<?xml version="1.0" encoding="UTF-8"?>
<!DOCTYPE article PUBLIC "-//NLM//DTD Journal Publishing with OASIS Tables v3.0 20080202//EN" "journalpub-oasis3.dtd">
<article xmlns:xlink="http://www.w3.org/1999/xlink" xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:oasis="http://docs.oasis-open.org/ns/oasis-exchange/table" xml:lang="en" dtd-version="3.0" article-type="research-article"><?xmltex \bartext{Research article}?>
  <front>
    <journal-meta><journal-id journal-id-type="publisher">ACP</journal-id><journal-title-group>
    <journal-title>Atmospheric Chemistry and Physics</journal-title>
    <abbrev-journal-title abbrev-type="publisher">ACP</abbrev-journal-title><abbrev-journal-title abbrev-type="nlm-ta">Atmos. Chem. Phys.</abbrev-journal-title>
  </journal-title-group><issn pub-type="epub">1680-7324</issn><publisher>
    <publisher-name>Copernicus Publications</publisher-name>
    <publisher-loc>Göttingen, Germany</publisher-loc>
  </publisher></journal-meta>
    <article-meta>
      <article-id pub-id-type="doi">10.5194/acp-22-6393-2022</article-id><title-group><article-title>Inverse modeling of the 2021 spring <?xmltex \hack{\break}?> super dust storms in East Asia</article-title><alt-title>Inverse modeling of the 2021 spring super dust storms in East Asia</alt-title>
      </title-group><?xmltex \runningtitle{Inverse modeling of the 2021 spring super dust storms in East Asia}?><?xmltex \runningauthor{J. Jin et al.}?>
      <contrib-group>
        <contrib contrib-type="author" corresp="no" rid="aff1">
          <name><surname>Jin</surname><given-names>Jianbing</given-names></name>
          
        <ext-link>https://orcid.org/0000-0002-2868-9343</ext-link></contrib>
        <contrib contrib-type="author" corresp="no" rid="aff1">
          <name><surname>Pang</surname><given-names>Mijie</given-names></name>
          
        <ext-link>https://orcid.org/0000-0001-9773-0488</ext-link></contrib>
        <contrib contrib-type="author" corresp="no" rid="aff2">
          <name><surname>Segers</surname><given-names>Arjo</given-names></name>
          
        <ext-link>https://orcid.org/0000-0002-1319-0195</ext-link></contrib>
        <contrib contrib-type="author" corresp="no" rid="aff3">
          <name><surname>Han</surname><given-names>Wei</given-names></name>
          
        <ext-link>https://orcid.org/0000-0002-1966-446X</ext-link></contrib>
        <contrib contrib-type="author" corresp="no" rid="aff1">
          <name><surname>Fang</surname><given-names>Li</given-names></name>
          
        </contrib>
        <contrib contrib-type="author" corresp="no" rid="aff1">
          <name><surname>Li</surname><given-names>Baojie</given-names></name>
          
        <ext-link>https://orcid.org/0000-0002-3747-6534</ext-link></contrib>
        <contrib contrib-type="author" corresp="no" rid="aff4">
          <name><surname>Feng</surname><given-names>Haochuan</given-names></name>
          
        </contrib>
        <contrib contrib-type="author" corresp="no" rid="aff5">
          <name><surname>Lin</surname><given-names>Hai Xiang</given-names></name>
          
        <ext-link>https://orcid.org/0000-0002-1653-4854</ext-link></contrib>
        <contrib contrib-type="author" corresp="yes" rid="aff1">
          <name><surname>Liao</surname><given-names>Hong</given-names></name>
          <email>hongliao@nuist.edu.cn</email>
        </contrib>
        <aff id="aff1"><label>1</label><institution>Jiangsu Key Laboratory of Atmospheric Environment Monitoring and Pollution Control, Jiangsu Collaborative Innovation Center of Atmospheric Environment and Equipment Technology, School of Environmental Science and Engineering, Nanjing University of Information Science and Technology, Nanjing, Jiangsu, China</institution>
        </aff>
        <aff id="aff2"><label>2</label><institution>TNO, Department of Climate, Air and Sustainability, Utrecht, the Netherlands</institution>
        </aff>
        <aff id="aff3"><label>3</label><institution>Numerical Weather Prediction Center, Chinese Meteorological Administration, Beijing, China</institution>
        </aff>
        <aff id="aff4"><label>4</label><institution>Department of Sediment Research, China Institute of Water Resources and Hydropower Research, <?xmltex \hack{\break}?> Beijing, China</institution>
        </aff>
        <aff id="aff5"><label>5</label><institution>Delft Institute of Applied Mathematics, Delft University of Technology, Delft, the Netherlands</institution>
        </aff>
      </contrib-group>
      <author-notes><corresp id="corr1">Hong Liao (hongliao@nuist.edu.cn)</corresp></author-notes><pub-date><day>18</day><month>May</month><year>2022</year></pub-date>
      
      <volume>22</volume>
      <issue>10</issue>
      <fpage>6393</fpage><lpage>6410</lpage>
      <history>
        <date date-type="received"><day>18</day><month>December</month><year>2021</year></date>
           <date date-type="rev-request"><day>13</day><month>January</month><year>2022</year></date>
           <date date-type="rev-recd"><day>14</day><month>April</month><year>2022</year></date>
           <date date-type="accepted"><day>30</day><month>April</month><year>2022</year></date>
      </history>
      <permissions>
        <copyright-statement>Copyright: © 2022 </copyright-statement>
        <copyright-year>2022</copyright-year>
      <license license-type="open-access"><license-p>This work is licensed under the Creative Commons Attribution 4.0 International License. To view a copy of this licence, visit <ext-link ext-link-type="uri" xlink:href="https://creativecommons.org/licenses/by/4.0/">https://creativecommons.org/licenses/by/4.0/</ext-link></license-p></license></permissions><self-uri xlink:href="https://acp.copernicus.org/articles/.html">This article is available from https://acp.copernicus.org/articles/.html</self-uri><self-uri xlink:href="https://acp.copernicus.org/articles/.pdf">The full text article is available as a PDF file from https://acp.copernicus.org/articles/.pdf</self-uri>
      <abstract><title>Abstract</title>

      <p id="d1e187">Last spring, super dust storms reappeared in East Asia after being absent for one and a half decades. The event caused enormous losses in both Mongolia and China. Accurate simulation of such super sandstorms is valuable for the quantification of health damage, aviation risks, and profound impacts on the Earth system, but also to reveal the climatic driving force and the process of desertification. However, accurate simulation of dust life cycles is challenging, mainly due to imperfect knowledge of emissions.
In this study, the emissions that lead to the 2021 spring dust storms are estimated through assimilation of MODIS AOD and ground-based PM<inline-formula><mml:math id="M1" display="inline"><mml:msub><mml:mi/><mml:mn mathvariant="normal">10</mml:mn></mml:msub></mml:math></inline-formula> concentration data simultaneously. With this, the dust concentrations during these super storms could be reproduced and validated with concentration observations. The multi-observation assimilation is also compared against emission inversion that assimilates AOD or PM<inline-formula><mml:math id="M2" display="inline"><mml:msub><mml:mi/><mml:mn mathvariant="normal">10</mml:mn></mml:msub></mml:math></inline-formula> concentration measurements alone, and the added values are analyzed.
The emission inversion results reveal that wind-blown dust emissions originated from both China and Mongolia during spring 2021. Specifically, <inline-formula><mml:math id="M3" display="inline"><mml:mrow><mml:mn mathvariant="normal">19.9</mml:mn><mml:mo>×</mml:mo><mml:msup><mml:mn mathvariant="normal">10</mml:mn><mml:mn mathvariant="normal">6</mml:mn></mml:msup></mml:mrow></mml:math></inline-formula> and <inline-formula><mml:math id="M4" display="inline"><mml:mrow><mml:mn mathvariant="normal">37.5</mml:mn><mml:mo>×</mml:mo><mml:msup><mml:mn mathvariant="normal">10</mml:mn><mml:mn mathvariant="normal">6</mml:mn></mml:msup></mml:mrow></mml:math></inline-formula> t of particles were released in the Chinese and Mongolian Gobi, respectively, during these severe dust events. By source apportionment it was revealed that the Mongolian Gobi poses more severe threats to the densely populated regions of the Fenwei Plain (FWP) and the North China Plain (NCP) located in northern China than does the Chinese Gobi. It was estimated that 63 % of the dust deposited in FWP was due to transnational transport from Mongolia. For NCP, the long-distance transport dust from Mongolia contributes about 69 % to the dust deposition.</p>
  </abstract>
    </article-meta>
  </front>
<body>
      

<sec id="Ch1.S1" sec-type="intro">
  <label>1</label><title>Introduction</title>
      <p id="d1e247">Dust storms occur as a result of wind erosion liberating dust particles from dry and barren surfaces <xref ref-type="bibr" rid="bib1.bibx49" id="paren.1"/>.
They are relatively common meteorological hazards in arid or semi-arid regions <xref ref-type="bibr" rid="bib1.bibx58" id="paren.2"/>.
Fine dust particles released from the ground can be lifted several kilometers high, and subsequently carried over long distances, sometimes even across continents <xref ref-type="bibr" rid="bib1.bibx65" id="paren.3"/>.
The substantial amounts of dust particles as well as irritating spores, bacteria, and viruses carried by the dust storms pose great threats to human health and agriculture <xref ref-type="bibr" rid="bib1.bibx57" id="paren.4"/>.
Next to these adverse health effects and property losses, the visibility reduction would cause severe disruption or disorders to transportation and aviation systems <xref ref-type="bibr" rid="bib1.bibx59" id="paren.5"/>.
The dust cycle itself also plays a key role in the Earth system by influencing radiative balance <xref ref-type="bibr" rid="bib1.bibx60" id="paren.6"/> but also forest and ocean
ecosystems <xref ref-type="bibr" rid="bib1.bibx50" id="paren.7"/>.</p>
      <p id="d1e272">Driven by strong cyclones, dust storms forming in the Gobi desert could affect East Asia following long-distance transport patterns, as can be seen in, for example,  Fig. <xref ref-type="fig" rid="Ch1.F1"/>a. These events usually occur in springtime, when particles are more erodible under the circumstance of a dry season and sparse vegetation. The frequency and intensity of spring dust storms reached a peak from the 1950s to 1970s, and declined steadily thereafter <xref ref-type="bibr" rid="bib1.bibx61" id="paren.8"/>. In the past decade, the dust storms were persistently rare, and only two strong events occurred in 2015 <xref ref-type="bibr" rid="bib1.bibx20" id="paren.9"/> and 2017 <xref ref-type="bibr" rid="bib1.bibx22" id="paren.10"/>. The declining trend in dust storm occurrence was co-driven by strict greenness controls in northern China over the past decades <xref ref-type="bibr" rid="bib1.bibx51" id="paren.11"/>, climate anomalies <xref ref-type="bibr" rid="bib1.bibx61" id="paren.12"/>, synoptic disturbances, and other factors not yet explored. However, the ongoing land degradation and desertification in Mongolia <xref ref-type="bibr" rid="bib1.bibx15" id="paren.13"/> is reported to aggravate the regional dust storms.
In spring 2021, however, East Asia experienced an outbreak of severe dust storms after an absence of one and a half decades. Specifically, three super events occurred during 14 to 16 March,  26 to 28 March, and 14 to 15 April, which will be described in Sect. <xref ref-type="sec" rid="Ch1.S2.SS1"/> later. The re-occurring spring super dust storms resulted in enormous losses directly. Taking the dust event in 14–16 March for example, 10 people were reported dead and hundreds of people reported missing in Mongolia <xref ref-type="bibr" rid="bib1.bibx3" id="paren.14"/>; the PM<inline-formula><mml:math id="M5" display="inline"><mml:msub><mml:mi/><mml:mn mathvariant="normal">10</mml:mn></mml:msub></mml:math></inline-formula> pollutant level in Beijing was brought over 8000 <inline-formula><mml:math id="M6" display="inline"><mml:mrow class="unit"><mml:mi mathvariant="normal">µ</mml:mi></mml:mrow></mml:math></inline-formula>g m<inline-formula><mml:math id="M7" display="inline"><mml:msup><mml:mi/><mml:mrow><mml:mo>-</mml:mo><mml:mn mathvariant="normal">3</mml:mn></mml:mrow></mml:msup></mml:math></inline-formula> and 12 provinces in northern China were affected, with thousands of flights grounded and public transportation systems halted <xref ref-type="bibr" rid="bib1.bibx18" id="paren.15"/>. Fully understanding the reappearing dust storms is of great interest to health professionals, aviation authorities, and policy- makers, not only to help evaluate property losses, adverse health effects, and Earth system impacts, but also to reveal the synoptic climatic driving forces and finally to build the next generation of the dust early-warning system.</p>
      <p id="d1e334">Numerical simulation models are commonly used to study dust lifecycles. Such models usually consist of an atmospheric transport model that is able to simulate aerosol concentrations, coupled to a dust-emission module. The underlying model is often a chemical transport model (CTM) which is already used to calculate concentrations of air pollutants, with dust just one of the aerosol species included.
Substantial effort has been invested to develop dust models, especially for dust-emission parameterization which is the most uncertain element in simulating the dust concentrations. Since the early 1990s, several dust-emission schemes have been introduced, e.g., MB95 <xref ref-type="bibr" rid="bib1.bibx32" id="paren.16"/>, Shao96/Shao04 <xref ref-type="bibr" rid="bib1.bibx52 bib1.bibx48" id="paren.17"/>, Ginoux01 <xref ref-type="bibr" rid="bib1.bibx11" id="paren.18"/>, and Zender03 <xref ref-type="bibr" rid="bib1.bibx64" id="paren.19"/>.
Such emission parameterizations have been included in many global or regional atmospheric transport models, e.g., ECMWF's Integrated Forecast System (IFS) <xref ref-type="bibr" rid="bib1.bibx37 bib1.bibx36 bib1.bibx38" id="paren.20"/>,
BSC-DREAM8b <xref ref-type="bibr" rid="bib1.bibx42 bib1.bibx35" id="paren.21"/>, CUACE/Dust <xref ref-type="bibr" rid="bib1.bibx67 bib1.bibx14" id="paren.22"/>, GEOS-Chem <xref ref-type="bibr" rid="bib1.bibx10" id="paren.23"/>, and LOTOS-EUROS <xref ref-type="bibr" rid="bib1.bibx54 bib1.bibx31" id="paren.24"/>, as will be used in this study.
Due to insufficient knowledge of the actual dust emission and transport, and due to limitations to computing resources to resolve finest-scale variabilities, huge difference might occur when simulations are compared to observations <xref ref-type="bibr" rid="bib1.bibx41 bib1.bibx17" id="paren.25"/>.</p>
      <p id="d1e368">Recent advances in sensor technologies and the continuously decreasing cost of electronic devices
have made large-scale measurements of dust feasible.
For dust storms over East Asia, important observation includes ground-based aerosol optical properties from
the AErosol RObotic NETwork (AERONET) <xref ref-type="bibr" rid="bib1.bibx7" id="paren.26"/> and
Sun-sky radiometer Observation NETwork (SONET) <xref ref-type="bibr" rid="bib1.bibx27" id="paren.27"/>, PM<inline-formula><mml:math id="M8" display="inline"><mml:msub><mml:mi/><mml:mn mathvariant="normal">10</mml:mn></mml:msub></mml:math></inline-formula> concentration from
the China Ministry of Environmental Protection (MEP) air-quality monitoring network <xref ref-type="bibr" rid="bib1.bibx20" id="paren.28"/>,
satellite remote-sensing data from polar orbiting instruments (e.g., Moderate-Resolution Imaging Spectroradiometer (MODIS), <xref ref-type="bibr" rid="bib1.bibx43" id="altparen.29"/>, and Cloud-Aerosol Lidar and Infrared Pathfinder Satellite Observations (CALIPSO), <xref ref-type="bibr" rid="bib1.bibx56" id="altparen.30"/>),
and from geostationary platforms (e.g., Himawari-8, <xref ref-type="bibr" rid="bib1.bibx1" id="altparen.31"/>, and FY-4 AOPs, <xref ref-type="bibr" rid="bib1.bibx34" id="altparen.32"/>).
These various types of observations have been used alongside the simulation models to analyze the dust storms <xref ref-type="bibr" rid="bib1.bibx12 bib1.bibx13" id="paren.33"/>.
Despite their significant roles in characterizing the atmospheric dust load,
using only the measurements is not sufficient to obtain a complete four-dimensional insight into the dust plumes, because either the measurements do not cover all areas (surface network), only observe once or twice a day (polar orbiting satellites), or observe only vertically integrated quantities (AOD).</p>
      <p id="d1e406">Instead of studying dust storms with either simulation models or observations only,
it is useful to combine the measurements and the simulations through data assimilation <xref ref-type="bibr" rid="bib1.bibx26" id="paren.34"/>.
For the purpose of dust storm simulations, the measurements could be used to decrease the uncertainty in the emissions such that the optimized simulation is in better agreement with those measurements <xref ref-type="bibr" rid="bib1.bibx14 bib1.bibx28" id="paren.35"/>.
For instance, aerosol products from the MODIS instrument onboard the polar orbiting satellites
Terra and Aqua have been widely applied in global or
East Asian dust storm assimilation <xref ref-type="bibr" rid="bib1.bibx5 bib1.bibx62" id="paren.36"/>.
The geostationary Himawari-8 data have recently also gained popularity and been used in dust storm detection and assimilation <xref ref-type="bibr" rid="bib1.bibx63" id="paren.37"/>. In our previous studies described in <xref ref-type="bibr" rid="bib1.bibx20 bib1.bibx21" id="text.38"/>, the ground-based PM<inline-formula><mml:math id="M9" display="inline"><mml:msub><mml:mi/><mml:mn mathvariant="normal">10</mml:mn></mml:msub></mml:math></inline-formula> concentration data were assimilated to estimate the dust source emission for a sandstorm in 2015 over East Asia. Himawari-8 AODs were also assimilated to nudge the dust emission, and the necessity of removing those AODs in cloudy environment is emphasized in <xref ref-type="bibr" rid="bib1.bibx22" id="text.39"/>. An adjoint model was applied to construct a background covariance that was better able to describe the potential source regions for dust emission <xref ref-type="bibr" rid="bib1.bibx23" id="paren.40"/>. Meanwhile, an imaging morphing-based assimilation method was designed which effectively corrected the position error caused in the long-distance dust plume transport <xref ref-type="bibr" rid="bib1.bibx24" id="paren.41"/>.</p>
      <p id="d1e443">While various kinds of dust or aerosol measurements were used in these assimilation experiments,
most of them assimilated only one type of observation at a time. For the sandstorms in East Asia,
however, none of the aforementioned measurements can provide sufficient information to track these short-term and fast-changing dust plumes completely. For a case in point, the China MEP air-quality monitoring network has over 1700 sites for PM<inline-formula><mml:math id="M10" display="inline"><mml:msub><mml:mi/><mml:mn mathvariant="normal">10</mml:mn></mml:msub></mml:math></inline-formula> concentrations by now, but these are mainly located in the downwind and densely populated regions, which are far away from the source region of dust in East Asia (the Gobi deserts that can be seen in Fig. <xref ref-type="fig" rid="Ch1.F1"/>a).
The network therefore only measures the dust plume when it has already been transported to the downwind urban regions; hence, it is of limited help to characterize the plume near the source regions in rural areas.
Besides, the ground-based PM<inline-formula><mml:math id="M11" display="inline"><mml:msub><mml:mi/><mml:mn mathvariant="normal">10</mml:mn></mml:msub></mml:math></inline-formula> data only represent the surface dust concentrations and lack information on the vertical structure.
AERONET and SONET instruments provide some information on the vertical structure via column-integrated observations, but the network is much sparser than the PM<inline-formula><mml:math id="M12" display="inline"><mml:msub><mml:mi/><mml:mn mathvariant="normal">10</mml:mn></mml:msub></mml:math></inline-formula> network, and most of the sites are also far away from the source regions.
The MODIS satellite products provide global coverage, but again only information on the total column and therefore no estimate of the plume height or thickness. These polar orbiting instruments also have a limited temporal coverage; for example, the MODIS Aqua and Terra platforms pass by only around 10:30 and 13:30 (local time).
Designed with the wide-observing coverage and high temporal resolution, geostationary measuring instruments provide valuable information.
However, large uncertainties were found in the Himawari-8 product due to
uncertainty in assumptions on aerosol models and surface reflectance estimation in the retrieval algorithm
<xref ref-type="bibr" rid="bib1.bibx66" id="paren.42"/>.
This was also found in <xref ref-type="bibr" rid="bib1.bibx22" id="text.43"/>: a strict observation selection was necessary when assimilating the Himawari-8 AOD values.</p>
      <p id="d1e482">Another challenge for dust assimilation is the proper definition of the observations. In general, the commonly used data assimilation schemes all rely on the basic assumption of an unbiased observation. However, all the aforementioned observations, e.g., AOD and PM<inline-formula><mml:math id="M13" display="inline"><mml:msub><mml:mi/><mml:mn mathvariant="normal">10</mml:mn></mml:msub></mml:math></inline-formula> concentration, not only measure dust aerosols, but rather actually the sum of the dust and other fine particles. These originate from, for example, anthropogenic activities, e.g., industry, vehicles, and households, and from natural sources such as wildfires and sea spray. These particles are referred to the non-dust fraction in the total aerosols in our study. In the presence of non-dust bias, it is impossible to attribute the difference between the a priori simulation and an observation to either this bias or to model deficiencies. The non-dust bias might lead to assimilations that diverge from reality <xref ref-type="bibr" rid="bib1.bibx30" id="paren.44"/>. However, aerosol measurements are usually directly assimilated, and little progress has been made in bias correction of fully aerosol measurements for their use in dust storm assimilation. <xref ref-type="bibr" rid="bib1.bibx28" id="text.45"/> selected only PM<inline-formula><mml:math id="M14" display="inline"><mml:msub><mml:mi/><mml:mn mathvariant="normal">10</mml:mn></mml:msub></mml:math></inline-formula> observations for assimilation when at least one occurrence of dust clouds was reported by the local stations. Both machine learning tools and chemical transport models were used for modeling the dynamic non-dust aerosol levels in the PM<inline-formula><mml:math id="M15" display="inline"><mml:msub><mml:mi/><mml:mn mathvariant="normal">10</mml:mn></mml:msub></mml:math></inline-formula> concentration measurements; the bias-corrected observations subsequently resulted in more promising assimilation analysis in <xref ref-type="bibr" rid="bib1.bibx21" id="text.46"/>. Some efforts have been made to exclude those “polluted” AODs induced by cloud scenes in the dust assimilation <xref ref-type="bibr" rid="bib1.bibx22" id="paren.47"/>; nevertheless, the issue of a non-dust AOD bias remains to large extent unresolved.</p>
      <p id="d1e525">To analyze the outbreak of super dust storms in spring 2021, an emission inversion will be performed through assimilation of multiple observation types. Both the MODIS AOD and the ground-based PM<inline-formula><mml:math id="M16" display="inline"><mml:msub><mml:mi/><mml:mn mathvariant="normal">10</mml:mn></mml:msub></mml:math></inline-formula> concentration observations will be used, each providing a different view on the dust plumes.
To use the AODs for representing the dust load, an data quality control is designed, which consists of an Ångström-based data screening and a non-dust AOD bias correction. The former is capable of selecting those coarse-mode dust AOD values and excludes the pixels dominated by fine-mode aerosol from non-dust sources, while the latter focuses on removing the non-dust baseline in the selected AOD. Similar for PM<inline-formula><mml:math id="M17" display="inline"><mml:msub><mml:mi/><mml:mn mathvariant="normal">10</mml:mn></mml:msub></mml:math></inline-formula>, a non-dust bias correction as used in <xref ref-type="bibr" rid="bib1.bibx21 bib1.bibx24" id="text.48"/> is adopted in this work.
Based on the posterior emission field obtained in the inversion, the spatial pattern of the emission analyzed for the three events is studied, and the active source regions are identified. In addition, source apportionment simulations are performed, in which the contribution of dust emissions from either Mongolia or China to the dust deposition is calculated for two mega-city clusters, the North China Plain (NCP) and the Fenwei Plain (FWP).</p>
      <p id="d1e549">The paper is organized as follows. Section <xref ref-type="sec" rid="Ch1.S2"/> introduces the three super dust events that occurred last spring, as well as the available dust observations from MODIS and ground-based PM<inline-formula><mml:math id="M18" display="inline"><mml:msub><mml:mi/><mml:mn mathvariant="normal">10</mml:mn></mml:msub></mml:math></inline-formula> networks measuring the dust loading from different perspectives. Before AOD is assimilated, quality control combining an Ångström-based screen and a non-dust bias correction are designed to ensure that the observations are representative for the dust load. A brief description of our dust simulation model (LOTOS-EUROS/dust) is presented as well. Section <xref ref-type="sec" rid="Ch1.S3"/> reviews the reduced-tangent-linearization four-dimensional variational (4DVar) data assimilation-based emission inversion system. In Sect. <xref ref-type="sec" rid="Ch1.S4"/>, the posterior emission field and the estimated AOD and surface dust concentration simulation are evaluated and discussed. Posterior emission obtained through the multi-observation assimilation is compared against the result from emission inversion that only assimilates PM<inline-formula><mml:math id="M19" display="inline"><mml:msub><mml:mi/><mml:mn mathvariant="normal">10</mml:mn></mml:msub></mml:math></inline-formula> or AOD data, and the added values are illustrated. A source apportionment study is performed to obtain information about dust pollution sources and the extent to which they contribute to ambient air pollution in northern China. Focusing on NCP and FWP, their dominant dust sources are identified herein.</p>
</sec>
<sec id="Ch1.S2">
  <label>2</label><title>Dust events, measurements, and model</title>
<sec id="Ch1.S2.SS1">
  <label>2.1</label><title>Dust events of spring 2021</title>
      <p id="d1e591">The Fenwei Plain (FWP) and the North China Plain (NCP) experienced the most dust affection in China since they face the dust source regions as shown in Fig. <xref ref-type="fig" rid="Ch1.F1"/>a. Figure <xref ref-type="fig" rid="Ch1.F1"/>b shows the recorded PM<inline-formula><mml:math id="M20" display="inline"><mml:msub><mml:mi/><mml:mn mathvariant="normal">10</mml:mn></mml:msub></mml:math></inline-formula> concentration average during spring (March–May) in NCP and FWP over 2019–2021. In spite of ongoing air pollutant reduction measures, the PM<inline-formula><mml:math id="M21" display="inline"><mml:msub><mml:mi/><mml:mn mathvariant="normal">10</mml:mn></mml:msub></mml:math></inline-formula> levels in NCP and FWP both reached their highest average values of 123 and 119 <inline-formula><mml:math id="M22" display="inline"><mml:mrow class="unit"><mml:mi mathvariant="normal">µ</mml:mi></mml:mrow></mml:math></inline-formula>g m<inline-formula><mml:math id="M23" display="inline"><mml:msup><mml:mi/><mml:mrow><mml:mo>-</mml:mo><mml:mn mathvariant="normal">3</mml:mn></mml:mrow></mml:msup></mml:math></inline-formula> in 2021. These mean values are 34 % and 27 % higher than the means of the previous 2 years. The reason for this is that East Asia suffered an outbreak of dust storms in the 2021 spring season, including three large-scale and severe ones. These dust events swept across northern China, as can be seen from the time series of the hourly PM<inline-formula><mml:math id="M24" display="inline"><mml:msub><mml:mi/><mml:mn mathvariant="normal">10</mml:mn></mml:msub></mml:math></inline-formula> concentrations reported by the monitoring stations in NCP and FWP in Fig. <xref ref-type="fig" rid="Ch1.F1"/>c–d. The timeline in Table <xref ref-type="table" rid="Ch1.T1"/> shows that these three dust events occurred around 15 March, 28 March, and 15 April, and each of them lasted 3 to 4 d. The highest PM<inline-formula><mml:math id="M25" display="inline"><mml:msub><mml:mi/><mml:mn mathvariant="normal">10</mml:mn></mml:msub></mml:math></inline-formula> concentrations reported by the ground-based air quality monitoring network reached values of 9993, 9985, and 4113 <inline-formula><mml:math id="M26" display="inline"><mml:mrow class="unit"><mml:mi mathvariant="normal">µ</mml:mi></mml:mrow></mml:math></inline-formula>g m<inline-formula><mml:math id="M27" display="inline"><mml:msup><mml:mi/><mml:mrow><mml:mo>-</mml:mo><mml:mn mathvariant="normal">3</mml:mn></mml:mrow></mml:msup></mml:math></inline-formula>; in this study, the three events are referred to as SD1, SD2, and SD3, respectively.</p>
      <p id="d1e680">The three dust storms will be studied by combining model simulations and AOD and PM<inline-formula><mml:math id="M28" display="inline"><mml:msub><mml:mi/><mml:mn mathvariant="normal">10</mml:mn></mml:msub></mml:math></inline-formula> concentration observations using assimilation, in order to identify the emission sources that affect densely populated regions in northern China.</p>

<?xmltex \floatpos{t}?><table-wrap id="Ch1.T1" specific-use="star"><?xmltex \currentcnt{1}?><label>Table 1</label><caption><p id="d1e695">Descriptions of the three severe dust storm events that occurred in China in spring 2021. Time zone is China Standard Time (CST).</p></caption><oasis:table frame="topbot"><oasis:tgroup cols="5">
     <oasis:colspec colnum="1" colname="col1" align="left"/>
     <oasis:colspec colnum="2" colname="col2" align="left"/>
     <oasis:colspec colnum="3" colname="col3" align="right"/>
     <oasis:colspec colnum="4" colname="col4" align="left"/>
     <oasis:colspec colnum="5" colname="col5" align="left"/>
     <oasis:thead>
       <oasis:row>
         <oasis:entry colname="col1">Dust event</oasis:entry>
         <oasis:entry colname="col2">Affected regions</oasis:entry>
         <oasis:entry colname="col3">Highest PM<inline-formula><mml:math id="M29" display="inline"><mml:msub><mml:mi/><mml:mn mathvariant="normal">10</mml:mn></mml:msub></mml:math></inline-formula></oasis:entry>
         <oasis:entry colname="col4">Assimilation window</oasis:entry>
         <oasis:entry colname="col5">Source apportionment</oasis:entry>
       </oasis:row>
       <oasis:row rowsep="1">
         <oasis:entry colname="col1"/>
         <oasis:entry colname="col2"/>
         <oasis:entry colname="col3">[<inline-formula><mml:math id="M30" display="inline"><mml:mrow class="unit"><mml:mi mathvariant="normal">µ</mml:mi></mml:mrow></mml:math></inline-formula>g m<inline-formula><mml:math id="M31" display="inline"><mml:msup><mml:mi/><mml:mrow><mml:mo>-</mml:mo><mml:mn mathvariant="normal">3</mml:mn></mml:mrow></mml:msup></mml:math></inline-formula>]</oasis:entry>
         <oasis:entry colname="col4"/>
         <oasis:entry colname="col5">simulation timeline</oasis:entry>
       </oasis:row>
     </oasis:thead>
     <oasis:tbody>
       <oasis:row>
         <oasis:entry colname="col1">SD1</oasis:entry>
         <oasis:entry colname="col2">NCP</oasis:entry>
         <oasis:entry colname="col3">9993</oasis:entry>
         <oasis:entry colname="col4">13 March 00:00 to 15 March 23:00</oasis:entry>
         <oasis:entry colname="col5">13 to 17 March</oasis:entry>
       </oasis:row>
       <oasis:row>
         <oasis:entry colname="col1">SD2</oasis:entry>
         <oasis:entry colname="col2">NCP, FWP</oasis:entry>
         <oasis:entry colname="col3">9985</oasis:entry>
         <oasis:entry colname="col4">26 March 00:00 to 28 March 23:00</oasis:entry>
         <oasis:entry colname="col5">26 to 29 March</oasis:entry>
       </oasis:row>
       <oasis:row>
         <oasis:entry colname="col1">SD3</oasis:entry>
         <oasis:entry colname="col2">NCP, FWP</oasis:entry>
         <oasis:entry colname="col3">4113</oasis:entry>
         <oasis:entry colname="col4">14 April 00:00 to 15 April 23:00</oasis:entry>
         <oasis:entry colname="col5">14 to 16 April</oasis:entry>
       </oasis:row>
     </oasis:tbody>
   </oasis:tgroup></oasis:table></table-wrap>

      <?xmltex \floatpos{t}?><fig id="Ch1.F1" specific-use="star"><?xmltex \currentcnt{1}?><?xmltex \def\figurename{Figure}?><label>Figure 1</label><caption><p id="d1e838"><bold>(a)</bold> Distribution of the potential dust emission source (barren and sparse vegetation land cover) over East Asia and the China MEP observing network over northern China. <bold>(b)</bold> The spring (March–May) mean PM<inline-formula><mml:math id="M32" display="inline"><mml:msub><mml:mi/><mml:mn mathvariant="normal">10</mml:mn></mml:msub></mml:math></inline-formula> concentration observations over NCP and FWP from 2019 to 2021. Time series of the hourly PM<inline-formula><mml:math id="M33" display="inline"><mml:msub><mml:mi/><mml:mn mathvariant="normal">10</mml:mn></mml:msub></mml:math></inline-formula> concentration measurements reported by stations in NCP <bold>(c)</bold> and FWP regions <bold>(d)</bold> during SD1 <bold>(i)</bold>, SD2 <bold>(ii)</bold>, and SD3 <bold>(iii)</bold>.</p></caption>
          <?xmltex \igopts{width=369.885827pt}?><graphic xlink:href="https://acp.copernicus.org/articles/22/6393/2022/acp-22-6393-2022-f01.png"/>

        </fig>

</sec>
<sec id="Ch1.S2.SS2">
  <label>2.2</label><?xmltex \opttitle{Ground-based PM${}_{{10}}$ observations}?><title>Ground-based PM<inline-formula><mml:math id="M34" display="inline"><mml:msub><mml:mi/><mml:mn mathvariant="normal">10</mml:mn></mml:msub></mml:math></inline-formula> observations</title>
      <p id="d1e904">A huge number of ground-based stations measuring air-quality indicators have been established by the China Ministry of Environmental Protection (MEP) since 2013. At present, the monitoring network has grown up to 1800 monitoring sites covering China, of which a part is shown in Fig. <xref ref-type="fig" rid="Ch1.F1"/>a. The PM<inline-formula><mml:math id="M35" display="inline"><mml:msub><mml:mi/><mml:mn mathvariant="normal">10</mml:mn></mml:msub></mml:math></inline-formula> concentrations observed by the network hence provide valuable information on the dust storms. Snapshots of the PM<inline-formula><mml:math id="M36" display="inline"><mml:msub><mml:mi/><mml:mn mathvariant="normal">10</mml:mn></mml:msub></mml:math></inline-formula> measurements for SD1, SD2, and SD3 can be found in Figs. <xref ref-type="fig" rid="Ch1.F2"/>a.i, <xref ref-type="fig" rid="Ch1.F3"/>a.i, and <xref ref-type="fig" rid="Ch1.F4"/>a.i, respectively. In the observations it is clearly visible where the dust plume is located and how it moves through the regions.</p>
      <p id="d1e934">It should be kept in mind that these PM<inline-formula><mml:math id="M37" display="inline"><mml:msub><mml:mi/><mml:mn mathvariant="normal">10</mml:mn></mml:msub></mml:math></inline-formula> measurements are actually a sum of dust and other airborne particles like black carbon, sulfate, etc. The PM<inline-formula><mml:math id="M38" display="inline"><mml:msub><mml:mi/><mml:mn mathvariant="normal">10</mml:mn></mml:msub></mml:math></inline-formula> data therefore cannot be used directly to represent the dust load. In a previous study <xref ref-type="bibr" rid="bib1.bibx21" id="paren.49"/> it has been shown that removing the non-dust baseline from PM<inline-formula><mml:math id="M39" display="inline"><mml:msub><mml:mi/><mml:mn mathvariant="normal">10</mml:mn></mml:msub></mml:math></inline-formula> observations will lead to a more accurate assimilation result, especially if the dust aerosol is not dominant. In this study, the same observational bias correction is performed to make the PM<inline-formula><mml:math id="M40" display="inline"><mml:msub><mml:mi/><mml:mn mathvariant="normal">10</mml:mn></mml:msub></mml:math></inline-formula> measurements fully representative of the dust loads.</p>
      <p id="d1e976">The PM<inline-formula><mml:math id="M41" display="inline"><mml:msub><mml:mi/><mml:mn mathvariant="normal">10</mml:mn></mml:msub></mml:math></inline-formula> bias correction takes two steps. First, non-dust aerosol levels are calculated using a model simulation with a configuration for air-quality simulation (as will be described in Sect. <xref ref-type="sec" rid="Ch1.S2.SS4"/>), but with the dust tracers disabled. Second, using these simulations, bias-corrected dust observations were calculated by subtracting the non-dust loads from the original PM<inline-formula><mml:math id="M42" display="inline"><mml:msub><mml:mi/><mml:mn mathvariant="normal">10</mml:mn></mml:msub></mml:math></inline-formula> concentration observations.</p>
      <p id="d1e1000">The non-dust aerosol surface concentrations from the simulations, and the dust-only bias-corrected observations of PM<inline-formula><mml:math id="M43" display="inline"><mml:msub><mml:mi/><mml:mn mathvariant="normal">10</mml:mn></mml:msub></mml:math></inline-formula> accompanying the original PM<inline-formula><mml:math id="M44" display="inline"><mml:msub><mml:mi/><mml:mn mathvariant="normal">10</mml:mn></mml:msub></mml:math></inline-formula> scenes shown in Figs. <xref ref-type="fig" rid="Ch1.F2"/>a.i, <xref ref-type="fig" rid="Ch1.F3"/>a.i, and <xref ref-type="fig" rid="Ch1.F4"/>a.i are available in Figs. S1, S3, and S5 in the Supplement. As an example, as shown in Fig. S1a–b, during the SD1 event, the non-dust aerosols were also carried southward and ahead of the dust plume due to the strong winds, and the bias-corrected PM<inline-formula><mml:math id="M45" display="inline"><mml:msub><mml:mi/><mml:mn mathvariant="normal">10</mml:mn></mml:msub></mml:math></inline-formula> data show the shape of the dust plume. The shape of the simulated a priori dust plume at 15 March 10:00 shown in Fig. <xref ref-type="fig" rid="Ch1.F2"/>a.ii matches with the observed shape, although the dust concentrations are sometimes very different; this will be discussed in more detail in Sect. <xref ref-type="sec" rid="Ch1.S2.SS4"/>. Similar features are seen in Figs. S3 and S5 for the other events as well.</p>
</sec>
<sec id="Ch1.S2.SS3">
  <label>2.3</label><title>MODIS AOD observations</title>
<sec id="Ch1.S2.SS3.SSS1">
  <label>2.3.1</label><title>MODIS Deep Blue AOD product</title>
      <p id="d1e1056">The Moderate-Resolution Imaging Spectroradiometer (MODIS) satellite instruments <xref ref-type="bibr" rid="bib1.bibx25" id="paren.50"/> on board of the polar orbiting satellites Terra and Aqua measure atmosphere reflectance at several wavelengths in the visible to near-infrared range. The aerosol properties that are retrieved from the MODIS observations have provided high-quality data since 2000 (Terra) and 2002 (Aqua). Designed with a wide swath (<inline-formula><mml:math id="M46" display="inline"><mml:mo lspace="0mm">≈</mml:mo></mml:math></inline-formula> 2330 km), MODIS provides near-global observations on an almost daily basis. In this study, the Deep Blue dataset in the newest MODIS Collection 6.1 Operational Level-2 Aerosol Product is used. The Deep Blue dataset is generated by the Enhanced Deep Blue algorithm <xref ref-type="bibr" rid="bib1.bibx16 bib1.bibx45" id="paren.51"/>, reporting retrievals over all cloud-free and snow-free land surfaces at a resolution of 10 km.
The MODIS Deep Blue dataset is widely used in dust source identification
<xref ref-type="bibr" rid="bib1.bibx12" id="paren.52"/>, dust model calibration <xref ref-type="bibr" rid="bib1.bibx65" id="paren.53"/> and emission inversion <xref ref-type="bibr" rid="bib1.bibx62 bib1.bibx5" id="paren.54"/>, and dust field reanalysis <xref ref-type="bibr" rid="bib1.bibx6" id="paren.55"/> through data assimilation.</p>
      <p id="d1e1085">Snapshots of the MODIS Deep Blue AOD for SD1, SD2, and SD3 can be found in
Figs. <xref ref-type="fig" rid="Ch1.F2"/>b.i–c.i, <xref ref-type="fig" rid="Ch1.F3"/>b.i, and <xref ref-type="fig" rid="Ch1.F4"/>b.i. For each of the three events, MODIS AOD provides valuable information for a large part of the domain of interest. For instance, the SD2 plume position and intensity at 28 March 11:00 are clearly identified in Fig. <xref ref-type="fig" rid="Ch1.F3"/>b.i, while the SD3 plume at 15 April 11:00 is observed clearly in Fig. <xref ref-type="fig" rid="Ch1.F4"/>b.i. The two AOD scans in Fig. <xref ref-type="fig" rid="Ch1.F2"/>b.i–c.i together also provide the general spatial pattern of the dust plume.</p>
      <p id="d1e1101">Compared to the ground-based PM<inline-formula><mml:math id="M47" display="inline"><mml:msub><mml:mi/><mml:mn mathvariant="normal">10</mml:mn></mml:msub></mml:math></inline-formula> measurements, satellite instruments are designed with a larger observing coverage. However, they also have higher uncertainties in representing the aerosol/dust load <xref ref-type="bibr" rid="bib1.bibx22" id="paren.56"/>. In most of the dust model evaluation/calibration using the satellite data, AOD measurements are directly used for comparing to simulated dust AODs and analyzing the dust strengths. For those regions affected by the severe dust storm, it would be reliable to approximate the dust AODs using the AOD measurements, since the amounts of non-dust aerosols are negligible compared to the dust loading; for those areas where dust is not the dominant aerosol, it is then necessary to perform the dust/non-dust AOD discrimination.</p>

      <?xmltex \floatpos{t}?><fig id="Ch1.F2" specific-use="star"><?xmltex \currentcnt{2}?><?xmltex \def\figurename{Figure}?><label>Figure 2</label><caption><p id="d1e1119">Snapshots of observations and simulations during severe dust event 1 (SD1) at 15 March 2021. Row <bold>(a)</bold> shows PM<inline-formula><mml:math id="M48" display="inline"><mml:msub><mml:mi/><mml:mn mathvariant="normal">10</mml:mn></mml:msub></mml:math></inline-formula> concentrations observed by ground network at 10:00 CST <bold>(i)</bold>, and a priori <bold>(ii)</bold> and posterior <bold>(iii)</bold> simulations by the LOTOS-EUROS/dust model. Similarly, row <bold>(b)</bold> shows for the same hour AOD at 550 nm from MODIS Deep Blue observations <bold>(i)</bold> and simulations <bold>(ii–iii)</bold> at 10:00; row <bold>(c)</bold> shows the same for the MODIS overpass at 12:00.</p></caption>
            <?xmltex \igopts{width=497.923228pt}?><graphic xlink:href="https://acp.copernicus.org/articles/22/6393/2022/acp-22-6393-2022-f02.png"/>

          </fig>

      <?xmltex \floatpos{t}?><fig id="Ch1.F3" specific-use="star"><?xmltex \currentcnt{3}?><?xmltex \def\figurename{Figure}?><label>Figure 3</label><caption><p id="d1e1164">Snapshots of observations and simulations during severe dust event 2 (SD2) at 28 March 2021. Row <bold>(a)</bold> shows PM<inline-formula><mml:math id="M49" display="inline"><mml:msub><mml:mi/><mml:mn mathvariant="normal">10</mml:mn></mml:msub></mml:math></inline-formula> concentrations observed by ground network at 11:00 CST <bold>(i)</bold>, a priori <bold>(ii)</bold>, and posterior <bold>(iii)</bold> simulations by the LOTOS-EUROS/dust model. Row <bold>(b)</bold> shows for the same hour AOD at 550 nm from MODIS Deep Blue observations <bold>(i)</bold> and simulations <bold>(ii–iii)</bold>.</p></caption>
            <?xmltex \igopts{width=497.923228pt}?><graphic xlink:href="https://acp.copernicus.org/articles/22/6393/2022/acp-22-6393-2022-f03.png"/>

          </fig>

      <?xmltex \floatpos{t}?><fig id="Ch1.F4" specific-use="star"><?xmltex \currentcnt{4}?><?xmltex \def\figurename{Figure}?><label>Figure 4</label><caption><p id="d1e1206">Snapshots of observations and simulations during severe dust event 3 (SD3) at 15 April 2021. Row <bold>(a)</bold> shows PM<inline-formula><mml:math id="M50" display="inline"><mml:msub><mml:mi/><mml:mn mathvariant="normal">10</mml:mn></mml:msub></mml:math></inline-formula> concentrations observed by ground network at 11:00 CST <bold>(i)</bold>, a priori <bold>(ii)</bold>, and posterior <bold>(iii)</bold> simulations by the LOTOS-EUROS/dust model. Row <bold>(b)</bold> shows for the same hour AOD at 550 nm from MODIS Deep Blue observations <bold>(i)</bold> and simulations <bold>(ii–iii)</bold>.</p></caption>
            <?xmltex \igopts{width=497.923228pt}?><graphic xlink:href="https://acp.copernicus.org/articles/22/6393/2022/acp-22-6393-2022-f04.png"/>

          </fig>

</sec>
<sec id="Ch1.S2.SS3.SSS2">
  <label>2.3.2</label><title>AOD quality control</title>
      <p id="d1e1254">While aerosols originating from biomass-burning, urban pollution, and biogenic sources consist mainly of fine-mode particles (with a radius less than 1 <inline-formula><mml:math id="M51" display="inline"><mml:mrow class="unit"><mml:mi mathvariant="normal">µ</mml:mi></mml:mrow></mml:math></inline-formula>m), the aerosols in a dust plume are mainly in the coarse modes (particle radius larger than 1 <inline-formula><mml:math id="M52" display="inline"><mml:mrow class="unit"><mml:mi mathvariant="normal">µ</mml:mi></mml:mrow></mml:math></inline-formula>m) <xref ref-type="bibr" rid="bib1.bibx8 bib1.bibx21" id="paren.57"/>. The dominance of one mode over the other can be measured with the Ångström wavelength exponent <inline-formula><mml:math id="M53" display="inline"><mml:mi mathvariant="italic">α</mml:mi></mml:math></inline-formula> <xref ref-type="bibr" rid="bib1.bibx47 bib1.bibx44 bib1.bibx29" id="paren.58"/>, which describes how the optical thickness depends on the wavelength of the incident light. The Ångström exponent is in the range of <inline-formula><mml:math id="M54" display="inline"><mml:mrow><mml:mo>-</mml:mo><mml:mn mathvariant="normal">0.5</mml:mn></mml:mrow></mml:math></inline-formula> to 2.5, and inversely related to the average size of the measured aerosol: the smaller the particles, the larger the exponent.
<xref ref-type="bibr" rid="bib1.bibx9" id="text.59"/> used <inline-formula><mml:math id="M55" display="inline"><mml:mrow><mml:mi mathvariant="italic">α</mml:mi><mml:mo>=</mml:mo><mml:mn mathvariant="normal">0.5</mml:mn></mml:mrow></mml:math></inline-formula> as the threshold for an aerosol mixture dominated by dust, while  <xref ref-type="bibr" rid="bib1.bibx46" id="text.60"/> used <inline-formula><mml:math id="M56" display="inline"><mml:mrow><mml:mi mathvariant="italic">α</mml:mi><mml:mo>=</mml:mo><mml:mn mathvariant="normal">0.6</mml:mn></mml:mrow></mml:math></inline-formula> to detect the presence of dust particles.
<xref ref-type="bibr" rid="bib1.bibx12" id="text.61"/> used <inline-formula><mml:math id="M57" display="inline"><mml:mrow><mml:mi mathvariant="italic">α</mml:mi><mml:mo>=</mml:mo><mml:mn mathvariant="normal">0</mml:mn></mml:mrow></mml:math></inline-formula> to select areas dominated by dust in a single-mode distribution of coarse particles.</p>
      <p id="d1e1343">Figure <xref ref-type="fig" rid="Ch1.F5"/> shows the four snapshots of retrieved Ångström exponents corresponding to the AOD values of Figs. <xref ref-type="fig" rid="Ch1.F2"/>b.i–c.i, <xref ref-type="fig" rid="Ch1.F3"/>b.i, and <xref ref-type="fig" rid="Ch1.F4"/>b.i.
In general, the locations of low Ångström exponent values correspond to the high AOD values of the dust plumes.
For example, for the snapshot in Fig. <xref ref-type="fig" rid="Ch1.F5"/>d, it indicates that a plume of coarse aerosols stays in the Inner Mongolia provinces of China, while fine aerosols are more dominant in the southern regions. This matches with the dust simulation that is shown in Fig. <xref ref-type="fig" rid="Ch1.F4"/>b.ii.
This gives confidence in using the Ångström exponent values for the discrimination between dust and non-dust AOD.</p>
      <p id="d1e1359">To be able to use the AOD observations in an assimilation focusing on dust only, an Ångström-based data screening and a non-dust AOD bias correction has been developed and applied. The screening and bias-correction procedures are performed after each other.
First the Ångström-based screening selects the pixels with <inline-formula><mml:math id="M58" display="inline"><mml:mrow><mml:mi mathvariant="italic">α</mml:mi><mml:mo>&lt;</mml:mo><mml:mn mathvariant="normal">0.5</mml:mn></mml:mrow></mml:math></inline-formula>, assuming that these are the ones that are dominated by (coarse-mode) dust. For these pixels, a bias-corrected AOD is calculated by subtracting a non-dust AOD fraction from the selected AODs. Similar to the non-dust PM<inline-formula><mml:math id="M59" display="inline"><mml:msub><mml:mi/><mml:mn mathvariant="normal">10</mml:mn></mml:msub></mml:math></inline-formula> simulation, these non-dust AOD baselines are also calculated using a full-chemistry LOTOS-EUROS simulation with the dust tracers disabled.
Finally, to make the AOD data resolution consistent with the model, the MODIS Deep Blue AODs are coarsened by taking the average over the 0.25 <inline-formula><mml:math id="M60" display="inline"><mml:mo>×</mml:mo></mml:math></inline-formula> 0.25 model grid cell. Snapshots of the non-dust AOD simulation and the bias-corrected AOD measurements for assimilation can be found in Figs. S2, S4, and S6.</p>

      <?xmltex \floatpos{t}?><fig id="Ch1.F5" specific-use="star"><?xmltex \currentcnt{5}?><?xmltex \def\figurename{Figure}?><label>Figure 5</label><caption><p id="d1e1393">Snapshots of MODIS Deep Blue Ångström exponent accompanying AOD scenes shown in Figs. <xref ref-type="fig" rid="Ch1.F2"/>a.i–b.i,  <xref ref-type="fig" rid="Ch1.F3"/>a.i, and  <xref ref-type="fig" rid="Ch1.F4"/>a.i.</p></caption>
            <?xmltex \igopts{width=369.885827pt}?><graphic xlink:href="https://acp.copernicus.org/articles/22/6393/2022/acp-22-6393-2022-f05.png"/>

          </fig>

</sec>
</sec>
<sec id="Ch1.S2.SS4">
  <label>2.4</label><title>Dust simulation model</title>
      <p id="d1e1417">The regional chemical transport model LOTOS-EUROS v2.1 <xref ref-type="bibr" rid="bib1.bibx31" id="paren.62"/> is used to simulate dust concentrations.
LOTOS-EUROS has been used for a wide range of applications supporting scientific research and operational air-quality forecasts over Europe, China, and other regions.
Daily operational forecasts over China used to be released via the MarcoPolo–Panda projects <xref ref-type="bibr" rid="bib1.bibx54 bib1.bibx2" id="paren.63"/>. Additionally, it is also implemented in the World Meteorological Organization (WMO) Sand and Dust Storm Warning Advisory and Assessment System to provide short-term forecasting of the dust load over the areas of North Africa, Middle East, and Europe; the online forecast product is delivered via <uri>http://sds-was.aemet.es/forecast-products/dust-forecasts</uri> (last access: July 2021).</p>
      <p id="d1e1429">To simulate dust concentrations over East Asia, the model is configured on a domain from 15 to 50<inline-formula><mml:math id="M61" display="inline"><mml:msup><mml:mi/><mml:mo>∘</mml:mo></mml:msup></mml:math></inline-formula> N  and 70 to 140<inline-formula><mml:math id="M62" display="inline"><mml:msup><mml:mi/><mml:mo>∘</mml:mo></mml:msup></mml:math></inline-formula> E, with a resolution of 0.25<inline-formula><mml:math id="M63" display="inline"><mml:msup><mml:mi/><mml:mo>∘</mml:mo></mml:msup></mml:math></inline-formula> <inline-formula><mml:math id="M64" display="inline"><mml:mo>×</mml:mo></mml:math></inline-formula> 0.25<inline-formula><mml:math id="M65" display="inline"><mml:msup><mml:mi/><mml:mo>∘</mml:mo></mml:msup></mml:math></inline-formula>.
Vertically, the model consists of eight layers, with a top at 10 km. Our regional model has zero boundary conditions by assuming all dust aerosols are emitted regionally and the external dust flows can be ignored. Although BSC-DREAM8b simulation, which is released via <uri>https://ess.bsc.es/bsc-dust-daily-forecast</uri> (last access: January 2022), indicates that part of the dust plume released in the Middle East is likely to have been transported to western China during SD1, those particles did not go as far as FWP or NCP.
Boundary conditions from a global dust model are in demand if we focus on western China in our future work.
The dust simulation is driven by the European Centre for Medium-Ranged Weather Forecasts' (ECMWF) operational forecasts over 3–12 h, retrieved at a regular longitude–latitude grid with a resolution of about 7 km. An interface to the ECMWF output set is designed, which not only interpolates the default 3 h ECMWF short-term forecast meteorology to hourly values but also averages the forecast to fit the LOTOS-EUROS spatial resolutions <xref ref-type="bibr" rid="bib1.bibx31" id="paren.64"/>. Physical processes included are wind-blown dust emission, diffusion, advection, dry and wet deposition, and sedimentation.</p>
</sec>
<sec id="Ch1.S2.SS5">
  <label>2.5</label><title>Dust emission and uncertainty</title>
      <p id="d1e1491">The goal of this work is to calculate the optimal emission field that best fits the a priori model simulation and the observation.
It is then necessary to define and quantify the uncertainty in the dust simulations.
In this study, we define the main model uncertainty to be in the parameterization of the dust emissions.
Although other model processes such as transport and deposition are uncertain too, for the events studied here, these are assumed to be of less importance than the location and the intensity of dust emission.</p>
      <p id="d1e1494">The physical basis of the dust-emission model adopted in LOTOS-EUROS is the parameterization scheme by <xref ref-type="bibr" rid="bib1.bibx64" id="text.65"/>. The dust flux <inline-formula><mml:math id="M66" display="inline"><mml:mi>f</mml:mi></mml:math></inline-formula> is calculated as a function of horizontal saltation <inline-formula><mml:math id="M67" display="inline"><mml:mrow><mml:msub><mml:mi>f</mml:mi><mml:mi mathvariant="normal">h</mml:mi></mml:msub></mml:mrow></mml:math></inline-formula> <xref ref-type="bibr" rid="bib1.bibx32" id="paren.66"/>, the sandblasting efficiency <inline-formula><mml:math id="M68" display="inline"><mml:mi>a</mml:mi></mml:math></inline-formula>, a terrain preference <inline-formula><mml:math id="M69" display="inline"><mml:mi mathvariant="script">S</mml:mi></mml:math></inline-formula>, and an erodible surface fraction <inline-formula><mml:math id="M70" display="inline"><mml:mi mathvariant="script">C</mml:mi></mml:math></inline-formula>:
            <disp-formula id="Ch1.E1" content-type="numbered"><label>1</label><mml:math id="M71" display="block"><mml:mrow><mml:mi>f</mml:mi><mml:mo>=</mml:mo><mml:msub><mml:mi>f</mml:mi><mml:mi mathvariant="normal">h</mml:mi></mml:msub><mml:mo>⋅</mml:mo><mml:mi>a</mml:mi><mml:mo>⋅</mml:mo><mml:mi mathvariant="script">S</mml:mi><mml:mo>⋅</mml:mo><mml:mi mathvariant="script">C</mml:mi><mml:mspace width="0.125em" linebreak="nobreak"/><mml:mo>.</mml:mo></mml:mrow></mml:math></disp-formula>
          The dust saltation rate <inline-formula><mml:math id="M72" display="inline"><mml:mrow><mml:msub><mml:mi>f</mml:mi><mml:mi mathvariant="normal">h</mml:mi></mml:msub></mml:mrow></mml:math></inline-formula> is proportional to the third power of the wind friction velocity <inline-formula><mml:math id="M73" display="inline"><mml:mi>u</mml:mi></mml:math></inline-formula>, as long as this exceeds a certain (surface-dependent) friction velocity threshold <inline-formula><mml:math id="M74" display="inline"><mml:mrow><mml:msub><mml:mi>u</mml:mi><mml:mi mathvariant="normal">t</mml:mi></mml:msub></mml:mrow></mml:math></inline-formula>:
            <disp-formula id="Ch1.E2" content-type="numbered"><label>2</label><mml:math id="M75" display="block"><mml:mrow><mml:msub><mml:mi>f</mml:mi><mml:mi mathvariant="normal">h</mml:mi></mml:msub><mml:mo>=</mml:mo><mml:mfenced open="{" close=""><mml:mtable class="cases" columnspacing="1em" rowspacing="0.2ex" columnalign="left left" framespacing="0em"><mml:mtr><mml:mtd><mml:mn mathvariant="normal">0</mml:mn></mml:mtd><mml:mtd><mml:mrow><mml:mi>u</mml:mi><mml:mo>≤</mml:mo><mml:msub><mml:mi>u</mml:mi><mml:mi mathvariant="normal">t</mml:mi></mml:msub></mml:mrow></mml:mtd></mml:mtr><mml:mtr><mml:mtd><mml:mrow><mml:mstyle displaystyle="false"><mml:mstyle displaystyle="false"><mml:mfrac style="text"><mml:mrow><mml:msub><mml:mi>p</mml:mi><mml:mi>a</mml:mi></mml:msub></mml:mrow><mml:mi>g</mml:mi></mml:mfrac></mml:mstyle></mml:mstyle><mml:mspace width="0.25em" linebreak="nobreak"/><mml:msup><mml:mi>u</mml:mi><mml:mn mathvariant="normal">3</mml:mn></mml:msup><mml:mspace width="0.25em" linebreak="nobreak"/><mml:mfenced close=")" open="("><mml:mrow><mml:mn mathvariant="normal">1</mml:mn><mml:mo>+</mml:mo><mml:mstyle displaystyle="false"><mml:mstyle displaystyle="false"><mml:mfrac style="text"><mml:mrow><mml:msub><mml:mi>u</mml:mi><mml:mi mathvariant="normal">t</mml:mi></mml:msub></mml:mrow><mml:mi>u</mml:mi></mml:mfrac></mml:mstyle></mml:mstyle></mml:mrow></mml:mfenced><mml:mspace width="0.25em" linebreak="nobreak"/><mml:mfenced close=")" open="("><mml:mrow><mml:mn mathvariant="normal">1</mml:mn><mml:mo>-</mml:mo><mml:mstyle displaystyle="false"><mml:mstyle displaystyle="false"><mml:mfrac style="text"><mml:mrow><mml:msubsup><mml:mi>u</mml:mi><mml:mi mathvariant="normal">t</mml:mi><mml:mn mathvariant="normal">2</mml:mn></mml:msubsup></mml:mrow><mml:mrow><mml:msup><mml:mi>u</mml:mi><mml:mn mathvariant="normal">2</mml:mn></mml:msup></mml:mrow></mml:mfrac></mml:mstyle></mml:mstyle></mml:mrow></mml:mfenced></mml:mrow></mml:mtd><mml:mtd><mml:mrow><mml:mi>u</mml:mi><mml:mo>&gt;</mml:mo><mml:msub><mml:mi>u</mml:mi><mml:mi mathvariant="normal">t</mml:mi></mml:msub><mml:mspace linebreak="nobreak" width="0.125em"/><mml:mo>.</mml:mo></mml:mrow></mml:mtd></mml:mtr></mml:mtable></mml:mfenced></mml:mrow></mml:math></disp-formula>
          The friction velocity threshold controls whether dust is released from a surface at all, and how strong the emission is.
In <xref ref-type="bibr" rid="bib1.bibx20" id="text.67"/>, it was shown that the uncertainty in the friction velocity threshold parameterization is the main factor that prohibits accurate dust-emission forecasts. Although other controlling factors, e.g., the wind-field uncertainty, will also introduce uncertainty in the emission partially, these were found to be of less important in the emission error quantification <xref ref-type="bibr" rid="bib1.bibx22" id="paren.68"/>.</p><?xmltex \hack{\newpage}?>
</sec>
</sec>
<sec id="Ch1.S3">
  <label>3</label><title>Data assimilation algorithms</title>
<sec id="Ch1.S3.SS1">
  <label>3.1</label><title>Assimilation method</title>
      <p id="d1e1732">The assimilation system that will be used to combine PM<inline-formula><mml:math id="M76" display="inline"><mml:msub><mml:mi/><mml:mn mathvariant="normal">10</mml:mn></mml:msub></mml:math></inline-formula> and AOD measurements with dust simulations is based on the reduced-tangent-linearization four-dimensional variational (4DVar) data assimilation developed in <xref ref-type="bibr" rid="bib1.bibx20" id="text.69"/>. The goal of the 4DVar technique is to find the maximum likelihood estimation of a state vector, which is here the dust emission field <inline-formula><mml:math id="M77" display="inline"><mml:mi mathvariant="bold-italic">f</mml:mi></mml:math></inline-formula>, given both the available AOD and PM<inline-formula><mml:math id="M78" display="inline"><mml:msub><mml:mi/><mml:mn mathvariant="normal">10</mml:mn></mml:msub></mml:math></inline-formula> measurements over a time window.
The optimal emission <inline-formula><mml:math id="M79" display="inline"><mml:mi mathvariant="bold-italic">f</mml:mi></mml:math></inline-formula> is calculated by minimizing the cost function:
            <disp-formula id="Ch1.E3" content-type="numbered"><label>3</label><mml:math id="M80" display="block"><mml:mrow><mml:mi mathvariant="script">J</mml:mi><mml:mo>(</mml:mo><mml:mi mathvariant="bold-italic">f</mml:mi><mml:mo>)</mml:mo><mml:mo>=</mml:mo><mml:msub><mml:mi mathvariant="script">J</mml:mi><mml:mi mathvariant="normal">b</mml:mi></mml:msub><mml:mo>(</mml:mo><mml:mi mathvariant="bold-italic">f</mml:mi><mml:mo>)</mml:mo><mml:mspace linebreak="nobreak" width="0.25em"/><mml:mo>+</mml:mo><mml:mspace linebreak="nobreak" width="0.25em"/><mml:msubsup><mml:mi mathvariant="script">J</mml:mi><mml:mi mathvariant="normal">o</mml:mi><mml:mtext>PM</mml:mtext></mml:msubsup><mml:mo>(</mml:mo><mml:mi mathvariant="bold-italic">f</mml:mi><mml:mo>)</mml:mo><mml:mspace linebreak="nobreak" width="0.25em"/><mml:mo>+</mml:mo><mml:mspace linebreak="nobreak" width="0.25em"/><mml:msubsup><mml:mi mathvariant="script">J</mml:mi><mml:mi mathvariant="normal">o</mml:mi><mml:mtext>AOD</mml:mtext></mml:msubsup><mml:mo>(</mml:mo><mml:mi mathvariant="bold-italic">f</mml:mi><mml:mo>)</mml:mo><mml:mspace width="0.125em" linebreak="nobreak"/><mml:mo>.</mml:mo></mml:mrow></mml:math></disp-formula>
          In here, <inline-formula><mml:math id="M81" display="inline"><mml:mrow><mml:msub><mml:mi mathvariant="script">J</mml:mi><mml:mi mathvariant="normal">b</mml:mi></mml:msub></mml:mrow></mml:math></inline-formula> represents the background term as follows:
            <disp-formula id="Ch1.E4" content-type="numbered"><label>4</label><mml:math id="M82" display="block"><mml:mrow><mml:msub><mml:mi mathvariant="script">J</mml:mi><mml:mi mathvariant="normal">b</mml:mi></mml:msub><mml:mo>(</mml:mo><mml:mi mathvariant="bold-italic">f</mml:mi><mml:mo>)</mml:mo><mml:mo>=</mml:mo><mml:mstyle displaystyle="true"><mml:mfrac style="display"><mml:mn mathvariant="normal">1</mml:mn><mml:mn mathvariant="normal">2</mml:mn></mml:mfrac></mml:mstyle><mml:mo>(</mml:mo><mml:mi mathvariant="bold-italic">f</mml:mi><mml:mo>-</mml:mo><mml:msub><mml:mi mathvariant="bold-italic">f</mml:mi><mml:mi mathvariant="normal">b</mml:mi></mml:msub><mml:msup><mml:mo>)</mml:mo><mml:mtext>T</mml:mtext></mml:msup><mml:msup><mml:mtext mathvariant="bold">B</mml:mtext><mml:mrow><mml:mo>-</mml:mo><mml:mn mathvariant="normal">1</mml:mn></mml:mrow></mml:msup><mml:mo>(</mml:mo><mml:mi mathvariant="bold-italic">f</mml:mi><mml:mo>-</mml:mo><mml:msub><mml:mi mathvariant="bold-italic">f</mml:mi><mml:mi mathvariant="normal">b</mml:mi></mml:msub><mml:mo>)</mml:mo><mml:mo>,</mml:mo></mml:mrow></mml:math></disp-formula>
          where <inline-formula><mml:math id="M83" display="inline"><mml:mrow><mml:msub><mml:mi mathvariant="bold-italic">f</mml:mi><mml:mi mathvariant="normal">b</mml:mi></mml:msub></mml:mrow></mml:math></inline-formula> represents the a priori or background dust emission vector which follows the calculation described in Sect. <xref ref-type="sec" rid="Ch1.S2.SS4"/>. As in <xref ref-type="bibr" rid="bib1.bibx20" id="text.70"/>, the errors in dust emission field are assumed to be only caused by the uncertainty in the friction velocity threshold in Eq. (<xref ref-type="disp-formula" rid="Ch1.E2"/>),
and it can be compensated by introducing a spatially varying multiplicative factor <inline-formula><mml:math id="M84" display="inline"><mml:mi mathvariant="italic">β</mml:mi></mml:math></inline-formula>:
            <disp-formula id="Ch1.E5" content-type="numbered"><label>5</label><mml:math id="M85" display="block"><mml:mrow><mml:msubsup><mml:mi>u</mml:mi><mml:mi mathvariant="normal">t</mml:mi><mml:mtext>true</mml:mtext></mml:msubsup><mml:mo>(</mml:mo><mml:mi>i</mml:mi><mml:mo>,</mml:mo><mml:mi>t</mml:mi><mml:mo>)</mml:mo><mml:mo>=</mml:mo><mml:mi mathvariant="italic">β</mml:mi><mml:mo>(</mml:mo><mml:mi>i</mml:mi><mml:mo>)</mml:mo><mml:mo>⋅</mml:mo><mml:msub><mml:mi>u</mml:mi><mml:mi mathvariant="normal">t</mml:mi></mml:msub><mml:mo>(</mml:mo><mml:mi>i</mml:mi><mml:mo>,</mml:mo><mml:mi>t</mml:mi><mml:mo>)</mml:mo><mml:mo>,</mml:mo></mml:mrow></mml:math></disp-formula>
          where <inline-formula><mml:math id="M86" display="inline"><mml:mrow><mml:msub><mml:mi>u</mml:mi><mml:mi mathvariant="normal">t</mml:mi></mml:msub><mml:mo>(</mml:mo><mml:mi>i</mml:mi><mml:mo>,</mml:mo><mml:mi>t</mml:mi><mml:mo>)</mml:mo></mml:mrow></mml:math></inline-formula> denotes the model parameterized friction velocity threshold in a given grid cell <inline-formula><mml:math id="M87" display="inline"><mml:mi>i</mml:mi></mml:math></inline-formula> at instant <inline-formula><mml:math id="M88" display="inline"><mml:mi>t</mml:mi></mml:math></inline-formula>, while <inline-formula><mml:math id="M89" display="inline"><mml:mrow><mml:msubsup><mml:mi>u</mml:mi><mml:mi mathvariant="normal">t</mml:mi><mml:mtext>true</mml:mtext></mml:msubsup><mml:mo>(</mml:mo><mml:mi>i</mml:mi><mml:mo>,</mml:mo><mml:mi>t</mml:mi><mml:mo>)</mml:mo></mml:mrow></mml:math></inline-formula> denotes the true value.
The <inline-formula><mml:math id="M90" display="inline"><mml:mi mathvariant="italic">β</mml:mi></mml:math></inline-formula> values are defined to be random variables with a mean of 1.0 and a standard deviation <inline-formula><mml:math id="M91" display="inline"><mml:mrow><mml:msub><mml:mi mathvariant="italic">σ</mml:mi><mml:mi mathvariant="italic">β</mml:mi></mml:msub><mml:mo>=</mml:mo><mml:mn mathvariant="normal">0.1</mml:mn></mml:mrow></mml:math></inline-formula>.
This empirical standard deviation was found to provide sufficient variations to resolve the observation–simulation difference. A background covariance matrix <inline-formula><mml:math id="M92" display="inline"><mml:mrow><mml:msub><mml:mi mathvariant="bold">B</mml:mi><mml:mi mathvariant="italic">β</mml:mi></mml:msub></mml:mrow></mml:math></inline-formula> is
then formulated by combining the constant standard deviation and a correlation matrix <inline-formula><mml:math id="M93" display="inline"><mml:mi mathvariant="bold">C</mml:mi></mml:math></inline-formula>:
            <disp-formula id="Ch1.E6" content-type="numbered"><label>6</label><mml:math id="M94" display="block"><mml:mrow><mml:msub><mml:mi mathvariant="bold">B</mml:mi><mml:mi mathvariant="italic">β</mml:mi></mml:msub><mml:mo>(</mml:mo><mml:mi>i</mml:mi><mml:mo>,</mml:mo><mml:mi>j</mml:mi><mml:mo>)</mml:mo><mml:mo>=</mml:mo><mml:msubsup><mml:mi mathvariant="italic">σ</mml:mi><mml:mi mathvariant="italic">β</mml:mi><mml:mn mathvariant="normal">2</mml:mn></mml:msubsup><mml:mo>⋅</mml:mo><mml:mi mathvariant="bold">C</mml:mi><mml:mo>(</mml:mo><mml:mi>i</mml:mi><mml:mo>,</mml:mo><mml:mi>j</mml:mi><mml:mo>)</mml:mo><mml:mspace linebreak="nobreak" width="0.125em"/><mml:mo>.</mml:mo></mml:mrow></mml:math></disp-formula>
          Here, <inline-formula><mml:math id="M95" display="inline"><mml:mrow><mml:mi mathvariant="bold">C</mml:mi><mml:mo>(</mml:mo><mml:mi>i</mml:mi><mml:mo>,</mml:mo><mml:mi>j</mml:mi><mml:mo>)</mml:mo></mml:mrow></mml:math></inline-formula> denotes a distance-based spatial correlation between the <inline-formula><mml:math id="M96" display="inline"><mml:mi mathvariant="italic">β</mml:mi></mml:math></inline-formula>s in two grid cells <inline-formula><mml:math id="M97" display="inline"><mml:mi>i</mml:mi></mml:math></inline-formula> and <inline-formula><mml:math id="M98" display="inline"><mml:mi>j</mml:mi></mml:math></inline-formula>, which is defined as
            <disp-formula id="Ch1.E7" content-type="numbered"><label>7</label><mml:math id="M99" display="block"><mml:mrow><mml:mi mathvariant="bold">C</mml:mi><mml:mo>(</mml:mo><mml:mi>i</mml:mi><mml:mo>,</mml:mo><mml:mi>j</mml:mi><mml:mo>)</mml:mo><mml:mo>=</mml:mo><mml:msup><mml:mi>e</mml:mi><mml:mrow><mml:mo>-</mml:mo><mml:mo>(</mml:mo><mml:msub><mml:mi>d</mml:mi><mml:mrow><mml:mi>i</mml:mi><mml:mo>,</mml:mo><mml:mi>j</mml:mi></mml:mrow></mml:msub><mml:mo>/</mml:mo><mml:mi>L</mml:mi><mml:msup><mml:mo>)</mml:mo><mml:mn mathvariant="normal">2</mml:mn></mml:msup><mml:mo>/</mml:mo><mml:mn mathvariant="normal">2</mml:mn></mml:mrow></mml:msup><mml:mo>,</mml:mo></mml:mrow></mml:math></disp-formula>
          where <inline-formula><mml:math id="M100" display="inline"><mml:mrow><mml:msub><mml:mi>d</mml:mi><mml:mrow><mml:mi>i</mml:mi><mml:mo>,</mml:mo><mml:mi>j</mml:mi></mml:mrow></mml:msub></mml:mrow></mml:math></inline-formula> represents the distance between two grid cells <inline-formula><mml:math id="M101" display="inline"><mml:mi>i</mml:mi></mml:math></inline-formula> and <inline-formula><mml:math id="M102" display="inline"><mml:mi>j</mml:mi></mml:math></inline-formula>. In <xref ref-type="bibr" rid="bib1.bibx20" id="text.71"/>, the correlation length scale <inline-formula><mml:math id="M103" display="inline"><mml:mi>L</mml:mi></mml:math></inline-formula> was configured to be 800 km, which was found to be suitable to simulate the main characteristics of the dust event studied there. In this study, however, a smaller length scale <inline-formula><mml:math id="M104" display="inline"><mml:mrow><mml:mi>L</mml:mi><mml:mo>=</mml:mo><mml:mn mathvariant="normal">300</mml:mn></mml:mrow></mml:math></inline-formula> km is used, which gives a higher spatial degree of freedom, while it can still be resolved by the assimilation due to the larger number of MODIS AOD and PM<inline-formula><mml:math id="M105" display="inline"><mml:msub><mml:mi/><mml:mn mathvariant="normal">10</mml:mn></mml:msub></mml:math></inline-formula> concentration measurements.</p>
      <p id="d1e2294">With the covariance matrix <inline-formula><mml:math id="M106" display="inline"><mml:mrow><mml:msub><mml:mi mathvariant="bold">B</mml:mi><mml:mi mathvariant="italic">β</mml:mi></mml:msub></mml:mrow></mml:math></inline-formula>, an ensemble (<inline-formula><mml:math id="M107" display="inline"><mml:mrow><mml:mi>N</mml:mi><mml:mo>=</mml:mo><mml:mn mathvariant="normal">200</mml:mn></mml:mrow></mml:math></inline-formula>) of samples of <inline-formula><mml:math id="M108" display="inline"><mml:mi mathvariant="italic">β</mml:mi></mml:math></inline-formula> is generated randomly. These ensemble samples are then applied in our dust model, and each of them produces an emission forecast. The covariance of these emission fields are approximated as follows:
            <disp-formula id="Ch1.E8" content-type="numbered"><label>8</label><mml:math id="M109" display="block"><mml:mrow><mml:mi mathvariant="bold">B</mml:mi><mml:mo>≈</mml:mo><mml:mstyle displaystyle="true"><mml:mfrac style="display"><mml:mn mathvariant="normal">1</mml:mn><mml:mrow><mml:mi>N</mml:mi><mml:mo>-</mml:mo><mml:mn mathvariant="normal">1</mml:mn></mml:mrow></mml:mfrac></mml:mstyle><mml:munderover><mml:mo movablelimits="false">∑</mml:mo><mml:mrow><mml:mi>i</mml:mi><mml:mo>=</mml:mo><mml:mn mathvariant="normal">1</mml:mn></mml:mrow><mml:mi>N</mml:mi></mml:munderover><mml:mo>(</mml:mo><mml:msub><mml:mi mathvariant="bold-italic">f</mml:mi><mml:mrow><mml:msub><mml:mi>u</mml:mi><mml:mi mathvariant="normal">t</mml:mi></mml:msub><mml:mo>,</mml:mo><mml:mi>i</mml:mi></mml:mrow></mml:msub><mml:mo>-</mml:mo><mml:msub><mml:mover accent="true"><mml:mi mathvariant="bold-italic">f</mml:mi><mml:mo mathvariant="normal">‾</mml:mo></mml:mover><mml:mrow><mml:msub><mml:mi>u</mml:mi><mml:mi mathvariant="normal">t</mml:mi></mml:msub></mml:mrow></mml:msub><mml:mo>)</mml:mo><mml:mo>(</mml:mo><mml:msub><mml:mi mathvariant="bold-italic">f</mml:mi><mml:mrow><mml:msub><mml:mi>u</mml:mi><mml:mi mathvariant="normal">t</mml:mi></mml:msub><mml:mo>,</mml:mo><mml:mi>i</mml:mi></mml:mrow></mml:msub><mml:mo>-</mml:mo><mml:msub><mml:mover accent="true"><mml:mi mathvariant="bold-italic">f</mml:mi><mml:mo mathvariant="normal">‾</mml:mo></mml:mover><mml:mrow><mml:msub><mml:mi>u</mml:mi><mml:mi mathvariant="normal">t</mml:mi></mml:msub></mml:mrow></mml:msub><mml:msup><mml:mo>)</mml:mo><mml:mi mathvariant="normal">T</mml:mi></mml:msup><mml:mo>,</mml:mo></mml:mrow></mml:math></disp-formula>
          where <inline-formula><mml:math id="M110" display="inline"><mml:mrow><mml:msub><mml:mi mathvariant="bold-italic">f</mml:mi><mml:mrow><mml:msub><mml:mi>u</mml:mi><mml:mi mathvariant="normal">t</mml:mi></mml:msub><mml:mo>,</mml:mo><mml:mi>i</mml:mi></mml:mrow></mml:msub></mml:mrow></mml:math></inline-formula> represents the emission vector computed using the friction velocity threshold ensemble member <inline-formula><mml:math id="M111" display="inline"><mml:mi>i</mml:mi></mml:math></inline-formula>, while <inline-formula><mml:math id="M112" display="inline"><mml:mrow><mml:msub><mml:mover accent="true"><mml:mi mathvariant="bold-italic">f</mml:mi><mml:mo mathvariant="normal">‾</mml:mo></mml:mover><mml:mrow><mml:msub><mml:mi>u</mml:mi><mml:mi mathvariant="normal">t</mml:mi></mml:msub></mml:mrow></mml:msub></mml:mrow></mml:math></inline-formula> is the ensemble mean.</p>
      <p id="d1e2473">The observation term of the cost function, <inline-formula><mml:math id="M113" display="inline"><mml:mrow><mml:msubsup><mml:mi mathvariant="script">J</mml:mi><mml:mi mathvariant="normal">o</mml:mi><mml:mtext>PM</mml:mtext></mml:msubsup></mml:mrow></mml:math></inline-formula> and <inline-formula><mml:math id="M114" display="inline"><mml:mrow><mml:msubsup><mml:mi mathvariant="script">J</mml:mi><mml:mi mathvariant="normal">o</mml:mi><mml:mtext>AOD</mml:mtext></mml:msubsup></mml:mrow></mml:math></inline-formula> quantify mismatch between dust simulation and PM<inline-formula><mml:math id="M115" display="inline"><mml:msub><mml:mi/><mml:mn mathvariant="normal">10</mml:mn></mml:msub></mml:math></inline-formula> and between dust simulation and AOD measurements, respectively.

                <disp-formula specific-use="gather" content-type="numbered"><mml:math id="M116" display="block"><mml:mtable displaystyle="true"><mml:mlabeledtr id="Ch1.E9"><mml:mtd><mml:mtext>9</mml:mtext></mml:mtd><mml:mtd><mml:mrow><mml:mstyle class="stylechange" displaystyle="true"/><mml:mtable class="split" rowspacing="0.2ex" displaystyle="true" columnalign="right left"><mml:mtr><mml:mtd><mml:mrow><mml:msubsup><mml:mi mathvariant="script">J</mml:mi><mml:mi mathvariant="normal">o</mml:mi><mml:mtext>PM</mml:mtext></mml:msubsup><mml:mo>(</mml:mo><mml:mi mathvariant="bold-italic">f</mml:mi><mml:mo>)</mml:mo></mml:mrow></mml:mtd><mml:mtd><mml:mrow><mml:mo>=</mml:mo><mml:mstyle displaystyle="true"><mml:mfrac style="display"><mml:mn mathvariant="normal">1</mml:mn><mml:mn mathvariant="normal">2</mml:mn></mml:mfrac></mml:mstyle><mml:munderover><mml:mo movablelimits="false">∑</mml:mo><mml:mrow><mml:mi>i</mml:mi><mml:mo>=</mml:mo><mml:mn mathvariant="normal">1</mml:mn></mml:mrow><mml:mi>m</mml:mi></mml:munderover><mml:msup><mml:mfenced close="}" open="{"><mml:mrow><mml:msubsup><mml:mi mathvariant="bold-italic">y</mml:mi><mml:mi>i</mml:mi><mml:mtext>PM</mml:mtext></mml:msubsup><mml:mo>-</mml:mo><mml:msubsup><mml:mi mathvariant="script">H</mml:mi><mml:mi>i</mml:mi><mml:mtext>PM</mml:mtext></mml:msubsup><mml:msub><mml:mi mathvariant="script">M</mml:mi><mml:mi>i</mml:mi></mml:msub><mml:mo>(</mml:mo><mml:mi mathvariant="bold-italic">f</mml:mi><mml:mo>)</mml:mo></mml:mrow></mml:mfenced><mml:mtext>T</mml:mtext></mml:msup><mml:msubsup><mml:msup><mml:mi mathvariant="bold">O</mml:mi><mml:mtext>PM</mml:mtext></mml:msup><mml:mi>i</mml:mi><mml:mrow><mml:mo>-</mml:mo><mml:mn mathvariant="normal">1</mml:mn></mml:mrow></mml:msubsup></mml:mrow></mml:mtd></mml:mtr><mml:mtr><mml:mtd/><mml:mtd><mml:mrow><mml:mo>⋅</mml:mo><mml:mfenced open="{" close="}"><mml:mrow><mml:msubsup><mml:mi mathvariant="bold-italic">y</mml:mi><mml:mi>i</mml:mi><mml:mtext>PM</mml:mtext></mml:msubsup><mml:mo>-</mml:mo><mml:msubsup><mml:mi mathvariant="script">H</mml:mi><mml:mi>i</mml:mi><mml:mtext>PM</mml:mtext></mml:msubsup><mml:msub><mml:mi mathvariant="script">M</mml:mi><mml:mi>i</mml:mi></mml:msub><mml:mo>(</mml:mo><mml:mi mathvariant="bold-italic">f</mml:mi><mml:mo>)</mml:mo><mml:mspace linebreak="nobreak" width="0.25em"/></mml:mrow></mml:mfenced></mml:mrow></mml:mtd></mml:mtr></mml:mtable></mml:mrow></mml:mtd></mml:mlabeledtr><mml:mlabeledtr id="Ch1.E10"><mml:mtd><mml:mtext>10</mml:mtext></mml:mtd><mml:mtd><mml:mrow><mml:mstyle class="stylechange" displaystyle="true"/><mml:mtable class="split" rowspacing="0.2ex" displaystyle="true" columnalign="right left"><mml:mtr><mml:mtd><mml:mrow><mml:msubsup><mml:mi mathvariant="script">J</mml:mi><mml:mi mathvariant="normal">o</mml:mi><mml:mtext>AOD</mml:mtext></mml:msubsup><mml:mo>(</mml:mo><mml:mi mathvariant="bold-italic">f</mml:mi><mml:mo>)</mml:mo></mml:mrow></mml:mtd><mml:mtd><mml:mrow><mml:mo>=</mml:mo><mml:mstyle displaystyle="true"><mml:mfrac style="display"><mml:mn mathvariant="normal">1</mml:mn><mml:mn mathvariant="normal">2</mml:mn></mml:mfrac></mml:mstyle><mml:munderover><mml:mo movablelimits="false">∑</mml:mo><mml:mrow><mml:mi>i</mml:mi><mml:mo>=</mml:mo><mml:mn mathvariant="normal">1</mml:mn></mml:mrow><mml:mi>n</mml:mi></mml:munderover><mml:msup><mml:mfenced close="}" open="{"><mml:mrow><mml:msubsup><mml:mi mathvariant="bold-italic">y</mml:mi><mml:mi>i</mml:mi><mml:mtext>AOD</mml:mtext></mml:msubsup><mml:mo>-</mml:mo><mml:msubsup><mml:mi mathvariant="script">H</mml:mi><mml:mi>i</mml:mi><mml:mtext>AOD</mml:mtext></mml:msubsup><mml:msub><mml:mi mathvariant="script">M</mml:mi><mml:mi>i</mml:mi></mml:msub><mml:mo>(</mml:mo><mml:mi mathvariant="bold-italic">f</mml:mi><mml:mo>)</mml:mo></mml:mrow></mml:mfenced><mml:mtext>T</mml:mtext></mml:msup></mml:mrow></mml:mtd></mml:mtr><mml:mtr><mml:mtd/><mml:mtd><mml:mrow><mml:mo>⋅</mml:mo><mml:msubsup><mml:msup><mml:mi mathvariant="bold">O</mml:mi><mml:mtext>AOD</mml:mtext></mml:msup><mml:mi>i</mml:mi><mml:mrow><mml:mo>-</mml:mo><mml:mn mathvariant="normal">1</mml:mn></mml:mrow></mml:msubsup><mml:mfenced open="{" close="}"><mml:mrow><mml:msubsup><mml:mi mathvariant="bold-italic">y</mml:mi><mml:mi>i</mml:mi><mml:mtext>AOD</mml:mtext></mml:msubsup><mml:mo>-</mml:mo><mml:msubsup><mml:mi mathvariant="script">H</mml:mi><mml:mi>i</mml:mi><mml:mtext>AOD</mml:mtext></mml:msubsup><mml:msub><mml:mi mathvariant="script">M</mml:mi><mml:mi>i</mml:mi></mml:msub><mml:mo>(</mml:mo><mml:mi mathvariant="bold-italic">f</mml:mi><mml:mo>)</mml:mo></mml:mrow></mml:mfenced></mml:mrow></mml:mtd></mml:mtr></mml:mtable><mml:mo>,</mml:mo></mml:mrow></mml:mtd></mml:mlabeledtr></mml:mtable></mml:math></disp-formula>

            where <inline-formula><mml:math id="M117" display="inline"><mml:mi>m</mml:mi></mml:math></inline-formula> and <inline-formula><mml:math id="M118" display="inline"><mml:mi>n</mml:mi></mml:math></inline-formula> are the number of time steps within the assimilation window; <inline-formula><mml:math id="M119" display="inline"><mml:mrow><mml:msup><mml:mi mathvariant="bold-italic">y</mml:mi><mml:mtext>PM</mml:mtext></mml:msup></mml:mrow></mml:math></inline-formula> and <inline-formula><mml:math id="M120" display="inline"><mml:mrow><mml:msup><mml:mi mathvariant="bold-italic">y</mml:mi><mml:mtext>AOD</mml:mtext></mml:msup></mml:mrow></mml:math></inline-formula> contain the pre-processed PM<inline-formula><mml:math id="M121" display="inline"><mml:msub><mml:mi/><mml:mn mathvariant="normal">10</mml:mn></mml:msub></mml:math></inline-formula> and MODIS AOD measurements; <inline-formula><mml:math id="M122" display="inline"><mml:mi mathvariant="script">M</mml:mi></mml:math></inline-formula> denotes the LOTOS-EUROS/dust transport model that is driven by the emission <inline-formula><mml:math id="M123" display="inline"><mml:mi mathvariant="bold-italic">f</mml:mi></mml:math></inline-formula>, and <inline-formula><mml:math id="M124" display="inline"><mml:mrow><mml:msup><mml:mi mathvariant="script">H</mml:mi><mml:mtext>PM</mml:mtext></mml:msup></mml:mrow></mml:math></inline-formula> and <inline-formula><mml:math id="M125" display="inline"><mml:mrow><mml:msup><mml:mi mathvariant="script">H</mml:mi><mml:mtext>AOD</mml:mtext></mml:msup></mml:mrow></mml:math></inline-formula> are the observation operators that convert simulated dust concentrations into PM<inline-formula><mml:math id="M126" display="inline"><mml:msub><mml:mi/><mml:mn mathvariant="normal">10</mml:mn></mml:msub></mml:math></inline-formula> and AOD observation space. The PM<inline-formula><mml:math id="M127" display="inline"><mml:msub><mml:mi/><mml:mn mathvariant="normal">10</mml:mn></mml:msub></mml:math></inline-formula> and AOD observation mismatch terms are weighted by observation error covariance <inline-formula><mml:math id="M128" display="inline"><mml:mrow><mml:msup><mml:mi mathvariant="bold">O</mml:mi><mml:mtext>PM</mml:mtext></mml:msup></mml:mrow></mml:math></inline-formula> and <inline-formula><mml:math id="M129" display="inline"><mml:mrow><mml:msup><mml:mi mathvariant="bold">O</mml:mi><mml:mtext>AOD</mml:mtext></mml:msup></mml:mrow></mml:math></inline-formula>. The uncertainties in the observations are assumed to be independent, and hence both <inline-formula><mml:math id="M130" display="inline"><mml:mrow><mml:msup><mml:mi mathvariant="bold">O</mml:mi><mml:mtext>PM</mml:mtext></mml:msup></mml:mrow></mml:math></inline-formula> and <inline-formula><mml:math id="M131" display="inline"><mml:mrow><mml:msup><mml:mi mathvariant="bold">O</mml:mi><mml:mtext>AOD</mml:mtext></mml:msup></mml:mrow></mml:math></inline-formula> are diagonal matrices.</p>
      <p id="d1e2918">Both the instrument and representing errors are considered when the observation error covariance, <inline-formula><mml:math id="M132" display="inline"><mml:mrow><mml:msup><mml:mi mathvariant="bold">O</mml:mi><mml:mtext>PM</mml:mtext></mml:msup></mml:mrow></mml:math></inline-formula> and <inline-formula><mml:math id="M133" display="inline"><mml:mrow><mml:msup><mml:mi mathvariant="bold">O</mml:mi><mml:mtext>AOD</mml:mtext></mml:msup></mml:mrow></mml:math></inline-formula>, are designed. The uncertainty (square root of the individual diagonal element in <inline-formula><mml:math id="M134" display="inline"><mml:mrow><mml:msup><mml:mi mathvariant="bold">O</mml:mi><mml:mtext>PM</mml:mtext></mml:msup></mml:mrow></mml:math></inline-formula>) of the pre-processed PM<inline-formula><mml:math id="M135" display="inline"><mml:msub><mml:mi/><mml:mn mathvariant="normal">10</mml:mn></mml:msub></mml:math></inline-formula> measurements for assimilation is assumed to be due to uncertainty in the PM<inline-formula><mml:math id="M136" display="inline"><mml:msub><mml:mi/><mml:mn mathvariant="normal">10</mml:mn></mml:msub></mml:math></inline-formula> data and the non-dust PM<inline-formula><mml:math id="M137" display="inline"><mml:msub><mml:mi/><mml:mn mathvariant="normal">10</mml:mn></mml:msub></mml:math></inline-formula> bias correction. We have used <inline-formula><mml:math id="M138" display="inline"><mml:mrow><mml:msup><mml:mi mathvariant="italic">σ</mml:mi><mml:mtext>PM</mml:mtext></mml:msup><mml:mo>=</mml:mo><mml:mo>max⁡</mml:mo><mml:mo>(</mml:mo><mml:mn mathvariant="normal">200</mml:mn><mml:mo>,</mml:mo><mml:mspace linebreak="nobreak" width="0.25em"/><mml:mn mathvariant="normal">10</mml:mn><mml:mspace width="0.125em" linebreak="nobreak"/><mml:mi mathvariant="italic">%</mml:mi><mml:mo>⋅</mml:mo><mml:msup><mml:mi>y</mml:mi><mml:mtext>PM</mml:mtext></mml:msup><mml:mo>+</mml:mo><mml:mn mathvariant="normal">180</mml:mn><mml:mo>)</mml:mo></mml:mrow></mml:math></inline-formula> to characterize the uncertainty of PM<inline-formula><mml:math id="M139" display="inline"><mml:msub><mml:mi/><mml:mn mathvariant="normal">10</mml:mn></mml:msub></mml:math></inline-formula> data. It follows the choice of 10 % in our previous study <xref ref-type="bibr" rid="bib1.bibx20" id="paren.72"/>, with uncertainty inflated for this application. This is mainly to prevent the posterior from getting too close to the low-value PM<inline-formula><mml:math id="M140" display="inline"><mml:msub><mml:mi/><mml:mn mathvariant="normal">10</mml:mn></mml:msub></mml:math></inline-formula> observations and hence being model divergent.
In addition, the uncertainty of the non-dust PM<inline-formula><mml:math id="M141" display="inline"><mml:msub><mml:mi/><mml:mn mathvariant="normal">10</mml:mn></mml:msub></mml:math></inline-formula> simulation <inline-formula><mml:math id="M142" display="inline"><mml:mrow><mml:msup><mml:mi mathvariant="italic">σ</mml:mi><mml:mtext>BC</mml:mtext></mml:msup></mml:mrow></mml:math></inline-formula> that is introduced in Sect. <xref ref-type="sec" rid="Ch1.S2.SS2"/> is set to 40 % following the aerosol simulation analysis over China using LOTOS-EUROS <xref ref-type="bibr" rid="bib1.bibx54" id="paren.73"/>.
The integrated uncertainty <inline-formula><mml:math id="M143" display="inline"><mml:mrow><mml:msup><mml:mi mathvariant="italic">σ</mml:mi><mml:mtext>integrated</mml:mtext></mml:msup></mml:mrow></mml:math></inline-formula> for using the bias-corrected PM<inline-formula><mml:math id="M144" display="inline"><mml:msub><mml:mi/><mml:mn mathvariant="normal">10</mml:mn></mml:msub></mml:math></inline-formula> to represent the dust load is then calculated as
            <disp-formula id="Ch1.E11" content-type="numbered"><label>11</label><mml:math id="M145" display="block"><mml:mrow><mml:msup><mml:mi mathvariant="italic">σ</mml:mi><mml:mtext>integrated</mml:mtext></mml:msup><mml:mo>=</mml:mo><mml:mo mathvariant="italic">{</mml:mo><mml:mspace linebreak="nobreak" width="0.25em"/><mml:mo>(</mml:mo><mml:msup><mml:mi mathvariant="italic">σ</mml:mi><mml:mtext>PM</mml:mtext></mml:msup><mml:msup><mml:mo>)</mml:mo><mml:mn mathvariant="normal">2</mml:mn></mml:msup><mml:mo>+</mml:mo><mml:mo>(</mml:mo><mml:msup><mml:mi mathvariant="italic">σ</mml:mi><mml:mtext>BC</mml:mtext></mml:msup><mml:msup><mml:mo>)</mml:mo><mml:mn mathvariant="normal">2</mml:mn></mml:msup><mml:mspace linebreak="nobreak" width="0.25em"/><mml:msup><mml:mo mathvariant="italic">}</mml:mo><mml:mn mathvariant="normal">0.5</mml:mn></mml:msup><mml:mspace linebreak="nobreak" width="0.125em"/><mml:mo>.</mml:mo></mml:mrow></mml:math></disp-formula>
          Snapshots of <inline-formula><mml:math id="M146" display="inline"><mml:mrow><mml:msup><mml:mi mathvariant="italic">σ</mml:mi><mml:mtext>PM</mml:mtext></mml:msup></mml:mrow></mml:math></inline-formula> and <inline-formula><mml:math id="M147" display="inline"><mml:mrow><mml:msup><mml:mi mathvariant="italic">σ</mml:mi><mml:mtext>integrated</mml:mtext></mml:msup></mml:mrow></mml:math></inline-formula> distribution accompanying the PM<inline-formula><mml:math id="M148" display="inline"><mml:msub><mml:mi/><mml:mn mathvariant="normal">10</mml:mn></mml:msub></mml:math></inline-formula> measurements shown in Figs. <xref ref-type="fig" rid="Ch1.F2"/>a.i, <xref ref-type="fig" rid="Ch1.F3"/>a.i, and <xref ref-type="fig" rid="Ch1.F4"/>a.i in the three dust events are shown in Figs. S1c–d,  S3c–d, and  S5c–d, respectively.</p>
      <p id="d1e3181">The integrated uncertainty of AOD measurements for assimilation is also calculated as the sum of the instrument error and the error of the non-dust AOD bias correction.
The former is taken directly from the MODIS Deep Blue product, while the uncertainty of non-dust AOD simulation is set to 40 % as well. Snapshots of the AOD instrument uncertainty and integrated uncertainty with respect to the AOD observations in the three dust events can be found in Figs. S2c–d,  S4c–d, and S6c–d.</p>
</sec>
<sec id="Ch1.S3.SS2">
  <label>3.2</label><title>Assimilation window</title>
      <p id="d1e3192">Table <xref ref-type="table" rid="Ch1.T1"/> shows the timeline of the three severe dust events studied here.
These dust events have a short duration, and therefore a single assimilation window of length 72, 72, and 48 h is used, respectively. Dust emissions occur at the start of the assimilation window when a plume is lifted high to be carried downwind. The PM<inline-formula><mml:math id="M149" display="inline"><mml:msub><mml:mi/><mml:mn mathvariant="normal">10</mml:mn></mml:msub></mml:math></inline-formula> network is only able to observe the plume when it has moved downwind, and is already far away from the source region. Therefore, assimilation windows covering both the time of dust emission and the moment of observation are necessary. When the dust plume is carried further southward or eastward, the error of the simulated dust concentration grows steadily due to the accumulation of dust transport and deposition uncertainty. Therefore, AOD or PM<inline-formula><mml:math id="M150" display="inline"><mml:msub><mml:mi/><mml:mn mathvariant="normal">10</mml:mn></mml:msub></mml:math></inline-formula> measurements out of the assimilation window help little in emission inversion and are therefore not used here.</p>
</sec>
</sec>
<sec id="Ch1.S4">
  <label>4</label><title>Results and discussions</title>
<sec id="Ch1.S4.SS1">
  <label>4.1</label><title>Dust storm inverse modeling</title>
      <p id="d1e3231">Using the assimilation system introduced in Sect. <xref ref-type="sec" rid="Ch1.S3"/>, the emission inversions were performed by assimilating the bias-corrected PM<inline-formula><mml:math id="M151" display="inline"><mml:msub><mml:mi/><mml:mn mathvariant="normal">10</mml:mn></mml:msub></mml:math></inline-formula> and MODIS AODs processed in Sect. <xref ref-type="sec" rid="Ch1.S2.SS2"/> and Sect. <xref ref-type="sec" rid="Ch1.S2.SS3"/>, which is referred to as <italic>multi-observation</italic> assimilation in this study.
First, the posterior emission analysis is carried out in Sect. <xref ref-type="sec" rid="Ch1.S4.SS1.SSS1"/>, then the dust simulation driven by the posterior emission result is illustrated in Sect. <xref ref-type="sec" rid="Ch1.S4.SS1.SSS2"/>.</p>
<sec id="Ch1.S4.SS1.SSS1">
  <label>4.1.1</label><title>Dust emission analysis</title>
      <p id="d1e3264">To assess the spatial pattern of the dust emission, the accumulated emissions over the assimilation window are calculated for each of the dust events.
Figure <xref ref-type="fig" rid="Ch1.F6"/> shows the map of the a priori and posterior accumulated dust emissions in SD1, SD2, and SD3 over the potential source regions. The a priori model simulation indicates that dust emission took place over both the Alxa desert (part of the Chinese Gobi) and the central region of the Mongolian Gobi desert during SD1, and their maximum emission flux exceeded 200 g m<inline-formula><mml:math id="M152" display="inline"><mml:msup><mml:mi/><mml:mrow><mml:mo>-</mml:mo><mml:mn mathvariant="normal">2</mml:mn></mml:mrow></mml:msup></mml:math></inline-formula>.
Through assimilating the MODIS AODs and ground-based PM<inline-formula><mml:math id="M153" display="inline"><mml:msub><mml:mi/><mml:mn mathvariant="normal">10</mml:mn></mml:msub></mml:math></inline-formula> concentration measurements, the emission field is estimated. It indicates that emission took place in more grid cells that are located in the Alxa desert and in eastern Mongolia.
However, the dust plume released from the Alxa desert did not move far towards the south or the east, while Mongolian dust was the main source of dust affection in northern China as will be discussed in Sect. <xref ref-type="sec" rid="Ch1.S4.SS3"/>.</p>

      <?xmltex \floatpos{t}?><fig id="Ch1.F6" specific-use="star"><?xmltex \currentcnt{6}?><?xmltex \def\figurename{Figure}?><label>Figure 6</label><caption><p id="d1e3294">Distribution of the a priori <bold>(i)</bold> and posterior <bold>(ii)</bold> accumulated dust emission for the SD1 <bold>(a)</bold>, SD2 <bold>(b)</bold>, and SD3 <bold>(c)</bold>; the total priori and posterior emission either from China or from Mongolia during SD1, SD2, and SD3 <bold>(d)</bold>.</p></caption>
            <?xmltex \igopts{width=369.885827pt}?><graphic xlink:href="https://acp.copernicus.org/articles/22/6393/2022/acp-22-6393-2022-f06.png"/>

          </fig>

      <p id="d1e3322">For the SD2 event, the accumulated a priori emissions are rather high, especially around the border between China and Mongolia. The emission accumulations in several grid cells here are in the order of 300 g m<inline-formula><mml:math id="M154" display="inline"><mml:msup><mml:mi/><mml:mrow><mml:mo>-</mml:mo><mml:mn mathvariant="normal">2</mml:mn></mml:mrow></mml:msup></mml:math></inline-formula>, which results in an overestimation of surface dust concentration and AOD simulation as shown in Fig. <xref ref-type="fig" rid="Ch1.F3"/>a.ii–b.ii. In contrast, the posterior simulation estimates that the border region is almost free of dust emission during the event, and the dust plume is actually attributed to the dust emission from northern Mongolia. The Alxa desert also contributed partially to the dust plume that effected the FWP region.</p>
      <p id="d1e3340">For the SD3 event, the a priori emission simulation suggests that most of the dust originated from the source regions in China rather than from the bare lands in Mongolia. Especially in the Tengger (also part of the Chinese Gobi) and Alxa deserts, the accumulated emissions reach values over 200 g m<inline-formula><mml:math id="M155" display="inline"><mml:msup><mml:mi/><mml:mrow><mml:mo>-</mml:mo><mml:mn mathvariant="normal">2</mml:mn></mml:mrow></mml:msup></mml:math></inline-formula>. By assimilating the MODIS AOD and PM<inline-formula><mml:math id="M156" display="inline"><mml:msub><mml:mi/><mml:mn mathvariant="normal">10</mml:mn></mml:msub></mml:math></inline-formula> concentration data, the posterior dust emission field is updated and the assimilated estimate of the accumulation map is plotted in Fig. <xref ref-type="fig" rid="Ch1.F6"/>c.ii. What is interesting about the result in the posterior map is that a much smaller amount of dust is estimated to be emitted from the Tengger and Alxa deserts, while sparsely vegetated regions in the northeastern part of the Inner Mongolia province are estimated to be a significant source. Dust from these regions is estimated to be transported towards NCP and northeast China, as will be discussed later. Close inspection of the posterior emission map shows that the emission over the Mongolian and Chinese border region was also one of the main sources in the SD3 event.</p>
      <p id="d1e3366">Apart from the spatial patterns of dust emission, the total mass of the emissions from Mongolia and China was calculated and shown in Fig. <xref ref-type="fig" rid="Ch1.F6"/>d, which helps to evaluate the emission intensity of dust in these two countries.
Although the spatial patterns have been strongly changed by assimilation of the MODIS and PM<inline-formula><mml:math id="M157" display="inline"><mml:msub><mml:mi/><mml:mn mathvariant="normal">10</mml:mn></mml:msub></mml:math></inline-formula> concentration measurements, the posterior emission sums per country and event are in most cases close to the a priori values. An exception is the value for China in SD3, for which the total emission is decreased from <inline-formula><mml:math id="M158" display="inline"><mml:mrow><mml:mn mathvariant="normal">15.4</mml:mn><mml:mo>×</mml:mo><mml:msup><mml:mn mathvariant="normal">10</mml:mn><mml:mn mathvariant="normal">6</mml:mn></mml:msup></mml:mrow></mml:math></inline-formula> to <inline-formula><mml:math id="M159" display="inline"><mml:mrow><mml:mn mathvariant="normal">7.1</mml:mn><mml:mo>×</mml:mo><mml:msup><mml:mn mathvariant="normal">10</mml:mn><mml:mn mathvariant="normal">6</mml:mn></mml:msup></mml:mrow></mml:math></inline-formula> t by the assimilation. The emission sums show that the Mongolian Gobi is a stronger source of dust (<inline-formula><mml:math id="M160" display="inline"><mml:mrow><mml:mn mathvariant="normal">37.5</mml:mn><mml:mo>×</mml:mo><mml:msup><mml:mn mathvariant="normal">10</mml:mn><mml:mn mathvariant="normal">6</mml:mn></mml:msup></mml:mrow></mml:math></inline-formula> t) than the Chinese Gobi (<inline-formula><mml:math id="M161" display="inline"><mml:mrow><mml:mn mathvariant="normal">19.9</mml:mn><mml:mo>×</mml:mo><mml:msup><mml:mn mathvariant="normal">10</mml:mn><mml:mn mathvariant="normal">6</mml:mn></mml:msup></mml:mrow></mml:math></inline-formula> t).</p>
      <p id="d1e3441">The dust emission inversion successfully optimized the dust simulation using different types of observations of the dust plumes by adjusting the emission fields. The posterior emission field not only helps us exploit the spatial pattern of active dust sources, but also to simulate long-distance dust transport more accurately. This could be used for evaluation of the threats that dust imposes on human health, the transport system, and the Earth system.</p>
</sec>
<sec id="Ch1.S4.SS1.SSS2">
  <label>4.1.2</label><title>Simulated dust field</title>
      <p id="d1e3452">The impact of the assimilation was evaluated by comparing the simulated observations from the a priori and posterior simulations with the (bias corrected) observations. The purpose of this paper is to have an emission field that is as accurate as possible; therefore, we assimilated all available PM<inline-formula><mml:math id="M162" display="inline"><mml:msub><mml:mi/><mml:mn mathvariant="normal">10</mml:mn></mml:msub></mml:math></inline-formula> and MODIS AOD observations instead of leaving a subset of them for independent validation.
To quantify the performance, the root mean square error (RMSE) that calculates the deviation of the simulation with respect to the bias-corrected measurements is calculated.</p>
      <p id="d1e3464">A snapshot of the a priori and posterior surface dust concentration and AOD during SD1 are shown in the middle and right panels in Fig. <xref ref-type="fig" rid="Ch1.F2"/>, respectively. Both of the a priori and posterior simulations show a similar pattern as the (bias-corrected) PM<inline-formula><mml:math id="M163" display="inline"><mml:msub><mml:mi/><mml:mn mathvariant="normal">10</mml:mn></mml:msub></mml:math></inline-formula> and AOD observations that are shown in Figs. S1 and S2, or the raw data shown in the left panel of Fig. <xref ref-type="fig" rid="Ch1.F2"/>. However, the dust concentrations are underestimated in the a priori simulation in the entire region, resulting in a PM<inline-formula><mml:math id="M164" display="inline"><mml:msub><mml:mi/><mml:mn mathvariant="normal">10</mml:mn></mml:msub></mml:math></inline-formula> RMSE of 833 <inline-formula><mml:math id="M165" display="inline"><mml:mrow class="unit"><mml:mi mathvariant="normal">µ</mml:mi></mml:mrow></mml:math></inline-formula>g m<inline-formula><mml:math id="M166" display="inline"><mml:msup><mml:mi/><mml:mrow><mml:mo>-</mml:mo><mml:mn mathvariant="normal">3</mml:mn></mml:mrow></mml:msup></mml:math></inline-formula> and AOD RMSE of 1.36 and 1.53. Compared to the a priori simulation, the posterior emission field simulated a more severe dust plume, and the PM<inline-formula><mml:math id="M167" display="inline"><mml:msub><mml:mi/><mml:mn mathvariant="normal">10</mml:mn></mml:msub></mml:math></inline-formula> RMSE is reduced to 743 <inline-formula><mml:math id="M168" display="inline"><mml:mrow class="unit"><mml:mi mathvariant="normal">µ</mml:mi></mml:mrow></mml:math></inline-formula>g m<inline-formula><mml:math id="M169" display="inline"><mml:msup><mml:mi/><mml:mrow><mml:mo>-</mml:mo><mml:mn mathvariant="normal">3</mml:mn></mml:mrow></mml:msup></mml:math></inline-formula> and the AOD RMSE declined to 1.30 and 1.34, simultaneously. One important reason for this high error residue is the position mismatch among the simulation and observations. For instance, the plume front (red dashed line) that is visible in the MODIS AOD retrievals in Fig. <xref ref-type="fig" rid="Ch1.F2"/>b.i is about 100 km ahead of the front line shown in PM<inline-formula><mml:math id="M170" display="inline"><mml:msub><mml:mi/><mml:mn mathvariant="normal">10</mml:mn></mml:msub></mml:math></inline-formula> measurements in Fig. <xref ref-type="fig" rid="Ch1.F2"/>a.i.
It is possible that the dust plume in the higher layers moved faster and was further southeast than the dust cloud at the bottom layer.
However, this feature is not correctly captured by our LOTOS-EUROS/dust model.
Both the simulated plume of AOD and surface dust concentration are in the same position; the simulated plume fronts indicated in Fig. <xref ref-type="fig" rid="Ch1.F2"/>a.ii and b.ii moved faster than the front line indicated by PM<inline-formula><mml:math id="M171" display="inline"><mml:msub><mml:mi/><mml:mn mathvariant="normal">10</mml:mn></mml:msub></mml:math></inline-formula> measurements, but slower than the front line in the MODIS AOD. The mismatch in simulated vertical structure is mostly likely caused by uncertainty in the advection transport. The two-dimensional grid distortion technique <xref ref-type="bibr" rid="bib1.bibx24" id="paren.74"/>, which is independent of emission inversion, could adjust the horizontal position of the dust cloud simulation to better fit the available measurements, but is not yet able to adjust the vertical structure as is required here. A three-dimensional grid distortion with data measuring the vertical profile of the dust cloud is planned to solve this issue in our future research.</p>
      <p id="d1e3567">Figure <xref ref-type="fig" rid="Ch1.F3"/> shows snapshots of original AOD and PM<inline-formula><mml:math id="M172" display="inline"><mml:msub><mml:mi/><mml:mn mathvariant="normal">10</mml:mn></mml:msub></mml:math></inline-formula> measurements, as well as the a priori and posterior AOD and surface dust concentration simulation at 28 March 11:00 (CST) within SD2. What stands out in the a priori AOD simulation in Fig. <xref ref-type="fig" rid="Ch1.F3"/>a.ii is the overestimation, especially at the center of the plume in the NCP region. The standard model simulated AOD values even larger than 4, while the measurements were around 2 to 3. It should be noted that all the AOD measurements shown in Fig. <xref ref-type="fig" rid="Ch1.F3"/>a.i contain both the dust and non-dust fractions, and the baseline-removed AODs which can be found in Fig. S4 are therefore a bit lower. By assimilating these bias-corrected measurements, posterior AOD simulations in Fig. <xref ref-type="fig" rid="Ch1.F3"/>a.iii are now in better agreement with the AOD observations. The AOD RMSE is therefore reduced from 0.99 in the a priori simulation to 0.81 in the posterior simulation. The improvement is also seen in the surface PM<inline-formula><mml:math id="M173" display="inline"><mml:msub><mml:mi/><mml:mn mathvariant="normal">10</mml:mn></mml:msub></mml:math></inline-formula> simulation, where the RMSE score decreases from 471 <inline-formula><mml:math id="M174" display="inline"><mml:mrow class="unit"><mml:mi mathvariant="normal">µ</mml:mi></mml:mrow></mml:math></inline-formula>g m<inline-formula><mml:math id="M175" display="inline"><mml:msup><mml:mi/><mml:mrow><mml:mo>-</mml:mo><mml:mn mathvariant="normal">3</mml:mn></mml:mrow></mml:msup></mml:math></inline-formula> in the a priori to 359 <inline-formula><mml:math id="M176" display="inline"><mml:mrow class="unit"><mml:mi mathvariant="normal">µ</mml:mi></mml:mrow></mml:math></inline-formula>g m<inline-formula><mml:math id="M177" display="inline"><mml:msup><mml:mi/><mml:mrow><mml:mo>-</mml:mo><mml:mn mathvariant="normal">3</mml:mn></mml:mrow></mml:msup></mml:math></inline-formula> in the posterior simulation.</p>
      <p id="d1e3637">Scenes of original AOD and PM<inline-formula><mml:math id="M178" display="inline"><mml:msub><mml:mi/><mml:mn mathvariant="normal">10</mml:mn></mml:msub></mml:math></inline-formula> measurements and the a priori and posterior simulated AOD and surface dust concentration at 15 April 11:00 (CST) during SD3 are plotted in Fig. <xref ref-type="fig" rid="Ch1.F4"/>. It is apparent from this figure that dust concentrations are overestimated by the model within this event as well. For a case in point, the PM<inline-formula><mml:math id="M179" display="inline"><mml:msub><mml:mi/><mml:mn mathvariant="normal">10</mml:mn></mml:msub></mml:math></inline-formula> measurements in Fig. <xref ref-type="fig" rid="Ch1.F4"/>a.i show that FWP is almost free of dust at that moment, and this is confirmed by the baseline-removed PM<inline-formula><mml:math id="M180" display="inline"><mml:msub><mml:mi/><mml:mn mathvariant="normal">10</mml:mn></mml:msub></mml:math></inline-formula> measurements shown in Fig. S5. The multi-observation assimilation successfully resolves the PM<inline-formula><mml:math id="M181" display="inline"><mml:msub><mml:mi/><mml:mn mathvariant="normal">10</mml:mn></mml:msub></mml:math></inline-formula> and AOD measurements. The simulation driven by the posterior emission field is in much better agreement with the measurements, with the PM<inline-formula><mml:math id="M182" display="inline"><mml:msub><mml:mi/><mml:mn mathvariant="normal">10</mml:mn></mml:msub></mml:math></inline-formula> RMSE reduced from 891 to 144 <inline-formula><mml:math id="M183" display="inline"><mml:mrow class="unit"><mml:mi mathvariant="normal">µ</mml:mi></mml:mrow></mml:math></inline-formula>g m<inline-formula><mml:math id="M184" display="inline"><mml:msup><mml:mi/><mml:mrow><mml:mo>-</mml:mo><mml:mn mathvariant="normal">3</mml:mn></mml:mrow></mml:msup></mml:math></inline-formula> and the AOD RMSE drastically decreased from 1.79 to 0.73.</p>
</sec>
</sec>
<sec id="Ch1.S4.SS2">
  <label>4.2</label><?xmltex \opttitle{AOD-only or PM${}_{{10}}$-only assimilation evaluation}?><title>AOD-only or PM<inline-formula><mml:math id="M185" display="inline"><mml:msub><mml:mi/><mml:mn mathvariant="normal">10</mml:mn></mml:msub></mml:math></inline-formula>-only assimilation evaluation</title>
      <p id="d1e3729">Next to the multi-observation (AOD and PM<inline-formula><mml:math id="M186" display="inline"><mml:msub><mml:mi/><mml:mn mathvariant="normal">10</mml:mn></mml:msub></mml:math></inline-formula> together) emission inversion described in Sect. <xref ref-type="sec" rid="Ch1.S4.SS1"/>, assimilation tests with the same configurations but using only AOD or PM<inline-formula><mml:math id="M187" display="inline"><mml:msub><mml:mi/><mml:mn mathvariant="normal">10</mml:mn></mml:msub></mml:math></inline-formula> observations are carried out as well. They are referred to as <italic>AOD-only</italic> and <italic>PM</italic><inline-formula><mml:math id="M188" display="inline"><mml:msub><mml:mi/><mml:mn mathvariant="normal">10</mml:mn></mml:msub></mml:math></inline-formula><italic>-only</italic> assimilation in this study. The difference and added value of the multi-observation assimilation vs. AOD-only/PM<inline-formula><mml:math id="M189" display="inline"><mml:msub><mml:mi/><mml:mn mathvariant="normal">10</mml:mn></mml:msub></mml:math></inline-formula>-only assimilation are analyzed.
In addition, once AOD or PM<inline-formula><mml:math id="M190" display="inline"><mml:msub><mml:mi/><mml:mn mathvariant="normal">10</mml:mn></mml:msub></mml:math></inline-formula> measurements are assimilated alone, the other one will be used as the independent data for validation.</p>
      <p id="d1e3787">Figure <xref ref-type="fig" rid="Ch1.F7"/> shows snapshots of the dust AOD and surface concentration simulation driven by the AOD-only (panels a.ii and b.ii) and PM<inline-formula><mml:math id="M191" display="inline"><mml:msub><mml:mi/><mml:mn mathvariant="normal">10</mml:mn></mml:msub></mml:math></inline-formula>-only (panels a.iii and b.iii) posterior emission field with the SD3.
It clearly illustrates the typical results that are observed in the other two events. As mentioned above, the AOD RMSE is reduced from 1.79 (the a priori) to 0.72 in the multi-observation assimilation. Once the MODIS AODs are assimilated alone, better dust AOD simulation performance is obtained, with the AOD RMSE further reduced to 0.69. The effectiveness of emission estimation through assimilating AOD is validated using the independent PM<inline-formula><mml:math id="M192" display="inline"><mml:msub><mml:mi/><mml:mn mathvariant="normal">10</mml:mn></mml:msub></mml:math></inline-formula> measurements; the PM<inline-formula><mml:math id="M193" display="inline"><mml:msub><mml:mi/><mml:mn mathvariant="normal">10</mml:mn></mml:msub></mml:math></inline-formula> RMSE decreased from 891 <inline-formula><mml:math id="M194" display="inline"><mml:mrow class="unit"><mml:mi mathvariant="normal">µ</mml:mi></mml:mrow></mml:math></inline-formula>g m<inline-formula><mml:math id="M195" display="inline"><mml:msup><mml:mi/><mml:mrow><mml:mo>-</mml:mo><mml:mn mathvariant="normal">3</mml:mn></mml:mrow></mml:msup></mml:math></inline-formula> (the a priori) to 210 <inline-formula><mml:math id="M196" display="inline"><mml:mrow class="unit"><mml:mi mathvariant="normal">µ</mml:mi></mml:mrow></mml:math></inline-formula>g m<inline-formula><mml:math id="M197" display="inline"><mml:msup><mml:mi/><mml:mrow><mml:mo>-</mml:mo><mml:mn mathvariant="normal">3</mml:mn></mml:mrow></mml:msup></mml:math></inline-formula> simultaneously. However, the PM<inline-formula><mml:math id="M198" display="inline"><mml:msub><mml:mi/><mml:mn mathvariant="normal">10</mml:mn></mml:msub></mml:math></inline-formula> RMSE stays at a slightly higher level compared to 143 <inline-formula><mml:math id="M199" display="inline"><mml:mrow class="unit"><mml:mi mathvariant="normal">µ</mml:mi></mml:mrow></mml:math></inline-formula>g m<inline-formula><mml:math id="M200" display="inline"><mml:msup><mml:mi/><mml:mrow><mml:mo>-</mml:mo><mml:mn mathvariant="normal">3</mml:mn></mml:mrow></mml:msup></mml:math></inline-formula> as obtained in the multi-observation assimilation. Similarly, the PM<inline-formula><mml:math id="M201" display="inline"><mml:msub><mml:mi/><mml:mn mathvariant="normal">10</mml:mn></mml:msub></mml:math></inline-formula>-only assimilation provides a further lower PM<inline-formula><mml:math id="M202" display="inline"><mml:msub><mml:mi/><mml:mn mathvariant="normal">10</mml:mn></mml:msub></mml:math></inline-formula> RMSE 133 <inline-formula><mml:math id="M203" display="inline"><mml:mrow class="unit"><mml:mi mathvariant="normal">µ</mml:mi></mml:mrow></mml:math></inline-formula>g m<inline-formula><mml:math id="M204" display="inline"><mml:msup><mml:mi/><mml:mrow><mml:mo>-</mml:mo><mml:mn mathvariant="normal">3</mml:mn></mml:mrow></mml:msup></mml:math></inline-formula>, but the simulated AOD field (RMSE <inline-formula><mml:math id="M205" display="inline"><mml:mo>=</mml:mo></mml:math></inline-formula> 0.77) is not as accurate as the one from the multi-observation assimilation.</p>

      <?xmltex \floatpos{t}?><fig id="Ch1.F7" specific-use="star"><?xmltex \currentcnt{7}?><?xmltex \def\figurename{Figure}?><label>Figure 7</label><caption><p id="d1e3937">The posterior dust surface dust concentration and AOD simulation either driven by the AOD-only or driven by the PM<inline-formula><mml:math id="M206" display="inline"><mml:msub><mml:mi/><mml:mn mathvariant="normal">10</mml:mn></mml:msub></mml:math></inline-formula>-only assimilation emission result during SD3. <bold>(a.i, b.i)</bold> The bias-corrected (BC) PM<inline-formula><mml:math id="M207" display="inline"><mml:msub><mml:mi/><mml:mn mathvariant="normal">10</mml:mn></mml:msub></mml:math></inline-formula> and AOD observations; <bold>(a.ii, b.ii)</bold> the dust surface concentration and AOD simulation driven by the posterior emission from AOD-only assimilation; <bold>(a.iii, b.iii)</bold> the dust surface concentration and AOD result driven by the posterior emission from PM<inline-formula><mml:math id="M208" display="inline"><mml:msub><mml:mi/><mml:mn mathvariant="normal">10</mml:mn></mml:msub></mml:math></inline-formula>-only assimilation at 11:00, 15 April.</p></caption>
          <?xmltex \igopts{width=497.923228pt}?><graphic xlink:href="https://acp.copernicus.org/articles/22/6393/2022/acp-22-6393-2022-f07.png"/>

        </fig>

      <p id="d1e3984">Scenes of dust AOD and surface concentration simulation from the AOD-only and PM<inline-formula><mml:math id="M209" display="inline"><mml:msub><mml:mi/><mml:mn mathvariant="normal">10</mml:mn></mml:msub></mml:math></inline-formula>-only assimilation in SD1 and SD2 can be found in Figs. S7 and S8. The corresponding AOD and PM<inline-formula><mml:math id="M210" display="inline"><mml:msub><mml:mi/><mml:mn mathvariant="normal">10</mml:mn></mml:msub></mml:math></inline-formula> RMSEs are calculated and shown in Table <xref ref-type="table" rid="Ch1.T2"/>. Very similar trends are found in the assimilation tests within SD2.
For SD3, the AOD-only or PM<inline-formula><mml:math id="M211" display="inline"><mml:msub><mml:mi/><mml:mn mathvariant="normal">10</mml:mn></mml:msub></mml:math></inline-formula>-only assimilation also results in closer dust simulation to the assimilated measurements. However, poor performance (compared to the a priori) is obtained against the independent observations. For instance, driven by the emission through assimilating the AOD, the posterior simulated dust concentrations are very different from the PM<inline-formula><mml:math id="M212" display="inline"><mml:msub><mml:mi/><mml:mn mathvariant="normal">10</mml:mn></mml:msub></mml:math></inline-formula> measurements, and the RMSE is lifted up to 887 <inline-formula><mml:math id="M213" display="inline"><mml:mrow class="unit"><mml:mi mathvariant="normal">µ</mml:mi></mml:mrow></mml:math></inline-formula>g m<inline-formula><mml:math id="M214" display="inline"><mml:msup><mml:mi/><mml:mrow><mml:mo>-</mml:mo><mml:mn mathvariant="normal">3</mml:mn></mml:mrow></mml:msup></mml:math></inline-formula> from 833 <inline-formula><mml:math id="M215" display="inline"><mml:mrow class="unit"><mml:mi mathvariant="normal">µ</mml:mi></mml:mrow></mml:math></inline-formula>g m<inline-formula><mml:math id="M216" display="inline"><mml:msup><mml:mi/><mml:mrow><mml:mo>-</mml:mo><mml:mn mathvariant="normal">3</mml:mn></mml:mrow></mml:msup></mml:math></inline-formula> (a priori). This is mainly because the AOD and PM<inline-formula><mml:math id="M217" display="inline"><mml:msub><mml:mi/><mml:mn mathvariant="normal">10</mml:mn></mml:msub></mml:math></inline-formula> measurements indicate the different dust plume position in the simulation space as described in Sect. <xref ref-type="sec" rid="Ch1.S4.SS1.SSS2"/>, and therefore the assimilation would lead the dust simulation into mismatch with the independent one.</p>
      <p id="d1e4077">Therefore, any AOD-only or PM<inline-formula><mml:math id="M218" display="inline"><mml:msub><mml:mi/><mml:mn mathvariant="normal">10</mml:mn></mml:msub></mml:math></inline-formula>-only assimilation would only result in the posterior being closer to the assimilated data, but the posterior simulation in the independent observation space is not ensured to be improved. It is even possible that the posterior simulation is misled when the dust vertical profile is not well reproduced (such as in our simulation over SD1) or systematical observation bias is present. In this case, multi-observation assimilation as used in this study is a safe choice to avoid the model divergence.</p>

<?xmltex \floatpos{t}?><table-wrap id="Ch1.T2" specific-use="star"><?xmltex \currentcnt{2}?><label>Table 2</label><caption><p id="d1e4092">Evaluation (RMSE) of the posterior dust simulation either by assimilating AOD, or by assimilating PM<inline-formula><mml:math id="M219" display="inline"><mml:msub><mml:mi/><mml:mn mathvariant="normal">10</mml:mn></mml:msub></mml:math></inline-formula>, or by assimilating both of them.</p></caption><oasis:table frame="topbot"><oasis:tgroup cols="6">
     <oasis:colspec colnum="1" colname="col1" align="left"/>
     <oasis:colspec colnum="2" colname="col2" align="left"/>
     <oasis:colspec colnum="3" colname="col3" align="right"/>
     <oasis:colspec colnum="4" colname="col4" align="right"/>
     <oasis:colspec colnum="5" colname="col5" align="right"/>
     <oasis:colspec colnum="6" colname="col6" align="right"/>
     <oasis:thead>
       <oasis:row rowsep="1">
         <oasis:entry colname="col1"/>
         <oasis:entry colname="col2">Timeline</oasis:entry>
         <oasis:entry colname="col3">Priori</oasis:entry>
         <oasis:entry colname="col4">Multi-observation</oasis:entry>
         <oasis:entry colname="col5">AOD only</oasis:entry>
         <oasis:entry colname="col6">PM<inline-formula><mml:math id="M220" display="inline"><mml:msub><mml:mi/><mml:mn mathvariant="normal">10</mml:mn></mml:msub></mml:math></inline-formula></oasis:entry>
       </oasis:row>
     </oasis:thead>
     <oasis:tbody>
       <oasis:row>
         <oasis:entry colname="col1">SD1</oasis:entry>
         <oasis:entry colname="col2">PM<inline-formula><mml:math id="M221" display="inline"><mml:msub><mml:mi/><mml:mn mathvariant="normal">10</mml:mn></mml:msub></mml:math></inline-formula> at 10:00, 15 March</oasis:entry>
         <oasis:entry colname="col3">833 <inline-formula><mml:math id="M222" display="inline"><mml:mrow class="unit"><mml:mi mathvariant="normal">µ</mml:mi></mml:mrow></mml:math></inline-formula>g m<inline-formula><mml:math id="M223" display="inline"><mml:msup><mml:mi/><mml:mrow><mml:mo>-</mml:mo><mml:mn mathvariant="normal">3</mml:mn></mml:mrow></mml:msup></mml:math></inline-formula></oasis:entry>
         <oasis:entry colname="col4">743 <inline-formula><mml:math id="M224" display="inline"><mml:mrow class="unit"><mml:mi mathvariant="normal">µ</mml:mi></mml:mrow></mml:math></inline-formula>g m<inline-formula><mml:math id="M225" display="inline"><mml:msup><mml:mi/><mml:mrow><mml:mo>-</mml:mo><mml:mn mathvariant="normal">3</mml:mn></mml:mrow></mml:msup></mml:math></inline-formula></oasis:entry>
         <oasis:entry colname="col5">887 <inline-formula><mml:math id="M226" display="inline"><mml:mrow class="unit"><mml:mi mathvariant="normal">µ</mml:mi></mml:mrow></mml:math></inline-formula>g m<inline-formula><mml:math id="M227" display="inline"><mml:msup><mml:mi/><mml:mrow><mml:mo>-</mml:mo><mml:mn mathvariant="normal">3</mml:mn></mml:mrow></mml:msup></mml:math></inline-formula></oasis:entry>
         <oasis:entry colname="col6">692 <inline-formula><mml:math id="M228" display="inline"><mml:mrow class="unit"><mml:mi mathvariant="normal">µ</mml:mi></mml:mrow></mml:math></inline-formula>g m<inline-formula><mml:math id="M229" display="inline"><mml:msup><mml:mi/><mml:mrow><mml:mo>-</mml:mo><mml:mn mathvariant="normal">3</mml:mn></mml:mrow></mml:msup></mml:math></inline-formula></oasis:entry>
       </oasis:row>
       <oasis:row>
         <oasis:entry colname="col1"/>
         <oasis:entry colname="col2">AOD at 10:00, 15 March</oasis:entry>
         <oasis:entry colname="col3">1.36</oasis:entry>
         <oasis:entry colname="col4">1.30</oasis:entry>
         <oasis:entry colname="col5">1.31</oasis:entry>
         <oasis:entry colname="col6">1.66</oasis:entry>
       </oasis:row>
       <oasis:row rowsep="1">
         <oasis:entry colname="col1"/>
         <oasis:entry colname="col2">AOD at 12:00, 15 March</oasis:entry>
         <oasis:entry colname="col3">1.53</oasis:entry>
         <oasis:entry colname="col4">1.34</oasis:entry>
         <oasis:entry colname="col5">1.26</oasis:entry>
         <oasis:entry colname="col6">1.63</oasis:entry>
       </oasis:row>
       <oasis:row>
         <oasis:entry colname="col1">SD2</oasis:entry>
         <oasis:entry colname="col2">PM<inline-formula><mml:math id="M230" display="inline"><mml:msub><mml:mi/><mml:mn mathvariant="normal">10</mml:mn></mml:msub></mml:math></inline-formula> at 11:00, 28 March</oasis:entry>
         <oasis:entry colname="col3">471 <inline-formula><mml:math id="M231" display="inline"><mml:mrow class="unit"><mml:mi mathvariant="normal">µ</mml:mi></mml:mrow></mml:math></inline-formula>g m<inline-formula><mml:math id="M232" display="inline"><mml:msup><mml:mi/><mml:mrow><mml:mo>-</mml:mo><mml:mn mathvariant="normal">3</mml:mn></mml:mrow></mml:msup></mml:math></inline-formula></oasis:entry>
         <oasis:entry colname="col4">359 <inline-formula><mml:math id="M233" display="inline"><mml:mrow class="unit"><mml:mi mathvariant="normal">µ</mml:mi></mml:mrow></mml:math></inline-formula>g m<inline-formula><mml:math id="M234" display="inline"><mml:msup><mml:mi/><mml:mrow><mml:mo>-</mml:mo><mml:mn mathvariant="normal">3</mml:mn></mml:mrow></mml:msup></mml:math></inline-formula></oasis:entry>
         <oasis:entry colname="col5">366 <inline-formula><mml:math id="M235" display="inline"><mml:mrow class="unit"><mml:mi mathvariant="normal">µ</mml:mi></mml:mrow></mml:math></inline-formula>g m<inline-formula><mml:math id="M236" display="inline"><mml:msup><mml:mi/><mml:mrow><mml:mo>-</mml:mo><mml:mn mathvariant="normal">3</mml:mn></mml:mrow></mml:msup></mml:math></inline-formula></oasis:entry>
         <oasis:entry colname="col6">351 <inline-formula><mml:math id="M237" display="inline"><mml:mrow class="unit"><mml:mi mathvariant="normal">µ</mml:mi></mml:mrow></mml:math></inline-formula>g m<inline-formula><mml:math id="M238" display="inline"><mml:msup><mml:mi/><mml:mrow><mml:mo>-</mml:mo><mml:mn mathvariant="normal">3</mml:mn></mml:mrow></mml:msup></mml:math></inline-formula></oasis:entry>
       </oasis:row>
       <oasis:row rowsep="1">
         <oasis:entry colname="col1"/>
         <oasis:entry colname="col2">AOD at 11:00, 28 March</oasis:entry>
         <oasis:entry colname="col3">0.99</oasis:entry>
         <oasis:entry colname="col4">0.81</oasis:entry>
         <oasis:entry colname="col5">0.79</oasis:entry>
         <oasis:entry colname="col6">0.89</oasis:entry>
       </oasis:row>
       <oasis:row>
         <oasis:entry colname="col1">SD3</oasis:entry>
         <oasis:entry colname="col2">PM<inline-formula><mml:math id="M239" display="inline"><mml:msub><mml:mi/><mml:mn mathvariant="normal">10</mml:mn></mml:msub></mml:math></inline-formula> at 11:00, 15 April</oasis:entry>
         <oasis:entry colname="col3">891 <inline-formula><mml:math id="M240" display="inline"><mml:mrow class="unit"><mml:mi mathvariant="normal">µ</mml:mi></mml:mrow></mml:math></inline-formula>g m<inline-formula><mml:math id="M241" display="inline"><mml:msup><mml:mi/><mml:mrow><mml:mo>-</mml:mo><mml:mn mathvariant="normal">3</mml:mn></mml:mrow></mml:msup></mml:math></inline-formula></oasis:entry>
         <oasis:entry colname="col4">143 <inline-formula><mml:math id="M242" display="inline"><mml:mrow class="unit"><mml:mi mathvariant="normal">µ</mml:mi></mml:mrow></mml:math></inline-formula>g m<inline-formula><mml:math id="M243" display="inline"><mml:msup><mml:mi/><mml:mrow><mml:mo>-</mml:mo><mml:mn mathvariant="normal">3</mml:mn></mml:mrow></mml:msup></mml:math></inline-formula></oasis:entry>
         <oasis:entry colname="col5">210 <inline-formula><mml:math id="M244" display="inline"><mml:mrow class="unit"><mml:mi mathvariant="normal">µ</mml:mi></mml:mrow></mml:math></inline-formula>g m<inline-formula><mml:math id="M245" display="inline"><mml:msup><mml:mi/><mml:mrow><mml:mo>-</mml:mo><mml:mn mathvariant="normal">3</mml:mn></mml:mrow></mml:msup></mml:math></inline-formula></oasis:entry>
         <oasis:entry colname="col6">133 <inline-formula><mml:math id="M246" display="inline"><mml:mrow class="unit"><mml:mi mathvariant="normal">µ</mml:mi></mml:mrow></mml:math></inline-formula>g m<inline-formula><mml:math id="M247" display="inline"><mml:msup><mml:mi/><mml:mrow><mml:mo>-</mml:mo><mml:mn mathvariant="normal">3</mml:mn></mml:mrow></mml:msup></mml:math></inline-formula></oasis:entry>
       </oasis:row>
       <oasis:row>
         <oasis:entry colname="col1"/>
         <oasis:entry colname="col2">AOD at 11:00, 15 April</oasis:entry>
         <oasis:entry colname="col3">1.79</oasis:entry>
         <oasis:entry colname="col4">0.72</oasis:entry>
         <oasis:entry colname="col5">0.69</oasis:entry>
         <oasis:entry colname="col6">0.77</oasis:entry>
       </oasis:row>
     </oasis:tbody>
   </oasis:tgroup></oasis:table></table-wrap>

</sec>
<sec id="Ch1.S4.SS3">
  <label>4.3</label><title>Source apportionment</title>
      <p id="d1e4571">China's government has launched several large-scale ecological engineering projects to combat the environmental problems in northern China during recent decades. One of the largest is the Three-North Shelter Forest Program, which aims at increasing the vegetation cover up to 15 % by 2050 <xref ref-type="bibr" rid="bib1.bibx40" id="paren.75"/>. Several studies <xref ref-type="bibr" rid="bib1.bibx51 bib1.bibx53" id="paren.76"/> reported that the vegetation recovery weakened dust storms substantially. In contrast, Mongolia has experienced ever-increasing land degradation and desertification <xref ref-type="bibr" rid="bib1.bibx33" id="paren.77"/>, which aggravates the spring dust storms <xref ref-type="bibr" rid="bib1.bibx15" id="paren.78"/>. To evaluate the roles of Mongolian and Chinese Gobi deserts in the 2021 super sandstorms quantitatively, source apportionment tests based on the estimated emission field are carried out. These source apportionment tests focus on the two dust-affected mega-city clusters in northern China, namely the North China Plain (NCP) and the Fenwei Plain (FWP), and aim to calculate whether the dust originates from transnational transport from Mongolia or from domestic sources in China.</p>
      <p id="d1e4586">Two LOTOS-EUROS/dust simulations were conducted with the posterior emission field obtained in the multi-observation assimilation in Sect. <xref ref-type="sec" rid="Ch1.S4.SS1.SSS1"/>, but with either only the emissions in Mongolia or only the emissions in China enabled. As can be seen in the time series of the hourly PM<inline-formula><mml:math id="M248" display="inline"><mml:msub><mml:mi/><mml:mn mathvariant="normal">10</mml:mn></mml:msub></mml:math></inline-formula> concentration measurements in NCP and FWP in Fig. <xref ref-type="fig" rid="Ch1.F1"/>c–d, after the peak of the dust had passed by the NCP and FWP regions, they still suffered from some less severe dust affection. Therefore, longer simulation windows are used in these source apportionment tests to simulate the full life cycle of these dust events as can be seen in Table <xref ref-type="table" rid="Ch1.T1"/>. Different from cases caused by other pollutants, the spring dust events in East Asia are usually short-term events, with concentrations of dust that quickly increase to huge levels, but also drop down quickly after the storm has passed by. Metrics such as daily average dust concentration do not reflect the dust intensity directly. Therefore, a dust-deposition index is introduced that measures the sum of dry and wet deposition to quantify the impact of the dust in the studied regions.</p>
      <p id="d1e4604">Figure <xref ref-type="fig" rid="Ch1.F8"/> shows the spatial pattern of the deposition for dust originating from China (left panels) and originating from Mongolia (right panels) in SD1, SD2, and SD3. The most interesting finding is that the sources in Mongolia play a much more important role in dust pollution in northern China than the sources in China itself. As shown in Fig. <xref ref-type="fig" rid="Ch1.F8"/>a.i, a huge quantity of particles were released in the Alxa desert during SD1, but these were mainly transported westward and only a small fraction of them moved to the densely populated areas.
During SD2, the deposition of dust released from China is non-negligible, but the total deposition is still dominated by dust released from Mongolia. Within SD3, the dust particles emitted from the Chinese Gobi were spread all over northern China and hence played a more significant role.</p>

      <?xmltex \floatpos{t}?><fig id="Ch1.F8" specific-use="star"><?xmltex \currentcnt{8}?><?xmltex \def\figurename{Figure}?><label>Figure 8</label><caption><p id="d1e4614">Spatial patterns of deposition of dust originating from China <bold>(i)</bold> and from Mongolia <bold>(ii)</bold>. Panel <bold>(d)</bold> shows the total mass of dust deposition on NCP and FWP either from China or from Mongolia during SD1 <bold>(a)</bold>, SD2 <bold>(a)</bold>, and SD3 <bold>(c)</bold>.</p></caption>
          <?xmltex \igopts{width=369.885827pt}?><graphic xlink:href="https://acp.copernicus.org/articles/22/6393/2022/acp-22-6393-2022-f08.png"/>

        </fig>

      <p id="d1e4642">The total deposition in the NCP and FWP regions was calculated and is shown in Fig. <xref ref-type="fig" rid="Ch1.F8"/>d. For the NCP region, <inline-formula><mml:math id="M249" display="inline"><mml:mrow><mml:mn mathvariant="normal">81</mml:mn><mml:mo>×</mml:mo><mml:msup><mml:mn mathvariant="normal">10</mml:mn><mml:mn mathvariant="normal">3</mml:mn></mml:msup></mml:mrow></mml:math></inline-formula>, 118 <inline-formula><mml:math id="M250" display="inline"><mml:mrow><mml:mo>×</mml:mo><mml:msup><mml:mn mathvariant="normal">10</mml:mn><mml:mn mathvariant="normal">3</mml:mn></mml:msup></mml:mrow></mml:math></inline-formula>, and 70 <inline-formula><mml:math id="M251" display="inline"><mml:mrow><mml:mo>×</mml:mo><mml:msup><mml:mn mathvariant="normal">10</mml:mn><mml:mn mathvariant="normal">3</mml:mn></mml:msup></mml:mrow></mml:math></inline-formula> t of Mongolian dust was deposited during SD1, SD2, and SD3 events, while the total deposition from Chinese desert was about 8.3 <inline-formula><mml:math id="M252" display="inline"><mml:mrow><mml:mo>×</mml:mo><mml:msup><mml:mn mathvariant="normal">10</mml:mn><mml:mn mathvariant="normal">3</mml:mn></mml:msup></mml:mrow></mml:math></inline-formula>, 20 <inline-formula><mml:math id="M253" display="inline"><mml:mrow><mml:mo>×</mml:mo><mml:msup><mml:mn mathvariant="normal">10</mml:mn><mml:mn mathvariant="normal">3</mml:mn></mml:msup></mml:mrow></mml:math></inline-formula>, and 93 <inline-formula><mml:math id="M254" display="inline"><mml:mrow><mml:mo>×</mml:mo><mml:msup><mml:mn mathvariant="normal">10</mml:mn><mml:mn mathvariant="normal">3</mml:mn></mml:msup></mml:mrow></mml:math></inline-formula> t, respectively. For the other important cluster FWP, 4.3 <inline-formula><mml:math id="M255" display="inline"><mml:mrow><mml:mo>×</mml:mo><mml:msup><mml:mn mathvariant="normal">10</mml:mn><mml:mn mathvariant="normal">3</mml:mn></mml:msup></mml:mrow></mml:math></inline-formula>, 22 <inline-formula><mml:math id="M256" display="inline"><mml:mrow><mml:mo>×</mml:mo><mml:msup><mml:mn mathvariant="normal">10</mml:mn><mml:mn mathvariant="normal">3</mml:mn></mml:msup></mml:mrow></mml:math></inline-formula>, and 24 <inline-formula><mml:math id="M257" display="inline"><mml:mrow><mml:mo>×</mml:mo><mml:msup><mml:mn mathvariant="normal">10</mml:mn><mml:mn mathvariant="normal">3</mml:mn></mml:msup></mml:mrow></mml:math></inline-formula> t were attributed to domestic sources, and 20 <inline-formula><mml:math id="M258" display="inline"><mml:mrow><mml:mo>×</mml:mo><mml:msup><mml:mn mathvariant="normal">10</mml:mn><mml:mn mathvariant="normal">3</mml:mn></mml:msup></mml:mrow></mml:math></inline-formula>, 57 <inline-formula><mml:math id="M259" display="inline"><mml:mrow><mml:mo>×</mml:mo><mml:msup><mml:mn mathvariant="normal">10</mml:mn><mml:mn mathvariant="normal">3</mml:mn></mml:msup></mml:mrow></mml:math></inline-formula>, and 7.5 <inline-formula><mml:math id="M260" display="inline"><mml:mrow><mml:mo>×</mml:mo><mml:msup><mml:mn mathvariant="normal">10</mml:mn><mml:mn mathvariant="normal">3</mml:mn></mml:msup></mml:mrow></mml:math></inline-formula> t of dust were attributed to transnational transport from Mongolia.
In general, the Mongolian Gobi poses a more severe threat to the FWP and NCP regions than the Chinese Gobi. About 63 % of the dust deposition in FWP is attributed to transnational transport. Over NCP, this value further increases up to 69 %.</p>
</sec>
</sec>
<sec id="Ch1.S5" sec-type="conclusions">
  <label>5</label><title>Summary and future work</title>
      <p id="d1e4817">In spring 2021, three super dust storms occurred in East Asia after being absent for one and a half decades, which brought enormous health damages and property losses.
To exploit the reappearing super sandstorms, inverse modeling was conducted through optimizing the dust simulation with observations of MODIS AODs and ground-based PM<inline-formula><mml:math id="M261" display="inline"><mml:msub><mml:mi/><mml:mn mathvariant="normal">10</mml:mn></mml:msub></mml:math></inline-formula> concentration from the Chinese MEP air-quality monitoring network. Data quality controls were designed and applied in order to use the AOD and PM<inline-formula><mml:math id="M262" display="inline"><mml:msub><mml:mi/><mml:mn mathvariant="normal">10</mml:mn></mml:msub></mml:math></inline-formula> measurements for representing the dust load. Based on the most likely emission field calculated by the inversion, source apportionment was further performed to derive the contribution of transnational transport from Mongolia and domestic dust emission to the dust pollutant level in northern China.</p>
      <p id="d1e4838">Emission inversion was successfully performed by assimilating the AOD and PM<inline-formula><mml:math id="M263" display="inline"><mml:msub><mml:mi/><mml:mn mathvariant="normal">10</mml:mn></mml:msub></mml:math></inline-formula> concentration measurements. The multi-observation assimilation showed that windblown dust emission occurred actively both in Chinese and Mongolian Gobi deserts during the events studied. Overall, about <inline-formula><mml:math id="M264" display="inline"><mml:mrow><mml:mn mathvariant="normal">37.5</mml:mn><mml:mo>×</mml:mo><mml:msup><mml:mn mathvariant="normal">10</mml:mn><mml:mn mathvariant="normal">6</mml:mn></mml:msup></mml:mrow></mml:math></inline-formula> t of dust was released in Mongolia, and the total emission from the Chinese Gobi was also as high as <inline-formula><mml:math id="M265" display="inline"><mml:mrow><mml:mn mathvariant="normal">19.9</mml:mn><mml:mo>×</mml:mo><mml:msup><mml:mn mathvariant="normal">10</mml:mn><mml:mn mathvariant="normal">6</mml:mn></mml:msup></mml:mrow></mml:math></inline-formula> t. The simulated AOD and surface dust concentration driven by the posterior emission fields have been validated to be in better agreement with the observations. To obtain a further accurate dust field analysis, however, vertical structure-adjusting techniques such as 3D grid distortion are then in demand for fully resolving the ground-based PM<inline-formula><mml:math id="M266" display="inline"><mml:msub><mml:mi/><mml:mn mathvariant="normal">10</mml:mn></mml:msub></mml:math></inline-formula> and column-integrated AOD together, and will be explored in our future work.</p>
      <p id="d1e4889">Emission inversions that only assimilated the AOD or PM<inline-formula><mml:math id="M267" display="inline"><mml:msub><mml:mi/><mml:mn mathvariant="normal">10</mml:mn></mml:msub></mml:math></inline-formula> concentration were also carried out. The comparison against the multi-observation assimilation showed that AOD-only or PM<inline-formula><mml:math id="M268" display="inline"><mml:msub><mml:mi/><mml:mn mathvariant="normal">10</mml:mn></mml:msub></mml:math></inline-formula>-only assimilation would result in the posterior that is closer to the assimilated data, but the improvement of the simulation in the independent observation space is not ensured. Especially when vertical structure is not well reproduced (such as in the simulation over SD1), the posterior simulation would be divergent.
Under this circumstance, multi-observation assimilation as used in this study is the more reliable choice.</p>
      <p id="d1e4910">A source apportionment study was then performed based on the multi-observation assimilation-estimated emission by estimating the origin of the dust that was deposited in regions of northern China. It indicated that the Mongolian Gobi posed more severe threats to the Fenwei Plain (FWP) and the North China Plain (NCP) than the Chinese Gobi within the three 2021 spring dust storms. For FWP, about 63 % of the dust deposition originated from transnational transport from Mongolia. In NCP, the Mongolian dust contribution was also as high as 69 %. To further explore the roles of specific deserts (such as the Alxa and Tengger deserts in Chinese Gobi) and long-distance transport patterns on the dust affection in northern China, more complex source apportionment tests are planned in our future research.</p>
</sec>

      
      </body>
    <back><notes notes-type="codedataavailability"><title>Code and data availability</title>

      <p id="d1e4917">The source code and user guide of the LOTOS-EUROS CTM could be obtained from <uri>https://lotos-euros.tno.nl</uri> <xref ref-type="bibr" rid="bib1.bibx55" id="paren.79"/>. The real-time PM<inline-formula><mml:math id="M269" display="inline"><mml:msub><mml:mi/><mml:mn mathvariant="normal">10</mml:mn></mml:msub></mml:math></inline-formula> data are from the network established by the China Ministry of Environmental Protection and available to the public via <uri>http://106.37.208.233:20035/</uri> <xref ref-type="bibr" rid="bib1.bibx4" id="paren.80"/> (Internet Explorer 10 or 11), the PM<inline-formula><mml:math id="M270" display="inline"><mml:msub><mml:mi/><mml:mn mathvariant="normal">10</mml:mn></mml:msub></mml:math></inline-formula> data used in this paper is also archived on Zendo (<uri>https://doi.org/10.5281/zenodo.6459866</uri>, <xref ref-type="bibr" rid="bib1.bibx19" id="altparen.81"/>). The MODIS Deep Blue C6 data suites are available at <uri>https://ladsweb.modaps.eosdis.nasa.gov/</uri> <xref ref-type="bibr" rid="bib1.bibx39" id="paren.82"/>. The datasets including measurements and model simulations can also be accessed by contacting the corresponding author.</p>
  </notes><app-group>
        <supplementary-material position="anchor"><p id="d1e4963">The supplement related to this article is available online at: <inline-supplementary-material xlink:href="https://doi.org/10.5194/acp-22-6393-2022-supplement" xlink:title="pdf">https://doi.org/10.5194/acp-22-6393-2022-supplement</inline-supplementary-material>.</p></supplementary-material>
        </app-group><notes notes-type="authorcontribution"><title>Author contributions</title>

      <p id="d1e4972">JJ conceived the study and designed the dust storm data assimilation.
JJ, MP, and AS performed the control and assimilation tests and carried out the data analysis.
AS, HL, WH, BL, and HXL provided useful comments on the paper.
JJ prepared the manuscript with contributions from AS and all others co-authors.</p>
  </notes><notes notes-type="competinginterests"><title>Competing interests</title>

      <p id="d1e4978">The contact author has declared that neither they nor their co-authors have any competing interests.</p>
  </notes><notes notes-type="disclaimer"><title>Disclaimer</title>

      <p id="d1e4984">Publisher’s note: Copernicus Publications remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.</p>
  </notes><notes notes-type="sistatement"><title>Special issue statement</title>

      <p id="d1e4990">This article is part of the special issue “Dust aerosol measurements, modeling and multidisciplinary effects (AMT/ACP inter-journal SI)”. It is not associated with a conference.</p>
  </notes><ack><title>Acknowledgements</title><p id="d1e4996">We would like to thank the referee Thomas Sekiyama for the in-depth comments.</p></ack><notes notes-type="financialsupport"><title>Financial support</title>

      <p id="d1e5001">This research has been supported by the National Natural Science Foundation of China (grant no. 42105109, 42021004) and the Natural Science Foundation of Jiangsu Province (grant no. BK20210664).</p>
  </notes><notes notes-type="reviewstatement"><title>Review statement</title>

      <p id="d1e5007">This paper was edited by Toshihiko Takemura and reviewed by Thomas Sekiyama and one anonymous referee.</p>
  </notes><ref-list>
    <title>References</title>

      <ref id="bib1.bibx1"><?xmltex \def\ref@label{{Bessho et~al.(2016)}}?><label>Bessho et al.(2016)</label><?label Bessho2016Introduction?><mixed-citation>Bessho, K., Date, K., Hayashi, M., Ikeda, A., Imai, T., Inoue, H., Kumagai, Y.,
Miyakawa, T., Murata, H., Ohno, T., Okuyama, A., Oyama, R., Sasaki, Y.,
Shimazu, Y., Shimoji, K., Sumida, Y., Suzuki, M., Taniguchi, H., Tsuchiyama,
H., Uesawa, D., Yokota, H., and Yoshida, R.: An Introduction to
Himawari-8/9 – Japan's New-Generation Geostationary
Meteorological Satellites, J. Meteorol. Soc. Jpn.
Ser. II, 94, 151–183, <ext-link xlink:href="https://doi.org/10.2151/jmsj.2016-009" ext-link-type="DOI">10.2151/jmsj.2016-009</ext-link>, 2016.</mixed-citation></ref>
      <ref id="bib1.bibx2"><?xmltex \def\ref@label{{Brasseur et~al.(2019)}}?><label>Brasseur et al.(2019)</label><?label Brasseur2019Ensemble?><mixed-citation>Brasseur, G. P., Xie, Y., Petersen, A. K., Bouarar, I., Flemming, J., Gauss, M., Jiang, F., Kouznetsov, R., Kranenburg, R., Mijling, B., Peuch, V.-H., Pommier, M., Segers, A., Sofiev, M., Timmermans, R., van der A, R., Walters, S., Xu, J., and Zhou, G.: Ensemble forecasts of air quality in eastern China – Part 1: Model description and implementation of the MarcoPolo–Panda prediction system, version 1, Geosci. Model Dev., 12, 33–67, <ext-link xlink:href="https://doi.org/10.5194/gmd-12-33-2019" ext-link-type="DOI">10.5194/gmd-12-33-2019</ext-link>, 2019.</mixed-citation></ref>
      <ref id="bib1.bibx3"><?xmltex \def\ref@label{{Chen and Walsh(2021)}}?><label>Chen and Walsh(2021)</label><?label Mongolia2021dust?><mixed-citation>Chen, L. and Walsh, M.: Vast sandstorms expose Mongolia's long-ignored
ecological crisis, <uri>https://asia.nikkei.com/Spotlight/Caixin/Vast-sandstorms-expose-Mongolia-s-long-ignored-ecological-crisis</uri> (last access: 14 May 2022),
2021.</mixed-citation></ref>
      <ref id="bib1.bibx4"><?xmltex \def\ref@label{{China Ministry of Environmental Protection(2022)}}?><label>China Ministry of Environmental Protection(2022)</label><?label CMEP2022?><mixed-citation>China Ministry of Environmental Protection: Ground-based air quality monitoring measurements, China Ministry of Environmental Protection [data set], <uri>http://106.37.208.233:20035/</uri>, last access: 14 May 2022.</mixed-citation></ref>
      <ref id="bib1.bibx5"><?xmltex \def\ref@label{{Di~Tomaso et~al.(2017)}}?><label>Di Tomaso et al.(2017)</label><?label DiTomaso2017Assimilation?><mixed-citation>Di Tomaso, E., Schutgens, N. A. J., Jorba, O., and Pérez García-Pando, C.: Assimilation of MODIS Dark Target and Deep Blue observations in the dust aerosol component of NMMB-MONARCH version 1.0, Geosci. Model Dev., 10, 1107–1129, <ext-link xlink:href="https://doi.org/10.5194/gmd-10-1107-2017" ext-link-type="DOI">10.5194/gmd-10-1107-2017</ext-link>, 2017.</mixed-citation></ref>
      <ref id="bib1.bibx6"><?xmltex \def\ref@label{{Di~Tomaso et~al.(2021)}}?><label>Di Tomaso et al.(2021)</label><?label Di2021MONARCH?><mixed-citation>Di Tomaso, E., Escribano, J., Basart, S., Ginoux, P., Macchia, F., Barnaba, F., Benincasa, F., Bretonnière, P.-A., Buñuel, A., Castrillo, M., Cuevas, E., Formenti, P., Gonçalves, M., Jorba, O., Klose, M., Mona, L., Montané, G., Mytilinaios, M., Obiso, V., Olid, M., Schutgens, N., Votsis, A., Werner, E., and Pérez García-Pando, C.: The MONARCH high-resolution reanalysis of desert dust aerosol over Northern Africa, the Middle East and Europe (2007–2016), Earth Syst. Sci. Data Discuss. [preprint], <ext-link xlink:href="https://doi.org/10.5194/essd-2021-358" ext-link-type="DOI">10.5194/essd-2021-358</ext-link>, in review, 2021.</mixed-citation></ref>
      <ref id="bib1.bibx7"><?xmltex \def\ref@label{{Dubovik et~al.(2000)}}?><label>Dubovik et al.(2000)</label><?label dubovik2000accuracy?><mixed-citation>
Dubovik, O., Smirnov, A., Holben, B., King, M., Kaufman, Y., Eck, T., and
Slutsker, I.: Accuracy assessments of aerosol optical properties retrieved
from Aerosol Robotic Network (AERONET) Sun and sky radiance measurements,
J. Geophys. Res.-Atmos., 105, 9791–9806, 2000.</mixed-citation></ref>
      <ref id="bib1.bibx8"><?xmltex \def\ref@label{{Dubovik et~al.(2002)}}?><label>Dubovik et al.(2002)</label><?label dubovik2002variability?><mixed-citation>
Dubovik, O., Holben, B., Eck, T. F., Smirnov, A., Kaufman, Y. J., King, M. D.,
Tanré, D., and Slutsker, I.: Variability of absorption and optical
properties of key aerosol types observed in worldwide locations, J. Atmos. Sci., 59, 590–608, 2002.</mixed-citation></ref>
      <ref id="bib1.bibx9"><?xmltex \def\ref@label{{Eck et~al.(1999)}}?><label>Eck et al.(1999)</label><?label eck1999wavelength?><mixed-citation>
Eck, T. F., Holben, B., Reid, J., Dubovik, O., Smirnov, A., O'neill, N.,
Slutsker, I., and Kinne, S.: Wavelength dependence of the optical depth of
biomass burning, urban, and desert dust aerosols, J. Geophys. Res.-Atmos., 104, 31333–31349, 1999.</mixed-citation></ref>
      <ref id="bib1.bibx10"><?xmltex \def\ref@label{{Fairlie et~al.(2007)}}?><label>Fairlie et al.(2007)</label><?label DUNCANFAIRLIE2007GEOS?><mixed-citation>Fairlie, T. D., Jacob, D. J., and Park, R. J.: The impact of transpacific
transport of mineral dust in the United States, Atmos. Environ., 41,
1251–1266, <ext-link xlink:href="https://doi.org/10.1016/j.atmosenv.2006.09.048" ext-link-type="DOI">10.1016/j.atmosenv.2006.09.048</ext-link>, 2007.</mixed-citation></ref>
      <ref id="bib1.bibx11"><?xmltex \def\ref@label{{Ginoux et~al.(2001)}}?><label>Ginoux et al.(2001)</label><?label Ginoux2001Sources?><mixed-citation>Ginoux, P., Chin, M., Tegen, I., Prospero, J. M., Holben, B., Dubovik, O., and
Lin, S.-J.: Sources and distributions of dust aerosols simulated with the
GOCART model, J. Geophys. Res., 106, 20255–20273,
<ext-link xlink:href="https://doi.org/10.1029/2000jd000053" ext-link-type="DOI">10.1029/2000jd000053</ext-link>, 2001.</mixed-citation></ref>
      <ref id="bib1.bibx12"><?xmltex \def\ref@label{{Ginoux et~al.(2012)}}?><label>Ginoux et al.(2012)</label><?label Ginoux2012Global?><mixed-citation>Ginoux, P., Prospero, J. M., Gill, T. E., Hsu, N. C., and Zhao, M.:
Global-scale attribution of anthropogenic and natural dust sources and their
emission rates based on MODIS Deep Blue aerosol products, Rev.
Geophys., 50,  RG3005, <ext-link xlink:href="https://doi.org/10.1029/2012RG000388" ext-link-type="DOI">10.1029/2012RG000388</ext-link>, 2012.</mixed-citation></ref>
      <ref id="bib1.bibx13"><?xmltex \def\ref@label{{Gkikas et~al.(2021)}}?><label>Gkikas et al.(2021)</label><?label Gkikas2021MIDAS?><mixed-citation>Gkikas, A., Proestakis, E., Amiridis, V., Kazadzis, S., Di Tomaso, E., Tsekeri, A., Marinou, E., Hatzianastassiou, N., and Pérez García-Pando, C.: ModIs Dust AeroSol (MIDAS): a global fine-resolution dust optical depth data set, Atmos. Meas. Tech., 14, 309–334, <ext-link xlink:href="https://doi.org/10.5194/amt-14-309-2021" ext-link-type="DOI">10.5194/amt-14-309-2021</ext-link>, 2021.</mixed-citation></ref>
      <ref id="bib1.bibx14"><?xmltex \def\ref@label{{Gong and Zhang(2008)}}?><label>Gong and Zhang(2008)</label><?label Gong2008CUACEDust?><mixed-citation>Gong, S. L. and Zhang, X. Y.: CUACE/Dust – an integrated system of observation and modeling systems for operational dust forecasting in Asia, Atmos. Chem. Phys., 8, 2333–2340, <ext-link xlink:href="https://doi.org/10.5194/acp-8-2333-2008" ext-link-type="DOI">10.5194/acp-8-2333-2008</ext-link>, 2008.</mixed-citation></ref>
      <ref id="bib1.bibx15"><?xmltex \def\ref@label{{Han et~al.(2021)}}?><label>Han et al.(2021)</label><?label han2021sandstorms?><mixed-citation>
Han, J., Dai, H., and Gu, Z.: Sandstorms and desertification in Mongolia, an
example of future climate events: A review, Environ. Chem. Lett., 19,
4063–4073, 2021.</mixed-citation></ref>
      <ref id="bib1.bibx16"><?xmltex \def\ref@label{{Hsu et~al.(2013)}}?><label>Hsu et al.(2013)</label><?label Hsu2013EhanceDB?><mixed-citation>Hsu, N. C., Jeong, M.-J., Bettenhausen, C., Sayer, A. M., Hansell, R., Seftor,
C. S., Huang, J., and Tsay, S.-C.: Enhanced Deep Blue aerosol retrieval
algorithm: The second generation, J. Geophys. Res.-Atmos., 118, 9296–9315, <ext-link xlink:href="https://doi.org/10.1002/jgrd.50712" ext-link-type="DOI">10.1002/jgrd.50712</ext-link>, 2013.</mixed-citation></ref>
      <ref id="bib1.bibx17"><?xmltex \def\ref@label{{Huneeus et~al.(2011)}}?><label>Huneeus et al.(2011)</label><?label Huneeus2011Global?><mixed-citation>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, <ext-link xlink:href="https://doi.org/10.5194/acp-11-7781-2011" ext-link-type="DOI">10.5194/acp-11-7781-2011</ext-link>, 2011.</mixed-citation></ref>
      <ref id="bib1.bibx18"><?xmltex \def\ref@label{{Jin(2021)}}?><label>Jin(2021)</label><?label China2021dust?><mixed-citation>Jin, G.: The most severe Sandstorm in a decade,
<uri>https://www.chinanews.com/m/sh/2021/03-15/9433005.shtml</uri> (last access: 14 May 2022),
2021.</mixed-citation></ref>
      <ref id="bib1.bibx19"><?xmltex \def\ref@label{{Jin(2022)}}?><label>Jin(2022)</label><?label jin22?><mixed-citation>Jin, J.: ground-based air quality measurements during the 2021 spring super dust storms, Zenodo [data set], <ext-link xlink:href="https://doi.org/10.5281/zenodo.6459866" ext-link-type="DOI">10.5281/zenodo.6459866</ext-link>, 2022.</mixed-citation></ref>
      <ref id="bib1.bibx20"><?xmltex \def\ref@label{{Jin et~al.(2018)}}?><label>Jin et al.(2018)</label><?label Jin2018Spatially?><mixed-citation>Jin, J., Lin, H. X., Heemink, A., and Segers, A.: Spatially varying parameter
estimation for dust emissions using reduced-tangent-linearization 4DVar,
Atmos. Environ., 187, 358–373,
<ext-link xlink:href="https://doi.org/10.1016/j.atmosenv.2018.05.060" ext-link-type="DOI">10.1016/j.atmosenv.2018.05.060</ext-link>, 2018.</mixed-citation></ref>
      <ref id="bib1.bibx21"><?xmltex \def\ref@label{{Jin et~al.(2019a)}}?><label>Jin et al.(2019a)</label><?label Jin2019Machine?><mixed-citation>Jin, J., Lin, H. X., Segers, A., Xie, Y., and Heemink, A.: Machine learning for observation bias correction with application to dust storm data assimilation, Atmos. Chem. Phys., 19, 10009–10026, <ext-link xlink:href="https://doi.org/10.5194/acp-19-10009-2019" ext-link-type="DOI">10.5194/acp-19-10009-2019</ext-link>, 2019a.</mixed-citation></ref>
      <ref id="bib1.bibx22"><?xmltex \def\ref@label{{Jin et~al.(2019b)}}?><label>Jin et al.(2019b)</label><?label Jin2019Dust?><mixed-citation>Jin, J., Segers, A., Heemink, A., Yoshida, M., Han, W., and Lin, H.-X.: Dust
Emission Inversion Using Himawari‐8 AODs Over East Asia: An Extreme Dust
Event in May 2017, J. Adv. Model. Earth Sy., 11,
446–467, <ext-link xlink:href="https://doi.org/10.1029/2018MS001491" ext-link-type="DOI">10.1029/2018MS001491</ext-link>, 2019b.</mixed-citation></ref>
      <ref id="bib1.bibx23"><?xmltex \def\ref@label{{Jin et~al.(2020)}}?><label>Jin et al.(2020)</label><?label Jin2020Source?><mixed-citation>Jin, J., Segers, A., Liao, H., Heemink, A., Kranenburg, R., and Lin, H. X.: Source backtracking for dust storm emission inversion using an adjoint method: case study of Northeast China, Atmos. Chem. Phys., 20, 15207–15225, <ext-link xlink:href="https://doi.org/10.5194/acp-20-15207-2020" ext-link-type="DOI">10.5194/acp-20-15207-2020</ext-link>, 2020.</mixed-citation></ref>
      <ref id="bib1.bibx24"><?xmltex \def\ref@label{{Jin et~al.(2021)}}?><label>Jin et al.(2021)</label><?label Jin2021Position?><mixed-citation>Jin, J., Segers, A., Lin, H. X., Henzing, B., Wang, X., Heemink, A., and Liao, H.: Position correction in dust storm forecasting using LOTOS-EUROS v2.1: grid-distorted data assimilation v1.0, Geosci. Model Dev., 14, 5607–5622, <ext-link xlink:href="https://doi.org/10.5194/gmd-14-5607-2021" ext-link-type="DOI">10.5194/gmd-14-5607-2021</ext-link>, 2021.</mixed-citation></ref>
      <ref id="bib1.bibx25"><?xmltex \def\ref@label{{Justice et~al.(1998)}}?><label>Justice et al.(1998)</label><?label Justice1990MODIS?><mixed-citation>Justice, C., Vermote, E., Townshend, J., Defries, R., Roy, D., Hall, D.,
Salomonson, V., Privette, J., Riggs, G., Strahler, A., Lucht, W., Myneni, R.,
Knyazikhin, Y., Running, S., Nemani, R., Wan, Z., Huete, A., van Leeuwen, W.,
Wolfe, R., Giglio, L., Muller, J., Lewis, P., and Barnsley, M.: The Moderate
Resolution Imaging Spectroradiometer (MODIS): land remote sensing for global
change research, IEEE T. Geosci. Remote Sens., 36,
1228–1249, <ext-link xlink:href="https://doi.org/10.1109/36.701075" ext-link-type="DOI">10.1109/36.701075</ext-link>, 1998.</mixed-citation></ref>
      <ref id="bib1.bibx26"><?xmltex \def\ref@label{{Kalnay(2002)}}?><label>Kalnay(2002)</label><?label Kalnay2002AtmosphereDA?><mixed-citation>Kalnay, E.: Atmospheric Modeling, Data Assimilation and Predictability, edited by: Kalnay, E.,
Cambridge University Press, <ext-link xlink:href="https://doi.org/10.1017/CBO9780511802270" ext-link-type="DOI">10.1017/CBO9780511802270</ext-link>, 2002.</mixed-citation></ref>
      <ref id="bib1.bibx27"><?xmltex \def\ref@label{{Li et~al.(2018)}}?><label>Li et al.(2018)</label><?label li2018comprehensive?><mixed-citation>
Li, Z., Xu, H., Li, K., Li, D., Xie, Y., Li, L., Zhang, Y., Gu, X., Zhao, W., Tian, Q., Deng, R., Su, X., Huang, B., Qiao, Y., Cui, W., Hu, Y., Gong, C., Wang, Y., Wang, X., Wang, J., Du, W., Pan, Z., Li, Z., and Bu, D.: Comprehensive study of optical, physical, chemical, and
radiative properties of total columnar atmospheric aerosols over China: An
overview of Sun–Sky Radiometer Observation Network (SONET) measurements,
B. Am. Meteorol. Soc., 99, 739–755, 2018.</mixed-citation></ref>
      <ref id="bib1.bibx28"><?xmltex \def\ref@label{{Lin et~al.(2008)}}?><label>Lin et al.(2008)</label><?label Lin2008Ensemble?><mixed-citation>Lin, C., Wang, Z., and Zhu, J.: An Ensemble Kalman Filter for severe dust storm data assimilation over China, Atmos. Chem. Phys., 8, 2975–2983, <ext-link xlink:href="https://doi.org/10.5194/acp-8-2975-2008" ext-link-type="DOI">10.5194/acp-8-2975-2008</ext-link>, 2008.</mixed-citation></ref>
      <ref id="bib1.bibx29"><?xmltex \def\ref@label{{Liu et~al.(2019)}}?><label>Liu et al.(2019)</label><?label LIU201990Quantifying?><mixed-citation>Liu, M., Zhou, G., Saari, R. K., Li, S., Liu, X., and Li, J.: Quantifying PM<inline-formula><mml:math id="M271" display="inline"><mml:msub><mml:mi/><mml:mn mathvariant="normal">2.5</mml:mn></mml:msub></mml:math></inline-formula>
mass concentration and particle radius using satellite data and an
optical-mass conversion algorithm, ISPRS J. Photogramm., 158, 90–98,
<ext-link xlink:href="https://doi.org/10.1016/j.isprsjprs.2019.10.010" ext-link-type="DOI">10.1016/j.isprsjprs.2019.10.010</ext-link>, 2019.</mixed-citation></ref>
      <ref id="bib1.bibx30"><?xmltex \def\ref@label{{Lorente-Plazas and Hacker(2017)}}?><label>Lorente-Plazas and Hacker(2017)</label><?label LorentePlazas2017Observation?><mixed-citation>Lorente-Plazas, R. and Hacker, J. P.: Observation and Model Bias Estimation in
the Presence of Either or Both Sources of Error, Mon. Weather Rev.,
145, 2683–2696, <ext-link xlink:href="https://doi.org/10.1175/MWR-D-16-0273.1" ext-link-type="DOI">10.1175/MWR-D-16-0273.1</ext-link>, 2017.</mixed-citation></ref>
      <ref id="bib1.bibx31"><?xmltex \def\ref@label{{Manders et~al.(2017)}}?><label>Manders et al.(2017)</label><?label Manders2017Curriculum?><mixed-citation>Manders, A. M. M., Builtjes, P. J. H., Curier, L., Denier van der Gon, H. A. C., Hendriks, C., Jonkers, S., Kranenburg, R., Kuenen, J. J. P., Segers, A. J., Timmermans, R. M. A., Visschedijk, A. J. H., Wichink Kruit, R. J., van Pul, W. A. J., Sauter, F. J., van der Swaluw, E., Swart, D. P. J., Douros, J., Eskes, H., van Meijgaard, E., van Ulft, B., van Velthoven, P., Banzhaf, S., Mues, A. C., Stern, R., Fu, G., Lu, S., Heemink, A., van Velzen, N., and Schaap, M.: Curriculum vitae of the LOTOS–EUROS (v2.0) chemistry transport model, Geosci. Model Dev., 10, 4145–4173, <ext-link xlink:href="https://doi.org/10.5194/gmd-10-4145-2017" ext-link-type="DOI">10.5194/gmd-10-4145-2017</ext-link>, 2017.</mixed-citation></ref>
      <ref id="bib1.bibx32"><?xmltex \def\ref@label{{Marticorena and Bergametti(1995)}}?><label>Marticorena and Bergametti(1995)</label><?label Marticorena1995Modeling?><mixed-citation>Marticorena, B. and Bergametti, G.: Modeling the atmospheric dust cycle: 1.
Design of a soil-derived dust emission scheme, J. Geophys. Res., 100,
16415–16430, <ext-link xlink:href="https://doi.org/10.1029/95JD00690" ext-link-type="DOI">10.1029/95JD00690</ext-link>, 1995.</mixed-citation></ref>
      <ref id="bib1.bibx33"><?xmltex \def\ref@label{{Meng et~al.(2020)}}?><label>Meng et al.(2020)</label><?label Meng2020Mongolia?><mixed-citation>Meng, X., Gao, X., Li, S., and Lei, J.: Spatial and Temporal Characteristics of
Vegetation NDVI Changes and the Driving Forces in Mongolia during
1982–2015, Remote Sens., 12, 603, <ext-link xlink:href="https://doi.org/10.3390/rs12040603" ext-link-type="DOI">10.3390/rs12040603</ext-link>, 2020.</mixed-citation></ref>
      <ref id="bib1.bibx34"><?xmltex \def\ref@label{{Min et~al.(2017)}}?><label>Min et al.(2017)</label><?label min2017developing?><mixed-citation>
Min, M., Wu, C., Li, C., Liu, H., Xu, N., Wu, X., Chen, L., Wang, F., Sun, F., Qin, D., Wang, X., Li, B., Zheng, Z., Cao, G., and Dong., L.: Developing the science product algorithm testbed for Chinese
next-generation geostationary meteorological satellites: Fengyun-4 series,
J. Meteorol. Res., 31, 708–719, 2017.</mixed-citation></ref>
      <ref id="bib1.bibx35"><?xmltex \def\ref@label{{Mona et~al.(2014)}}?><label>Mona et al.(2014)</label><?label Mona2014BSCDREAM8b?><mixed-citation>Mona, L., Papagiannopoulos, N., Basart, S., Baldasano, J., Binietoglou, I., Cornacchia, C., and Pappalardo, G.: EARLINET dust observations vs. BSC-DREAM8b modeled profiles: 12-year-long systematic comparison at Potenza, Italy, Atmos. Chem. Phys., 14, 8781–8793, <ext-link xlink:href="https://doi.org/10.5194/acp-14-8781-2014" ext-link-type="DOI">10.5194/acp-14-8781-2014</ext-link>, 2014.</mixed-citation></ref>
      <ref id="bib1.bibx36"><?xmltex \def\ref@label{{Morcrette et~al.(2008a)}}?><label>Morcrette et al.(2008a)</label><?label Morcrette2008Sea?><mixed-citation>Morcrette, J.-J., Beljaars, A., Benedetti, A., Jones, L., and Boucher, O.:
Sea-salt and dust aerosols in the ECMWF IFS model, Geophys. Res.
Lett., 35, L24813, <ext-link xlink:href="https://doi.org/10.1029/2008GL036041" ext-link-type="DOI">10.1029/2008GL036041</ext-link>, 2008a.</mixed-citation></ref>
      <ref id="bib1.bibx37"><?xmltex \def\ref@label{{Morcrette et~al.(2008b)}}?><label>Morcrette et al.(2008b)</label><?label ECMWF2008Aerosol_analysis?><mixed-citation>Morcrette, J.-J., Boucher, O., Jones, L., Salmond, D., Bechtold, P., Beljaars,
A., Benedetti, A., Bonet, A., Kaiser, J., Razinger, M., Schulz, M., Serrar,
S., Simmons, A., Sofiev, M., Suttie, M., Tompkins, A., Untch, A., and the
GEMS-AER team: Aerosol analysis and forecast in the ECMWF Integrated Forecast
System: Forward modelling, ECMWF Technical Memoranda, ECMWF, p. 35,
<ext-link xlink:href="https://doi.org/10.21957/jxqm3uq6j" ext-link-type="DOI">10.21957/jxqm3uq6j</ext-link>, 2008b.</mixed-citation></ref>
      <ref id="bib1.bibx38"><?xmltex \def\ref@label{{Morcrette et~al.(2009)}}?><label>Morcrette et al.(2009)</label><?label Morcrette2009Aerosol?><mixed-citation>Morcrette, J.-J., Boucher, O., Jones, L., Salmond, D., Bechtold, P., Beljaars,
A., Benedetti, A., Bonet, A., Kaiser, J. W., Razinger, M., Schulz, M.,
Serrar, S., Simmons, A. J., Sofiev, M., Suttie, M., Tompkins, A. M., and
Untch, A.: Aerosol analysis and forecast in the European Centre for
Medium-Range Weather Forecasts Integrated Forecast System: Forward modeling,
J. Geophys. Res.-Atmos., 114, D06206,
<ext-link xlink:href="https://doi.org/10.1029/2008JD011235" ext-link-type="DOI">10.1029/2008JD011235</ext-link>, 2009.</mixed-citation></ref>
      <ref id="bib1.bibx39"><?xmltex \def\ref@label{{NASA(2022)}}?><label>NASA(2022)</label><?label NASA2022?><mixed-citation>NASA: MODIS Data Collection, NASA [data set], <uri>https://ladsweb.modaps.eosdis.nasa.gov/</uri>,  last access: 14 May 2022.</mixed-citation></ref>
      <ref id="bib1.bibx40"><?xmltex \def\ref@label{{Niu et~al.(2019)}}?><label>Niu et al.(2019)</label><?label Niu2019ecological?><mixed-citation>Niu, Q., Xiao, X., Zhang, Y., Qin, Y., Dang, X., Wang, J., Zou, Z., Doughty,
R. B., Brandt, M., Tong, X., Horion, S., Fensholt, R., Chen, C., Myneni,
R. B., Xu, W., Di, G., and Zhou, X.: Ecological engineering projects
increased vegetation cover, production, and biomass in semiarid and subhumid
Northern China, Land Degrad. Dev., 30, 1620–1631,
<ext-link xlink:href="https://doi.org/10.1002/ldr.3351" ext-link-type="DOI">10.1002/ldr.3351</ext-link>, 2019.</mixed-citation></ref>
      <ref id="bib1.bibx41"><?xmltex \def\ref@label{{Niu et~al.(2008)}}?><label>Niu et al.(2008)</label><?label Niu2008Data?><mixed-citation>Niu, T., Gong, S. L., Zhu, G. F., Liu, H. L., Hu, X. Q., Zhou, C. H., and Wang, Y. Q.: Data assimilation of dust aerosol observations for the CUACE/dust forecasting system, Atmos. Chem. Phys., 8, 3473–3482, <ext-link xlink:href="https://doi.org/10.5194/acp-8-3473-2008" ext-link-type="DOI">10.5194/acp-8-3473-2008</ext-link>, 2008.</mixed-citation></ref>
      <ref id="bib1.bibx42"><?xmltex \def\ref@label{{Pérez et~al.(2006)}}?><label>Pérez et al.(2006)</label><?label Perez2006BSC?><mixed-citation>Pérez, C., Nickovic, S., Baldasano, J. M., Sicard, M., Rocadenbosch, F., and
Cachorro, V. E.: A long Saharan dust event over the western Mediterranean:
Lidar, Sun photometer observations, and regional dust modeling, J. Geophys. Res.-Atmos., 111, D15214, <ext-link xlink:href="https://doi.org/10.1029/2005JD006579" ext-link-type="DOI">10.1029/2005JD006579</ext-link>, 2006.</mixed-citation></ref>
      <ref id="bib1.bibx43"><?xmltex \def\ref@label{{Remer et~al.(2005)R}}?><label>Remer et al.(2005)R</label><?label Remer2005MODIS?><mixed-citation>
Remer, L. A., Kaufman, Y. J., Tanré, D., Mattoo, S., Chu, D. A., Martins,
J. V., Li, R. R., Ichoku, C., Levy, R. C., Kleidman, R. G., Eck, T. F.,
Vermote, E., and Holben, B. N.: The MODIS Aerosol Algorithm, Products, and
Validation, J. Atmos. Sci., 62, 947–973, 2005.</mixed-citation></ref>
      <ref id="bib1.bibx44"><?xmltex \def\ref@label{{Saide et~al.(2013)}}?><label>Saide et al.(2013)</label><?label Saide2013AerosolOptcalDepth?><mixed-citation>Saide, P. E., Carmichael, G. R., Liu, Z., Schwartz, C. S., Lin, H. C., da Silva, A. M., and Hyer, E.: Aerosol optical depth assimilation for a size-resolved sectional model: impacts of observationally constrained, multi-wavelength and fine mode retrievals on regional scale analyses and forecasts, Atmos. Chem. Phys., 13, 10425–10444, <ext-link xlink:href="https://doi.org/10.5194/acp-13-10425-2013" ext-link-type="DOI">10.5194/acp-13-10425-2013</ext-link>, 2013.</mixed-citation></ref>
      <ref id="bib1.bibx45"><?xmltex \def\ref@label{{Sayer et~al.(2014)}}?><label>Sayer et al.(2014)</label><?label Sayer2014DB?><mixed-citation>Sayer, A. M., Munchak, L. A., Hsu, N. C., Levy, R. C., Bettenhausen, C., and
Jeong, M.-J.: MODIS Collection 6 aerosol products: Comparison between Aqua's
e-Deep Blue, Dark Target, and “merged” data sets, and usage
recommendations, J. Geophys. Res.-Atmos., 119,
13965–13989, <ext-link xlink:href="https://doi.org/10.1002/2014JD022453" ext-link-type="DOI">10.1002/2014JD022453</ext-link>, 2014.</mixed-citation></ref>
      <ref id="bib1.bibx46"><?xmltex \def\ref@label{{Schepanski et~al.(2007)}}?><label>Schepanski et al.(2007)</label><?label schepanski2007new?><mixed-citation>Schepanski, K., Tegen, I., Laurent, B., Heinold, B., and Macke, A.: A new
Saharan dust source activation frequency map derived from MSG-SEVIRI
IR-channels, Geophys. Res. Lett., 34, L18803, <ext-link xlink:href="https://doi.org/10.1029/2007GL030168" ext-link-type="DOI">10.1029/2007GL030168</ext-link>, 2007.</mixed-citation></ref>
      <ref id="bib1.bibx47"><?xmltex \def\ref@label{{Schuster et~al.(2006)}}?><label>Schuster et al.(2006)</label><?label Schuster2006Angstrom?><mixed-citation>Schuster, G. L., Dubovik, O., and Holben, B. N.: Angstrom exponent and bimodal
aerosol size distributions, J. Geophys. Res.-Atmos.,
111, D07207, <ext-link xlink:href="https://doi.org/10.1029/2005JD006328" ext-link-type="DOI">10.1029/2005JD006328</ext-link>, 2006.</mixed-citation></ref>
      <ref id="bib1.bibx48"><?xmltex \def\ref@label{{Shao(2004)}}?><label>Shao(2004)</label><?label Shao2004Simplification?><mixed-citation>Shao, Y.: Simplification of a dust emission scheme and comparison with data,
J. Geophys. Res., 109, D10202, <ext-link xlink:href="https://doi.org/10.1029/2003jd004372" ext-link-type="DOI">10.1029/2003jd004372</ext-link>, 2004.</mixed-citation></ref>
      <ref id="bib1.bibx49"><?xmltex \def\ref@label{{Shao et~al.(1993)}}?><label>Shao et al.(1993)</label><?label Shao1993Effect?><mixed-citation>Shao, Y., Raupach, M. R., and Findlater, P. A.: Effect of saltation
bombardment on the entrainment of dust by wind, J. Geophys. Res., 98, 12719–12726,
<ext-link xlink:href="https://doi.org/10.1029/93jd00396" ext-link-type="DOI">10.1029/93jd00396</ext-link>, 1993.</mixed-citation></ref>
      <ref id="bib1.bibx50"><?xmltex \def\ref@label{{Shao et~al.(2011)}}?><label>Shao et al.(2011)</label><?label SHAO2011Dust?><mixed-citation>Shao, Y., Wyrwoll, K.-H., Chappell, A., Huang, J., Lin, Z., McTainsh, G. H.,
Mikami, M., Tanaka, T. Y., Wang, X., and Yoon, S.: Dust cycle: An emerging
core theme in Earth system science, Aeolian Res., 2, 181–204,
<ext-link xlink:href="https://doi.org/10.1016/j.aeolia.2011.02.001" ext-link-type="DOI">10.1016/j.aeolia.2011.02.001</ext-link>, 2011.</mixed-citation></ref>
      <ref id="bib1.bibx51"><?xmltex \def\ref@label{{Shao et~al.(2013)}}?><label>Shao et al.(2013)</label><?label Shao2013Global?><mixed-citation>Shao, Y., Klose, M., and Wyrwoll, K.-H.: Recent global dust trend and
connections to climate forcing, J. Geophys. Res.-Atmos.,
118, 11107–11118, <ext-link xlink:href="https://doi.org/10.1002/jgrd.50836" ext-link-type="DOI">10.1002/jgrd.50836</ext-link>, 2013.</mixed-citation></ref>
      <ref id="bib1.bibx52"><?xmltex \def\ref@label{{Shao et~al.(1996)}}?><label>Shao et al.(1996)</label><?label Shao1996Model?><mixed-citation>Shao, Y. P., Raupach, M. R., and Leys, J. F.: A model for predicting aeolian
sand drift and dust entrainment on scales from paddock to region, Australian
J. Soil Res., 34, 309–342, <ext-link xlink:href="https://doi.org/10.1071/sr9960309" ext-link-type="DOI">10.1071/sr9960309</ext-link>, 1996.</mixed-citation></ref>
      <ref id="bib1.bibx53"><?xmltex \def\ref@label{{Tan and Li(2015)}}?><label>Tan and Li(2015)</label><?label TAN201542Greengreatwall?><mixed-citation>Tan, M. and Li, X.: Does the Green Great Wall effectively decrease dust storm
intensity in China? A study based on NOAA NDVI and weather station data, Land
Use Policy, 43, 42–47,
<ext-link xlink:href="https://doi.org/10.1016/j.landusepol.2014.10.017" ext-link-type="DOI">10.1016/j.landusepol.2014.10.017</ext-link>, 2015.</mixed-citation></ref>
      <ref id="bib1.bibx54"><?xmltex \def\ref@label{{Timmermans et~al.(2017)}}?><label>Timmermans et al.(2017)</label><?label Timmermans2017Source?><mixed-citation>Timmermans, R., Kranenburg, R., Manders, A., Hendriks, C., Segers, A., Dammers,
E., Zhang, Q., Wang, L., Liu, Z., Zeng, L., Denier van der Gon, H., and
Schaap, M.: Source apportionment of PM<inline-formula><mml:math id="M272" display="inline"><mml:msub><mml:mi/><mml:mn mathvariant="normal">2.5</mml:mn></mml:msub></mml:math></inline-formula> across China using LOTOS-EUROS,
Atmos. Environ., 164, 370–386, <ext-link xlink:href="https://doi.org/10.1016/j.atmosenv.2017.06.003" ext-link-type="DOI">10.1016/j.atmosenv.2017.06.003</ext-link>, 2017.</mixed-citation></ref>
      <ref id="bib1.bibx55"><?xmltex \def\ref@label{{TNO(2022)}}?><label>TNO(2022)</label><?label TNO2022?><mixed-citation>TNO: LOTOS-EUROS source code, TNO [code], <uri>https://lotos-euros.tno.nl</uri>, last access: 14 May 2022.</mixed-citation></ref>
      <ref id="bib1.bibx56"><?xmltex \def\ref@label{{Winker et~al.(2007)}}?><label>Winker et al.(2007)</label><?label Winker2007Initial?><mixed-citation>Winker, D. M., Hunt, W. H., and McGill, M. J.: Initial performance assessment
of CALIOP, Geophys. Res. Lett., 34, L19803,
<ext-link xlink:href="https://doi.org/10.1029/2007gl030135" ext-link-type="DOI">10.1029/2007gl030135</ext-link>, 2007.</mixed-citation></ref>
      <ref id="bib1.bibx57"><?xmltex \def\ref@label{{{World Meteorological Organization}(2017)}}?><label>World Meteorological Organization(2017)</label><?label Organization2017WMO?><mixed-citation>World Meteorological Organization: WMO AIRBORNE DUST BULLETIN: Sand and Dust
Storm Warning Advisory and Assessment System,
<uri>https://library.wmo.int/doc_num.php?explnum_id=3416</uri> (last access: 14 May 2022), 2017.</mixed-citation></ref>
      <ref id="bib1.bibx58"><?xmltex \def\ref@label{{{World Meteorological Organization}(2019)}}?><label>World Meteorological Organization(2019)</label><?label WMO2019?><mixed-citation>World Meteorological Organization: WMO AIRBORNE DUST BULLETIN: Sand and Dust
Storm Warning Advisory and Assessment System, Tech. rep.,
<ext-link xlink:href="https://public.wmo.int/en/resources/library/wmo-airborne-dust-bulletin-sand-and-dust-storm-warning-advisory-and-assessment">https://public.wmo.int/en/resources/library/wmo-airborne-dust-bulletin-sand-and-dust-storm-warning-advisory-and-assessment</ext-link> (last access: 14 May 2022),
2019.</mixed-citation></ref>
      <ref id="bib1.bibx59"><?xmltex \def\ref@label{{{World Meteorological Organization}(2020)}}?><label>World Meteorological Organization(2020)</label><?label WMO2020?><mixed-citation>World Meteorological Organization: Sand and Dust Storm Warning Advisory and
Assessment System: Science Progress Report, Tech. rep., World Meteorological
Organization, <ext-link xlink:href="https://public.wmo.int/en/resources/library/sand-and-dust-storm-warning-advisory-and-assessment-system-science-progress-report">https://public.wmo.int/en/resources/library/sand-and-dust-storm-warning-advisory-and-assessment-system-science-progress-report</ext-link> (last access: 14 May 2022),
2020.</mixed-citation></ref>
      <ref id="bib1.bibx60"><?xmltex \def\ref@label{{Wu et~al.(2016)}}?><label>Wu et al.(2016)</label><?label Wu2016Processoriented?><mixed-citation>Wu, C., Lin, Z., He, J., Zhang, M., Liu, X., Zhang, R., and Brown, H.: A
process-oriented evaluation of dust emission parameterizations in CESM:
Simulation of a typical severe dust storm in East Asia, J. Adv. Model. Earth Sy., 8, 1432–1452, <ext-link xlink:href="https://doi.org/10.1002/2016MS000723" ext-link-type="DOI">10.1002/2016MS000723</ext-link>, 2016.</mixed-citation></ref>
      <ref id="bib1.bibx61"><?xmltex \def\ref@label{{Yin et~al.(2021)}}?><label>Yin et al.(2021)</label><?label Yin2021Sandstorm?><mixed-citation>Yin, Z., Wan, Y., Zhang, Y., and Wang, H.: Why super sandstorm 2021 in North
China, Natl. Sci. Rev., 9, nwab165, <ext-link xlink:href="https://doi.org/10.1093/nsr/nwab165" ext-link-type="DOI">10.1093/nsr/nwab165</ext-link>, 2021.</mixed-citation></ref>
      <ref id="bib1.bibx62"><?xmltex \def\ref@label{{Yumimoto and Takemura(2015)}}?><label>Yumimoto and Takemura(2015)</label><?label Yumimoto2015Longterm?><mixed-citation>Yumimoto, K. and Takemura, T.: Long-term inverse modeling of Asian dust:
Interannual variations of its emission, transport, deposition, and radiative
forcing, J. Geophys. Res.-Atmos., 120, 2014JD022390+,
<ext-link xlink:href="https://doi.org/10.1002/2014jd022390" ext-link-type="DOI">10.1002/2014jd022390</ext-link>, 2015.
</mixed-citation></ref><?xmltex \hack{\newpage}?>
      <ref id="bib1.bibx63"><?xmltex \def\ref@label{{Yumimoto et~al.(2016)}}?><label>Yumimoto et al.(2016)</label><?label Yumimoto2016Aerosol?><mixed-citation>Yumimoto, K., Nagao, T. M., Kikuchi, M., Sekiyama, T. T., Murakami, H., Tanaka,
T. Y., Ogi, A., Irie, H., Khatri, P., Okumura, H., Arai, K., Morino, I.,
Uchino, O., and Maki, T.: Aerosol data assimilation using data from
Himawari-8, a next-generation geostationary meteorological satellite,
Geophys. Res. Lett., 43, 2016GL069298, <ext-link xlink:href="https://doi.org/10.1002/2016gl069298" ext-link-type="DOI">10.1002/2016gl069298</ext-link>, 2016.</mixed-citation></ref>
      <ref id="bib1.bibx64"><?xmltex \def\ref@label{{Zender(2003)}}?><label>Zender(2003)</label><?label Zender2003Mineral?><mixed-citation>Zender, C. S.: Mineral Dust Entrainment and Deposition (DEAD) model:
Description and 1990s dust climatology, J. Geophys. Res.,
108, 4416, <ext-link xlink:href="https://doi.org/10.1029/2002jd002775" ext-link-type="DOI">10.1029/2002jd002775</ext-link>, 2003.</mixed-citation></ref>
      <ref id="bib1.bibx65"><?xmltex \def\ref@label{{Zhang et~al.(2018)}}?><label>Zhang et al.(2018)</label><?label Zhang2018East?><mixed-citation>Zhang, X.-X., Sharratt, B., Liu, L.-Y., Wang, Z.-F., Pan, X.-L., Lei, J.-Q., Wu, S.-X., Huang, S.-Y., Guo, Y.-H., Li, J., Tang, X., Yang, T., Tian, Y., Chen, X.-S., Hao, J.-Q., Zheng, H.-T., Yang, Y.-Y., and Lyu, Y.-L.: East Asian dust storm in May 2017: observations, modelling, and its influence on the Asia-Pacific region, Atmos. Chem. Phys., 18, 8353–8371, <ext-link xlink:href="https://doi.org/10.5194/acp-18-8353-2018" ext-link-type="DOI">10.5194/acp-18-8353-2018</ext-link>, 2018.</mixed-citation></ref>
      <ref id="bib1.bibx66"><?xmltex \def\ref@label{{Zhang et~al.(2019)}}?><label>Zhang et al.(2019)</label><?label zhang2019Himawarivalidation?><mixed-citation>
Zhang, Z., Wu, W., Fan, M., Tao, M., Wei, J., Jin, J., Tan, Y., and Wang, Q.:
Validation of Himawari-8 aerosol optical depth retrievals over China,
Atmos. Environ., 199, 32–44, 2019.</mixed-citation></ref>
      <ref id="bib1.bibx67"><?xmltex \def\ref@label{{Zhou et~al.(2008)}}?><label>Zhou et al.(2008)</label><?label Zhou2008Development?><mixed-citation>Zhou, C. H., Gong, S. L., Zhang, X. Y., Wang, Y. Q., Niu, T., Liu, H. L., Zhao, T. L., Yang, Y. Q., and Hou, Q.: Development and evaluation of an operational SDS forecasting system for East Asia: CUACE/Dust, Atmos. Chem. Phys., 8, 787–798, <ext-link xlink:href="https://doi.org/10.5194/acp-8-787-2008" ext-link-type="DOI">10.5194/acp-8-787-2008</ext-link>, 2008.</mixed-citation></ref>

  </ref-list></back>
    <!--<article-title-html>Inverse modeling of the 2021 spring  super dust storms in East Asia</article-title-html>
<abstract-html/>
<ref-html id="bib1.bib1"><label>Bessho et al.(2016)</label><mixed-citation>
Bessho, K., Date, K., Hayashi, M., Ikeda, A., Imai, T., Inoue, H., Kumagai, Y.,
Miyakawa, T., Murata, H., Ohno, T., Okuyama, A., Oyama, R., Sasaki, Y.,
Shimazu, Y., Shimoji, K., Sumida, Y., Suzuki, M., Taniguchi, H., Tsuchiyama,
H., Uesawa, D., Yokota, H., and Yoshida, R.: An Introduction to
Himawari-8/9 – Japan's New-Generation Geostationary
Meteorological Satellites, J. Meteorol. Soc. Jpn.
Ser. II, 94, 151–183, <a href="https://doi.org/10.2151/jmsj.2016-009" target="_blank">https://doi.org/10.2151/jmsj.2016-009</a>, 2016.
</mixed-citation></ref-html>
<ref-html id="bib1.bib2"><label>Brasseur et al.(2019)</label><mixed-citation>
Brasseur, G. P., Xie, Y., Petersen, A. K., Bouarar, I., Flemming, J., Gauss, M., Jiang, F., Kouznetsov, R., Kranenburg, R., Mijling, B., Peuch, V.-H., Pommier, M., Segers, A., Sofiev, M., Timmermans, R., van der A, R., Walters, S., Xu, J., and Zhou, G.: Ensemble forecasts of air quality in eastern China – Part 1: Model description and implementation of the MarcoPolo–Panda prediction system, version 1, Geosci. Model Dev., 12, 33–67, <a href="https://doi.org/10.5194/gmd-12-33-2019" target="_blank">https://doi.org/10.5194/gmd-12-33-2019</a>, 2019.
</mixed-citation></ref-html>
<ref-html id="bib1.bib3"><label>Chen and Walsh(2021)</label><mixed-citation>
Chen, L. and Walsh, M.: Vast sandstorms expose Mongolia's long-ignored
ecological crisis, <a href="https://asia.nikkei.com/Spotlight/Caixin/Vast-sandstorms-expose-Mongolia-s-long-ignored-ecological-crisis" target="_blank"/> (last access: 14 May 2022),
2021.
</mixed-citation></ref-html>
<ref-html id="bib1.bib4"><label>China Ministry of Environmental Protection(2022)</label><mixed-citation>
China Ministry of Environmental Protection: Ground-based air quality monitoring measurements, China Ministry of Environmental Protection [data set], <a href="http://106.37.208.233:20035/" target="_blank"/>, last access: 14 May 2022.
</mixed-citation></ref-html>
<ref-html id="bib1.bib5"><label>Di Tomaso et al.(2017)</label><mixed-citation>
Di Tomaso, E., Schutgens, N. A. J., Jorba, O., and Pérez García-Pando, C.: Assimilation of MODIS Dark Target and Deep Blue observations in the dust aerosol component of NMMB-MONARCH version 1.0, Geosci. Model Dev., 10, 1107–1129, <a href="https://doi.org/10.5194/gmd-10-1107-2017" target="_blank">https://doi.org/10.5194/gmd-10-1107-2017</a>, 2017.
</mixed-citation></ref-html>
<ref-html id="bib1.bib6"><label>Di Tomaso et al.(2021)</label><mixed-citation>
Di Tomaso, E., Escribano, J., Basart, S., Ginoux, P., Macchia, F., Barnaba, F., Benincasa, F., Bretonnière, P.-A., Buñuel, A., Castrillo, M., Cuevas, E., Formenti, P., Gonçalves, M., Jorba, O., Klose, M., Mona, L., Montané, G., Mytilinaios, M., Obiso, V., Olid, M., Schutgens, N., Votsis, A., Werner, E., and Pérez García-Pando, C.: The MONARCH high-resolution reanalysis of desert dust aerosol over Northern Africa, the Middle East and Europe (2007–2016), Earth Syst. Sci. Data Discuss. [preprint], <a href="https://doi.org/10.5194/essd-2021-358" target="_blank">https://doi.org/10.5194/essd-2021-358</a>, in review, 2021.
</mixed-citation></ref-html>
<ref-html id="bib1.bib7"><label>Dubovik et al.(2000)</label><mixed-citation>
Dubovik, O., Smirnov, A., Holben, B., King, M., Kaufman, Y., Eck, T., and
Slutsker, I.: Accuracy assessments of aerosol optical properties retrieved
from Aerosol Robotic Network (AERONET) Sun and sky radiance measurements,
J. Geophys. Res.-Atmos., 105, 9791–9806, 2000.
</mixed-citation></ref-html>
<ref-html id="bib1.bib8"><label>Dubovik et al.(2002)</label><mixed-citation>
Dubovik, O., Holben, B., Eck, T. F., Smirnov, A., Kaufman, Y. J., King, M. D.,
Tanré, D., and Slutsker, I.: Variability of absorption and optical
properties of key aerosol types observed in worldwide locations, J. Atmos. Sci., 59, 590–608, 2002.
</mixed-citation></ref-html>
<ref-html id="bib1.bib9"><label>Eck et al.(1999)</label><mixed-citation>
Eck, T. F., Holben, B., Reid, J., Dubovik, O., Smirnov, A., O'neill, N.,
Slutsker, I., and Kinne, S.: Wavelength dependence of the optical depth of
biomass burning, urban, and desert dust aerosols, J. Geophys. Res.-Atmos., 104, 31333–31349, 1999.
</mixed-citation></ref-html>
<ref-html id="bib1.bib10"><label>Fairlie et al.(2007)</label><mixed-citation>
Fairlie, T. D., Jacob, D. J., and Park, R. J.: The impact of transpacific
transport of mineral dust in the United States, Atmos. Environ., 41,
1251–1266, <a href="https://doi.org/10.1016/j.atmosenv.2006.09.048" target="_blank">https://doi.org/10.1016/j.atmosenv.2006.09.048</a>, 2007.
</mixed-citation></ref-html>
<ref-html id="bib1.bib11"><label>Ginoux et al.(2001)</label><mixed-citation>
Ginoux, P., Chin, M., Tegen, I., Prospero, J. M., Holben, B., Dubovik, O., and
Lin, S.-J.: Sources and distributions of dust aerosols simulated with the
GOCART model, J. Geophys. Res., 106, 20255–20273,
<a href="https://doi.org/10.1029/2000jd000053" target="_blank">https://doi.org/10.1029/2000jd000053</a>, 2001.
</mixed-citation></ref-html>
<ref-html id="bib1.bib12"><label>Ginoux et al.(2012)</label><mixed-citation>
Ginoux, P., Prospero, J. M., Gill, T. E., Hsu, N. C., and Zhao, M.:
Global-scale attribution of anthropogenic and natural dust sources and their
emission rates based on MODIS Deep Blue aerosol products, Rev.
Geophys., 50,  RG3005, <a href="https://doi.org/10.1029/2012RG000388" target="_blank">https://doi.org/10.1029/2012RG000388</a>, 2012.
</mixed-citation></ref-html>
<ref-html id="bib1.bib13"><label>Gkikas et al.(2021)</label><mixed-citation>
Gkikas, A., Proestakis, E., Amiridis, V., Kazadzis, S., Di Tomaso, E., Tsekeri, A., Marinou, E., Hatzianastassiou, N., and Pérez García-Pando, C.: ModIs Dust AeroSol (MIDAS): a global fine-resolution dust optical depth data set, Atmos. Meas. Tech., 14, 309–334, <a href="https://doi.org/10.5194/amt-14-309-2021" target="_blank">https://doi.org/10.5194/amt-14-309-2021</a>, 2021.
</mixed-citation></ref-html>
<ref-html id="bib1.bib14"><label>Gong and Zhang(2008)</label><mixed-citation>
Gong, S. L. and Zhang, X. Y.: CUACE/Dust – an integrated system of observation and modeling systems for operational dust forecasting in Asia, Atmos. Chem. Phys., 8, 2333–2340, <a href="https://doi.org/10.5194/acp-8-2333-2008" target="_blank">https://doi.org/10.5194/acp-8-2333-2008</a>, 2008.
</mixed-citation></ref-html>
<ref-html id="bib1.bib15"><label>Han et al.(2021)</label><mixed-citation>
Han, J., Dai, H., and Gu, Z.: Sandstorms and desertification in Mongolia, an
example of future climate events: A review, Environ. Chem. Lett., 19,
4063–4073, 2021.
</mixed-citation></ref-html>
<ref-html id="bib1.bib16"><label>Hsu et al.(2013)</label><mixed-citation>
Hsu, N. C., Jeong, M.-J., Bettenhausen, C., Sayer, A. M., Hansell, R., Seftor,
C. S., Huang, J., and Tsay, S.-C.: Enhanced Deep Blue aerosol retrieval
algorithm: The second generation, J. Geophys. Res.-Atmos., 118, 9296–9315, <a href="https://doi.org/10.1002/jgrd.50712" target="_blank">https://doi.org/10.1002/jgrd.50712</a>, 2013.
</mixed-citation></ref-html>
<ref-html id="bib1.bib17"><label>Huneeus et al.(2011)</label><mixed-citation>
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, <a href="https://doi.org/10.5194/acp-11-7781-2011" target="_blank">https://doi.org/10.5194/acp-11-7781-2011</a>, 2011.
</mixed-citation></ref-html>
<ref-html id="bib1.bib18"><label>Jin(2021)</label><mixed-citation>
Jin, G.: The most severe Sandstorm in a decade,
<a href="https://www.chinanews.com/m/sh/2021/03-15/9433005.shtml" target="_blank"/> (last access: 14 May 2022),
2021.
</mixed-citation></ref-html>
<ref-html id="bib1.bib19"><label>Jin(2022)</label><mixed-citation>
Jin, J.: ground-based air quality measurements during the 2021 spring super dust storms, Zenodo [data set], <a href="https://doi.org/10.5281/zenodo.6459866" target="_blank">https://doi.org/10.5281/zenodo.6459866</a>, 2022.
</mixed-citation></ref-html>
<ref-html id="bib1.bib20"><label>Jin et al.(2018)</label><mixed-citation>
Jin, J., Lin, H. X., Heemink, A., and Segers, A.: Spatially varying parameter
estimation for dust emissions using reduced-tangent-linearization 4DVar,
Atmos. Environ., 187, 358–373,
<a href="https://doi.org/10.1016/j.atmosenv.2018.05.060" target="_blank">https://doi.org/10.1016/j.atmosenv.2018.05.060</a>, 2018.
</mixed-citation></ref-html>
<ref-html id="bib1.bib21"><label>Jin et al.(2019a)</label><mixed-citation>
Jin, J., Lin, H. X., Segers, A., Xie, Y., and Heemink, A.: Machine learning for observation bias correction with application to dust storm data assimilation, Atmos. Chem. Phys., 19, 10009–10026, <a href="https://doi.org/10.5194/acp-19-10009-2019" target="_blank">https://doi.org/10.5194/acp-19-10009-2019</a>, 2019a.
</mixed-citation></ref-html>
<ref-html id="bib1.bib22"><label>Jin et al.(2019b)</label><mixed-citation>
Jin, J., Segers, A., Heemink, A., Yoshida, M., Han, W., and Lin, H.-X.: Dust
Emission Inversion Using Himawari‐8 AODs Over East Asia: An Extreme Dust
Event in May 2017, J. Adv. Model. Earth Sy., 11,
446–467, <a href="https://doi.org/10.1029/2018MS001491" target="_blank">https://doi.org/10.1029/2018MS001491</a>, 2019b.
</mixed-citation></ref-html>
<ref-html id="bib1.bib23"><label>Jin et al.(2020)</label><mixed-citation>
Jin, J., Segers, A., Liao, H., Heemink, A., Kranenburg, R., and Lin, H. X.: Source backtracking for dust storm emission inversion using an adjoint method: case study of Northeast China, Atmos. Chem. Phys., 20, 15207–15225, <a href="https://doi.org/10.5194/acp-20-15207-2020" target="_blank">https://doi.org/10.5194/acp-20-15207-2020</a>, 2020.
</mixed-citation></ref-html>
<ref-html id="bib1.bib24"><label>Jin et al.(2021)</label><mixed-citation>
Jin, J., Segers, A., Lin, H. X., Henzing, B., Wang, X., Heemink, A., and Liao, H.: Position correction in dust storm forecasting using LOTOS-EUROS v2.1: grid-distorted data assimilation v1.0, Geosci. Model Dev., 14, 5607–5622, <a href="https://doi.org/10.5194/gmd-14-5607-2021" target="_blank">https://doi.org/10.5194/gmd-14-5607-2021</a>, 2021.
</mixed-citation></ref-html>
<ref-html id="bib1.bib25"><label>Justice et al.(1998)</label><mixed-citation>
Justice, C., Vermote, E., Townshend, J., Defries, R., Roy, D., Hall, D.,
Salomonson, V., Privette, J., Riggs, G., Strahler, A., Lucht, W., Myneni, R.,
Knyazikhin, Y., Running, S., Nemani, R., Wan, Z., Huete, A., van Leeuwen, W.,
Wolfe, R., Giglio, L., Muller, J., Lewis, P., and Barnsley, M.: The Moderate
Resolution Imaging Spectroradiometer (MODIS): land remote sensing for global
change research, IEEE T. Geosci. Remote Sens., 36,
1228–1249, <a href="https://doi.org/10.1109/36.701075" target="_blank">https://doi.org/10.1109/36.701075</a>, 1998.
</mixed-citation></ref-html>
<ref-html id="bib1.bib26"><label>Kalnay(2002)</label><mixed-citation>
Kalnay, E.: Atmospheric Modeling, Data Assimilation and Predictability, edited by: Kalnay, E.,
Cambridge University Press, <a href="https://doi.org/10.1017/CBO9780511802270" target="_blank">https://doi.org/10.1017/CBO9780511802270</a>, 2002.
</mixed-citation></ref-html>
<ref-html id="bib1.bib27"><label>Li et al.(2018)</label><mixed-citation>
Li, Z., Xu, H., Li, K., Li, D., Xie, Y., Li, L., Zhang, Y., Gu, X., Zhao, W., Tian, Q., Deng, R., Su, X., Huang, B., Qiao, Y., Cui, W., Hu, Y., Gong, C., Wang, Y., Wang, X., Wang, J., Du, W., Pan, Z., Li, Z., and Bu, D.: Comprehensive study of optical, physical, chemical, and
radiative properties of total columnar atmospheric aerosols over China: An
overview of Sun–Sky Radiometer Observation Network (SONET) measurements,
B. Am. Meteorol. Soc., 99, 739–755, 2018.
</mixed-citation></ref-html>
<ref-html id="bib1.bib28"><label>Lin et al.(2008)</label><mixed-citation>
Lin, C., Wang, Z., and Zhu, J.: An Ensemble Kalman Filter for severe dust storm data assimilation over China, Atmos. Chem. Phys., 8, 2975–2983, <a href="https://doi.org/10.5194/acp-8-2975-2008" target="_blank">https://doi.org/10.5194/acp-8-2975-2008</a>, 2008.
</mixed-citation></ref-html>
<ref-html id="bib1.bib29"><label>Liu et al.(2019)</label><mixed-citation>
Liu, M., Zhou, G., Saari, R. K., Li, S., Liu, X., and Li, J.: Quantifying PM<sub>2.5</sub>
mass concentration and particle radius using satellite data and an
optical-mass conversion algorithm, ISPRS J. Photogramm., 158, 90–98,
<a href="https://doi.org/10.1016/j.isprsjprs.2019.10.010" target="_blank">https://doi.org/10.1016/j.isprsjprs.2019.10.010</a>, 2019.
</mixed-citation></ref-html>
<ref-html id="bib1.bib30"><label>Lorente-Plazas and Hacker(2017)</label><mixed-citation>
Lorente-Plazas, R. and Hacker, J. P.: Observation and Model Bias Estimation in
the Presence of Either or Both Sources of Error, Mon. Weather Rev.,
145, 2683–2696, <a href="https://doi.org/10.1175/MWR-D-16-0273.1" target="_blank">https://doi.org/10.1175/MWR-D-16-0273.1</a>, 2017.
</mixed-citation></ref-html>
<ref-html id="bib1.bib31"><label>Manders et al.(2017)</label><mixed-citation>
Manders, A. M. M., Builtjes, P. J. H., Curier, L., Denier van der Gon, H. A. C., Hendriks, C., Jonkers, S., Kranenburg, R., Kuenen, J. J. P., Segers, A. J., Timmermans, R. M. A., Visschedijk, A. J. H., Wichink Kruit, R. J., van Pul, W. A. J., Sauter, F. J., van der Swaluw, E., Swart, D. P. J., Douros, J., Eskes, H., van Meijgaard, E., van Ulft, B., van Velthoven, P., Banzhaf, S., Mues, A. C., Stern, R., Fu, G., Lu, S., Heemink, A., van Velzen, N., and Schaap, M.: Curriculum vitae of the LOTOS–EUROS (v2.0) chemistry transport model, Geosci. Model Dev., 10, 4145–4173, <a href="https://doi.org/10.5194/gmd-10-4145-2017" target="_blank">https://doi.org/10.5194/gmd-10-4145-2017</a>, 2017.
</mixed-citation></ref-html>
<ref-html id="bib1.bib32"><label>Marticorena and Bergametti(1995)</label><mixed-citation>
Marticorena, B. and Bergametti, G.: Modeling the atmospheric dust cycle: 1.
Design of a soil-derived dust emission scheme, J. Geophys. Res., 100,
16415–16430, <a href="https://doi.org/10.1029/95JD00690" target="_blank">https://doi.org/10.1029/95JD00690</a>, 1995.
</mixed-citation></ref-html>
<ref-html id="bib1.bib33"><label>Meng et al.(2020)</label><mixed-citation>
Meng, X., Gao, X., Li, S., and Lei, J.: Spatial and Temporal Characteristics of
Vegetation NDVI Changes and the Driving Forces in Mongolia during
1982–2015, Remote Sens., 12, 603, <a href="https://doi.org/10.3390/rs12040603" target="_blank">https://doi.org/10.3390/rs12040603</a>, 2020.
</mixed-citation></ref-html>
<ref-html id="bib1.bib34"><label>Min et al.(2017)</label><mixed-citation>
Min, M., Wu, C., Li, C., Liu, H., Xu, N., Wu, X., Chen, L., Wang, F., Sun, F., Qin, D., Wang, X., Li, B., Zheng, Z., Cao, G., and Dong., L.: Developing the science product algorithm testbed for Chinese
next-generation geostationary meteorological satellites: Fengyun-4 series,
J. Meteorol. Res., 31, 708–719, 2017.
</mixed-citation></ref-html>
<ref-html id="bib1.bib35"><label>Mona et al.(2014)</label><mixed-citation>
Mona, L., Papagiannopoulos, N., Basart, S., Baldasano, J., Binietoglou, I., Cornacchia, C., and Pappalardo, G.: EARLINET dust observations vs. BSC-DREAM8b modeled profiles: 12-year-long systematic comparison at Potenza, Italy, Atmos. Chem. Phys., 14, 8781–8793, <a href="https://doi.org/10.5194/acp-14-8781-2014" target="_blank">https://doi.org/10.5194/acp-14-8781-2014</a>, 2014.
</mixed-citation></ref-html>
<ref-html id="bib1.bib36"><label>Morcrette et al.(2008a)</label><mixed-citation>
Morcrette, J.-J., Beljaars, A., Benedetti, A., Jones, L., and Boucher, O.:
Sea-salt and dust aerosols in the ECMWF IFS model, Geophys. Res.
Lett., 35, L24813, <a href="https://doi.org/10.1029/2008GL036041" target="_blank">https://doi.org/10.1029/2008GL036041</a>, 2008a.
</mixed-citation></ref-html>
<ref-html id="bib1.bib37"><label>Morcrette et al.(2008b)</label><mixed-citation>
Morcrette, J.-J., Boucher, O., Jones, L., Salmond, D., Bechtold, P., Beljaars,
A., Benedetti, A., Bonet, A., Kaiser, J., Razinger, M., Schulz, M., Serrar,
S., Simmons, A., Sofiev, M., Suttie, M., Tompkins, A., Untch, A., and the
GEMS-AER team: Aerosol analysis and forecast in the ECMWF Integrated Forecast
System: Forward modelling, ECMWF Technical Memoranda, ECMWF, p. 35,
<a href="https://doi.org/10.21957/jxqm3uq6j" target="_blank">https://doi.org/10.21957/jxqm3uq6j</a>, 2008b.
</mixed-citation></ref-html>
<ref-html id="bib1.bib38"><label>Morcrette et al.(2009)</label><mixed-citation>
Morcrette, J.-J., Boucher, O., Jones, L., Salmond, D., Bechtold, P., Beljaars,
A., Benedetti, A., Bonet, A., Kaiser, J. W., Razinger, M., Schulz, M.,
Serrar, S., Simmons, A. J., Sofiev, M., Suttie, M., Tompkins, A. M., and
Untch, A.: Aerosol analysis and forecast in the European Centre for
Medium-Range Weather Forecasts Integrated Forecast System: Forward modeling,
J. Geophys. Res.-Atmos., 114, D06206,
<a href="https://doi.org/10.1029/2008JD011235" target="_blank">https://doi.org/10.1029/2008JD011235</a>, 2009.
</mixed-citation></ref-html>
<ref-html id="bib1.bib39"><label>NASA(2022)</label><mixed-citation>
NASA: MODIS Data Collection, NASA [data set], <a href="https://ladsweb.modaps.eosdis.nasa.gov/" target="_blank"/>,  last access: 14 May 2022.
</mixed-citation></ref-html>
<ref-html id="bib1.bib40"><label>Niu et al.(2019)</label><mixed-citation>
Niu, Q., Xiao, X., Zhang, Y., Qin, Y., Dang, X., Wang, J., Zou, Z., Doughty,
R. B., Brandt, M., Tong, X., Horion, S., Fensholt, R., Chen, C., Myneni,
R. B., Xu, W., Di, G., and Zhou, X.: Ecological engineering projects
increased vegetation cover, production, and biomass in semiarid and subhumid
Northern China, Land Degrad. Dev., 30, 1620–1631,
<a href="https://doi.org/10.1002/ldr.3351" target="_blank">https://doi.org/10.1002/ldr.3351</a>, 2019.
</mixed-citation></ref-html>
<ref-html id="bib1.bib41"><label>Niu et al.(2008)</label><mixed-citation>
Niu, T., Gong, S. L., Zhu, G. F., Liu, H. L., Hu, X. Q., Zhou, C. H., and Wang, Y. Q.: Data assimilation of dust aerosol observations for the CUACE/dust forecasting system, Atmos. Chem. Phys., 8, 3473–3482, <a href="https://doi.org/10.5194/acp-8-3473-2008" target="_blank">https://doi.org/10.5194/acp-8-3473-2008</a>, 2008.
</mixed-citation></ref-html>
<ref-html id="bib1.bib42"><label>Pérez et al.(2006)</label><mixed-citation>
Pérez, C., Nickovic, S., Baldasano, J. M., Sicard, M., Rocadenbosch, F., and
Cachorro, V. E.: A long Saharan dust event over the western Mediterranean:
Lidar, Sun photometer observations, and regional dust modeling, J. Geophys. Res.-Atmos., 111, D15214, <a href="https://doi.org/10.1029/2005JD006579" target="_blank">https://doi.org/10.1029/2005JD006579</a>, 2006.
</mixed-citation></ref-html>
<ref-html id="bib1.bib43"><label>Remer et al.(2005)R</label><mixed-citation>
Remer, L. A., Kaufman, Y. J., Tanré, D., Mattoo, S., Chu, D. A., Martins,
J. V., Li, R. R., Ichoku, C., Levy, R. C., Kleidman, R. G., Eck, T. F.,
Vermote, E., and Holben, B. N.: The MODIS Aerosol Algorithm, Products, and
Validation, J. Atmos. Sci., 62, 947–973, 2005.
</mixed-citation></ref-html>
<ref-html id="bib1.bib44"><label>Saide et al.(2013)</label><mixed-citation>
Saide, P. E., Carmichael, G. R., Liu, Z., Schwartz, C. S., Lin, H. C., da Silva, A. M., and Hyer, E.: Aerosol optical depth assimilation for a size-resolved sectional model: impacts of observationally constrained, multi-wavelength and fine mode retrievals on regional scale analyses and forecasts, Atmos. Chem. Phys., 13, 10425–10444, <a href="https://doi.org/10.5194/acp-13-10425-2013" target="_blank">https://doi.org/10.5194/acp-13-10425-2013</a>, 2013.
</mixed-citation></ref-html>
<ref-html id="bib1.bib45"><label>Sayer et al.(2014)</label><mixed-citation>
Sayer, A. M., Munchak, L. A., Hsu, N. C., Levy, R. C., Bettenhausen, C., and
Jeong, M.-J.: MODIS Collection 6 aerosol products: Comparison between Aqua's
e-Deep Blue, Dark Target, and “merged” data sets, and usage
recommendations, J. Geophys. Res.-Atmos., 119,
13965–13989, <a href="https://doi.org/10.1002/2014JD022453" target="_blank">https://doi.org/10.1002/2014JD022453</a>, 2014.
</mixed-citation></ref-html>
<ref-html id="bib1.bib46"><label>Schepanski et al.(2007)</label><mixed-citation>
Schepanski, K., Tegen, I., Laurent, B., Heinold, B., and Macke, A.: A new
Saharan dust source activation frequency map derived from MSG-SEVIRI
IR-channels, Geophys. Res. Lett., 34, L18803, <a href="https://doi.org/10.1029/2007GL030168" target="_blank">https://doi.org/10.1029/2007GL030168</a>, 2007.
</mixed-citation></ref-html>
<ref-html id="bib1.bib47"><label>Schuster et al.(2006)</label><mixed-citation>
Schuster, G. L., Dubovik, O., and Holben, B. N.: Angstrom exponent and bimodal
aerosol size distributions, J. Geophys. Res.-Atmos.,
111, D07207, <a href="https://doi.org/10.1029/2005JD006328" target="_blank">https://doi.org/10.1029/2005JD006328</a>, 2006.
</mixed-citation></ref-html>
<ref-html id="bib1.bib48"><label>Shao(2004)</label><mixed-citation>
Shao, Y.: Simplification of a dust emission scheme and comparison with data,
J. Geophys. Res., 109, D10202, <a href="https://doi.org/10.1029/2003jd004372" target="_blank">https://doi.org/10.1029/2003jd004372</a>, 2004.
</mixed-citation></ref-html>
<ref-html id="bib1.bib49"><label>Shao et al.(1993)</label><mixed-citation>
Shao, Y., Raupach, M. R., and Findlater, P. A.: Effect of saltation
bombardment on the entrainment of dust by wind, J. Geophys. Res., 98, 12719–12726,
<a href="https://doi.org/10.1029/93jd00396" target="_blank">https://doi.org/10.1029/93jd00396</a>, 1993.
</mixed-citation></ref-html>
<ref-html id="bib1.bib50"><label>Shao et al.(2011)</label><mixed-citation>
Shao, Y., Wyrwoll, K.-H., Chappell, A., Huang, J., Lin, Z., McTainsh, G. H.,
Mikami, M., Tanaka, T. Y., Wang, X., and Yoon, S.: Dust cycle: An emerging
core theme in Earth system science, Aeolian Res., 2, 181–204,
<a href="https://doi.org/10.1016/j.aeolia.2011.02.001" target="_blank">https://doi.org/10.1016/j.aeolia.2011.02.001</a>, 2011.
</mixed-citation></ref-html>
<ref-html id="bib1.bib51"><label>Shao et al.(2013)</label><mixed-citation>
Shao, Y., Klose, M., and Wyrwoll, K.-H.: Recent global dust trend and
connections to climate forcing, J. Geophys. Res.-Atmos.,
118, 11107–11118, <a href="https://doi.org/10.1002/jgrd.50836" target="_blank">https://doi.org/10.1002/jgrd.50836</a>, 2013.
</mixed-citation></ref-html>
<ref-html id="bib1.bib52"><label>Shao et al.(1996)</label><mixed-citation>
Shao, Y. P., Raupach, M. R., and Leys, J. F.: A model for predicting aeolian
sand drift and dust entrainment on scales from paddock to region, Australian
J. Soil Res., 34, 309–342, <a href="https://doi.org/10.1071/sr9960309" target="_blank">https://doi.org/10.1071/sr9960309</a>, 1996.
</mixed-citation></ref-html>
<ref-html id="bib1.bib53"><label>Tan and Li(2015)</label><mixed-citation>
Tan, M. and Li, X.: Does the Green Great Wall effectively decrease dust storm
intensity in China? A study based on NOAA NDVI and weather station data, Land
Use Policy, 43, 42–47,
<a href="https://doi.org/10.1016/j.landusepol.2014.10.017" target="_blank">https://doi.org/10.1016/j.landusepol.2014.10.017</a>, 2015.
</mixed-citation></ref-html>
<ref-html id="bib1.bib54"><label>Timmermans et al.(2017)</label><mixed-citation>
Timmermans, R., Kranenburg, R., Manders, A., Hendriks, C., Segers, A., Dammers,
E., Zhang, Q., Wang, L., Liu, Z., Zeng, L., Denier van der Gon, H., and
Schaap, M.: Source apportionment of PM<sub>2.5</sub> across China using LOTOS-EUROS,
Atmos. Environ., 164, 370–386, <a href="https://doi.org/10.1016/j.atmosenv.2017.06.003" target="_blank">https://doi.org/10.1016/j.atmosenv.2017.06.003</a>, 2017.
</mixed-citation></ref-html>
<ref-html id="bib1.bib55"><label>TNO(2022)</label><mixed-citation>
TNO: LOTOS-EUROS source code, TNO [code], <a href="https://lotos-euros.tno.nl" target="_blank"/>, last access: 14 May 2022.
</mixed-citation></ref-html>
<ref-html id="bib1.bib56"><label>Winker et al.(2007)</label><mixed-citation>
Winker, D. M., Hunt, W. H., and McGill, M. J.: Initial performance assessment
of CALIOP, Geophys. Res. Lett., 34, L19803,
<a href="https://doi.org/10.1029/2007gl030135" target="_blank">https://doi.org/10.1029/2007gl030135</a>, 2007.
</mixed-citation></ref-html>
<ref-html id="bib1.bib57"><label>World Meteorological Organization(2017)</label><mixed-citation>
World Meteorological Organization: WMO AIRBORNE DUST BULLETIN: Sand and Dust
Storm Warning Advisory and Assessment System,
<a href="https://library.wmo.int/doc_num.php?explnum_id=3416" target="_blank"/> (last access: 14 May 2022), 2017.
</mixed-citation></ref-html>
<ref-html id="bib1.bib58"><label>World Meteorological Organization(2019)</label><mixed-citation>
World Meteorological Organization: WMO AIRBORNE DUST BULLETIN: Sand and Dust
Storm Warning Advisory and Assessment System, Tech. rep.,
<a href="https://public.wmo.int/en/resources/library/wmo-airborne-dust-bulletin-sand-and-dust-storm-warning-advisory-and-assessment" target="_blank">https://public.wmo.int/en/resources/library/wmo-airborne-dust-bulletin-sand-and-dust-storm-warning-advisory-and-assessment</a> (last access: 14 May 2022),
2019.
</mixed-citation></ref-html>
<ref-html id="bib1.bib59"><label>World Meteorological Organization(2020)</label><mixed-citation>
World Meteorological Organization: Sand and Dust Storm Warning Advisory and
Assessment System: Science Progress Report, Tech. rep., World Meteorological
Organization, <a href="https://public.wmo.int/en/resources/library/sand-and-dust-storm-warning-advisory-and-assessment-system-science-progress-report" target="_blank">https://public.wmo.int/en/resources/library/sand-and-dust-storm-warning-advisory-and-assessment-system-science-progress-report</a> (last access: 14 May 2022),
2020.
</mixed-citation></ref-html>
<ref-html id="bib1.bib60"><label>Wu et al.(2016)</label><mixed-citation>
Wu, C., Lin, Z., He, J., Zhang, M., Liu, X., Zhang, R., and Brown, H.: A
process-oriented evaluation of dust emission parameterizations in CESM:
Simulation of a typical severe dust storm in East Asia, J. Adv. Model. Earth Sy., 8, 1432–1452, <a href="https://doi.org/10.1002/2016MS000723" target="_blank">https://doi.org/10.1002/2016MS000723</a>, 2016.
</mixed-citation></ref-html>
<ref-html id="bib1.bib61"><label>Yin et al.(2021)</label><mixed-citation>
Yin, Z., Wan, Y., Zhang, Y., and Wang, H.: Why super sandstorm 2021 in North
China, Natl. Sci. Rev., 9, nwab165, <a href="https://doi.org/10.1093/nsr/nwab165" target="_blank">https://doi.org/10.1093/nsr/nwab165</a>, 2021.
</mixed-citation></ref-html>
<ref-html id="bib1.bib62"><label>Yumimoto and Takemura(2015)</label><mixed-citation>
Yumimoto, K. and Takemura, T.: Long-term inverse modeling of Asian dust:
Interannual variations of its emission, transport, deposition, and radiative
forcing, J. Geophys. Res.-Atmos., 120, 2014JD022390+,
<a href="https://doi.org/10.1002/2014jd022390" target="_blank">https://doi.org/10.1002/2014jd022390</a>, 2015.

</mixed-citation></ref-html>
<ref-html id="bib1.bib63"><label>Yumimoto et al.(2016)</label><mixed-citation>
Yumimoto, K., Nagao, T. M., Kikuchi, M., Sekiyama, T. T., Murakami, H., Tanaka,
T. Y., Ogi, A., Irie, H., Khatri, P., Okumura, H., Arai, K., Morino, I.,
Uchino, O., and Maki, T.: Aerosol data assimilation using data from
Himawari-8, a next-generation geostationary meteorological satellite,
Geophys. Res. Lett., 43, 2016GL069298, <a href="https://doi.org/10.1002/2016gl069298" target="_blank">https://doi.org/10.1002/2016gl069298</a>, 2016.
</mixed-citation></ref-html>
<ref-html id="bib1.bib64"><label>Zender(2003)</label><mixed-citation>
Zender, C. S.: Mineral Dust Entrainment and Deposition (DEAD) model:
Description and 1990s dust climatology, J. Geophys. Res.,
108, 4416, <a href="https://doi.org/10.1029/2002jd002775" target="_blank">https://doi.org/10.1029/2002jd002775</a>, 2003.
</mixed-citation></ref-html>
<ref-html id="bib1.bib65"><label>Zhang et al.(2018)</label><mixed-citation>
Zhang, X.-X., Sharratt, B., Liu, L.-Y., Wang, Z.-F., Pan, X.-L., Lei, J.-Q., Wu, S.-X., Huang, S.-Y., Guo, Y.-H., Li, J., Tang, X., Yang, T., Tian, Y., Chen, X.-S., Hao, J.-Q., Zheng, H.-T., Yang, Y.-Y., and Lyu, Y.-L.: East Asian dust storm in May 2017: observations, modelling, and its influence on the Asia-Pacific region, Atmos. Chem. Phys., 18, 8353–8371, <a href="https://doi.org/10.5194/acp-18-8353-2018" target="_blank">https://doi.org/10.5194/acp-18-8353-2018</a>, 2018.
</mixed-citation></ref-html>
<ref-html id="bib1.bib66"><label>Zhang et al.(2019)</label><mixed-citation>
Zhang, Z., Wu, W., Fan, M., Tao, M., Wei, J., Jin, J., Tan, Y., and Wang, Q.:
Validation of Himawari-8 aerosol optical depth retrievals over China,
Atmos. Environ., 199, 32–44, 2019.
</mixed-citation></ref-html>
<ref-html id="bib1.bib67"><label>Zhou et al.(2008)</label><mixed-citation>
Zhou, C. H., Gong, S. L., Zhang, X. Y., Wang, Y. Q., Niu, T., Liu, H. L., Zhao, T. L., Yang, Y. Q., and Hou, Q.: Development and evaluation of an operational SDS forecasting system for East Asia: CUACE/Dust, Atmos. Chem. Phys., 8, 787–798, <a href="https://doi.org/10.5194/acp-8-787-2008" target="_blank">https://doi.org/10.5194/acp-8-787-2008</a>, 2008.
</mixed-citation></ref-html>--></article>
