future devel-
tial variance in mining and mining-related industrial activi- opment of the area and developing any possible mitigation
ties need to be considered when addressing the health and strategies.
environmental impact. This study aims to understand the spatial and tempo-
The port at Saldanha Bay is situated on the west coast ral variations of the dust emission around Saldanha Bay,
of South Africa and is an important port for the transport the specific sources of dust, and any possible natural and
of mining materials and includes an industrial zone that anthropogenic drivers behind this variation. To do so, dust
is responsible for processing this material. In this area, monitoring data from 2015 to 2022 by the Saldanha Bay
55.2 million tons of iron ore and 4.7 million tons of mag- Municipality will be analysed in combination with weather
nesium ore have been stored, processed, and transported in data. We will furthermore address the potential impact of the
2021/2022 (AEC 2022). Furthermore, lead, copper, zinc, dust on public health and determine the possible impacted
and steel, which include galvanized steel, have been pro- environmental areas. These results can function as an initial
duced in this area. In smaller amounts, titanium slag, rutile, understanding of the high health-risk areas of this region

13
Content courtesy of Springer Nature, terms of use apply. Rights reserved.
Air Quality, Atmosphere & Health (2024) 17:2445–2459 2447

and temporal variation caused by meteorological effects. It the hourly rainfall, wind velocity, and wind direction. The
will furthermore give insight into the effects of expanding study area received an average annual rainfall of 234 mm
industrial activities on the dust load in the area. between 2015 and 2022 and is characterised by winter rain-
fall (Fig. 2). The winds generally come from south to south-
west, albeit occasionally winds come from north to northeast
Methods (Fig. 2). The wind velocity is generally higher during winter
with a more pronounced north-northeastern wind direction
Study area during this period (Figures S1 and S2). From 2015 to 2022,
the wind velocity distribution did not significantly change
The Saldanha Bay Municipality is located on the west coast (Figures S2 and S3).
of South Africa and holds the Saldanha Bay port and the sur-
rounding industrial area (Fig. 1). The Saldanha Bay port area Dust monitoring
is roughly 5 km long and 5 km wide and hosts the dry bulk
stockpile, where iron ore is being stored, the multipurpose PM2.5 monitoring
and iron ore terminal, and the dry and liquid bulk terminal.
The industrial area surrounding it, called the Besaansklip Ambient concentrations of particulates of less than 2.5 μm
Industrial Area, is much larger and includes, among others, in diameter (PM2.5) were measured in the town of Saldanha
various smelters and steel plants, the industrial develop- (Fig. 1). The PM2.5 content is generally used to assess the
ment zone, and a gravel quarry. South of the port area lies impact of dust on public health (WHO 2016, 2021; Cui et
the West Coast National Park, part of the UNESCO Cape al. 2019; Khreis et al. 2023). The PM2.5 concentration was
West Coast Biosphere Reserve. Two towns are in proxim- measured according to the South African Ambient Air Qual-
ity to the industrial area: Saldanha and Vredenburg, where ity Standards methods (DEA 2013) using a Tapered Element
respectively 28,000 and 38,000 people live. In total, includ- Oscillating Microbalance (TEOM). The concentration was
ing informal settlements, around 123,000 people live in this measured at hourly intervals from January 2015 to Septem-
municipality (SEP 2021). ber 2018. Due to vandalization, no measurements were per-
formed after this period. To further analyse the measured
Climate and weather monitoring PM2.5 concentration and to compare the data to national and
international health standards (see Sect. 2.3.3), the daily
Daily weather data from 2015 to 2022 has been provided average of the PM2.5 concentration was calculated as the
by the South African Weather Service (SAWS) from a cli- sum of the hourly measurements divided by the duration.
mate station at Langebaanweg (Fig. 1). This data includes

Fig. 1 Map of southern Africa

(a), Saldanha Bay located in
the West Coast District (b), and
locations of installed monitoring
equipment (c). See c for legend
and Sect. 2.3.3. for monitoring
station abbreviation. Credits: Esri
South Africa, Esri, HERE Gar-
min, Foursquare, METI/NASA,
and USGS

13
Content courtesy of Springer Nature, terms of use apply. Rights reserved.
2448 Air Quality, Atmosphere & Health (2024) 17:2445–2459

Fig. 2 The average monthly (a)

and yearly rainfall (b) and the
windrose (c) from 2015 to 2022
as measured by the climate sta-
tion at Langebaanweg (Fig. 1)
showing the dominating SW
wind

Dust deposition monitoring The dust buckets were installed at a height of 2 to 3 m

from the ground. The buckets were sampled approximately
In addition to the PM2.5 monitoring, seven dust buckets once a month (Table S1) and from the sample weight, the
have been installed (Fig. 1). These buckets are named as surface area of the dust buckets, and the sampling period,
follows: N-A (north-A), N-B (north-B), NE (northeast), the depositional flux was calculated in mg day− 1 m− 2. The
CN (central-north), CS (central-south), SW (southwest), equipment was installed from the beginning of 2015 until
and SE (southeast). These buckets capture both the gravi- the end of 2022. Unfortunately, during the COVID-19 lock-
metrical deposited and wet deposited dust, successfully down period from March 2020 until March 2022, the sam-
used in, among others, Audoux et al. (2022) to determine pling interval was longer. Furthermore, over time, two dust
the influence of cold pools on dust washout in the Sahel, buckets corroded (ARGOS 2022). Samples contaminated
Ganor and Foner (2001) for quantifying the yearly variation from bird droppings were excluded from further analyses.
of dust deposition and the local contribution of the dust load
over Israel, and Krah et al. (2004) in the Okavango Delta in National and international health standards for dust
Botswana to determine the spatial variation of dust deposi-
tion and the possible origin of this dust. Other studies that To lower the negative effect of air pollution, the World
used dust buckets successfully are McTainsh et al. (1997), Health Organisation (WHO) has set a guideline for a maxi-
where dust buckets were used to determine the input of dust mum average PM2.5 concentration of 15 µg m− 3 daily and
into soils in Mali and the importance of locally derived dust 5 µg m− 3 annually (WHO 2021). However, since 99% of
in this process, and Rasmussen et al. (2018), which showed the world’s population lives above these guidelines (WHO
that dust deposition data are useful indicators for the indoor 2021), the WHO has introduced four interim targets that
exposure to several elements. The advantages of these dust offer a more gradual shift to lower air pollutant concentra-
buckets are the low-cost and low- maintenance aspects of tions (WHO 2016, 2021). In addition, many nations have
this equipment, which enables assessing the spatial variabil- their own air quality standards (Vallack and Shillito 1998;
ity. However, dust buckets can be contaminated by drop- You 2014; Kelly et al. 2017). South Africa has introduced
pings or insects. There is furthermore a difference between the National Ambient Air Quality Standards (NAAQS)
the grain size distribution of the total suspended particles which prescribe a maximum average concentration of PM2.5.
and the deposited particles, whereby deposited particles are Concerns for public health led to a change in the NAAQS:
expected to be larger on average (McTainsh et al. 1997), in 2030, the maximum daily average will be lowered from
meaning that relating the dust deposition to the suspended 40 µg m− 3 to 25 µg m− 3 and the maximum annual average
particles should be done with caution. will be lowered from 20 µg m− 3 to 15 µg m− 3 (DEA 2012).

13
Content courtesy of Springer Nature, terms of use apply. Rights reserved.
Air Quality, Atmosphere & Health (2024) 17:2445–2459 2449

South Africa has furthermore standards for the dust depo- data, the Global Data Assimilation System (GDAS1) was
sition flux described by the South African National Dust used, a meteorological dataset that has a 3-hour interval and
Control Regulations (DEA 2013). The standard for residen- a 1-degree latitude-longitude grid. The location of the par-
tial and light commercial areas is below 600 mg day− 1 m− 2 ticles at 24 h and 120 h were visualised.
averaged over 30 days, but the Saldanha Bay Municipality
has set 300 mg day− 1 m− 2 as a local goal (ARGOS 2022,
2023). Results

Data analyses and visualisation Spatial variation of dust deposition and origin of
PM2.5
The temporal change in daily PM2.5 concentration and
dust deposition flux was visualised using linear regression The average dust deposition fluxes from 2015 to 2022 is
from the package “ggplot2” in R. To visualise the change 114 mg day− 1 m− 2, but the average over this period from the
in dust deposition flux, the date halfway through the sam- individual stations ranges between 56 mg day− 1 m− 2 and
pling period has been used to plot the measured flux value. 262 mg day− 1 m− 2 (Fig. 3). The variation occurs within a
Additionally, due to the high measurement interval, a gen- 7.5 km radius, whereby the stations with the highest and
eralized additive mode smoothing with penalized regression lowest dust deposition flux are only 6 km away from each
was used for the daily average PM2.5 concentration. This other. This illustrates the large spatial variability of dust
method was preferred over the local polynomial regression fluxes in the region. The deposition fluxes are the smallest in
method due to the high number of data points. To determine the urban areas of Saldanha and Vredenburg (69 mg day− 1
and visualise the relationship between the dust deposition m− 2 and 56 mg day− 1 m− 2 respectively), and highest adja-
flux from the different stations and the PM2.5 concentration, cent to the north-western part of the industrial area (262 mg
a correlation matrix was created with a Pearson correlation, day− 1 m− 2). The areas close to the industrial region experi-
using the package “corrplot” in R. To do so, the average of ence a generally higher dust deposition flux, except for the
the PM2.5 concentration was calculated over the sampling CS station west of the industrial area (Fig. 3).
periods of the dust depositional fluxes. Due to the skewness The hourly measurements were used to create the polar
of the data, the hourly and daily PM2.5 concentrations might plot of the PM2.5 concentration in the town of Saldanha
be non-parametric. For this reason, Wilcoxon tests (also (Fig. 3). In total, most dust was brought by strong north-
called the Mann-Whitney test) were used to compare the eastern and southern winds. This indicates that the indus-
value differences between the two sets. When multiple data- trial area is likely the origin of this dust, although no exact
sets were compared, ANOVA tests were utilised. For any point of origin can be derived from this graph. Furthermore,
statistical test, an alpha value of 0.01 was chosen to signify a high dust centration was associated with low wind speeds
statistical significance. For boxplot graphs, values that are (below 2 m/s), which could indicate local dust sources such
more than 1.5 times the interquartile range are visualised as as traffic. The dust coming from the southeast to the south-
separate data points. west could originate from the coastal regions situated south
To visualise the relationship between wind velocity, and west of Saldanha (Fig. 1). The yearly resolved polar
wind direction, and PM2.5 concentration, polar plots were plots (Figure S4) show interannual variability in origin in
created using the package “openair” in R (Carslaw and Rop- the dust. Eastern dust sources dominated from 2016 to 2018,
kins 2012). To create these plots, the hourly wind data from whereas dust from strong southern winds was most promi-
the Langebaanweg climate station was combined with the nent in 2015.
PM2.5 data sampled at the same time. The lower sampling Since the predominant wind direction is south to south-
interval of the wind conditions means that not the whole west, the dust emitted from the industrial area is most of the
PM2.5 dataset was used to create the polar plots. time not transported in the direction of the PM2.5 monitoring
To determine which areas beyond the Saldanha Bay in the town of Saldanha, but in the north- to northeastern
Municipality could be impacted by the dust, the Hybrid direction (Fig. 1). It is therefore possible that the areas north
Single Particle Lagrangian Integrated Trajectory Model to northeast of the port, such as the town of Vredenburg,
(HYSPLIT, Stein et al. 2015; Rolph et al. 2017) was used have a higher PM2.5 flux than the point where the PM2.5 is
to calculate the trajectory of particles from the port (Fig. 1). being monitored. The measurement in Saldanha may there-
The trajectory was calculated for each day from 2015 to fore not capture the main transport of dust from the indus-
2022 at 12:00 (n = 2557). The trajectory was calculated trial area and might also not represent the highest health risk
for 120 h, with a starting height of 100 m, following the that the PM2.5 emission poses.
methods from Neff and Bertler (2015). For the weather

13
Content courtesy of Springer Nature, terms of use apply. Rights reserved.
2450 Air Quality, Atmosphere & Health (2024) 17:2445–2459

Fig. 3 Polar plot for the PM2.5

concentration from the moni-
toring station in the town of
Saldanha from 2015 to 2018 (a)
and the average dust deposition
fluxes in mg day−1 m−2 at the
seven monitoring stations (b).
The colour bar in the polar plot
(a) indicates the PM2.5 concen-
tration related to certain wind
speeds and wind directions, while
the size and colour of the circles
in the map (b) are an indication
of the dust deposition flux

The large difference between the dust deposition at sta- weak but mostly positively correlated (-0.08 < r < 0.72).
tion CN, and N-A and N-B could be explained by the dif- Surprisingly, the adjacent stations N-A and N-B showed no
ference in elevation between these stations (Fig. 4). When significant relationship with each other. This suggests that
sediment-carrying winds move over a hill, the dust deposi- either the dust sources are local, or that additional drivers
tion is strongest on the windward side and weakest at the of dust transportation, such as elevation or influence from
peak (Goossens 1989, 2006; Zufall et al. 1999; Parker and buildings in Vredenburg, drive a local imprint.
Kinnersley 2004). This would explain the low dust deposi-
tion for stations N-A and N-B, considering their location on Temporal variation of PM2.5
top of a hill, at 136 m (N-A) and 154 m (N-B) elevation,
which is higher than station CN at 26 m elevation, but simi- The average PM2.5 concentration was 7.1 µg m− 3 reach-
lar elevation as SE, NE, CS, and SW (Fig. 4). ing peak hourly concentrations of 95 µg m− 3 (Fig. 6a). The
hourly values were transformed into daily average values to
Temporal variation of dust deposition determine a temporal pattern (Fig. 6b). The daily average
PM2.5 concentration did show a weak decrease from 2015
The dust deposition fluxes from the individual monitoring to 2018. Per year, the daily concentration decreased signifi-
stations from 2015 to 2022 show a large variance (Fig. 5 and cantly from 2017 to 2018 (Fig. 7a, p = 8.8 *10− 4), the other
Figure S5). The median dust deposition increases over time years the change was non-significant. A seasonal cycle in
(R2 = 0.19) whereby the peak median dust deposition flux in the PM2.5 concentration can be discerned by a generalized
2022 is mainly driven by the high deposition flux at station additive mode smoothing regression (Sect. 2.4) whereby
SE which has been attributed to building activities within the PM2.5 concentration peaks around winter and is low-
100 m west of the measuring station (ARGOS 2023). The est around mid-summer. The PM2.5 concentration differed
individual dust deposition fluxes from the monitoring sta- significantly (Fig. 7b, p = 0.006) between summer (Octo-
tions and SE, SW, CS, and N-A show a weak, positive rela- ber to March) and winter (April to Sep) with an average
tionship with the year, which confirms the small increase concentration around 8% higher in winter. To understand
in dust deposition over time, whereas CN and NE show no the diurnal pattern of the PM2.5 concentration, the time of
change (R2 < 0.01) and N-B shows a weak decrease over the maximum daily PM2.5 concentration was determined
time (R2 = 0.03). There is furthermore no consistent seasonal (Fig. 8a). The daily peak PM2.5 concentration generally
trend in the dust deposition flux (Figure S6). The relation- occurs between 06:00 and 09:00 and between 18:00 and
ship between the different monitoring stations, the average 21:00. The difference in PM2.5 between the high and the low
deposition fluxes, the year, and the PM2.5 concentration are concentration periods is statistically significant (p < 0.10− 6)
shown in a Pearson correlation matrix (Figure S7). Dust (Fig. 8b) and between 26% and 42% higher during the high
deposition fluxes among the different stations are generally concentration periods.

13
Content courtesy of Springer Nature, terms of use apply. Rights reserved.
Air Quality, Atmosphere & Health (2024) 17:2445–2459 2451

Fig. 4 Map of the elevation

of the area and the location of
the monitoring equipment, the
towns, the roads and railroad,
and the industrial area. Eleva-
tion data credits: NASA, USGS,
SERVIR-RCMRD

Fig. 5 The monthly dust

deposition as measured by the
individual monitoring station
grouped per year. Note the break
in the axis at 800 mg m-2 day-1 to
improve the visualisation of the
outliers

13
Content courtesy of Springer Nature, terms of use apply. Rights reserved.
2452 Air Quality, Atmosphere & Health (2024) 17:2445–2459

Fig. 6 PM2.5 concentrations illustrated as (a) measured hourly data (black line) and calculated 24-hour floating average (blue line), and (b) the
calculated daily averages (black dots) with a generalized additive mode smoothing regression (blue line), and a linear regression (red line)

13
Content courtesy of Springer Nature, terms of use apply. Rights reserved.
Air Quality, Atmosphere & Health (2024) 17:2445–2459 2453

Fig. 7 Daily average PM2.5

concentration for each year (a)
and summer and winter (b).
The diamond and the associated
number reflect the average of the
PM2.5 concentration per group

Fig. 8 Time of the day of peak

concentration from the days
between 2015–2018 (a), and
averages of hourly PM2.5 concen-
trations during different periods
of the day. In (b), the diamond
and the associated number reflect
the average of the PM2.5 concen-
tration per group

The dust deposition fluxes did not show a significant cor-

relation to the PM2.5 concentrations (-0.22 < r < 0.20, Figure Discussion
S7). This lack of a significant correlation is especially note-
worthy at station SW (r = 0.002) since PM2.5 is measured Spatiotemporal variation and controls of dust
at this location. This disconnect in the temporal pattern can
also be observed in the difference in the seasonal pattern Spatial
of the dust deposition and the PM2.5 concentration, and the
fact that the PM2.5 concentration shows a small decrease The PM2.5 particles captured at the Saldanha monitoring sta-
over time, whereas the dust deposition flux shows a small tion, southwest of the industrial area, have a predominant
increase. northeastern to eastern origin (Fig. 3), which corresponds to
the location of the industrial area, although an exact point
Rainfall impact source cannot be identified. Furthermore, a southwestern
transport likely carried particles originating from the coastal
There is a strong significant difference in the PM2.5 con- regions in 2015 (Figure S4). Such coastal regions are com-
centration on hours with and without rain (p < 2 * 10−−16) monly recognised sources of aerosols, including emissions
whereby the average is 54% higher for days without rain of sea spray and mineral dust from sandy beaches and dune
(Fig. 9a). However, there is no correlation between the areas (e.g. Engelbrecht et al. 2009; Dansie et al. 2017). In
hourly rainfall and PM2.5 concentration (R2 < 0.01, Fig. 9b). addition, the high concentration related to a low wind speed
The rainfall amount did furthermore not correlate with any could indicate emissions from local sources such as traffic
of the dust deposition fluxes observed at the seven monitor- or construction sites.
ing stations (r < 0.2, Figure S7). Determining the driver of dust deposition is more diffi-
cult. Relating the spatial variability of the dust deposition
fluxes to certain wind events or wind conditions remains

13
Content courtesy of Springer Nature, terms of use apply. Rights reserved.
2454 Air Quality, Atmosphere & Health (2024) 17:2445–2459

Fig. 9 The average daily PM2.5

concentration of day with and
without rain whereby the squares
and the number describes the
average of the total values (a)
and the relationship between the
daily PM2.5 concentration and
rainfall (b)

challenging due to the low sampling interval, as dust depo- Diurnal

sition was only assessed monthly. The highest dust deposi-
tion fluxes were measured at the monitoring station located Peak dust concentration occurred in the morning and the
north of the port (CN) and northwest of the industrial area evening, which could be caused by high emissions and/or
and railway (NE) (Fig. 3). This is in line with the PM2.5 favourable wind conditions. Windroses indicate that wind
particles at Saldanha that originate from the direction of the velocities are typically low in the morning and higher in
port and the industrial area and the southern wind direction the afternoon and evening (Figure S8). Hereby, the weaker
transporting a majority of the particles in a northern direc- morning winds tend to originate from both the southeastern
tion. Besides proximity to the industrial source, elevation (in and northern winds, whereas in the evening, southeastern
the case of N-A and N-B) and local dust sources caused by winds dominate. Polar plots determined for the morning
construction (for SE in 2022) influenced the dust deposition. hours between 07:00 to 09:00 and evening hours between
18:00 to 20:00 point to differences in source areas: the
Inter-annual highest dust concentration in the morning occurs when
there is less wind, which suggests the dust originates from
Despite the increase in industrial activities around the port, local sources near the measuring station, such as traffic
both the PM2.5 concentration and the dust deposition flux (Fig. 10a), whereas in the evening, the highest dust concen-
showed no significant increase over time. In fact, the daily tration occurs when there are eastern winds, which suggests
PM2.5 concentration decreased significantly from 2017 the dust originates from the industrial area (Fig. 10b). The
to 2018 (Fig. 7). The cause of this decrease in measured higher emission in the afternoon could be related to higher
PM2.5 per year is difficult to attribute to certain industrial activity during this time since the wind in the evening is
activities. Rainfall appears to have a strong influence on the dominantly from the southwest. This shift in emissivity
daily PM2.5 concentration, but since the wettest year is also could be confirmed by increasing the spatial resolution of
the year with the highest average PM2.5 concentration, this PM2.5 concentration measurements.
effect appears to not be dominating on longer timescales.
Rainfall
Seasonal
The daily PM2.5 concentration is related to days with and
The PM2.5 concentration and the dust deposition flux show without rainfall, which could be interpreted as a conse-
only weak seasonal fluctuations (Figs. 5 and 6, and 7). quence of the wet removal effect of rain (Liu et al. 2020;
Slightly higher dust concentrations were observed in winter Zhao et al. 2020). This relationship is, however, more com-
possibly due to the typical north-to-northeast winter winds plex as the rainfall amount is unrelated to PM2.5 concentra-
(Figure S1). These winds would transport suspended parti- tion, even though both rainfall and low PM2.5 concentrations
cles from the industrial area to the PM2.5 monitoring station can be associated with strong northern winds (Figure S9). It
located to the southwest, resulting in a higher wintertime is furthermore surprising that rainfall appears not to influ-
PM2.5 concentration. The dust deposition, in contrast, did ence dust deposition (Figure S7, r < 0.2). This goes against
not show any consistent seasonal trend. the generally assumed importance of wet deposition, as

13
Content courtesy of Springer Nature, terms of use apply. Rights reserved.
Air Quality, Atmosphere & Health (2024) 17:2445–2459 2455

Fig. 10 The polar plots of the

hourly data for the morning
between 07:00 to 09:00 (a), and
evening between 18:00 to 20:00
(b). These times were chosen
based on the periods of highest
measured PM2.5 concentrations

described in several studies that use dust buckets (Ganor lowest concentration of PM2.5 in the Saldanha Bay area is
and Foner 2001; Krah et al. 2004; Morales-Baquero et al. observed during strong (above 6 m s− 1) western to north-
2013; Audoux et al. 2022). However, these studies mainly western winds. In contrast, the highest PM2.5 concentration
propose the importance of wet deposition for finer particles occurred in conditions of weak (below 2 m s− 1) and strong
(PM10) whereas larger particles seemed to be less affected (above 10 m s− 1) winds from the northeast. However, such
by rainfall, which could explain the relationship of rainfall meteorological conditions could not explain the PM2.5 and
with the daily PM2.5 concentration and the lack of relation- dust distribution entirely. Additionally, these dynamics
ship with dust deposition. We therefore postulate that the appear to be affected by the complexity and diversity of dust
fact that PM2.5 is related to days with rainfall is a conse- emitted in these industrial areas, including traffic, as well
quence of air masses originating from the Atlantic Ocean as small-scale factors such as the elevation of the sampling
carrying low dust loads. This demonstrates the complex sites.
and dynamic influence of meteorological conditions on dust
transport and deposition. Monitoring improvements

Summary of the spatiotemporal variability and controls The dust buckets-based monitoring in the study area offers
an affordable opportunity to assess spatial variability. How-
Interannual changes and trends of PM2.5 and dust deposi- ever, there appears to be no significant correlation between
tion were largely lacking significance despite the progres- the dust deposition fluxes measured monthly at the seven
sive extension of industrial activities, but seasonal changes dust bucket stations and the PM2.5 concentrations for the
were apparent. Winter has on average an 8% higher PM2.5 same period. This is not in line with the general assump-
concentration than summer. In addition, diurnal fluctuations tion that there is a significant relationship between the total
occurred, with higher PM2.5 concentrations in the afternoon suspended particle concentration and the PM2.5 and PM10
and evening than in the morning. The polar plots show that concentrations (Tucker 2000; Bacon et al. 2011; Zhang et al.
certain wind directions and velocities carry more dust, a 2017). However, this disconnect between deposition fluxes
relationship that fluctuates during the day. Wind and rainfall and suspended particles has been observed previously in
patterns explain the inter-annual, seasonal, and diurnal vari- dust events over Israel from 1969 to 1995 (Ganor and Foner
ability of PM2.5 and dust deposition in the area to a certain 2001). Around Saldanha Bay, this lack of relationship could
extent. For example, days without rain have 54% higher be explained by the difference in the particle size assessed
PM2.5 concentration compared to days with rain. Further- by the dust buckets and PM2.5 sensor. The deposited dust
more, a low PM2.5 concentration measured was associated could include larger particles, such as silt and fine sand and
with southwestern to northern winds with a velocity above point to mineral sources even if the emission is due to local
4 m s− 1. The low dust loads of air masses originating from anthropogenic activities, while the fine particle (PM2.5)
the Atlantic Ocean are also evident considering that the concentration could predominantly be associated with

13
Content courtesy of Springer Nature, terms of use apply. Rights reserved.
2456 Air Quality, Atmosphere & Health (2024) 17:2445–2459

combustion and industrial activities (Saucy et al. 2018; Ji et The current WHO daily air quality standard of 15 µg m− 3
al. 2018; Cui et al. 2019). Despite the merit of dust buckets PM2.5 was exceeded 2.9% of the days (ca. 10 days per year).
as cheap, sustainable measurement methods for understand- The WHO standard for a yearly average PM2.5 concentration
ing the composition of dust, dust buckets do not selectively of 5 µg m− 3, is exceeded every year, including when consid-
capture the finer particles which are the general indicators of ering the annual floating average of the PM2.5 concentration.
the risks to public health. However, comparing this to the WHO global air quality data-
So far, only one fine particle sensor has been used in this set (WHO 2024), Saldanha belongs to the lowest 10% areas in
study. Understanding the high spatiotemporal variation of PM2.5 concentration. Despite the relatively low concentration,
PM2.5 in urban and industrial environments is challeng- exposure to a daily average PM2.5 concentration of 7.1 µg m− 3
ing with such a limited number of PM2.5 measurements could be associated with a mortality increase of approximately
(Petavratzi et al. 2005; Zhang et al. 2018). Therefore, an 0.5–5% (Schwartz et al. 2002; Shi et al. 2016).
important next step for the improvement in the understand- To understand the precise health effects of the industrial
ing of the origin of dusts and the associated health risks, area in the Saldanha Bay municipality, it should be considered
and developing mitigation strategies, would be to establish that the dust deposition data and the polar plots show that the
a higher spatial resolution network of PM2.5 measurements. location of the PM2.5 monitoring station may not receive the
The recent developments of low-cost sensors enable such highest dust concentration. The dust-bucket data informed that
high-resolution measurements (Pope et al. 2018; Awokola to improve the understanding of the high-risk area, it would be
et al. 2022) and the spatial variability assessed through dust necessary to place PM2.5 monitoring equipment closer to the
bucket dust deposition data has informed about the key sites dust plume source. It is likely that these areas receive higher
for future fine particle sensor sites. It should be noted that dust concentrations and that these regions should be the lead-
these sensor data are not flawless: they only represent a spe- ing areas in determining the threat of these industries on public
cific fraction (typically PM2.5 or PM10), and the measure- health.
ments can be disturbed by changes in humidity and larger
faction particles (Stavroulas et al. 2020; Barkjohn et al. Environmental impact
2021). Satellite imagery methods (Yu and Zahidi 2023) and
physical models in combination with computational tech- Besides the impact on public health, the impact on the envi-
niques (Zhang et al. 2018) could be explored as an addi- ronment should be considered. As mentioned in the introduc-
tion to the regular dust monitoring in the Saldanha Bay area. tion, this can both be a fertilizing effect of the dust (Peterson
Lastly, real-time inhalation monitoring needs to be included et al. 1993; Okin et al. 2011; Moore et al. 2013; Dansie et al.
for accurate health impact assessment (Khamraev et al. 2017; Barkley et al. 2019) and a toxic effect, especially in the
2021), as well as the chemical composition of the particles, case of anthropogenic dust (Farmer 1993; Griffin and Kellogg
especially considering that industrial and mining dust often 2004; Paytan et al. 2009). These effects can be regional, but
includes potentially toxic components. dust particles also have the potential to travel far distances. The
HYSPLIT model shows the location density of the particles
Potential impact 24 h and 120 h after suspension (Fig. 11). These figures reveal
that especially the area northwest of Saldanha Bay (e.g., most of
Public health impact the West Coast and Namakwa Districts) is affected by the dust
from Saldanha Bay within 24 h after suspension (Fig. 11a). The
The initial goal of the monitoring activities was to determine trajectories 120 h are characterised by a redistribution between
whether the air quality in the residential areas adhered to the particles transported further north, along the west coast of
national air standard. The municipality’s dust deposition Namibia, and particles transported to the southeast, over South
standard of 300 mg day− 1 m− 2 is exceeded 8.6% of the time Africa and to the Indian Ocean (Fig. 11b). The location of each
between 2015 and 2022 and the recommendation by the South individual run is shown in Figure S10 and shows that there are
African National Dust Control Regulations for residential areas days when particles have the potential to reach as far as the
of 600 mg day− 1 m− 2 is exceeded 2.1% of the time (Fig. 5). Australian or Antarctic continents. These trajectories indicate
The daily average PM2.5 concentrations remain far below the that both the impact on land and the ocean need to be con-
daily NAAQS of 40 µg m− 3, and the PM2.5 annual average sidered to fully understand the impact of the dust. To further
remains below the NAAQS of 20 µg m− 3. Even the stricter assess the potential consequences of the dust on the environ-
health standard envisaged for 2030 (25 µg m− 3 daily, and ment, the geochemistry, solubility, and presence of any toxic
15 µg m− 3 annually), which is comparable to the 2nd interim particles need to be analysed.
target set by the WHO, is only exceeded one day during the
four years of monitoring.

13
Content courtesy of Springer Nature, terms of use apply. Rights reserved.
Air Quality, Atmosphere & Health (2024) 17:2445–2459 2457

Fig. 11 The density of the

particles emitted daily from the
industrial hub of Saldanha Bay
from 2015 to 2022 calculated by
the HYSPLIT trajectory model-
ling. The figures show the spatial
distribution of the trajectories
after 24 h (a) and 120 h (b) of
suspension

awarded to PI Dr Susanne Fietz.

Conclusion Open access funding provided by Stellenbosch University.

Data availability The datasets generated during the current study are
The PM2.5 concentration has been monitored from 2015 to available in the Saldanha Bay Municipality repository: https://sbm.
2018 in the town of Saldanha, and the dust deposition flux has gov.za/environmental/. The hourly meteorological data from 2015 to
been monitored from 2015 to 2022 at seven different locations 2022 is available at https://zenodo.org/records/10926062.
surrounding the Saldanha Bay industrial area. Despite changes
in industrial activities, there have been only small changes in Declarations
the annual dust deposition and PM2.5 concentration. Most of
Ethics approval and consent to participate Not applicable.
the diurnal and seasonal variation in PM2.5 can be attributed to
a certain extent to wind and rain conditions, but there remains
Consent for publication Not applicable.
a large variability unexplained. It appeared furthermore chal-
lenging to identify a single source of dust, which demonstrated Competing interests The authors declare no competing interests.
the complexity of dust emission from a diverse industrial area.
The PM2.5 concentration remained below the national stan- Open Access This article is licensed under a Creative Commons
dard and the envisaged standard of 2030. However, the current Attribution 4.0 International License, which permits use, sharing,
adaptation, distribution and reproduction in any medium or format,
monitoring does not provide insight into the air pollution in as long as you give appropriate credit to the original author(s) and the
high-risk areas and would need improvement to better assess source, provide a link to the Creative Commons licence, and indicate
the public health risks. if changes were made. The images or other third party material in this
article are included in the article’s Creative Commons licence, unless
Supplementary Information The online version contains indicated otherwise in a credit line to the material. If material is not
supplementary material available at https://doi.org/10.1007/s11869- included in the article’s Creative Commons licence and your intended
024-01581-8. use is not permitted by statutory regulation or exceeds the permitted
use, you will need to obtain permission directly from the copyright
holder. To view a copy of this licence, visit http://creativecommons.
Acknowledgements This work is based on the research sup-
org/licenses/by/4.0/.
ported in part by the National Research Foundation of South Africa
(Grant Numbers CPRR150612119375, SRUG2204204036, and
MESA170607237905) awarded to PI Dr Susanne Fietz. The weather
data has been provided by the South African Weather Service (SAWS). References
Author contributions All authors contributed to the study’s conception AEC (2022) The state of Saldanha Bay and Langebaan Lagoon
and design. Material preparation and data collection were performed 2021/2022. Anchor Environmental Consultants, Cape Town
by René Toesie and Grant Ravenscroft and the analyses of data were Al-Swadi HA, Usman ARA, Al-Farraj AS et al (2022) Sources, toxic-
performed by Kaukurauee I. Kangueehi and Heleen C. Vos. The first ity potential, and human health risk assessment of heavy metals-
draft of the manuscript was written by Heleen C. Vos and all authors laden soil and dust of urban and suburban areas as affected by
commented on previous versions of the manuscript. All authors read industrial and mining activities. Sci Rep 12:8972. https://doi.
and approved the final manuscript. org/10.1038/s41598-022-12345-8
ARGOS (2022) Test Report: Dust Fallout Monitoring. Saldanha Bay
Funding This work is based on the research supported in part by Municipality, https://sbm.gov.za/environmental/
the National Research Foundation of South Africa (Grant Numbers ARGOS (2023) Test Report: Dust Fallout Monitoring. Saldanha Bay
CPRR150612119375, SRUG2204204036, and MESA170607237905) Municipality, https://sbm.gov.za/environmental/

13
Content courtesy of Springer Nature, terms of use apply. Rights reserved.
2458 Air Quality, Atmosphere & Health (2024) 17:2445–2459

Audoux T, Laurent B, Marticorena B et al (2022) Wet deposi- Ganor E, Foner HA (2001) Mineral dust concentrations, deposition
tion fluxes of Mineral Dust and their relation with Cold fluxes and deposition velocities in dust episodes over Israel.
pools in the Central Sahel. Geophys Res Lett 49. https://doi. J Geophys Research: Atmos 106:18431–18437. https://doi.
org/10.1029/2021GL095005. e2021GL095005 org/10.1029/2000JD900535
Awokola B, Okello G, Johnson O et al (2022) Longitudinal ambient Ginoux P, Prospero JM, Gill TE et al (2012) Global-scale attribution
PM2.5 measurement at fifteen locations in eight sub-saharan of anthropogenic and natural dust sources and their emission rates
African countries using low-cost sensors. Atmosphere 13:1593. based on MODIS Deep Blue aerosol products. Rev Geophys
https://doi.org/10.3390/atmos13101593 50:3005. https://doi.org/10.1029/2012RG000388
Bacon SN, McDonald EV, Amit R et al (2011) Total suspended par- Goossens D (1989) Height distortion and the sedimentation of dust
ticulate matter emissions at high friction velocities from desert on topographic scale models: considerations and simulations.
landforms. J Geophys Research: Earth Surf 116. https://doi. Earth Surf Proc Land 14:655–667. https://doi.org/10.1002/
org/10.1029/2011JF001965 esp.3290140619
Barkjohn KK, Gantt B, Clements AL (2021) Development and applica- Goossens D (2006) Aeolian deposition of dust over hills: the effect of
tion of a United States-wide correction for PM2.5 data collected dust grain size on the deposition pattern. Earth Surf Proc Land
with the PurpleAir sensor. Atmos Meas Tech 14:4617–4637. 31:762–776. https://doi.org/10.1002/esp.1272
https://doi.org/10.5194/amt-14-4617-2021 Goudie AS (2013) Desert dust and human health disorders. Environ Int
Barkley AE, Prospero JM, Mahowald N et al (2019) African biomass 63 C:101–113. https://doi.org/10.1016/j.envint.2013.10.011
burning is a substantial source of phosphorus deposition to the Grantz DA, Garner JHB, Johnson DW (2003) Ecological effects of par-
Amazon, Tropical Atlantic Ocean, and Southern Ocean. Proceed- ticulate matter. Environ Int 29:213–239. https://doi.org/10.1016/
ings of the National Academy of Sciences 116:16216–16221. S0160-4120(02)00181-2
https://doi.org/10.1073/pnas.1906091116 Griffin DW, Kellogg CA (2004) Dust storms and their impact on
Bauer SE, Im U, Mezuman K, Gao CY (2019) Desert Dust, Industri- Ocean and Human Health: Dust in Earth’s atmosphere. Eco-
alization, and agricultural fires: Health impacts of Outdoor Air Health 1:284–295. https://doi.org/10.1007/s10393-004-0120-8
Pollution in Africa. J Geophys Research: Atmos 124:4104–4120. Ji X, Yao Y, Long X (2018) What causes PM2.5 pollution? Cross-econ-
https://doi.org/10.1029/2018JD029336 omy empirical analysis from socioeconomic perspective. Energy
Bridges EM, Oldeman LR (1999) Global Assessment of Human- Policy 119:458–472. https://doi.org/10.1016/j.enpol.2018.04.040
Induced Soil Degradation. Arid Soil Res Rehabil 13:319–325. Kelly JT, Reff A, Gantt B (2017) A method to predict PM2.5 result-
https://doi.org/10.1080/089030699263212 ing from compliance with national ambient air quality standards.
Carslaw DC, Ropkins K (2012) Openair — an R package for air qual- https://doi.org/10.1016/j.atmosenv.2017.05.009. Atmospheric
ity data analysis. Environ Model Softw 27–28:52–61. https://doi. Environment 162:
org/10.1016/j.envsoft.2011.09.008 Khamraev K, Cheriyan D, Choi J (2021) A review on health risk
Csavina J, Landázuri A, Wonaschütz A et al (2011) Metal and Metal- assessment of PM in the construction industry – current situation
loid contaminants in Atmospheric aerosols from Mining Opera- and future directions. Sci Total Environ 758:143716. https://doi.
tions. Water Air Soil Pollut 221:145–157. https://doi.org/10.1007/ org/10.1016/j.scitotenv.2020.143716
s11270-011-0777-x Khreis H, Sanchez KA, Foster M et al (2023) Urban policy inter-
Csavina J, Field J, Taylor M et al (2012) A review on the importance ventions to reduce traffic-related emissions and air pollution: a
of metals and metalloids in Atmospheric Dust and Aerosol from systematic evidence map. Environ Int 172:107805. https://doi.
Mining Operations. Sci Total Environ 433:58–73. https://doi. org/10.1016/j.envint.2023.107805
org/10.1016/j.scitotenv.2012.06.013 Koch D, Bond TC, Streets D et al (2007) Global impacts of aerosols
Cui M, Lu H, Wiggs GFS et al (2019) Quantifying the effect of geo- from particular source regions and sectors. J Geophys Research:
morphology on aeolian dust emission potential in northern China. Atmos 112. https://doi.org/10.1029/2005JD007024
Earth Surf Proc Land 44:2872–2884. https://doi.org/10.1002/ Krah M, McCarthy TS, Annegarn H, Ramberg L (2004) Airborne
esp.4714 dust deposition in the Okavango Delta, Botswana, and its impact
Dansie AP, Wiggs GFS, Thomas DSG, Washington R (2017) Measure- on landforms. Earth Surf Proc Land 29:565–577. https://doi.
ments of windblown dust characteristics and ocean fertilization org/10.1002/esp.1051
potential: the ephemeral river valleys of Namibia. Aeolian Res Kristensen LJ, Taylor MP, Morrison AL (2015) Lead and zinc dust
29:30–41. https://doi.org/10.1016/j.aeolia.2017.08.002 depositions from ore trains characterised using lead isotopic com-
DEA (2012) National Ambient Air Quality Standard for Particulate positions. Environ Sci: Processes Impacts 17:631–637. https://
Matter with Aerodynamic Diameter Less Than 2.5 Micron Metres doi.org/10.1039/C4EM00572D
(PM2.5). Department of Environmental Affairs (DEA) Landrigan PJ, Baker EL (1981) Exposure of children to heavy metals
DEA (2013) National Environmental Management: Air Quality Act, from smelters: Epidemiology and toxic consequences. Environ
2004. Department Of Environmental Affairs Res 25:204–224. https://doi.org/10.1016/0013-9351(81)90090-6
DEA&DP (2019) Risk and Resilience Assessment of Natural Capital Li X-D, Jin L, Kan H (2019) Air pollution: a global problem needs
in the Greater Saldanha Bay Municipality: A Navigational Tool local fixes. Nature 570:437–439. https://doi.org/10.1038/
for Strategic-Level decision-making. Western Cape Department d41586-019-01960-7
of Environmental Affairs and Development Planning, Cape Town Liu Z, Shen L, Yan C et al (2020) Analysis of the influence of pre-
Engelbrecht JP, McDonald EV, Gillies JA et al (2009) Characterizing cipitation and wind on PM2. 5 and PM10 in the atmosphere. Adv
Mineral dusts and other Aerosols from the Middle East—Part 1: Meteorol 2020:1–13
ambient sampling. Inhalation Toxicol 21:297–326. https://doi. Mahowald NM, Albani S, Kok JF et al (2014) The size distribution of
org/10.1080/08958370802464273 desert dust aerosols and its impact on the Earth system. Aeolian
Entwistle JA, Hursthouse AS, Reis PAM, Stewart AG (2019) Metallif- Res 15:53–71. https://doi.org/10.1016/j.aeolia.2013.09.002
erous Mine Dust: Human Health impacts and the potential deter- Marcella MP, Eltahir EAB (2014) The role of mineral aerosols in
minants of Disease in Mining communities. Curr Pollution Rep shaping the regional climate of West Africa. J Geophys Research:
5:67–83. https://doi.org/10.1007/s40726-019-00108-5 Atmos 119:5806–5822. https://doi.org/10.1002/2012JD019394
Farmer AM (1993) The effects of dust on vegetation—a review. Environ McCartor A, Becker D (2010) Top six toxic threats, world’s worst pol-
Pollut 79:63–75. https://doi.org/10.1016/0269-7491(93)90179-R lution problems report. Blacksmith Institute, New York

13
Content courtesy of Springer Nature, terms of use apply. Rights reserved.
Air Quality, Atmosphere & Health (2024) 17:2445–2459 2459

McTainsh GH, Nickling WG, Lynch AW (1997) Dust deposition and J Environ Res Public Health 15:1452. https://doi.org/10.3390/
particle size in Mali, West Africa. CATENA 29:307–322. https:// ijerph15071452
doi.org/10.1016/S0341-8162(96)00075-6 Schwartz JD, Laden F, Zanobetti A (2002) The concentration-response
Moore CM, Mills MM, Arrigo KR et al (2013) Processes and patterns relation between PM(2.5) and daily deaths. Environ Health Per-
of oceanic nutrient limitation. Nat Geosci 6:701–710. https://doi. spect 110:1025–1029. https://doi.org/10.1289/ehp.021101025
org/10.1038/ngeo1765 SEP (2021) Socio-Economic Profile: Saldanha Bay Municipality.
Morales-Baquero R, Pulido-Villena E, Reche I (2013) Chemical sig- Western Cape Government
nature of Saharan dust on dry and wet atmospheric deposition Shi L, Zanobetti A, Kloog I et al (2016) Low-concentration PM2.5
in the south-western Mediterranean region. Tellus B: Chem Phys and mortality: estimating Acute and Chronic effects in a Popula-
Meteorol 65:18720. https://doi.org/10.3402/tellusb.v65i0.18720 tion-based study. Environ Health Perspect 124:46–52. https://doi.
Neff PD, Bertler NAN (2015) Trajectory modeling of modern dust trans- org/10.1289/ehp.1409111
port to the Southern Ocean and Antarctica. J Geophys Research: Stavroulas I, Grivas G, Michalopoulos P et al (2020) Field evalua-
Atmos 120:9303–9322. https://doi.org/10.1002/2015JD023304 tion of low-cost PM sensors (Purple Air PA-II) under Variable
Okin GS, Baker AR, Tegen I et al (2011) Impacts of atmospheric Urban Air Quality conditions, in Greece. https://doi.org/10.3390/
nutrient deposition on marine productivity: roles of nitrogen, atmos11090926. Atmosphere 11:
phosphorus, and iron. Glob Biogeochem Cycles 25. https://doi. Stein AF, Draxler RR, Rolph GD et al (2015) NOAA’s HYSPLIT
org/10.1029/2010GB003858 Atmospheric transport and dispersion modeling system. Bull
Pan B, Wang Y, Hu J et al (2018) Impacts of Saharan Dust on Atlantic Am Meteorol Soc 96:2059–2077. https://doi.org/10.1175/
Regional Climate and implications for Tropical cyclones. J Clim BAMS-D-14-00110.1
31:7621–7644. https://doi.org/10.1175/JCLI-D-16-0776.1 Tucker WG (2000) An overview of PM2.5 sources and control
Pandey A, Brauer M, Cropper ML et al (2021) Health and economic strategies. Fuel Process Technol 65–66:379–392. https://doi.
impact of air pollution in the States of India: the global burden of org/10.1016/S0378-3820(99)00105-8
Disease Study 2019. Lancet Planet Health 5:e25–e38. https://doi. Vallack HW, Shillito DE (1998) Suggested guidelines for depos-
org/10.1016/S2542-5196(20)30298-9 ited ambient dust. Atmos Environ 32:2737–2744. https://doi.
Parker ST, Kinnersley RP (2004) A computational and wind tun- org/10.1016/S1352-2310(98)00037-5
nel study of particle dry deposition in complex topography. WHO (2016) Ambient air pollution: a global assessment of exposure
Atmos Environ 38:3867–3878. https://doi.org/10.1016/j. and burden of disease. World Health Organization, Geneva
atmosenv.2004.03.046 WHO (2021) WHO global air quality guidelines: particulate matter
Paytan A, Mackey KRM, Chen Y et al (2009) Toxicity of atmospheric (PM2.5 and PM10), ozone, nitrogen dioxide, sulfur dioxide and
aerosols on marine phytoplankton. Proc Natl Acad Sci 106:4601– carbon monoxide. World Health Organization, Geneva
4605. https://doi.org/10.1073/pnas.0811486106 WHO (2024) WHO global ambient air quality database (update 2024).
Petavratzi E, Kingman S, Lowndes I (2005) Particulates from mining World Health Organization: Geneva, Switzerland
operations: a review of sources, effects and regulations. Miner Eng Xia W, Wang Y, Chen S et al (2022) Double trouble of Air Pollution
18:1183–1199. https://doi.org/10.1016/j.mineng.2005.06.017 by Anthropogenic Dust. Environ Sci Technol 56:761–769. https://
Peterson BJ, Deegan L, Helfrich J et al (1993) Biological responses of doi.org/10.1021/acs.est.1c04779
a Tundra River to Fertilization. Ecology 74:653–672. https://doi. You M (2014) Addition of PM2.5 into the National Ambient Air Qual-
org/10.2307/1940794 ity standards of China and the contribution to Air Pollution Con-
Pope P F, Gichuru M, Ng’ang’a D et al (2018) Airborne particulate trol: the Case Study of Wuhan, China. Sci World J 2014:768405.
matter monitoring in Kenya using calibrated low-cost sensors. https://doi.org/10.1155/2014/768405
Atmos Chem Phys 18:15403–15418. https://doi.org/10.5194/ Yu H, Zahidi I (2023) Environmental hazards posed by mine dust, and
acp-18-15403-2018 monitoring method of mine dust pollution using remote sensing
Querol X, Tobías A, Pérez N et al (2019) Monitoring the impact of technologies: an overview. Sci Total Environ 864:161135. https://
desert dust outbreaks for air quality for health studies. Environ Int doi.org/10.1016/j.scitotenv.2022.161135
130:104867. https://doi.org/10.1016/j.envint.2019.05.061 Zhang X-X, Sharratt B, Chen X et al (2017) Dust deposition and ambi-
Rasmussen PE, Levesque C, Chénier M, Gardner HD (2018) Con- ent PM10 concentration in northwest China: spatial and tem-
tribution of metals in resuspended dust to indoor and personal poral variability. Atmos Chem Phys 17:1699–1711. https://doi.
inhalation exposures: relationships between PM10 and settled org/10.5194/acp-17-1699-2017
dust. Build Environ 143:513–522. https://doi.org/10.1016/j. Zhang G, Rui X, Fan Y (2018) Critical review of methods to Estimate
buildenv.2018.07.044 PM2.5 concentrations within specified Research Region. ISPRS
Ritchie H, Roser M (2019) Outdoor Air Pollution. Our World in Data Int J Geo-Information 7. https://doi.org/10.3390/ijgi7090368
Rolph G, Stein A, Stunder B (2017) Real-time Environmental Appli- Zhao X, Sun Y, Zhao C, Jiang H (2020) Impact of precipitation with
cations and Display sYstem: READY. Environ Model Softw different intensity on PM2.5 over typical regions of China. https://
95:210–228. https://doi.org/10.1016/j.envsoft.2017.06.025 doi.org/10.3390/atmos11090906. Atmosphere 11:
Salawu-Rotimi A, Lebre P, Vos HC et al (2021) Gone with the wind: Zheng N, Liu J, Wang Q, Liang Z (2010) Health risk assessment of
Microbial communities Associated with Dust from Emis- heavy metal exposure to street dust in the zinc smelting district,
sive Farmlands. Microb Ecol 1–11. https://doi.org/10.1007/ Northeast of China. Sci Total Environ 408:726–733. https://doi.
s00248-021-01717-8 org/10.1016/j.scitotenv.2009.10.075
Saldanha Bay Municipality (2023) First Review and Amendment of Zufall MJ, Dai W, Davidson CI (1999) Dry deposition of particles
the 5th Generation Integrated Development Plan 2022–2027. Sal- to wave surfaces: II. Wind tunnel experiments. Atmos Environ
danha Bay Municipality 33:4283–4290. https://doi.org/10.1016/S1352-2310(99)00178-8
Samset BH, Sand M, Smith CJ et al (2018) Climate impacts from a
removal of Anthropogenic Aerosol emissions. Geophys Res Lett Publisher’s Note Springer Nature remains neutral with regard to juris-
45:1020–1029. https://doi.org/10.1002/2017GL076079 dictional claims in published maps and institutional affiliations.
Saucy A, Röösli M, Künzli N et al (2018) Land Use Regression Mod-
elling of Outdoor NO2 and PM2.5 concentrations in three low
income areas in the Western Cape Province, South Africa. Int

13
Content courtesy of Springer Nature, terms of use apply. Rights reserved.
Terms and Conditions
Springer Nature journal content, brought to you courtesy of Springer Nature Customer Service Center GmbH (“Springer Nature”).
Springer Nature supports a reasonable amount of sharing of research papers by authors, subscribers and authorised users (“Users”), for small-
scale personal, non-commercial use provided that all copyright, trade and service marks and other proprietary notices are maintained. By
accessing, sharing, receiving or otherwise using the Springer Nature journal content you agree to these terms of use (“Terms”). For these
purposes, Springer Nature considers academic use (by researchers and students) to be non-commercial.
These Terms are supplementary and will apply in addition to any applicable website terms and conditions, a relevant site licence or a personal
subscription. These Terms will prevail over any conflict or ambiguity with regards to the relevant terms, a site licence or a personal subscription
(to the extent of the conflict or ambiguity only). For Creative Commons-licensed articles, the terms of the Creative Commons license used will
apply.
We collect and use personal data to provide access to the Springer Nature journal content. We may also use these personal data internally within
ResearchGate and Springer Nature and as agreed share it, in an anonymised way, for purposes of tracking, analysis and reporting. We will not
otherwise disclose your personal data outside the ResearchGate or the Springer Nature group of companies unless we have your permission as
detailed in the Privacy Policy.
While Users may use the Springer Nature journal content for small scale, personal non-commercial use, it is important to note that Users may
not:

1. use such content for the purpose of providing other users with access on a regular or large scale basis or as a means to circumvent access
control;
2. use such content where to do so would be considered a criminal or statutory offence in any jurisdiction, or gives rise to civil liability, or is
otherwise unlawful;
3. falsely or misleadingly imply or suggest endorsement, approval , sponsorship, or association unless explicitly agreed to by Springer Nature in
writing;
4. use bots or other automated methods to access the content or redirect messages
5. override any security feature or exclusionary protocol; or
6. share the content in order to create substitute for Springer Nature products or services or a systematic database of Springer Nature journal
content.
In line with the restriction against commercial use, Springer Nature does not permit the creation of a product or service that creates revenue,
royalties, rent or income from our content or its inclusion as part of a paid for service or for other commercial gain. Springer Nature journal
content cannot be used for inter-library loans and librarians may not upload Springer Nature journal content on a large scale into their, or any
other, institutional repository.
These terms of use are reviewed regularly and may be amended at any time. Springer Nature is not obligated to publish any information or
content on this website and may remove it or features or functionality at our sole discretion, at any time with or without notice. Springer Nature
may revoke this licence to you at any time and remove access to any copies of the Springer Nature journal content which have been saved.
To the fullest extent permitted by law, Springer Nature makes no warranties, representations or guarantees to Users, either express or implied
with respect to the Springer nature journal content and all parties disclaim and waive any implied warranties or warranties imposed by law,
including merchantability or fitness for any particular purpose.
Please note that these rights do not automatically extend to content, data or other material published by Springer Nature that may be licensed
from third parties.
If you would like to use or distribute our Springer Nature journal content to a wider audience or on a regular basis or in any other manner not
expressly permitted by these Terms, please contact Springer Nature at

[email protected]