Mineral dust is an important component of the climate
system, interacting with radiation, clouds, and biogeochemical systems and
impacting atmospheric circulation, air quality, aviation, and solar energy
generation. These impacts are sensitive to dust particle size distribution
(PSD), yet models struggle or even fail to represent coarse (diameter (d)
>2.5µm) and giant (d>20µm) dust
particles and the evolution of the PSD with transport. Here we examine three
state-of-the-art airborne observational datasets, all of which measured the
full size range of dust (d=0.1 to >100µm) at different
stages during transport with consistent instrumentation. We quantify the
presence and evolution of coarse and giant particles and their contribution
to optical properties using airborne observations over the Sahara (from the
Fennec field campaign) and in the Saharan Air Layer (SAL) over the tropical
eastern Atlantic (from the AER-D field campaign).
Observations show significantly more abundant coarse and giant dust
particles over the Sahara compared to the SAL: effective diameters of up to
20 µm were observed over the Sahara compared to 4 µm in the
SAL. Excluding giant particles over the Sahara results in significant
underestimation of mass concentration (40 %), as well as underestimates of
both shortwave and longwave extinction (18 % and 26 %, respectively, from
scattering calculations), while the effects in the SAL are smaller but
non-negligible. The larger impact on longwave extinction compared to
shortwave implies a bias towards a radiative cooling effect in dust models,
which typically exclude giant particles and underestimate coarse-mode
concentrations.
A compilation of the new and published effective diameters against dust age since uplift
time suggests that two regimes of dust transport exist. During the initial
1.5 d, both coarse and giant particles are rapidly deposited. During the
subsequent 1.5 to 10 d, PSD barely changes with transport, and the coarse
mode is retained to a much greater degree than expected from estimates of
gravitational sedimentation alone. The reasons for this are unclear and
warrant further investigation in order to improve dust transport schemes
and the associated radiative effects of coarse and giant particles in
models.
Introduction
Mineral dust aerosol is an important component of the climate system.
Between 1000 and 4000 Tg yr-1 of dust is uplifted annually, with
around 57 % of this originating from North Africa (Huneeus et al.,
2011; IPCC, 2013). Atmospheric mineral dust is estimated to account for
70 % of the global aerosol mass burden and 25 % of the global aerosol
optical depth (AOD) (Kinne et al., 2006). During atmospheric transport
and through subsequent deposition, dust exerts an impact the climate
system by interacting with both shortwave and longwave radiation (Tegen
and Lacis, 1996; Liao and Seinfeld, 1998). These radiative effects can
impact on the global energy balance, land and sea surface temperatures,
atmospheric heating, and thus circulation patterns. Impacts can be
particularly strong regionally where dust loadings are high, such as the
Sahara where dust affects North African atmospheric dynamics through
the Saharan heat low, Sahelian precipitation, and North Atlantic hurricane
development (e.g. Colarco et al., 2014; Pan et al., 2018; Lavaysse et al., 2011; Strong et al., 2018). Additionally, dust particles can impact
cloud development by acting as cloud condensation nuclei and ice nuclei
(Kumar et al., 2011; Hoose and Mohler, 2012). Dust can affect atmospheric
chemistry by providing a surface for heterogeneous reactions (Bauer et
al., 2004). Dust is deposited to the oceans and Amazon rainforest, providing
nutrients to a variety of ecosystems (Jickells et al., 2005; Yu et al.,
2015). Finally, dust is a natural hazard, having a negative impact on
aviation and transport (Weinzierl et al., 2012), solar energy generation,
air quality, and hence human health (Middleton et al., 2018).
The annual economic cost of dust storms may reach into the billions of US
dollars for certain countries (Middleton, 2017).
All of these impacts are sensitive to dust particle size (Mahowald et
al., 2014). For example, dust size distribution can affect cloud
interactions since smaller dust particles can be more hygroscopic
(Ibrahim et al., 2018), while larger particles can be
more effective cloud condensation nuclei (Petters and Kreidenweis,
2007). Size distribution also affects surface area and therefore ice
nucleation (Diehl et al., 2014). Larger particles contribute
more to dust mass, which controls the impact of dust on ocean and tropical
rainforest ecosystems (Jickells et al., 2005; Yu et al., 2015). A higher
proportion of fine particles will lead to elevated PM2.5 and subsequent
impacts on respiratory health (Middleton, 2017).
Dust optical properties are influenced by several factors, including
chemical composition, mixing state, particle shape, and size. Dust size
distribution has a strong impact on its radiative interactions (Tegen
and Lacis, 1996). In the shortwave spectrum, a larger coarse mode reduces
the single-scattering albedo (SSA) of dust, causing more absorption of solar
radiation and atmospheric heating. For example, Ryder et al. (2013b)
found that including the coarse and giant modes over the Sahara resulted in
the SSA dropping from 0.92 to 0.80, with an associated increase in
atmospheric heating by up to a factor of 3. In the longwave spectrum, larger
particles are able to exert a stronger radiative effect. For example,
Otto et al. (2011) show that including particles larger than 5 µm more than doubles the longwave aerosol optical depth (AOD). Together
these radiative effects can change the sign of the net radiative effect of
dust and the impact of dust on atmospheric circulation (Woodage and
Woodward, 2014; Strong et al., 2018). Given these impacts of dust size
distribution on climate and particularly radiation, it is important to have
the best possible observations of dust particle size distribution (PSD)
across all sizes to understand its vertical distribution through the
atmosphere and how these change with transport.
Airborne campaigns measuring size distributions of Saharan mineral
dust since 2006, showing maximum particle size measured and size
restrictions by inlets when instruments were located inside the aircraft
cabin. OPC size ranges are nominal diameters. Table reproduced from Ryder
et al. (2018). APS: aerodynamic particle sampler; CAS-DPOL: cloud and
aerosol spectrometer with depolarization detection; FSSP: forward-scattering
spectrometer probe; OAP: optical array probe; OPC: optical particle counter;
SID: small ice detector.
CampaignAcronymFieldwork dateLocationMeasurement upper size limit (µm)Instrument typeIn-cabin or wing-mountedDetailsPublicationDust and Biomass Burning ExperimentDABEXJanuary–February 2006Niger10OPCIn-cabinPCASP-X, behind a counterflow virtual impactor with significant pipework; loss of majority of coarse particlesOsborne et al. (2008)10Filter samplesIn-cabinInlet restricted measurements to 35 % of coarse mode (d>1.4µm)Chou et al. (2008)Dust Outflow and Deposition to the Ocean 2DODO2August 2006Tropical eastern Atlantic40OPCWing-mountedCDP measurements on a few flights only; otherwise size distributions up to 3 µmMcConnell et al. (2008)African Monsoon Multidisci-plinary AnalysisAMMAJune–July 2006Niger and Benin20OPCIn-cabinGrimm OPC behind isokinetic inlet with 50 % passing efficiency at 9 µmFormenti et al. (2011a)NASA AMMANAMMAAugust–September 2006Tropical eastern Atlantic5APSIn-cabinAPS behind an inlet with 50 % sampling efficiency at 5 µmChen et al. (2011)Saharan Mineral Dust Experiment 1SAMUM1May–June 2006Morocco30 / 100OPCsWing-mountedFSSP-300/FSSP-100Weinzierl et al. (2009)Geostationary Earth Radiation Budget Intercomparison of Longwave and Shortwave RadiationGERBILSJune 2007Mali, southern Mauritania60OPCWing-mountedSID-2; PSDs represent aged, trans-ported dust events with light dust loadingsJohnson and Osborne (2011)Saharan Mineral Dust Experiment 2SAMUM2January–February 2008Tropical eastern Atlantic30OPCWing-mountedFSSP-300Weinzierl et al. (2011)Fennec – the Saharan Climate SystemFennec-SaharaJune 2011Mali, Mauritania50 / 60 / 930OPCs and OAPsWing-mountedCDP/SID2/CIP15Ryder et al. (2013b)Fennec – the Saharan Climate SystemFennec-SALJune 2011Canary Islands, Fuerteventura50 / 60 / 930OPCs and OAPsWing-mountedCDP/SID2/CIP15(Ryder et al., 2013a)Aerosol Direct Radiative Impact on the RegionalClimate in the MEDiterranean RegionADRIMEDJune–July 2013Mediterranean Sea20OPCWing-mountedFSSP-300Denjean et al. (2016)Saharan Aerosol Long-Range Transport and Aerosol–CloudInteraction ExperimentSALTRACEJune–July 2013Tropical western Atlantic50 / 100OPCsWing-mountedCAS-DPOL/FSSP-100; some mea-surements additionally taken over the eastern tropical AtlanticWeinzierl et al. (2017)AERosol Properties – DustAER-DAugust 2015Tropical eastern Atlantic100OPCs and OAPsWing-mountedCDP, CIP15, and 2DS(Ryder et al., 2018)
Typically, dust models do not include particles larger than 20 µm
in diameter (Huneeus et al., 2011). Historically this has been because
larger particles have been assumed to be rapidly deposited. However, recent
work has shown that climate models face serious challenges in representing
the dust cycle adequately, part of which stems from accurately representing
dust PSDs. For example, Evan et al. (2014) find that CMIP5 climate
models underestimate the dust mass path (dust mass loading per square metre) by
a factor of 3, 66 % of which is due to a bias in size distribution skewed
towards smaller particles. Kok et al. (2017) found that by using an
observationally constrained dust emission PSD, global model calculations of
dust radiative forcing were more positive (-0.48 to +0.20 W m-2)
compared to previous estimates from AeroCom models (-0.6 to -0.3 W m-2)
wherein smaller, more cooling particles were overrepresented and coarser,
more warming particles were underestimated. As a result, observations of
dust which include the coarse mode are in demand (Formenti et al., 2011;
Ansmann et al., 2011, 2017; Samset et al., 2018) for model
validation. There are also implications for satellite optical models and
retrievals since these also rely on accurate aerosol optical properties,
which are affected by PSD.
Airborne observations are an important tool for probing the vertical
distribution of dust size and concentration. Historically, optical
measurement techniques have frequently been utilized, which require a
conversion of scattered signal to particle size and therefore incorporate
uncertainties due to particle refractive index, shape, and non-monotonic Mie
scattering (Ryder et al., 2015, 2013b; Walser et al.,
2017). Many earlier measurements of dust were also limited by the maximum
size measured (often not more than 10 µm in diameter) or by sampling
behind inlets, which restricted the maximum particle size and passing
efficiency (e.g. Ryder et al., 2018 and Table 1). In the last 10 years, airborne observations of dust have progressed to
measuring significantly larger particle sizes, often on wing probes which do
not suffer from inlet loss effects (Weinzierl et al., 2009; Ryder et al.,
2013b). More recently, light shadowing measurement techniques, which do not
require a scattering to size conversion, have been applied to particles
larger than 10 µm in diameter (Ryder et al., 2013b,
2018). Finally, airborne observations have taken place in more remote
Saharan regions where larger dust particles are more likely to be
prevalent (Ryder et al., 2015; Weinzierl et al., 2009).
As a result of these developments, observational campaigns have now shown
that coarse and giant dust particles are far more prevalent and transported
further and higher than previously thought. Fennec, SAMUM1, SAMUM2,
SALTRACE, AER-D, and ADRIMED have all reported a significant presence of
coarse to giant dust particles, despite the sampling locations of Saharan
dust ranging from very close to sources to thousands of kilometres away (see
Table 1 for field campaign acronyms and references).
Here we contrast state-of-the art airborne observations of dust size at two
stages representative of Saharan dust transport. We compare observations
over the Sahara from the Fennec fieldwork to observations over the tropical
eastern Atlantic within the Saharan Air Layer (SAL) from both AER-D and
Fennec fieldwork campaigns. These observations fully include the coarse and giant
modes of dust, measuring up to 100 µm for AER-D and 300 µm for
Fennec. Both observational campaigns use consistent instrumentation,
with wing probes and light shadowing techniques for the giant mode,
thus evading some of the historical measurement challenges in dust
observations. The Fennec dataset is particularly novel since it includes
observations within 12 h of dust uplift in remote Saharan locations, where
few other airborne measurements (if any) have been taken.
We contrast dust characteristics close to sources with those at the beginning
of trans-Atlantic transport. We present mean size distributions, vertical
distributions of size metrics, and the vertical distribution of mass
concentration for different size ranges, for some of which Fennec data have
not previously been published. We then calculate optical properties as a
function of size using the ambient number concentrations measured to
illustrate the contribution of coarse and giant particles with a range of
the latest refractive indices from the literature. We include longwave
scattering, which is frequently neglected. Finally, we put the Fennec and
AER-D size distributions and dust age into context with published airborne
observations to show the wider context of the transport of coarse and giant
particles.
Methods
In the literature “coarse” and “giant” aerosol
particles are not well defined. This is because the origins of aerosol mode
size terminology relate to broad size modes, partly overlapping in size,
relating to the aerosol generation mechanism, composition, and/or measurement
technique (Whitby, 1978; Kulkarni et al., 2011). For example, the lower
bound of the coarse-mode diameter has been defined as particles larger than
the following: 1 µm (Lohmann et al., 2016; Mahowald et al., 2014),
2 µm (Kulkarni et al., 2011), 2.5 µm (often relating
to PM2.5) (Neff et al., 2013; Seinfeld and Pandis, 2006; NASA, 2018), 5 µm (Kok et al., 2017), and 10 µm (Renard et al., 2018).
Similarly, giant particles are referred to as covering a wide size range
upwards of 20 µm (Feingold et al., 1999), 37.5 µm (Ryder et al., 2013a), 40 µm (Jaenicke and Schutz,
1978), 62.5 µm (Goudie and Middleton, 2001), and 75 µm (Betzer et al., 1988; Stevenson et al., 2015). Weinzierl et al. (2011) do not define giant particles but start counting “large coarse-mode”
dust particles upwards of 10 µm. Often the definitions of coarse and
giant particles are relative and case study or instrument specific. In this
paper we define the accumulation mode as 0.1<d<2.5µm, the coarse mode as d>2.5µm, and the giant mode as
d>20µm, since this is the diameter above which models
rarely incorporate dust (Huneeus et al., 2011). Henceforth in this
article, particle size is referred to in terms of diameter (d).
This work exploits airborne observations taken during the Fennec project
during June 2011 over both the Sahara and in the SAL in the vicinity
of the Canary Islands (Washington et al., 2012; Ryder et al., 2015),
with more recent measurements over the tropical Atlantic Ocean within the SAL during the
AER-D project in August 2015 (Ryder et al., 2018).
Figure 1 shows the location of the fieldwork. During
both fieldwork projects, the FAAM BAe146 research aircraft was deployed, and
size distributions of the full particle size distribution were measured by
wing probes (up to 300 µm during Fennec and up to 100 µm
during AER-D) using a passive cavity aerosol spectrometer probe (PCASP),
cloud droplet probe (CDP), and cloud-imaging probe 15 (CIP15) during Fennec
and a PCASP, CDP, and 2-D stereo probe (2DS) instruments during AER-D. Size
distributions from both field campaigns have already been published: full
descriptions of the instrumentation, uncertainties, and findings are
available for the Fennec observations over the Sahara (Fennec-Sahara:
Ryder et al., 2013b), the Fennec observations in the SAL (Fennec-SAL:
Ryder et al., 2013a), and the AER-D observations in the SAL between
Cape Verde and the Canary Islands (AER-D SAL: Ryder et al., 2018), as well
as specific flight locations, tracks, and details of dust events sampled.
For Fennec-Sahara and AER-D-SAL, observations from horizontal flight legs
are available (117 from Fennec-Sahara, 19 from AER-D-SAL), which capture
some of the spatial variability in dust properties. Horizontal flight leg
data are not available for Fennec-SAL, during which only take-off and landing
profile observations were made. For all three campaigns observations from
aircraft profiles are available (21 from Fennec-Sahara, 31 from AER-D-SAL,
21 from Fennec-SAL), which capture a more complete altitude range.
Fennec-Sahara profiles do not extend all the way to the surface due to
aircraft operating restrictions. In addition, both the Fennec-Sahara
horizontal flight legs and profiles are separated into fresh, aged, or
uncategorized dust events (see Sect. 2.3). Although
each campaign lasted only around 3 weeks, the data captured by each have been
shown to be climatologically representative (Ryder et al., 2015, 2018).
Besides presenting the nature of the full size distributions, we calculate
two size metrics representing the full PSD. These are the maximum size detected
(dmax) and the effective diameter (deff) calculated directly from the
aircraft-measured PSDs during horizontal flight legs. Effective diameter
(deff) is a commonly used metric (Hansen and Travis, 1974),
representing an area-weighted mean diameter; dmax was initially used by
Weinzierl et al. (2009) and is a useful indicator of the transport of the
largest sizes which dominate the mass fraction. Here we use a simple
estimation of dmax as described in Ryder et al. (2018), wherein
dmax represents the maximum particle size during a flight leg for which at
least four particles were detected within a single size bin. This implicitly
represents the maximum size measured when concentrations of dust exceed
10-5cm-3 (or 10 m-3) for a 20 min flight segment for a
particle size of 30 µm. Full details are provided in Ryder et al. (2018). We also provide dust mass profiles calculated using the measured
PSDs and assuming a density of 2.65 g cm-3 (Hess et al.,
1998), which is representative of quartz particles (Woodward, 2001;
Haywood et al., 2001; Kandler et al., 2009; Chen et al., 2011), taking data
from aircraft profiles. Finally, we also calculate the dust mass path (DMP) as
in Ryder et al. (2018): the vertically integrated mass of dust per unit
surface area, which has been used in satellite and model evaluations
(Evan et al., 2014). All size distributions, size metrics, and mass
concentrations are provided at ambient conditions.
Campaign ambient mean log-fit size distributions for Fennec-Sahara
(orange), AER-D SAL (black), and Fennec-SAL (blue). Bold lines indicate field
campaign mean PSDs, and shading indicates min : max range for SAL data and 10th
percentile : maximum range for Fennec-Sahara.
We provide mean size distributions for each fieldwork campaign, utilizing
the lognormal size distributions (since they are easily reproducible), as
well as their uncertainty ranges. For Fennec-Sahara and AER-D SAL, the
lognormal PSDs are taken from horizontal flight legs representing the range
of observations encountered, as shown in Fig. 2.
For Fennec-Sahara, lognormal PSDs are provided in Ryder et al. (2013b).
Here we use the mean log-fit curves, and as bounds of uncertainty on the PSD
we also use the maximum and 10th percentile log-fit curves (orange
shading in Fig. 2). The 10th percentile PSD
(data given in the Supplement) is selected as the lower bound since the minimum
curve for Fennec-Sahara presented in Ryder et al. (2013b) is an outlier
of one case with extremely low dust loadings. For AER-D-SAL, we use the mean
log-fit curve bounded by the minimum and maximum given in Ryder et al. (2018). For Fennec-SAL, only profile data are available (not horizontal
flight legs). Therefore, a log-fit curve is fitted to the mean observational
profile data from Ryder et al. (2013a) as shown by the blue line in
Fig. 2 (data available in the Supplement). The spread
of PSDs for Fennec-SAL (blue shading) is narrower compared to the other two
PSDs because the minimum and maximum represent the standard error of the
mean as given in Ryder et al. (2013a).
This article expands on the existing published work and data from Fennec and
AER-D. Our emphasis is on using the combination of data in the context of
transport time and vertical distribution. New data specifically include
the Fennec-SAL lognormal mean PSD and uncertainties, vertical distributions
of dmax for Fennec-Sahara, vertical distributions of deff for
Fennec-Sahara separated by fresh and aged dust events, vertical
distributions of mass concentration, and DMP for Fennec-Sahara and
Fennec-SAL.
Optical property calculations
In order to calculate dust optical properties, the Fennec and AER-D mean
lognormal size distributions (Sect. 2.1) are used
in combination with a range of literature refractive index (RI) data and a
Mie scattering code, implying a spherical assumption. Although observations
show that dust is not spherical, here we retain this simplification in order
to allow for a range of fast calculations and also because many climate models
assume spherical properties. In the longwave spectrum, the non-sphericity
effects of dust are not significant (Yang et al., 2007). Kok et al. (2017) show that dust non-sphericity increases shortwave extinction
efficiency by around 50 % for coarse particles, so therefore our results
represent a lower bound on the impact of the coarse mode in the solar
spectrum.
Dust spectral refractive index datasets from the literature.
Vertical lines indicate wavelengths of 0.55 and 10.8 µm. See the text for
dataset descriptions. Partial lines only provide a subset of spectral
refractive indices.
Spectral RI data, for which the real part represents scattering and the
imaginary part represents absorption, are taken from a range of sources. For
the full spectrum, RI data are available from the OPAC database
(Hess et al., 1998) based on values from d'Almeida et al. (1991) and Shettle and Fenn (1979), Volz (1973), and Balkanski et al. (2007), assuming a 1.5 % hematite content, as well as the
World Meteorological Organization (WMO, 1983) and Fouquart et al. (1987). For the shortwave spectrum RI
data are also available from Colarco et al. (2014), and for the longwave
spectrum data are available from Di Biagio et al. (2017), from which we have
selected the Mauritania subset as it is representative of the
middle of the range for their North Africa samples. Values are shown in
Fig. 3. At 0.55 µm these datasets yield real
values of 1.52–1.53 and imaginary components of 0.0015 to 0.0080. The
Balkanski et al. (2007) and Colarco et al. (2014) datasets
represent significantly more recent estimates of refractive index:
Balkanski et al. (2007) estimate refractive indices assuming a
central (1.5 %) content of hematite when hematite is embedded in a matrix
of clay, and RIs are calculated assuming a dielectric mixture. Colarco et al. (2014) combine refractive indices from Colarco et al. (2002) from
Total Ozone Mapping Spectrometer satellite retrievals at ultraviolet
wavelengths and Kim et al. (2011) from the AERosol Robotic NETwork
(AERONET) at visible wavelengths. Both of the latter two produce
significantly lower imaginary parts, 0.0015 and 0.0024 at 0.55 µm,
respectively, widely considered to be more appropriate for accurately
representing dust properties and consistent with recent observations
(Rocha-Lima et al., 2018). In the longwave spectrum there is more
variability between the RI datasets compared to the shortwave. We highlight
the use of the much more recent and higher-spectral-resolution Di Biagio
et al. (2017) dataset. The older (pre-2000) longwave datasets were limited
in applicability due to the fact that (1) they were collected at limited geographic locations,
(2) they are based on unknown mineral composition, (3) they may have been subject
to unknown physiochemical ageing, and (4) only Fouquart
et al. (1987) satisfy the Kramers–Kronig relationship (Di Biagio et
al., 2017).
In order to illustrate the impact of coarse particles on dust optical
properties, we firstly calculate optical properties for the three mean PSDs
and their uncertainties, which are calculated from the shaded PSD range
shown in Fig. 2 for each campaign and which represent
the variability in the PSD, as well as each of the refractive index datasets
described above. Secondly, optical properties are calculated with a
gradually incrementing maximum cut-off diameter for each PSD in order to
show how the optical properties depend on the maximum size considered and
how this differs for the three different PSDs measured during Fennec and
AER-D. This enables the contribution of coarse and giant particles to the
optical properties to be quantified. For these calculations only two
wavelengths are selected, 0.55 and 10.8 µm: 0.55 µm since it
represents the peak intensity of the solar radiation spectrum and 10.8 µm since extinction from dust at this wavelength is typically quite
high, it falls within the atmospheric window in which dust is able to exert a
strong radiative effect, it avoids ozone and water vapour absorption
channels, and it is also representative of one of the Spinning Enhanced
Visible and Infrared Imager (SEVIRI) dust red–green–blue (RGB) channels
(Brindley et al., 2012). Different thermal infrared wavelengths were also
tested, and sensitivity to chosen wavelength in the results in Sect. 3.2.2 was found to be low.
Estimation of dust age
Estimates of dust age for Fennec-Sahara and AER-D since uplift are taken
from Ryder et al. (2013b) and Ryder et al. (2018), respectively.
Briefly, for both campaigns, broad geographic dust source locations have
been identified using the SEVIRI dust RGB thermal infrared satellite imagery
product (Lensky and Rosenfeld, 2008). Dust events sampled by the
aircraft are tracked backwards in time visually, which allows for the determination
of dust uplift time and location and therefore dust age. For Fennec, this
technique was combined with back-trajectory analysis from the Hybrid
Single-Particle Lagrangian Integrated Trajectory model (HYSPLIT)
(Draxler and Hess, 1998) and from the FLEXible PARTicle dispersion model
(FLEXPART) (Stohl et al., 2005). For AER-D, every dust event
sampled could be linked to a haboob originating from a mesoscale convective
system. For AER-D, only SEVIRI imagery was used for dust source
identification since for each case HYSPLIT back trajectories indicated
different dust source locations, likely due to poor meteorological
representation over the Sahara when convection was important (Ryder et
al., 2018). Dust ages for Fennec-SAL are not included here since their
values have been found to cover an extremely large range of times (Ryder
et al., 2013a).
As in Ryder et al. (2013a, b), Fennec-Sahara data are also separated into
“fresh” and “aged” categories, wherein fresh represents dust sampled in under
12 h since uplift time. Of the 119 sampling legs performed, 22 were fresh,
55 aged, and the remainder uncategorized. Of the 21 Fennec-Sahara profiles,
5 were fresh and 16 aged.
The ages of two SALTRACE dust samples from Weinzierl et al. (2017)
measured over the western and eastern Atlantic were derived from new
backward simulations with the Lagrangian particle dispersion model FLEXPART
(Stohl et al., 1998, 2005; Seibert and Frank, 2004) using
meteorological fields from the European Centre for Medium-Range Weather
Forecasts ERA5 reanalysis (0.25∘, 1 h resolution) as input. A
generic aerosol species with a mean mass diameter of 7.9 µm and
logarithmic standard deviation of 2.5 was tracked back from the five
selected flight segments in each location, including the effects of
gravitational settling and dry and wet deposition. The model produced
source–receptor sensitivity values for a 50 m layer adjacent to the ground.
These sensitivities were multiplied with gridded, time-dependent dust
emissions from the Copernicus Atmosphere Monitoring Service global natural
emissions dataset to obtain the corresponding contribution to the mass. The
sum of the contributions over all grid cells at each of the time steps
produced is thus the simulated age distribution of the sampled dust aerosol.
For both the eastern and western observations, the flight legs have been
separated into five segments and the ages calculated separately for each. The
best estimate of the SALTRACE dust age is given by the median for the
segment with the highest receptor mass concentration, while the
uncertainties are given by the minimum and maximum 25th and 75th
percentile ages across all five segments.
ResultsSize distributions, mass concentration, and vertical distribution
The mean log-fit volume size distributions from Fennec and AER-D and their
variability are shown in Fig. 2. Overall,
Fig. 2 shows the following features, which will be
important later in terms of optical properties: a strong giant mode for
Fennec-Sahara and subsequent loss of this by Fennec-SAL and AER-D SAL; an
enhanced accumulation and coarse mode for AER-D SAL relative to
Fennec-Sahara and Fennec-SAL.
Variation of dust size with altitude from Fennec-Sahara and
AER-D-SAL, showing (a) maximum size detected (dmax) and (b) effective
diameter (deff). The deff uncertainties are 5 %, and the dmax
uncertainties are 10 µm for AER-D and 15 µm for Fennec. Data are
from horizontal flight legs (and therefore not available for Fennec-SAL).
As expected, over the Sahara the giant mode (d>20µm) is
enhanced compared to the SAL. The Fennec-Sahara PSD peaks at 20–30 µm, while the AER-D-SAL PSD peaks at ∼5µm, and the
Fennec-SAL PSD peaks at 10–12 µm. In these cases, this can be
explained by a greater dust age and distance from dust sources contributing
to the loss of the giant mode.
The accumulation and coarse mode are enhanced in AER-D-SAL compared to
Fennec-Sahara and Fennec-SAL, with higher concentrations below 10 µm.
However, we did not observe this enhancement when the same dust events were
observed in Fennec-Sahara and Fennec-SAL; rather, the accumulation and coarse
modes decreased in concentration from Fennec-Sahara to Fennec-SAL. The
AER-D-SAL accumulation- and coarse-mode enhancement may occur because AER-D
simply sampled more intense dust events, though this seems unlikely given
that the Fennec dust events were also often very intense and AODs were
mostly higher than AER-D (Ryder et al., 2015). This enhancement of the
accumulation mode is similar to differences between SAMUM1 (Morocco) and
SAMUM2 (Cape Verde region), for which enhancements in number concentration
between 0.3 and 4 µm during SAMUM2 were attributed to coagulational
growth (Weinzierl et al., 2011). A number of the AER-D data segments were
collected further south, closer to the intertropical convergence zone in
moister conditions. Therefore, another possibility is that hygroscopic growth
took place, although generally dust is considered unlikely to react
hygroscopically in this way (Denjean et al., 2015). Satellite imagery
indicated that clouds developed in the vicinity of every dust event sampled
during AER-D-SAL during transport over the Sahara. Therefore, there is a
possibility that the dust was affected by cloud or water vapour recycling
during its transport journey, which may have allowed for some form of
coagulation, potentially impacting the size distribution (Ryder et al.,
2015; Diaz-Hernandez and Sanchez-Navas, 2016; Weinzierl et al., 2011).
Another possibility is that a slight difference in the dust sources
activated between Fennec and AER-D led to different size distributions being
mobilized initially.
Figure 4 demonstrates how dust size for Fennec-Sahara
and AER-D-SAL changes with altitude (z) over the desert and in the SAL. AER-D
data points at z<100 m are marine boundary layer samples and are not
discussed. Both deff and dmax show much larger values at all
altitudes in Fennec-Sahara compared to AER-D-SAL. Over the Sahara deff
and dmax drop off sharply with altitude, while in the SAL they are more
homogeneous in altitude. For Fennec-Sahara dmax varied from 90 to 300 µm beneath 600 m, while above 3.5 km dmax varied from 15 to 180 µm. Contrastingly, values for AER-D-SAL were 20 to 80 µm.
Particles sized over 20 µm (100 µm) were detected in 99 %
(89 %) of the Fennec-Sahara dust layers, while particles sized over 20 µm were always present during AER-D-SAL, though particles as large as
100 µm were never detected. The impact of decreasing size with
increased transport can also be seen in Fig. 4b;
AER-D-SAL deff values are much lower than those for Fennec-Sahara, with
a range of 3.6 to 4.0 µm in the SAL compared to 1.8 to 20.5 µm
over the Sahara.
The largest deff and dmax values in Fig. 4 are clearly dominated by fresh dust events (under 12 h since uplift).
However, even for aged dust events (over 12 h since uplift, circles) very
large particles were encountered, including at high altitudes: for
Fennec-Sahara aged dust dmax reached 195 µm beneath 1.5 km and
210 µm above 1.5 km, while deff reached 10.7 µm beneath
1.5 km and 10.5 µm above 1.5 km. Aged deff values over the
Sahara are fairly homogeneous in the vertical. These large values at high
altitudes indicate that the coarse and giant dust particles are entrained
and transported in the atmosphere on longer than superficial timescales and
that for very fresh dust the coarse and giant mode are particularly enhanced
at low altitudes.
Weinzierl et al. (2011) performed a similar comparison of dmax between
SAMUM1 and SAMUM2. Their results are not directly comparable to ours due to
different instrumentation. However, relative altitude dependencies and
changes during transport can still be compared. During SAMUM1, dust was
well mixed vertically, showing no altitude dependence of size and being
similar to that of the aged dust from Fennec. Weinzierl et al. (2011) also
saw a decrease in dmax between dust closer to sources in SAMUM1 (90 %
of cases had particles larger than 20 µm) and low-altitude wintertime
dust sampled over the Atlantic in SAMUM2 (33 % of cases had particles
larger than 20 µm), similar to the dmax decreases between
Fennec-Sahara and AER-D-SAL.
Vertically resolved mass concentrations for Fennec-Sahara
(orange), Fennec-SAL (blue), and AER-D-SAL. (black) (a) Total mass
concentration across all sizes measured; (b) accumulation-mode mass
concentration d<2.5µm; (c) and (d) fraction of mass at
d>5µm (c) and d>20µm (d). Bold lines
and shading indicate the median and interquartile range, respectively. Data are
smoothed over 250 m intervals and for Fennec-Sahara are only available down to
350 m due to flight restrictions.
Figure 5 shows the vertically resolved mass
concentrations, since they are frequently used as a model diagnostic and
biogeochemical cycles and respiratory health are also impacted by dust mass.
Total mass concentrations (panel a) were notably higher at all altitudes
during Fennec-Sahara, gradually decreasing with altitude. In the SAL, mass
concentrations were lower, peaked between 2 and 4 km for
AER-D, and were extremely homogeneous in height for Fennec-SAL upwards of 1 km. Fennec-Sahara mass concentrations can be extremely high, especially at
lower altitudes, with the 75th percentile reaching values of up to 1940 µg m-3. Contrastingly, the mass concentration in the
accumulation mode (panel b) is highest during AER-D-SAL, which is a
reflection of the enhanced accumulation mode shown in
Fig. 2. For Fennec-Sahara, there is a sharp
increase in the accumulation-mode mass concentration beneath 1.4 km. Above
1.5 km, Fennec-SAL displays a similar profile to Fennec-Sahara, albeit in
lower concentrations in keeping with the reduced concentrations shown in
Fig. 2. Given that the World Health Organization
limits for air quality particulate matter for 24 h mean PM2.5
and PM10 are 25 and 50 µg m-3, respectively, the observations in
Fig. 5 are often well above these values,
reinforcing the hazardous nature of dust events.
In Fig. 5c and d the fraction of mass found at
sizes greater than 5 and 20 µm in diameter is shown. As in Ryder et al. (2018) these sizes are selected since they represent the diameters at which
models begin to underestimate the concentration of coarse particles (5 µm) and at which models have an upper limit (20 µm) (Kok
et al., 2017). It is clear in panel c that during Fennec-Sahara the vast
majority of dust mass was present at sizes greater than 5 µm (an
average of 93 % beneath 4.5 km), similar to Fennec-SAL (89 % between 1
and 5 km), and there is also a large amount during AER-D-SAL (61 % between 1 and 4 km in the SAL). Since models begin to underestimate dust concentration at
sizes above 5 µm in diameter, showing an underestimation by up to a
factor of 10 (Kok et al., 2017), a very large fraction of mass will be
neglected. Similarly, during Fennec-Sahara, sizes greater than 20 µm
in diameter were still found to contain 40 % of the dust mass beneath 4.5 km
(panel d) or up to 68 % for the 75th percentile. For AER-D-SAL and
Fennec-SAL, 2 % and 12 % of total mass, respectively, was found at these
large diameters, though the 75th percentile reaches up to 19 % and
56 %, respectively. Since 20 µm is typically the maximum diameter
represented by dust models, a large fraction of dust mass over the Sahara is
being completely excluded from models, and although the percentage of mass
found at sizes larger than 20 µm is fairly small on average,
individual event values can reach much higher values, which will also be
excluded by most models.
Mean DMPs are calculated at 3.2 g m-2 (0.8 to 12.1 g m-2) for
Fennec-Sahara, 1.5 g m-2 (0.2 to 6.2 g m-2) for AER-D-SAL, and 1.4 g m-2 (0.2 to 2.3 g m-2) for Fennec-SAL. As expected, mean values
over the Sahara are higher compared to the SAL. All these values are much
higher than those produced by models, such as the CMIP5 models analysed by
Evan et al. (2014) with values of 0.05 to 0.46 g m-2 and a
multi-model median of 0.26 g m-2 in the geographic region of the
AER-D-SAL observations. Although the aircraft data only represent periods of
around 3 weeks for each campaign, aerosol optical depths (AODs) were found
to be climatological (Ryder et al., 2013b, 2018), though
they do represent the dustier summer months, while the satellite and model
data referred to here are annual means. An unpublished analysis of
summertime-only DMPs from a subset of CMIP5 models suggests values higher by
around 35 % (personal communication, A. Evan, 2019) – not nearly enough to
reconcile the observational–model differences.
Calculated spectral extinction coefficient (Mm-1) (a) and
factor increase in extinction (b) between Fennec-Sahara (bold lines) and
AER-D-SAL (lightweight lines), calculated spectral absorption coefficient
(Mm-1) (c), and factor increase in absorption (d) between Fennec-Sahara
(bold lines) and AER-D-SAL (lightweight lines). Different colours indicate
different RI datasets as in the legend. Vertical lines indicate 0.55, 8.0,
9.6, and 10.8 µm wavelengths.
Optical propertiesSpectral optical properties
Figure 6a shows the spectral extinction coefficient
calculated from the campaign mean full PSDs shown in
Fig. 2 and the range of refractive index datasets
described in Sect. 2.2. For clarity only
Fennec-Sahara and AER-D-SAL are shown. In the shortwave spectrum, it is
clear that the size distribution difference between Fennec-Sahara and
AER-D-SAL dominates the impact on extinction, with the AER-D-SAL PSD
resulting in higher extinction due to the greater number concentration
between 0.5 and 8 µm in diameter in AER-D-SAL compared to Fennec-Sahara.
As a result, Fennec-Sahara extinction is a factor of 0.7 less than AER-D-SAL
(panel b). The extinction at these wavelengths is dominated by scattering
(as opposed to absorption). As the RI real parts (relevant for scattering)
are similar in all cases (even though the imaginary part varies) this causes
little difference to the total extinction, and therefore the size
distribution is the dominant influence on extinction.
However, in the longwave spectrum, both PSD and RI are important for
extinction. Different combinations of RI and PSD can give different spectral
variations in extinction. Overall, the Fennec-Sahara PSD produces a higher
extinction by up to a maximum factor of 3.3 for the Di Biagio RI dataset.
This is due to the increased scattering and absorption from the larger
particles in the Fennec-Sahara PSD. Interestingly, the application of the
Fennec-Sahara PSD rather than the AER-D-SAL PSD is to dampen the spectral
variability of extinction in the 7 to 12 µm spectral region: exactly
the region utilized by satellite retrievals to detect dust. Thus, similar to
Banks et al. (2018), we find that the coarsest dust may pose a challenge
to longwave satellite detection algorithms by allowing coarse dust to
effectively “hide”.
Figure 6c shows the spectral absorption coefficient
for the mean PSDs and each RI dataset. Across the shortwave spectrum in
general there is an increase in absorption for Fennec-Sahara compared to
AER-D-SAL, by up to a factor of 2 at a wavelength of 2 µm. This also
shows that in the shortwave, both RI and PSD impact the spectral SSA. In the
longwave spectrum, the sensitivity of absorption to variation in both PSD
and RI is similar to that seen for extinction: both are important. The
overall question of the relative contribution of PSD and RI uncertainty to
optical property uncertainty is a complex one and depends on the optical
property in question and the spectral range under consideration.
Size-resolved optical properties
So far, we have shown how the different PSDs contribute to different
spectral extinction properties. Next, we examine the size-resolved
contribution to the extinction coefficient at specific wavelengths (0.55 and
10.8 µm) in order to see how important the inclusion of a specific
size range is to the optical properties.
Size-resolved contribution to total absorption (thin lines) and
extinction coefficient (bold lines) calculated for AER-D-SAL (black),
Fennec-SAL (blue), and Fennec-Sahara (orange) at 0.55 µm using the
Colarco RI dataset. (a) Percentage contribution as a function of diameter,
(b) cumulative percentage extinction coefficient as a function of diameter,
and (c) cumulative percentage absorption coefficient as a function of diameter.
In (b) and (c), shading bounded by dashed lines shows the uncertainty due to
the range of RI datasets and PSD variability observed in each observational
campaign. Vertical lines indicate diameters of 2.5, 5, 10, 20, and 30 µm.
Figure 7 shows the shortwave size-resolved percentage
contribution to absorption (lightweight lines) and extinction (bold lines)
coefficients at 0.55 µm for three different PSDs (different colours).
In each case, the campaign mean PSD (as shown in
Fig. 2) and Colarco RI are used, as they represent
central values. This is shown both as a percentage contribution to the total
extinction (panel a) and cumulatively (panels b and c) to illustrate the
cut-off diameter at which the majority of the extinction is captured. Panel
a uses the Colarco RI exclusively, while in panels b and c the shading
represents the uncertainty for both the ranges of PSD shown in
Fig. 2 and the range of refractive indices tested.
For AER-D-SAL, Fig. 7a shows that the main
extinction contribution (thick black line) comes from particles sized around
1 and 3 µm. The scattering percentage contribution is not
shown because it is very similar to the extinction curve since the extinction
is dominated by scattering. However, the absorption (thin black line) is
dominated by a contribution from larger particles, with most absorption
coming from particles sized around 5 µm. The Fennec-Sahara PSD
(orange lines) shows an influence of much larger particles. In addition to
the peaks at 0.9 and 3 µm, the largest extinction comes from 14 µm diameter particles. Similarly for absorption (thin orange line),
the Fennec-Sahara optical properties are strongly dominated by the giant
mode, with a peak contribution from 20 µm diameter particles. The
properties of the Fennec-SAL dataset are between the other two datasets,
with peak contributions to extinction at the 10 µm diameter and peak
contributions to absorption at the 12 µm diameter. The size-resolved
extinction and absorption curves are a direct reflection of the shape and
abundance of the different PSDs shown in Fig. 2.
Percentage contribution to total shortwave scattering as well as absorption
and extinction coefficient at 0.55 µm, as a function of maximum
particle size considered, for the Fennec-Sahara, AER-D-SAL, and Fennec-SAL
mean size distributions using the Colarco et al. (2014) refractive index
dataset. Values correspond to data shown in Fig. 6. Uncertainties shown in
parentheses represent lower and upper values due to uncertainties in the PSD and
RI dataset.
Figure 7b and c clearly show that the cumulative
optical properties increase much more slowly as a function of diameter for
Fennec-Sahara compared to AER-D-SAL and Fennec-SAL due to the effect of the
greater concentration of giant particles in Fennec-Sahara. Only representing
dust particles sized up to 20 µm in diameter, as in many dust models,
results in 99 % (99 %–100 %) of extinction in AER-D-SAL and 96 %
(96 %–97 %) of extinction in Fennec-SAL but only 82 % (77 %–92 %) of the
extinction over the Sahara (Fennec-Sahara) (see also
Table 2). (Uncertainties are propagated from the
range of PSDs and RI datasets.) Besides the impacts on extinction, there are
impacts on absorption: representing only up to 20 µm diameter results
in 98 % (97 %–100 %) and 90 % (87 %–91 %) of absorption being
represented for AER-D-SAL and Fennec-SAL, respectively, but only 61 %
(52 %–82 %) of absorption being represented for Fennec-Sahara. Whilst total
extinction drives AOD, absorption drives shortwave atmospheric heating and
may subsequently impact regional circulation and the semi-direct effect. We
note that these figures are lower-bound estimates of the impact of neglected
absorption and extinction in dust models, since they only account for giant
particles being excluded and not any underestimation of the coarse mode,
which is included but poorly represented in models (e.g. Kok et al., 2017; Evan et al., 2014). It is also evident that by only representing
sizes up to 2.5 µm, the majority of extinction is omitted (only 27 %,
48 %, and 31 % of extinction for Fennec-Sahara, AER-D SAL, and Fennec-SAL,
respectively, is captured). This result emphasizes that it is crucial to
measure the coarse mode of dust aerosol in order to fully capture its
optical properties, and dust observations sampling only PM2.5 or behind
size-restricted aircraft inlets will not provide a realistic representation
of dust size and the associated optical properties.
Size-resolved contribution to total absorption (thin lines) and
extinction coefficient (bold lines) calculated for AER-D-SAL (black),
Fennec-SAL (blue), and Fennec-Sahara (orange) at 10.8 µm using the
Volz RI dataset. (a) Percentage contribution as a function of diameter, (b) cumulative percentage extinction as a function of diameter, and (c) cumulative
percentage absorption coefficient as a function of diameter. In (b) and (c), shading bounded by dashed lines shows the uncertainty due to the range
of RI datasets and PSD variability observed in each observational campaign.
Vertical lines indicate diameters of 2.5, 5, 10, 20, and 30 µm.
Figure 8 shows the size-resolved contribution to
optical properties but for a wavelength of 10.8 µm, representing the
longwave spectrum. As in Fig. 7, the three campaign
mean PSDs have been used (from Fig. 2) with the
Colarco RI. Panel a uses the Colarco RI exclusively, while in panels b and c
the shading represents the uncertainty for both the ranges of PSD shown in
Fig. 2 and the different RI datasets. In Fig. 8a,
for AER-D-SAL and Fennec-SAL, the main contribution to extinction comes from
particles sized around 6 and 10 µm in diameter, respectively,
while the main contribution for Fennec-Sahara comes from particles sized 13 µm in diameter. There is little difference in the relative contributions
from scattering and absorption at this wavelength, with both contributing
roughly equal amounts to the extinction (giving SSA values of 0.4–0.5).
Figure 8b shows the same results cumulatively for
extinction. As with the results from the shortwave spectrum, much of the
extinction for AER-D-SAL results from particles smaller than 10 µm
in diameter, while extinction for Fennec-SAL and Fennec-Sahara rises more
slowly as a function of maximum diameter. Representing particles up to 20 µm in diameter captures 98 % (98 %–100 %) and 94 % (91 %–94 %) of
the extinction for AER-D-SAL and Fennec-SAL, respectively, but only 74 %
(66 %–89 %) for Fennec-Sahara (see also Table 3);
i.e. 26 % (11 %–34 %) of extinction at a wavelength of 10.8 µm is
missed by not including any representation of giant dust particles over the
Sahara. Also, representing only up to 2.5 µm (such as done for PM2.5
observations or many observations behind aircraft inlets) results in only 2 %,
9 %, or 3 % (for Fennec-Sahara, AER-D SAL, and Fennec-SAL, respectively) of the
total extinction being captured.
Percentage contribution to total longwave scattering as well as absorption
and extinction coefficient at 10.8 µm, as a function of maximum
particle size considered, for the Fennec-Sahara, AER-D-SAL, and Fennec-SAL
mean size distributions using the Volz et al. (1973) refractive index
dataset. Values correspond to data shown in Fig. 7. Uncertainties shown in
parentheses represent lower and upper values due to uncertainties in the PSD and
RI dataset.
Lognormal ambient volume size distributions for recent airborne
campaigns measuring Saharan dust extending to sizes larger than 20 µm
diameter. Observations close to dust sources are coloured orange. AER-D SAL
mean and minimum–maximum envelope is shaded grey, Fennec-Sahara 10th
percentile–maximum envelope is shaded orange, and Fennec-SAL minimum–maximum
envelope is shaded blue as in Fig. 2. ADRIMED a and b represent dust above
3 km and beneath 3 km, respectively. SALTRACE E and W represent observations
over the eastern vs. western Atlantic. Lognormal curves are not shown at
sizes above which measurements were made. See Table 1 for references for each campaign. SAMUM2 data are provided at standard
temperature and pressure.
Sensitivity to the behaviour of the extinction curves at different wavelengths
was tested, but no significant differences in the size-resolved behaviour
was found, although the total extinction is different (as shown in
Fig. 6). The cumulative curves for extinction and
absorption at 10.8 µm (Fig. 8b and
c) are also very similar for the longwave, since the scattering curve is
similar to the absorption curve (in contrast to the shortwave spectrum).
This is consistent with Sicard et al. (2014), who showed that the effects
of dust LW scattering are significant and can cause up to a 50 %
underestimate in the dust radiative effect at the top of the atmosphere (TOA) if neglected (Dufresne
et al., 2002; Coelho, 2006).
The wider context of dust size and transport
Figure 9 compares the AER-D-SAL and Fennec PSDs to
previous aircraft observations of Saharan dust from the last 10 years, which
fully observed the presence of the coarse and giant modes, at least up to the 20 µm diameter: SAMUM1, SAMUM2, GERBILS, ADRIMED, and SALTRACE
observations over the eastern and western Atlantic (see
Table 1 for campaign references). For the SALTRACE
PSDs, the submicron and supermicron data shown in Weinzierl et al. (2017) have
been combined and collectively inverted, guaranteeing a consistent
propagation of measurement uncertainties (in optical particle counter-response, optical particle properties, etc.) for the complete size range.
Although other studies and fieldwork campaigns have also measured dust size
distributions, here we focus on the coarse and giant modes and therefore
only include studies which measured d>20µm (and
therefore do not include airborne observations from the DABEX, AMMA, and
NAMMA campaigns). Details of the instrumentation operated in each fieldwork
campaign, the relevant size limitations, and maximum size measured are
provided in Table 1. We do not extrapolate the PSD
modes beyond the size measured (e.g. 20 µm for ADRIMED).
Overall, although the size distribution of dust shown in
Fig. 9 varies, it is clear that there is always a
significant contribution from dust particles sized d>5µm, and when dust is closer to the source, there is also a strong
contribution from particles larger than 20 µm in diameter.
Clearly, the size distribution of Saharan dust can be highly variable.
However, the two campaigns measuring the greatest abundance of coarse and
giant particles with d>10µm were Fennec-Sahara and
SAMUM1, both taking observations in remote desert locations closer to dust
sources. Volume mean diameters (VMDs) calculated from the mean PSDs (or
envelope of PSDs for SAMUM) were also larger, at 21 µm for
Fennec-Sahara and 5–14 µm for SAMUM1. AER-D-SAL, GERBILS, SAMUM2,
Fennec-SAL, and SALTRACE, further afield from dust sources, measured fewer
giant particles, with maximum dV/ dlogD at around 3 to 5 µm. Giant
particles were present at 20–30 µm but vastly reduced in volume
concentration compared to Fennec-Sahara and SAMUM1. VMDs were lower at 3–4 µm (SAMUM2), 4 µm (GERBILS), 5.6 µm (AER-D-SAL), 12 µm (Fennec-SAL), and 10–12 µm (SALTRACE E and W). These values
represent the means of each campaign, and there will therefore be some
additional overlap due to instrumental uncertainties and spatial and
temporal variability within campaigns, though these data are not always
available from the individual publications.
SAMUM2 represents dust transported over the Atlantic during winter at low
altitudes. Although GERBILS observations were made over the west African
continent during summer, it is likely that the dust events sampled
represented aged regional dust with a depleted coarse mode (Haywood et al.,
2011; Johnson and Osborne, 2011). ADRIMED also represents transported dust,
but over the Mediterranean Sea. At diameters of 20 µm ADRIMED volume
concentrations are similar to AER-D-SAL and SAMUM2, with a suggestion of a
very large giant mode at even larger diameters (e.g. figures in Denjean
et al., 2016). AER-D-SAL also represents transported dust and accordingly
sits closer to GERBILS and SAMUM2 in Fig. 9 than to
Fennec-Sahara and SAMUM1.
Aircraft observations of effective diameter for the full size
distribution against dust age since uplift. (a) Fennec and AER-D: Fennec is
categorized by type of dust event (see Ryder et al., 2013b), and AER-D data are
separated by flight. (b) Saharan dust aircraft observations which fully
measured coarse-mode size distribution up to at least 20 µm in diameter;
deff is shown for the full size distribution or up to the maximum
measurement diameter. Fennec-Sahara data are from Ryder et al. (2013b) and
are identical to values shown in (a), but with data merged into fresh
and aged dust categories. AER-D-SAL data represent the range of
flight-by-flight data shown in (a). SAMUM1 data are from Weinzierl et al. (2009; Table 4). SAMUM2 data are from Weinzierl et al. (2011; Table 3).
ADRIMED data are calculated from lognormal size distribution parameters in
Denjean et al. (2016a) up to a maximum measurement size of 20 µm.
SALTRACE (E and W: east and west) data are new calculations based on flight
segments from Weinzierl et al. (2017). Data for (b) are given in
the Supplement.
Figure 10 shows dust effective diameters as a
function of estimated dust age since uplift. Firstly,
Fig. 10a shows Fennec-Sahara and AER-D-SAL
separated by dust events. Fennec-SAL is excluded because the range of dust
ages is too broad for it to be a useful addition (Ryder et al., 2013a).
During AER-D-SAL, the estimated dust age varied from 0.7 to 4.6 d, while
the range of effective diameters was very small, with flight means between
3.9 and 4.2 µm. Uncertainties in dust age for flights b928 and b934
are much larger due to the possibility of dust uplift from multiple sources
along the transport pathway. Despite AER-D-SAL flights measuring dust with a
range of transport times, the effective diameter showed only a variation of
5 % about the mean of 4.0 µm. This contrasts sharply with
observations of fresher dust from Fennec-Sahara for which deff showed a
decreasing trend with dust age. For Fennec-Sahara the freshest dust events
(under 12 h since uplift) had mean deff values of 8 to 13 µm,
dropping to a mean of 6 µm for dust aged around 2 d. The addition
of the data from AER-D-SAL suggests that in the bigger picture, dust size
distributions change rapidly following initial uplift and transport,
depositing some fraction of both coarse and giant particles, but after
around 2 d size distribution appears to stabilize.
Figure 10b shows deff against dust age since
uplift for a range of airborne fieldwork campaigns, after Ryder et al. (2013a) (their Fig. 11) and Denjean et al. (2016) (also their Fig. 11). However, here we show deff for the full size distribution (0.1 to
300 µm, or up to the maximum size measured in each campaign as shown
in Fig. 9), since dust particles are present in
both the submicron sizes (Formenti et al., 2011) and at d>20µm (in contrast to Denjean et al. (2016), wherein deff
representing solely 1–20 µm was presented, and consequently their
values are higher). GERBILS data yield a mean effective diameter of around 3 µm but are not included in Fig. 10b as no
estimate of dust age was provided, though dust was likely to be relatively
aged rather than fresh (personal communication, B. Johnson, 2017). This analysis is different
to previous compilations of dust size observations (e.g. Reid et al., 2008; Formenti et al., 2011) because we (1) relate dust size to time since
uplift, (2) only include airborne observations (since elevated dust
properties are often different to those at the surface), (3) only include
observations which measured at least up to the 20 µm diameter
unencumbered by inlet restrictions, and (4) incorporate more recent data,
particularly that from Fennec, which provides data from the remote Sahara
very close to dust uplift time, and SALTRACE, providing trans-Atlantic
observations.
Figure 10b shows that the stabilization of the size
distribution indicated in Fig. 10a still holds once
other airborne data are included. Very large particles are evident
immediately after uplift with high mean deff values of 6 to 10 µm; deff decreases rapidly until around 1.5 d after uplift, after
which the observations suggest little change in deff from around 2 d of transport onwards.
The range of deff values at over 1.5 d of transport in
Fig. 10b is fairly wide (from 1.4 to 5.2 µm). SAMUM2 data show a slightly lower mean deff value (2.4 µm)
compared to AER-D-SAL, ADRIMED, and SALTRACE (3.9 to 5.0 µm), though
this may be a result of SAMUM2 observations being taken in the winter season
when dust is transported by different meteorological mechanisms and uplifted
to lower altitudes over the Sahara (McConnell et al., 2008; Knippertz and
Todd, 2012; Tsamalis et al., 2013), which may influence size distribution
differences. Focusing solely on the summertime campaign data, the spread of
deff values is very narrow, even after 9 d of transport across the
Atlantic for SALTRACE-W, with deff of 4.1 µm.
The stabilization of the size distribution is contrary to what would be
expected from gravitational sedimentation theory. However, it is consistent
with the findings of now numerous publications of individual field campaign
dust size distributions, during which larger particles were observed than could be
explained by gravitational settling alone (Ryder et al., 2013a, 2018; Denjean
et al., 2016; Weinzierl et al., 2017; Stevenson et al., 2015; Gasteiger et
al., 2017; van der Does et al., 2018; Maring et al.,
2003). Ryder et al. (2013a) examined the mechanisms for transport between
fresh, aged, and SAL dust during Fennec-Sahara and found that sedimentation
and dispersion were able to account for the loss of the accumulation- and
giant-mode changes observed between the Saharan boundary layer and the SAL
during Fennec-Sahara but not for the coarse mode, which was retained to a
greater degree than expected. Gasteiger et al. (2017) developed a simplified
model for the long-range transport of Saharan dust aerosols over the
Atlantic Ocean that was consistent with observations. Their results suggest
that vertical mixing of the SAL air during the day (via convection caused by
the absorption of sunlight) was likely to be an important factor in
explaining the dust measurements at different stages of the transport.
Van der Does et al. (2018) examined potential mechanisms for the long-range
transport of giant dust particles and found it would be most likely under
highly optimal conditions incorporating high levels of turbulence and strong
winds, which may also allow for the electrical levitation of dust particles.
Recently, Harrison et al. (2018) have observed charged dust
during long-range transport to the UK, and Toth et al. (2019) and
Harrison et al. (2018) have shown that electric fields are able to
influence long-range-transported dust size distributions, enhancing the
coarse particle concentration. Long-range transport could be further
enhanced by repeated lifting of dust particles by deep convective clouds.
However, van der Does et al. (2018) stress that the details of these
mechanisms are mostly unquantified and require further research.
Denjean et al. (2016) suggest that during ADRIMED high turbulent updrafts and
downdrafts of up to 5 cm s-1 (from model simulations) enabled large-particle lifetime enhancement. During AER-D-SAL, measured vertical
velocities within the SAL were over ±30 cm s-1 in all cases and
sometimes up to ±80 cm s-1. During Fennec-Sahara, vertical
velocities were even larger: generally greater than 200 cm s-1 within
the convective boundary layer (consistent with values from Marsham et al., 2013) and frequently over 50 cm s-1 up to 5 km of altitude. The
gravitational settling velocity of a 10 µm diameter particle would be
1.1 and 28 cm s-1 for a 100 µm particle (Li
and Osada, 2007). Therefore, it appears possible that high levels of
atmospheric turbulence could have sustained the transport of larger particles
for longer than expected by gravitational sedimentation. Additionally,
during AER-D-SAL, vertical velocities were net positive in the SAL,
supporting the possibility of solar absorption by the dust particles
generating convection and daytime vertical mixing within the SAL
(Gasteiger et al., 2017). The more absorbing nature of coarser particles
in the solar spectrum would reinforce this mechanism.
Conclusions
Several airborne observational campaigns have recently revealed the
ubiquitous nature of coarse and giant dust particles within dusty air
masses. Here, we present mean PSDs and their uncertainties from one Saharan
dataset and two SAL datasets for which state-of-the art airborne measurements
with consistent instrumentation were performed. These have been used to
provide insights into how dust properties, particularly the coarse and
giant modes, change with transport and how this impacts optical properties.
We have contrasted the mean airborne ambient size distributions of dust
measured over the Sahara during the Fennec fieldwork (both over the Sahara
and in the SAL near the Canary Islands) with the more recent observations made
during the AER-D fieldwork within the SAL. The observations utilize light
shadowing techniques which allow for the measurement of giant-mode dust particles
and avert some of the historical challenges of airborne measurements of
dust. All datasets fully capture the coarse and giant dust particles up to
sizes of 100 µm (AER-D-SAL) and 300 µm (Fennec). As expected,
Fennec-Sahara shows a greater giant mode (d>20µm) than
AER-D-SAL and Fennec-SAL, but the AER-D-SAL mean PSD shows a greater volume
concentration at diameters smaller than 8 µm.
The vertical distribution of dust size shows that size distributions with an
extremely strong giant mode (displaying deff between 12 and 21 µm) are only observed at low altitudes over the Sahara (up to around 1 km)
and only for fresh events (under 12 h since uplift). However, for aged
events (longer than 12 h since uplift), giant particles are still present in
the PSD up to 5 km of altitude with large deff values of 5 to 10 µm. Effective diameters in AER-D-SAL were homogeneous at around 4 µm
throughout the SAL.
Models often use mass concentration as a diagnostic of aerosol amount,
and therefore we have provided these from observational data in order to
facilitate model validation studies. Mass concentration decreases with
height over the Sahara but is more homogeneous and well mixed in the
vertical in the SAL. Over the Sahara, 93 % of dust mass is constituted by
particles sized larger than 5 µm on average, and 40 % of dust mass
is constituted by particles sized larger than 20 µm. Since 5 and 20 µm are the diameters at which models begin to underestimate
coarse-mode concentrations and omit the giant mode, respectively, models will
be omitting a very large fraction of mass over the Sahara. During individual
events, models may be missing up to 60 % of mass by excluding dust sizes
greater than 20 µm. Over the SAL, the fraction of mass omitted is
smaller compared to the Sahara but potentially still important: 61 % to
89 % of dust mass is constituted by sizes over 5 µm and 2 % to
12 % by sizes over 20 µm. This misrepresentation of dust mass in
models will have a subsequent impact on the influence of dust in
biogeochemical cycles and on human health and air quality. Other processes,
which were not examined directly here, such as the role of coarse and giant
particles as ice-nucleating particles or cloud condensation nuclei, which
affect the impact of dust on cloud development, will also be affected by
model underrepresentation of coarse and giant dust particles.
The size-resolved contribution of the different PSDs to the extinction
coefficient has also been calculated. By excluding particles larger than 20 µm in diameter, as in many dust models, 18 % (8 %–23 %) of extinction
at a wavelength of 0.55 µm will be omitted over the Sahara and
1 %–4 % (0 %–4 %) will be omitted in the SAL. (Ranges correspond to mean
values for both SAL campaigns, and values in parentheses represent the range
of uncertainty due to both PSD variability and the RI dataset.) Similarly, for
absorption at 0.55 µm, excluding the giant mode will omit 39 %
(18 %–48 %) over the Sahara and 2 %–10 % (0 %–13 %) over the SAL. In the
longwave spectrum at 10.8 µm, we find that only representing
particles sized up to the 20 µm diameter omits 26 % (11 %–34 %) of the
extinction over the Sahara and 2 % to 6 % (0 %–9 %) of the extinction over
the SAL.
The extinction coefficient profile determines the aerosol optical depth and
the direct radiative effect of dust, while the absorption profile determines
the semi-direct effect, impacts dust-driven shortwave atmospheric heating,
and may subsequently impact regional circulation (Perlwitz and Miller,
2010; Solmon et al., 2012; Woodage and Woodward, 2014). Our results suggest
that the missing extinction and absorption in models will therefore alter
the impact of dust in models. Omitting the giant mode results in a greater
omission of the longwave extinction than of the shortwave. Additionally, in
the shortwave, the omission of absorption from the giant mode has the most impact.
Since both these processes lead to a warming of the Earth–atmosphere system,
this suggests that models are likely to be underestimating the warming
influence of dust, with the radiative forcing due to aerosol(dust)–radiation interactions estimated to be -0.1 W m-2 (-0.3 to +0.1)
in the latest IPCC report (IPCC, 2013).
Additionally, these figures are lower-bound estimates of the impact of
neglected absorption and extinction in dust models, since they only account
for giant particles being excluded and not any additional underestimation
of the coarse mode, which is included but poorly represented in models (e.g.
Kok et al., 2017; Evan et al., 2014). Both excluding giant particles and underrepresenting the concentrations of coarse and giant particles will
lead to more important consequences over the Sahara compared to in the SAL.
This work makes the assumption that dust particles are spherical for the
optical property calculations in order to enable multiple rapid
computations. This assumption is likely to have little impact in the
longwave spectrum, since the size parameter is smaller. In the shortwave,
our results represent a lower bound for the impact of the coarser dust:
Kok et al. (2017) show that non-spherical dust increases extinction
efficiency by 50 % for coarse particles. Additionally, most climate models
still assume spherical dust properties. Measuring the aspect ratio across the
full size range from in situ measurements remains a challenging process. For
the field campaigns studied here, aspect ratios were available only for a
few samples from AER-D (Ryder et al., 2018), and future work will consider
dust shape during Fennec. We emphasize the need for further work to obtain
observations of dust particle shape, particularly across the full size range
of dust as presented here, and to calculate the optical properties for
non-spherical dust across all size and spectral ranges, which requires
extensive computing resources.
Another important factor for consideration is that the Fennec and AER-D
observations are taken in summertime when Saharan and SAL dust loadings are
at a maximum, and coarse and giant particles are also present in a greater
fraction due to strong convection lifting dust up to high altitudes over
the Sahara, enabling further transport of the larger dust particles (e.g.
McConnell et al., 2008; van der Does et al., 2016). This is also
reflected in the slightly lower sizes seen in SAMUM2 during winter.
Therefore, the impact of coarse and giant dust particles on mass
concentrations and radiative effects presented here should be viewed as an
upper bound within the seasonal cycle of dust.
Overall, the three main uncertainties impacting this work are the exclusion
of any underestimation of the coarse mode (defined here as 2.5<d<20µm) by models (in addition to the exclusion of the giant
mode, d>20µm), a spherical assumption for scattering
calculations, and the use of data based on summertime dust transport. The
former two mean that our results of the impact of coarse and giant dust
particles are underestimates, while the latter means our results are
overestimates compared to an annual average.
Finally, we put the Fennec-Sahara and AER-D-SAL PSDs in the context of other
airborne campaigns of the last 10 years which have measured Saharan dust
and included measurements larger than 10 µm in diameter. The two sets of
dust observations closest to dust sources, Fennec-Sahara and SAMUM1, show a
clear presence of giant particles influencing the shape of the PSDs, while
those measuring transported dust showed a steeper drop-off of the PSD and
lower total concentrations. Despite this, there is still a significant
presence of coarse and giant particles in the “transported” size
distributions. Evaluating the effective diameter for each field campaign against
dust age since uplift time reveals what appear to be two regimes of dust
transport: firstly, deff drops off rapidly during initial transport
within the first 36 h, and secondly, deff appears very stable
despite significant amounts of transport between around 2 and 10 d.
It is clear that mineral dust coarse and giant modes are retained to a much
greater degree than expected from gravitational sedimentation alone. The
processes behind this are still unclear (e.g. van der Does et al., 2018). Potential explanations which warrant further study include
variations in fall speed dependent on particle composition, density, shape
and orientation, turbulent and convective mixing, triboelectric charging, and
radiative lofting impacts of the coarse and giant particles. Similar
processes and uncertainties also apply to the atmospheric transport of volcanic
ash, wherein similar unexplained long-range transport of coarse and giant
particles has been observed (e.g. Stevenson et al., 2015; Beckett et al., 2015; Saxby et al., 2018).
Overall, climate models generally do not incorporate dust particles sized
over 20 µm. Historically this has been because of the assumption that
larger particles are deposited rapidly. This work suggests that although
particles larger than 20 µm do exist up to high altitudes even in
transported dust, it is over the Sahara that the contribution of this size
range to total mass, absorption, and extinction are most significant. For
transported dust in the SAL, the size distribution has evolved such that the
giant particles contribute only a small amount to total extinction and dust
mass concentration. However, models begin to underestimate dust
concentrations at sizes well below this, from 5 µm upwards. Our
results show that dust particles in this size range (diameters 5 to 20 µm) are still highly prevalent and contribute a large amount to
extinction and dust mass in the SAL as well as over the Sahara, so
better representation of the coarse-mode size distribution within dust
models is also an area for improvement.
In the absence of other mechanisms and explanations, it is natural that to
date climate models have employed some form of gravitational settling for the dry
deposition of dust. However, other mechanisms must be occurring in the real
world in order to transport coarse and giant particles as far and for as
long as detected in observations. Therefore, further work, ideally combining
observations and modelling efforts, in order to explain this transport is
required.
Data availability
We are in the process of uploading the campaign mean data presented here to
the Centre for Environmental Data Analysis (CEDA). Flight-by-flight aircraft
data are publicly available at https://catalogue.ceda.ac.uk/uuid/affe775e8d8890a4556aec5bc4e0b45c, last access: 2 December 2019, Smith, 2004.
The supplement related to this article is available online at: https://doi.org/10.5194/acp-19-15353-2019-supplement.
Author contributions
CLR designed and carried out the analysis and wrote the paper. EJH
discussed the methodology and results. SALTRACE size distributions were
provided by AW and BW. SALTRACE dust age estimates were provided by PS and
AP. All authors read and commented on the paper.
Competing interests
The authors declare that they have no conflict of interest.
Acknowledgements
FLEXPART output was generated using ERA5 data (Copernicus Climate Change Service, 2018) accessed through the ECMWF's Meteorological Archival and
Retrieval System (MARS). SALTRACE dust age estimates were calculated using
Copernicus Atmosphere Monitoring Service (2018) information. Petra Seibert and Anne Philipp thank
the Austrian Meteorological Service, ZAMG, for access to MARS. Bernadett Weinzierl, Anne Philipp, and Adrian Walser
were funded by the European Research Council (ERC) under the European
Union's Horizon 2020 research and innovation framework programme under grant
agreement no. 640458 (A-LIFE). The SALTRACE research flights were funded by
the Helmholtz Association under grant VH-NG-606
(Helmholtz-Hochschul-Nachwuchsforschergruppe AerCARE) and by DLR. The
authors are grateful to Margaret Woodage for comments on the paper and James Banks for discussions relating to longwave dust radiative interactions.
Financial support
This research has been supported by a NERC independent research fellowship grant (grant no. NE/M018288/1).
Review statement
This paper was edited by Stelios Kazadzis and reviewed by three anonymous referees.
ReferencesAnsmann, A., Petzold, A., Kandler, K., Tegen, I., Wendisch, M., Müller,
D., Weinzierl, B., Müller, T., and Heintzenberg, J.: Saharan Mineral
Dust Experiments SAMUM-1 and SAMUM-2: what have we learned?, Tellus B, 63,
403–429, 10.1111/j.1600-0889.2011.00555.x, 2011.Ansmann, A., Rittmeister, F., Engelmann, R., Basart, S., Jorba, O., Spyrou, C., Remy, S., Skupin, A., Baars, H., Seifert, P., Senf, F., and Kanitz, T.: Profiling of Saharan dust from the Caribbean to western Africa – Part 2: Shipborne lidar measurements versus forecasts, Atmos. Chem. Phys., 17, 14987–15006, 10.5194/acp-17-14987-2017, 2017.Balkanski, Y., Schulz, M., Claquin, T., and Guibert, S.: Reevaluation of Mineral aerosol radiative forcings suggests a better agreement with satellite and AERONET data, Atmos. Chem. Phys., 7, 81–95, 10.5194/acp-7-81-2007, 2007.Banks, J. R., Schepanski, K., Heinold, B., Hünerbein, A., and Brindley, H. E.: The influence of dust optical properties on the colour of simulated MSG-SEVIRI Desert Dust infrared imagery, Atmos. Chem. Phys., 18, 9681–9703, 10.5194/acp-18-9681-2018, 2018.Bauer, S. E., Balkanski, Y., Schulz, M., Hauglustaine, D. A., and Dentener,
F.: Global modeling of heterogeneous chemistry on mineral aerosol surfaces:
Influence on tropospheric ozone chemistry and comparison to observations, J.
Geophys. Res.-Atmos., 109, D02304, 10.1029/2003jd003868, 2004.Beckett, F. M., Witham, C. S., Hort, M. C., Stevenson, J. A., Bonadonna, C.,
and Millington, S. C.: Sensitivity of dispersion model forecasts of volcanic
ash clouds to the physical characteristics of the particles, J. Geophys.
Res.-Atmos., 120, 11636–11652, 10.1002/2015jd023609, 2015.Betzer, P. R., Carder, K. L., Duce, R. A., Merrill, J. T., Tindale, N. W.,
Uematsu, M., Costello, D. K., Young, R. W., Feely, R. A., Breland, J. A.,
Bernstein, R. E., and Greco, A. M.: Long-Range Transport of Giant Mineral
Aerosol-Particles, Nature, 336, 568–571, 10.1038/336568a0, 1988.Brindley, H., Knippertz, P., Ryder, C., and Ashpole, I.: A critical
evaluation of the ability of the Spinning Enhanced Visible and Infrared
Imager (SEVIRI) thermal infrared red-green-blue rendering to identify dust
events: Theoretical analysis, J. Geophys. Res.-Atmos., 117, D07201, 10.1029/2011jd017326, 2012.Chen, G., Ziemba, L. D., Chu, D. A., Thornhill, K. L., Schuster, G. L., Winstead, E. L., Diskin, G. S., Ferrare, R. A., Burton, S. P., Ismail, S., Kooi, S. A., Omar, A. H., Slusher, D. L., Kleb, M. M., Reid, J. S., Twohy, C. H., Zhang, H., and Anderson, B. E.: Observations of Saharan dust microphysical and optical properties from the Eastern Atlantic during NAMMA airborne field campaign, Atmos. Chem. Phys., 11, 723–740, 10.5194/acp-11-723-2011, 2011.
Coelho, D. C.: A New Estimate of the Components of the Earth's Longwave
Radiation Budget, PhD, Department of Meteorology, University of Reading, 89–116,
2006.Colarco, P. R., Toon, O. B., Torres, O., and Rasch, P. J.: Determining the
UV imaginary index of refraction of Saharan dust particles from Total Ozone
Mapping Spectrometer data using a three-dimensional model of dust transport,
J. Geophys. Res.-Atmos., 107, D16, 4289, 10.1029/2001jd000903, 2002.Colarco, P. R., Nowottnick, E. P., Randles, C. A., Yi, B. Q., Yang, P., Kim,
K. M., Smith, J. A., and Bardeen, C. G.: Impact of radiatively interactive
dust aerosols in the NASA GEOS-5 climate model: Sensitivity to dust particle
shape and refractive index, J. Geophys. Res.-Atmos., 119, 753–786,
10.1002/2013jd020046, 2014.
d'Almeida, G. A., Koepke, P., and Shettle, E. P.: Atmospheric Aerosols:
Global Climatology and Radiation Characteristics, A Deepak Pub, Hampton, VA,
USA, 561 pp, 1991.Denjean, C., Caquineau, S., Desboeufs, K., Laurent, B., Maille, M., Rosado,
M. Q., Vallejo, P., Mayol-Bracero, O. L., and Formenti, P.: Long-range
transport across the Atlantic in summertime does not enhance the
hygroscopicity of African mineral dust, Geophys. Res. Lett., 42, 7835–7843,
10.1002/2015gl065693, 2015.Denjean, C., Cassola, F., Mazzino, A., Triquet, S., Chevaillier, S., Grand, N., Bourrianne, T., Momboisse, G., Sellegri, K., Schwarzenbock, A., Freney, E., Mallet, M., and Formenti, P.: Size distribution and optical properties of mineral dust aerosols transported in the western Mediterranean, Atmos. Chem. Phys., 16, 1081–1104, 10.5194/acp-16-1081-2016, 2016.Di Biagio, C., Formenti, P., Balkanski, Y., Caponi, L., Cazaunau, M., Pangui, E., Journet, E., Nowak, S., Caquineau, S., Andreae, M. O., Kandler, K., Saeed, T., Piketh, S., Seibert, D., Williams, E., and Doussin, J.-F.: Global scale variability of the mineral dust long-wave refractive index: a new dataset of in situ measurements for climate modeling and remote sensing, Atmos. Chem. Phys., 17, 1901–1929, 10.5194/acp-17-1901-2017, 2017.Diaz-Hernandez, J. L. and Sanchez-Navas, A.: Saharan dust outbreaks and
iberulite episodes, J. Geophys. Res.-Atmos., 121, 7064–7078,
10.1002/2016jd024913, 2016.Diehl, K., Debertshäuser, M., Eppers, O., Schmithüsen, H., Mitra, S. K., and Borrmann, S.: Particle surface area dependence of mineral dust in immersion freezing mode: investigations with freely suspended drops in an acoustic levitator and a vertical wind tunnel, Atmos. Chem. Phys., 14, 12343–12355, 10.5194/acp-14-12343-2014, 2014.
Draxler, R. R. and Hess, G. D.: An overview of the HYSPLIT_4
modeling system of trajectories, dispersion, and deposition, Aust. Meteorol. Mag., 47, 295–308, 1998.Dufresne, J. L., Gautier, C., Ricchiazzi, P., and Fouquart, Y.: Longwave
scattering effects of mineral aerosols, J. Atmos. Sci., 59, 1959–1966, 10.1175/1520-0469(2002)059<1959:Lseoma>2.0.Co;2, 2002.Evan, A. T., Flamant, C., Fiedler, S., and Doherty, O.: An analysis of
aeolian dust in climate models, Geophys. Res. Lett., 41, 5996–6001,
10.1002/2014gl060545, 2014.Feingold, G., Cotton, W. R., Kreidenweis, S. M., and Davis, J. T.: The
impact of giant cloud condensation nuclei on drizzle formation in
stratocumulus: Implications for cloud radiative properties, J. Atmos. Sci., 56,
4100–4117, 10.1175/1520-0469(1999)056<4100:Tiogcc>2.0.Co;2, 1999.Formenti, P., Schütz, L., Balkanski, Y., Desboeufs, K., Ebert, M., Kandler, K., Petzold, A., Scheuvens, D., Weinbruch, S., and Zhang, D.: Recent progress in understanding physical and chemical properties of African and Asian mineral dust, Atmos. Chem. Phys., 11, 8231–8256, 10.5194/acp-11-8231-2011, 2011.Fouquart, Y., Bonnel, B., Brogniez, G., Buriez, J. C., Smith, L., Morcrette,
J. J., and Cerf, A.: Observations of Saharan Aerosols - Results of Eclats
Field Experiment, 2. Broad-Band Radiative Characteristics of the Aerosols
and Vertical Radiative Flux Divergence, J. Clim. Appl. Meteorol., 26, 38–52, 10.1175/1520-0450(1987)026<0038:Oosaro>2.0.Co;2, 1987.Gasteiger, J., Groß, S., Sauer, D., Haarig, M., Ansmann, A., and Weinzierl, B.: Particle settling and vertical mixing in the Saharan Air Layer as seen from an integrated model, lidar, and in situ perspective, Atmos. Chem. Phys., 17, 297–311, 10.5194/acp-17-297-2017, 2017.Goudie, A. S. and Middleton, N. J.: Saharan dust storms: nature and
consequences, Earth.-Sci. Rev., 56, 179–204, 10.1016/S0012-8252(01)00067-8,
2001.Hansen, J. E. and Travis, L. D.: Light-Scattering in Planetary Atmospheres,
Space Sci. Rev., 16, 527–610, 10.1007/Bf00168069, 1974.Harrison, R. G., Nicoll, K. A., Marlton, G. J., Ryder, C. L., and Bennett,
A. J.: Saharan dust plume charging observed over the UK, Environ. Res. Lett.,
13, 054018, 10.1088/1748-9326/aabcd9, 2018.Haywood, J. M., Francis, P. N., Glew, M. D., and Taylor, J. P.: Optical
properties and direct radiative effect of Saharan dust: A case study of two
Saharan dust outbreaks using aircraft data, J. Geophys. Res.-Atmos., 106,
18417–18430, 10.1029/2000jd900319, 2001.Hess, M., Koepke, P., and Schult, I.: Optical properties of aerosols and
clouds: The software package OPAC, B. Am. Meteorol. Soc., 79, 831–844, 10.1175/1520-0477(1998)079<0831:Opoaac>2.0.Co;2, 1998.Hoose, C. and Möhler, O.: Heterogeneous ice nucleation on atmospheric aerosols: a review of results from laboratory experiments, Atmos. Chem. Phys., 12, 9817–9854, 10.5194/acp-12-9817-2012, 2012.Huneeus, N., Schulz, M., Balkanski, Y., Griesfeller, J., Prospero, J., Kinne, S., Bauer, S., Boucher, O., Chin, M., Dentener, F., Diehl, T., Easter, R., Fillmore, D., Ghan, S., Ginoux, P., Grini, A., Horowitz, L., Koch, D., Krol, M. C., Landing, W., Liu, X., Mahowald, N., Miller, R., Morcrette, J.-J., Myhre, G., Penner, J., Perlwitz, J., Stier, P., Takemura, T., and Zender, C. S.: Global dust model intercomparison in AeroCom phase I, Atmos. Chem. Phys., 11, 7781–7816, 10.5194/acp-11-7781-2011, 2011.Ibrahim, S., Romanias, M. N., Alleman, L. Y., Zeineddine, M. N., Angeli, G.
K., Trikalitis, P. N., and Thevenet, F.: Water Interaction with Mineral Dust
Aerosol: Particle Size and Hygroscopic Properties of Dust, ACS Earth Space
Chem., 2, 376–386, 10.1021/acsearthspacechem.7b00152, 2018.
IPCC: IPCC, 2013: Summary for Policymakers, in: Climate Change 2013: The
Physical Science Basis, Contribution of Working Group I to the Fifth
Assessment Report of the Intergovernmental Panel on Climate Change
Cambridge, 29 pp., 2013.Jaenicke, R. and Schutz, L.: Comprehensive Study of Physical and Chemical
Properties of Surface Aerosols in Cape-Verde-Islands Region, J. Geophys.
Res.-Oc. Atm., 83, 3585–3599, 10.1029/JC083iC07p03585, 1978.
Jickells, T. D., An, Z. S., Andersen, K. K., Baker, A. R., Bergametti, G.,
Brooks, N., Cao, J. J., Boyd, P. W., Duce, R. A., Hunter, K. A., Kawahata,
H., Kubilay, N., laRoche, J., Liss, P. S., Mahowald, N., Prospero, J. M.,
Ridgwell, A. J., Tegen, I., and Torres, R.: Global iron connections between
desert dust, ocean biogeochemistry, and climate, Science, 308, 67–71, 2005.Johnson, B. T. and Osborne, S. R.: Physical and optical properties of
mineral dust aerosol measured by aircraft during the GERBILS campaign, Q. J. Roy. Meteor. Soc., 137, 1117–1130, 10.1002/Qj.777, 2011.Kandler, K., Schutz, L., Deutscher, C., Ebert, M., Hofmann, H., Jäckel,
S., Jaenicke, R., Knippertz, P., Lieke, K., Massling, A., Petzold, A.,
Schladitz, A., Weinzierl, B., Wiedensohler, A., Zorn, S., and Weinbruch, S.:
Size distribution, mass concentration, chemical and mineralogical
composition and derived optical parameters of the boundary layer aerosol at
Tinfou, Morocco, during SAMUM 2006, Tellus B, 61, 32–50, 10.1111/j.1600-0889.2008.00385.x, 2009.Kim, D., Chin, M., Yu, H., Eck, T. F., Sinyuk, A., Smirnov, A., and Holben, B. N.: Dust optical properties over North Africa and Arabian Peninsula derived from the AERONET dataset, Atmos. Chem. Phys., 11, 10733–10741, 10.5194/acp-11-10733-2011, 2011.Kinne, S., Schulz, M., Textor, C., Guibert, S., Balkanski, Y., Bauer, S. E., Berntsen, T., Berglen, T. F., Boucher, O., Chin, M., Collins, W., Dentener, F., Diehl, T., Easter, R., Feichter, J., Fillmore, D., Ghan, S., Ginoux, P., Gong, S., Grini, A., Hendricks, J., Herzog, M., Horowitz, L., Isaksen, I., Iversen, T., Kirkevåg, A., Kloster, S., Koch, D., Kristjansson, J. E., Krol, M., Lauer, A., Lamarque, J. F., Lesins, G., Liu, X., Lohmann, U., Montanaro, V., Myhre, G., Penner, J., Pitari, G., Reddy, S., Seland, O., Stier, P., Takemura, T., and Tie, X.: An AeroCom initial assessment – optical properties in aerosol component modules of global models, Atmos. Chem. Phys., 6, 1815–1834, 10.5194/acp-6-1815-2006, 2006.Knippertz, P. and Todd, M. C.: Mineral Dust Aerosols over the Sahara:
Meteorological Controls on Emission and Transport and Implications for
Modeling, Rev. Geophys., 50, 2011RG000362, 10.1029/2011rg000362, 2012.Kok, J. F., Ridley, D. A., Zhou, Q., Miller, R. L., Zhao, C., Heald, C. L.,
Ward, D. S., Albani, S., and Haustein, K.: Smaller desert dust cooling
effect estimated from analysis of dust size and abundance, Nat. Geosci., 10,
274–278, 10.1038/Ngeo2912, 2017.
Kulkarni, P., Baron, P. A., and Willeke, K.: in: Aerosol Measurement:
Principles, Techniques, and Applications, 3 Edn., edited by: Kulkarni, P.,
John Wiley & Sons, Hoboken, New Jersey, 3–10, 2011.Kumar, P., Sokolik, I. N., and Nenes, A.: Cloud condensation nuclei activity and droplet activation kinetics of wet processed regional dust samples and minerals, Atmos. Chem. Phys., 11, 8661–8676, 10.5194/acp-11-8661-2011, 2011.Lavaysse, C., Chaboureau, J. P., and Flamant, C.: Dust impact on the West
African heat low in summertime, Q. J. Roy. Meteor. Soc., 137, 1227–1240, 10.1002/Qj.844, 2011.Lensky, I. M. and Rosenfeld, D.: Clouds-Aerosols-Precipitation Satellite Analysis Tool (CAPSAT), Atmos. Chem. Phys., 8, 6739–6753, 10.5194/acp-8-6739-2008, 2008.Li, J. M. and Osada, K. Z.: Preferential settling of elongated mineral dust
particles in the atmosphere, Geophys. Res. Lett., 34, L17807, 10.1029/2007gl030262,
2007.
Liao, H. and Seinfeld, J. H.: Radiative forcing by mineral dust aerosols:
sensitivity to key variables, J. Geophys. Res.-Atmos., 103, 31637–31645, 1998.
Lohmann, U., Luond, F., and Mahrt, F.: An Introduction to Clouds: From the
Microscale to Climate, Cambridge University Press, Cambridge, UK, 382 pp., 2016.Mahowald, N., Albani, S., Kok, J. F., Engelstaeder, S., Scanza, R., Ward, D.
S., and Flanner, M. G.: The size distribution of desert dust aerosols and
its impact on the Earth system, Aeolian. Res., 15, 53–71,
10.1016/j.aeolia.2013.09.002, 2014.Maring, H., Savoie, D. L., Izaguirre, M. A., Custals, L., and Reid, J. S.:
Mineral dust aerosol size distribution change during atmospheric transport,
J. Geophys. Res.-Atmos., 108, 8592, 10.1029/2002jd002536, 2003.
Marsham, J. H., Hobby, M., Allen, C. J. T., Banks, J. R., Bart, M., Brooks,
B. J., Cavazos-Guerra, C., Engelstaedter, S., Gascoyne, M., Lima, A. R.,
Martins, J. V., McQuaid, J. B., O'Leary, A., Ouchene, B., Ouladichir, A.,
Parker, D. J., Saci, A., Salah-Ferroudj, M., Todd, M. C., and Washington,
R.: Meteorology and dust in the central Sahara: Observations from Fennec
supersite-1 during the June 2011 Intensive Observation Period, J. Geophys. Res.-Atmos., 118, 4069–4089, 10.1002/jgrd.50211, 2013.McConnell, C. L., Highwood, E. J., Coe, H., Formenti, P., Anderson, B.,
Osborne, S., Nava, S., Desboeufs, K., Chen, G., and Harrison, M. A. J.:
Seasonal variations of the physical and optical characteristics of Saharan
dust: Results from the Dust Outflow and Deposition to the Ocean (DODO)
experiment, J. Geophys. Res.-Atmos., 113, D14S05, 10.1029/2007jd009606, 2008.Middleton, N., Tozer, P., and Tozer, B.: Sand and dust storms: underrated
natural hazards, Disasters, 43, 390–409, 10.1111/disa.12320,
2018.Middleton, N. J.: Desert dust hazards: A global review, Aeolian. Res., 24,
53–63, 10.1016/j.aeolia.2016.12.001, 2017.NASA GES DISC Glossary:
https://disc.gsfc.nasa.gov/information/glossary?title=AerosolEffectiveRadius, last access: 4 October 2018.Neff, J. C., Reynolds, R. L., Munson, S. M., Fernandez, D., and Belnap, J.:
The role of dust storms in total atmospheric particle concentrations at two
sites in the western US, J. Geophys. Res.-Atmos., 118, 11201–11212,
10.1002/jgrd.50855, 2013.Osborne, S. R., Johnson, B. T., Haywood, J. M., Baran, A. J., Harrison, M.
A. J., and McConnell, C. L.: Physical and optical properties of mineral dust
aerosol during the Dust and Biomass-burning Experiment, J. Geophys. Res.-Atmos.,
113, D00C03, 10.1029/2007jd009551, 2008.Otto, S., Trautmann, T., and Wendisch, M.: On realistic size equivalence and shape of spheroidal Saharan mineral dust particles applied in solar and thermal radiative transfer calculations, Atmos. Chem. Phys., 11, 4469–4490, 10.5194/acp-11-4469-2011, 2011.Pan, B. W., Wang, Y. A., Hu, J. X., Lin, Y., Hsieh, J. S., Logan, T., Feng,
X. D., Jiang, J. H., Yung, Y. L., and Zhang, R. Y.: Impacts of Saharan Dust
on Atlantic Regional Climate and Implications for Tropical Cyclones, J.
Climate, 31, 7621–7644, 10.1175/Jcli-D-16-0776.1, 2018.Perlwitz, J. and Miller, R. L.: Cloud cover increase with increasing
aerosol absorptivity: A counterexample to the conventional semidirect
aerosol effect, J. Geophys. Res.-Atmos., 115, D08203, 10.1029/2009jd012637, 2010.Petters, M. D. and Kreidenweis, S. M.: A single parameter representation of hygroscopic growth and cloud condensation nucleus activity, Atmos. Chem. Phys., 7, 1961–1971, 10.5194/acp-7-1961-2007, 2007.Reid, J. S., Reid, E. A., Walker, A., Piketh, S., Cliff, S., Al Mandoos, A.,
Tsay, S. C., and Eck, T. F.: Dynamics of southwest Asian dust particle size
characteristics with implications for global dust research, J. Geophys. Res.-Atmos., 113, D14212, 10.1029/2007jd009752, 2008.Renard, J.-B., Dulac, F., Durand, P., Bourgeois, Q., Denjean, C., Vignelles, D., Couté, B., Jeannot, M., Verdier, N., and Mallet, M.: In situ measurements of desert dust particles above the western Mediterranean Sea with the balloon-borne Light Optical Aerosol Counter/sizer (LOAC) during the ChArMEx campaign of summer 2013, Atmos. Chem. Phys., 18, 3677–3699, 10.5194/acp-18-3677-2018, 2018.Rocha-Lima, A., Martins, J. V., Remer, L. A., Todd, M., Marsham, J. H., Engelstaedter, S., Ryder, C. L., Cavazos-Guerra, C., Artaxo, P., Colarco, P., and Washington, R.: A detailed characterization of the Saharan dust collected during the Fennec campaign in 2011: in situ ground-based and laboratory measurements, Atmos. Chem. Phys., 18, 1023–1043, 10.5194/acp-18-1023-2018, 2018.Ryder, C. L., Highwood, E. J., Lai, T. M., Sodemann, H., and Marsham, J. H.:
Impact of atmospheric transport on the evolution of microphysical and
optical properties of Saharan dust, Geophys. Res. Lett., 40, 2433–2438, 10.1002/Grl.50482, 2013a.Ryder, C. L., Highwood, E. J., Rosenberg, P. D., Trembath, J., Brooke, J. K., Bart, M., Dean, A., Crosier, J., Dorsey, J., Brindley, H., Banks, J., Marsham, J. H., McQuaid, J. B., Sodemann, H., and Washington, R.: Optical properties of Saharan dust aerosol and contribution from the coarse mode as measured during the Fennec 2011 aircraft campaign, Atmos. Chem. Phys., 13, 303–325, 10.5194/acp-13-303-2013, 2013b.Ryder, C. L., McQuaid, J. B., Flamant, C., Rosenberg, P. D., Washington, R., Brindley, H. E., Highwood, E. J., Marsham, J. H., Parker, D. J., Todd, M. C., Banks, J. R., Brooke, J. K., Engelstaedter, S., Estelles, V., Formenti, P., Garcia-Carreras, L., Kocha, C., Marenco, F., Sodemann, H., Allen, C. J. T., Bourdon, A., Bart, M., Cavazos-Guerra, C., Chevaillier, S., Crosier, J., Darbyshire, E., Dean, A. R., Dorsey, J. R., Kent, J., O'Sullivan, D., Schepanski, K., Szpek, K., Trembath, J., and Woolley, A.: Advances in understanding mineral dust and boundary layer processes over the Sahara from Fennec aircraft observations, Atmos. Chem. Phys., 15, 8479–8520, 10.5194/acp-15-8479-2015, 2015.Ryder, C. L., Marenco, F., Brooke, J. K., Estelles, V., Cotton, R., Formenti, P., McQuaid, J. B., Price, H. C., Liu, D., Ausset, P., Rosenberg, P. D., Taylor, J. W., Choularton, T., Bower, K., Coe, H., Gallagher, M., Crosier, J., Lloyd, G., Highwood, E. J., and Murray, B. J.: Coarse-mode mineral dust size distributions, composition and optical properties from AER-D aircraft measurements over the tropical eastern Atlantic, Atmos. Chem. Phys., 18, 17225–17257, 10.5194/acp-18-17225-2018, 2018.Samset, B. H., Stjern, C. W., Andrews, E., Kahn, R. A., Myhre, G., Schulz,
M., and Schuster, G. L.: Aerosol Absorption: Progress Towards Global and
Regional Constraints, Curr. Clim. Change Rep., 4, 65–83,
10.1007/s40641-018-0091-4, 2018.Saxby, J., Beckett, F., Cashman, K., Rust, A., and Tennant, E.: The impact
of particle shape on fall velocity: Implications for volcanic ash dispersion
modelling, J. Volcanol. Geoth. Res., 362, 32–48,
10.1016/j.jvolgeores.2018.08.006, 2018.Seibert, P. and Frank, A.: Source-receptor matrix calculation with a Lagrangian particle dispersion model in backward mode, Atmos. Chem. Phys., 4, 51–63, 10.5194/acp-4-51-2004, 2004.
Seinfeld, J. H. and Pandis, S. N.: Properties of the Atmospheric Aerosol,
in: Atmospheric Chemistry and Physics: From Air Pollution to Climate Change,
2 Edn., John Wiley & Sons, New Jersey, USA, 350–388, 2006.
Shettle, E. P. and Fenn, R. W.: Models for the Aerosols of the Lower
Atmosphere and the Effects of Humidity Variations on Their Optical
Properties, Air Force Geophysics Laboratory Environmental Research Papers,
Hanscomb, MA, 25 pp., 1979.Sicard, M., Bertolín, S., Mallet, M., Dubuisson, P., and Comerón, A.: Estimation of mineral dust long-wave radiative forcing: sensitivity study to particle properties and application to real cases in the region of Barcelona, Atmos. Chem. Phys., 14, 9213–9231, 10.5194/acp-14-9213-2014, 2014.Smith, M.: Facility for Airborne Atmospheric Measurements; Met Office, Natural Environment Research Council, Facility for Airborne Atmospheric Measurements (FAAM) flights, NCAS British Atmospheric Data Centre, available at: https://catalogue.ceda.ac.uk/uuid/affe775e8d8890a4556aec5bc4e0b45c, last access: 2 December 2019, 2004.Solmon, F., Elguindi, N., and Mallet, M.: Radiative and climatic effects of
dust over West Africa, as simulated by a regional climate model, Clim. Res.,
52, 97–113, 10.3354/cr01039, 2012.Stevenson, J. A., Millington, S. C., Beckett, F. M., Swindles, G. T., and Thordarson, T.: Big grains go far: understanding the discrepancy between tephrochronology and satellite infrared measurements of volcanic ash, Atmos. Meas. Tech., 8, 2069–2091, 10.5194/amt-8-2069-2015, 2015.Stohl, A., Hittenberger, M., and Wotawa, G.: Validation of the Lagrangian
particle dispersion model FLEXPART against large-scale tracer experiment
data, Atmos. Environ., 32, 4245–4264, 10.1016/S1352-2310(98)00184-8, 1998.Stohl, A., Forster, C., Frank, A., Seibert, P., and Wotawa, G.: Technical note: The Lagrangian particle dispersion model FLEXPART version 6.2, Atmos. Chem. Phys., 5, 2461–2474, 10.5194/acp-5-2461-2005, 2005.Strong, J. D. O., Vecchi, G. A., and Ginoux, P.: The Climatological Effect
of Saharan Dust on Global Tropical Cyclones in a Fully Coupled GCM, J. Geophys. Res.-Atmos., 123, 5538–5559, 10.1029/2017jd027808, 2018.Tegen, I. and Lacis, A. A.: Modeling of particle size distribution and its
influence on the radiative properties of mineral dust aerosol, J. Geophys. Res.-Atmos., 101, 19237–19244, 10.1029/95jd03610, 1996.Toth III, J. R., Rajupet, S., Squire, H., Volbers, B., Zhou, J., Xie, L., Sankaran, R. M., and Lacks, D. J.: Electrostatic forces alter particle size distributions in atmospheric dust, Atmos. Chem. Phys. Discuss., 10.5194/acp-2019-650, in review, 2019.Tsamalis, C., Chédin, A., Pelon, J., and Capelle, V.: The seasonal vertical distribution of the Saharan Air Layer and its modulation by the wind, Atmos. Chem. Phys., 13, 11235–11257, 10.5194/acp-13-11235-2013, 2013.van der Does, M., Korte, L. F., Munday, C. I., Brummer, G.-J. A., and Stuut, J.-B. W.: Particle size traces modern Saharan dust transport and deposition across the equatorial North Atlantic, Atmos. Chem. Phys., 16, 13697–13710, 10.5194/acp-16-13697-2016, 2016.van der Does, M., Knippertz, P., Zschenderlein, P., Harrison, R. G., and
Stuut, J. B. W.: The mysterious long-range transport of giant mineral dust
particles, Sci. Adv., 4, eaau2768, 10.1126/sciadv.aau2768, 2018.Volz, F. E.: Infrared Optical-Constants of Ammonium Sulfate, Sahara Dust,
Volcanic Pumice, and Flyash, Appl. Opt., 12, 564–568, 10.1364/Ao.12.000564, 1973.Walser, A., Sauer, D., Spanu, A., Gasteiger, J., and Weinzierl, B.: On the parametrization of optical particle counter response including instrument-induced broadening of size spectra and a self-consistent evaluation of calibration measurements, Atmos. Meas. Tech., 10, 4341–4361, 10.5194/amt-10-4341-2017, 2017.
Washington, R., Flamant, C., Parker, D. J., Marsham, J. H., McQuaid, J. B.,
Brindley, H., Todd, M., Highwood, E. J., Ryder, C. L., Chaboureau, J.-P.,
Kocha, C., Bechir, M., and Saci, A.: Fennec - The Saharan Climate System,
CLIVAR Exchanges, 17, 31–32, 2012.Weinzierl, B., Petzold, A., Esselborn, M., Wirth, M., Rasp, K., Kandler, K.,
Schütz, L., Koepke, P., and Fiebig, M.: Airborne measurements of dust
layer properties, particle size distribution and mixing state of Saharan
dust during SAMUM 2006, Tellus B, 61, 96–117, 10.1111/j.1600-0889.2008.00392.x, 2009.Weinzierl, B., Sauer, D., Esselborn, M., Petzold, A., Veira, A., Rose, M.,
Mund, S., Wirth, M., Ansmann, A., Tesche, M., Gross, S., and Freudenthaler,
V.: Microphysical and optical properties of dust and tropical biomass
burning aerosol layers in the Cape Verde region-an overview of the airborne
in situ and lidar measurements during SAMUM-2, Tellus B, 63, 589–618, 10.1111/j.1600-0889.2011.00566.x, 2011.Weinzierl, B., Sauer, D., Minikin, A., Reitebuch, O., Dahlkotter, F., Mayer,
B., Emde, C., Tegen, I., Gasteiger, J., Petzold, A., Veira, A., Kueppers,
U., and Schumann, U.: On the visibility of airborne volcanic ash and mineral
dust from the pilot's perspective in flight, Phys. Chem. Earth, 45, 87–102,
10.1016/j.pce.2012.04.003, 2012.Weinzierl, B., Ansmann, A., Prospero, J. M., Althausen, D., Benker, N.,
Chouza, F., Dollner, M., Farrell, D., Fomba, W. K., Freudenthaler, V.,
Gasteiger, J., Gross, S., Haarig, M., Heinold, B., Kandler, K., Kristensen,
T. B., Mayol-Bracero, O. L., Muller, T., Reitebuch, O., Sauer, D., Schafler,
A., Schepanski, K., Spanu, A., Tegen, I., Toledano, C., and Walser, A.: The
Saharan Aerosol Long-range Transport and Aerosol-cloud-interaction
experiment: Overview and Selected Highlights, B. Am. Meteorol. Soc., 98,
1427–1451, 10.1175/Bams-D-15-00142.1, 2017.
Whitby, K. T.: Physical Characteristics of Sulfur Aerosols, Atmos. Environ.,
12, 135–159, 10.1016/0004-6981(78)90196-8, 1978.
WMO: World Meteorological Organization Report of the Experts Meeting on
Aerosols and Their Climatic Effects, Geneva, Switzerland, 1983.Woodage, M. J. and Woodward, S.: UK HiGEM: Impacts of Desert Dust Radiative
Forcing in a High-Resolution Atmospheric GCM, J. Climate, 27, 5907–5928,
10.1175/Jcli-D-13-00556.1, 2014.Woodward, S.: Modeling the atmospheric life cycle and radiative impact of
mineral dust in the Hadley Centre climate model, J. Geophys. Res.-Atmos., 106,
18155–18166, 10.1029/2000jd900795, 2001.Yang, P., Feng, Q., Hong, G., Kattawar, G. W., Wiscombe, W. J., Mishchenko,
M. I., Dubovik, O., Laszlo, I., and Sokolik, I. N.: Modeling of the
scattering and radiative properties of nonspherical dust-like aerosols, J.
Aerosol. Sci., 38, 995–1014, 10.1016/j.jacrosci.2007.07.001, 2007.Yu, H. B., Chin, M., Yuan, T. L., Bian, H. S., Remer, L. A., Prospero, J.
M., Omar, A., Winker, D., Yang, Y. K., Zhang, Y., Zhang, Z. B., and Zhao,
C.: The fertilizing role of African dust in the Amazon rainforest: A first
multiyear assessment based on data from Cloud-Aerosol Lidar and Infrared
Pathfinder Satellite Observations, Geophys. Res. Lett., 42, 1984–1991,
10.1002/2015gl063040, 2015.