<?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">
  <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-19-2135-2019</article-id><title-group><article-title>Mesospheric nitric oxide model from SCIAMACHY data</article-title><alt-title>SCIAMACHY mesosphere NO model</alt-title>
      </title-group><?xmltex \runningtitle{SCIAMACHY mesosphere NO model}?><?xmltex \runningauthor{S. Bender et al.}?>
      <contrib-group>
        <contrib contrib-type="author" corresp="yes" rid="aff1">
          <name><surname>Bender</surname><given-names>Stefan</given-names></name>
          <email>stefan.bender@ntnu.no</email>
        <ext-link>https://orcid.org/0000-0002-8728-053X</ext-link></contrib>
        <contrib contrib-type="author" corresp="no" rid="aff2">
          <name><surname>Sinnhuber</surname><given-names>Miriam</given-names></name>
          
        <ext-link>https://orcid.org/0000-0002-3527-9051</ext-link></contrib>
        <contrib contrib-type="author" corresp="no" rid="aff1">
          <name><surname>Espy</surname><given-names>Patrick J.</given-names></name>
          
        <ext-link>https://orcid.org/0000-0002-0793-6202</ext-link></contrib>
        <contrib contrib-type="author" corresp="no" rid="aff3">
          <name><surname>Burrows</surname><given-names>John P.</given-names></name>
          
        <ext-link>https://orcid.org/0000-0003-1547-8130</ext-link></contrib>
        <aff id="aff1"><label>1</label><institution>Department of Physics, Norwegian University of Science and Technology, Trondheim, Norway</institution>
        </aff>
        <aff id="aff2"><label>2</label><institution>Institute of Meteorology and Climate Research, Karlsruhe Institute of Technology, Karlsruhe, Germany</institution>
        </aff>
        <aff id="aff3"><label>3</label><institution>Institute of Environmental Physics, University of Bremen, Bremen, Germany</institution>
        </aff>
      </contrib-group>
      <author-notes><corresp id="corr1">Stefan Bender (stefan.bender@ntnu.no)</corresp></author-notes><pub-date><day>18</day><month>February</month><year>2019</year></pub-date>
      
      <volume>19</volume>
      <issue>4</issue>
      <fpage>2135</fpage><lpage>2147</lpage>
      <history>
        <date date-type="received"><day>23</day><month>August</month><year>2018</year></date>
           <date date-type="rev-request"><day>22</day><month>October</month><year>2018</year></date>
           <date date-type="rev-recd"><day>31</day><month>January</month><year>2019</year></date>
           <date date-type="accepted"><day>5</day><month>February</month><year>2019</year></date>
      </history>
      <permissions>
        
        
      <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>
    <p id="d1e121">We present an empirical model for nitric oxide (<inline-formula><mml:math id="M1" display="inline"><mml:mrow class="chem"><mml:mi mathvariant="normal">NO</mml:mi></mml:mrow></mml:math></inline-formula>) in the mesosphere
(<inline-formula><mml:math id="M2" display="inline"><mml:mrow><mml:mo>≈</mml:mo><mml:mn mathvariant="normal">60</mml:mn></mml:mrow></mml:math></inline-formula>–90 km) derived from SCIAMACHY (SCanning Imaging Absorption spectroMeter for Atmospheric CHartoghraphY) limb scan data.
This work complements and extends the NOEM
(Nitric Oxide Empirical Model; <xref ref-type="bibr" rid="bib1.bibx45" id="altparen.1"/>) and
SANOMA (SMR Acquired Nitric Oxide Model Atmosphere; <xref ref-type="bibr" rid="bib1.bibx40" id="altparen.2"/>)
empirical models in the lower thermosphere.
The regression ansatz builds on the heritage of studies
by <xref ref-type="bibr" rid="bib1.bibx37" id="text.3"/> and the superposed epoch
analysis by <xref ref-type="bibr" rid="bib1.bibx60" id="text.4"/> which
estimate <inline-formula><mml:math id="M3" display="inline"><mml:mrow class="chem"><mml:mi mathvariant="normal">NO</mml:mi></mml:mrow></mml:math></inline-formula> production from particle precipitation.</p>
    <p id="d1e163">Our model relates the daily (longitudinally) averaged <inline-formula><mml:math id="M4" display="inline"><mml:mrow class="chem"><mml:mi mathvariant="normal">NO</mml:mi></mml:mrow></mml:math></inline-formula> number densities from
SCIAMACHY <xref ref-type="bibr" rid="bib1.bibx15 bib1.bibx14" id="paren.5"/> as a function of geomagnetic latitude
to the solar Lyman-<inline-formula><mml:math id="M5" display="inline"><mml:mi mathvariant="italic">α</mml:mi></mml:math></inline-formula> and the geomagnetic
AE (auroral electrojet) indices.
We use a non-linear regression model, incorporating a finite and seasonally
varying lifetime for the geomagnetically induced <inline-formula><mml:math id="M6" display="inline"><mml:mrow class="chem"><mml:mi mathvariant="normal">NO</mml:mi></mml:mrow></mml:math></inline-formula>.
We estimate the parameters by finding the maximum posterior probability
and calculate the parameter uncertainties using
Markov chain Monte Carlo sampling.
In addition to providing an estimate of the <inline-formula><mml:math id="M7" display="inline"><mml:mrow class="chem"><mml:mi mathvariant="normal">NO</mml:mi></mml:mrow></mml:math></inline-formula> content in the mesosphere,
the regression coefficients indicate regions where certain processes dominate.</p>
  </abstract>
    </article-meta>
  </front>
<body>
      

<sec id="Ch1.S1" sec-type="intro">
  <title>Introduction</title>
      <p id="d1e207">It has been recognized in the past decades that the mesosphere and
stratosphere are coupled in various ways <xref ref-type="bibr" rid="bib1.bibx4" id="paren.6"/>. Consequently,
climate models have been evolving to extend to increasingly higher levels in
the atmosphere to improve the accuracy of medium- and long-term predictions.
Nowadays it is not unusual that these models include the mesosphere
(40–90 km) or the lower thermosphere
(90–120 km) <xref ref-type="bibr" rid="bib1.bibx47" id="paren.7"/>. It is therefore important to
understand the processes in the mesosphere and lower thermosphere and to find
the important drivers of chemistry and dynamics in that region. The
atmosphere above the stratosphere (<inline-formula><mml:math id="M8" display="inline"><mml:mrow><mml:mi mathvariant="italic">≳</mml:mi><mml:mn mathvariant="normal">40</mml:mn></mml:mrow></mml:math></inline-formula> km) is coupled to solar and
geomagnetic activity, also known as space weather <xref ref-type="bibr" rid="bib1.bibx59" id="paren.8"/>.
Electrons and protons from the solar wind and the radiation belts with
sufficient kinetic energy enter the atmosphere in that region. Since as
charged particles they move along the magnetic field, this precipitation
occurs primarily at high geomagnetic latitudes.</p>
      <?pagebreak page2136?><p id="d1e229">Previously the role of <inline-formula><mml:math id="M9" display="inline"><mml:mrow class="chem"><mml:mi mathvariant="normal">NO</mml:mi></mml:mrow></mml:math></inline-formula> in the mesosphere has been
identified as an important free radical and in this sense
a driver of the chemistry <xref ref-type="bibr" rid="bib1.bibx41 bib1.bibx5 bib1.bibx6 bib1.bibx7 bib1.bibx10 bib1.bibx54 bib1.bibx3" id="paren.9"/>,
particularly
during winter when it is long-lived because of reduced photodissociation.
<inline-formula><mml:math id="M10" display="inline"><mml:mrow class="chem"><mml:mi mathvariant="normal">NO</mml:mi></mml:mrow></mml:math></inline-formula> generated in the region between 90 and 120 km at
auroral latitudes is strongly influenced by both solar and
geomagnetic activity <xref ref-type="bibr" rid="bib1.bibx45 bib1.bibx58 bib1.bibx60 bib1.bibx36 bib1.bibx37" id="paren.10"/>.
At high latitudes,
<inline-formula><mml:math id="M11" display="inline"><mml:mrow class="chem"><mml:mi mathvariant="normal">NO</mml:mi></mml:mrow></mml:math></inline-formula> is transported down to the upper stratosphere during winter,
usually down to 50 km and occasionally down to
30 km <xref ref-type="bibr" rid="bib1.bibx62 bib1.bibx51 bib1.bibx26 bib1.bibx32" id="paren.11"/>.
At those altitudes and also in the mesosphere, <inline-formula><mml:math id="M12" display="inline"><mml:mrow class="chem"><mml:mi mathvariant="normal">NO</mml:mi></mml:mrow></mml:math></inline-formula>
participates in the “odd oxygen catalytic cycle which depletes ozone” <xref ref-type="bibr" rid="bib1.bibx19" id="paren.12"/>.
Additional dynamical processes also result in the strong
downward transport of mesospheric air into the upper stratosphere,
such as the strong downwelling that
often occurs in the recovery phase of a sudden stratospheric warming (SSW) (<?xmltex \hack{\mbox\bgroup}?><xref ref-type="bibr" rid="bib1.bibx50" id="altparen.13"/><?xmltex \hack{\egroup}?>; <?xmltex \hack{\mbox\bgroup}?><xref ref-type="bibr" rid="bib1.bibx48" id="altparen.14"/><?xmltex \hack{\egroup}?>).
This downwelling is typically associated with the formation of an elevated stratopause.</p>
      <p id="d1e288">Different instruments have been measuring <inline-formula><mml:math id="M13" display="inline"><mml:mrow class="chem"><mml:mi mathvariant="normal">NO</mml:mi></mml:mrow></mml:math></inline-formula> in the mesosphere and lower
thermosphere, but at different altitudes and at different local times.
Measurements from solar occultation instruments such as Scisat-1/ACE-FTS
or AIM/SOFIE are limited in latitude and local time (sunrise and sunset).
Global observations from sun-synchronously orbiting satellites
are available from
Envisat/MIPAS below 70 km daily and 42–172 km every
10 days <xref ref-type="bibr" rid="bib1.bibx25 bib1.bibx27 bib1.bibx16" id="paren.15"/>; from
Odin/SMR between 45 and 115 km <xref ref-type="bibr" rid="bib1.bibx50 bib1.bibx40" id="paren.16"/>;
or from Envisat/SCIAMACHY (SCanning Imaging Absorption spectroMeter for Atmospheric CHartoghraphY) between 60 and 90 km
daily <xref ref-type="bibr" rid="bib1.bibx15" id="paren.17"/> and
60–160 km every 15 days <xref ref-type="bibr" rid="bib1.bibx13" id="paren.18"/>.
Because the Odin and Envisat orbits are sun-synchronous,
the measurement local times are fixed to around 06:00 and 18:00 (Odin) and 10:00 and 22:00 (Envisat).
While MIPAS has both daytime and night-time measurements,
SCIAMACHY provides daytime (10:00)
data because of the measurement principle (fluorescent UV scattering;
see <xref ref-type="bibr" rid="bib1.bibx13 bib1.bibx15" id="altparen.19"/>).
Unfortunately, Envisat stopped communicating in April 2012 ,and therefore
the data available from MIPAS and SCIAMACHY are limited to nearly
10 years from August 2002 to April 2012.
The other aforementioned instruments are still operational and
provide ongoing data as long as satellite operations continue.</p>
      <p id="d1e315">Chemistry–climate models struggle to simulate the
<inline-formula><mml:math id="M14" display="inline"><mml:mrow class="chem"><mml:mi mathvariant="normal">NO</mml:mi></mml:mrow></mml:math></inline-formula> amounts and distributions in the mesosphere and lower thermosphere
(see, for example, <xref ref-type="bibr" rid="bib1.bibx34 bib1.bibx52 bib1.bibx48 bib1.bibx38" id="altparen.20"/>).
To remedy the situation,
some models constrain the NO content at their top layer
through observation-based parametrizations.
For example, the next generation of climate simulations
(CMIP6; see <xref ref-type="bibr" rid="bib1.bibx47" id="altparen.21"/>) and other recent model
simulations <xref ref-type="bibr" rid="bib1.bibx61" id="paren.22"/> parametrizes particle effects
as derived partly from Envisat/MIPAS <inline-formula><mml:math id="M15" display="inline"><mml:mrow class="chem"><mml:mi mathvariant="normal">NO</mml:mi></mml:mrow></mml:math></inline-formula> measurements <xref ref-type="bibr" rid="bib1.bibx33" id="paren.23"/>.</p>
<sec id="Ch1.S1.SSx1" specific-use="unnumbered">
  <?xmltex \opttitle{{$\protect\chem{NO}$} in the mesosphere and lower thermosphere}?><title><inline-formula><mml:math id="M16" display="inline"><mml:mrow class="chem"><mml:mi mathvariant="normal">NO</mml:mi></mml:mrow></mml:math></inline-formula> in the mesosphere and lower thermosphere</title>
      <p id="d1e361"><inline-formula><mml:math id="M17" display="inline"><mml:mrow class="chem"><mml:mi mathvariant="normal">NO</mml:mi></mml:mrow></mml:math></inline-formula> in the mesosphere and lower thermosphere is produced by
<inline-formula><mml:math id="M18" display="inline"><mml:mrow class="chem"><mml:msub><mml:mi mathvariant="normal">N</mml:mi><mml:mn mathvariant="normal">2</mml:mn></mml:msub></mml:mrow></mml:math></inline-formula> dissociation,</p>
      <p id="d1e382">

                <disp-formula id="Ch1.E1" content-type="numbered reaction"><mml:math id="M19" display="block"><mml:mrow><mml:mstyle displaystyle="true" class="stylechange"/><mml:mrow class="chem"><mml:msub><mml:mi mathvariant="normal">N</mml:mi><mml:mn mathvariant="normal">2</mml:mn></mml:msub></mml:mrow><mml:mo>+</mml:mo><mml:mi>h</mml:mi><mml:mi mathvariant="italic">ν</mml:mi><mml:mo>→</mml:mo><mml:mrow class="chem"><mml:mi mathvariant="normal">N</mml:mi><mml:msup><mml:mo>(</mml:mo><mml:mn mathvariant="normal">2</mml:mn></mml:msup><mml:mi mathvariant="normal">D</mml:mi><mml:mo>)</mml:mo></mml:mrow><mml:mo>+</mml:mo><mml:mrow class="chem"><mml:mi mathvariant="normal">N</mml:mi><mml:msup><mml:mo>(</mml:mo><mml:mn mathvariant="normal">4</mml:mn></mml:msup><mml:mi mathvariant="normal">S</mml:mi><mml:mo>)</mml:mo></mml:mrow><mml:mspace linebreak="nobreak" width="1em"/><mml:mo>(</mml:mo><mml:mi mathvariant="italic">λ</mml:mi><mml:mo>&lt;</mml:mo><mml:mn mathvariant="normal">102</mml:mn><mml:mspace linebreak="nobreak" width="0.125em"/><mml:mi mathvariant="normal">nm</mml:mi><mml:mo>)</mml:mo><mml:mspace width="0.25em" linebreak="nobreak"/><mml:mo>,</mml:mo></mml:mrow></mml:math></disp-formula>

          followed by the reaction of the excited nitrogen atom <inline-formula><mml:math id="M20" display="inline"><mml:mrow class="chem"><mml:mi mathvariant="normal">N</mml:mi><mml:msup><mml:mo>(</mml:mo><mml:mn mathvariant="normal">2</mml:mn></mml:msup><mml:mi mathvariant="normal">D</mml:mi><mml:mo>)</mml:mo></mml:mrow></mml:math></inline-formula>
with molecular oxygen <xref ref-type="bibr" rid="bib1.bibx63 bib1.bibx5 bib1.bibx6" id="paren.24"/>:

                <disp-formula id="Ch1.E2" content-type="numbered reaction"><mml:math id="M21" display="block"><mml:mrow><mml:mstyle class="stylechange" displaystyle="true"/><mml:mrow class="chem"><mml:mi mathvariant="normal">N</mml:mi><mml:msup><mml:mo>(</mml:mo><mml:mn mathvariant="normal">2</mml:mn></mml:msup><mml:mi mathvariant="normal">D</mml:mi><mml:mo>)</mml:mo></mml:mrow><mml:mo>+</mml:mo><mml:mrow class="chem"><mml:msub><mml:mi mathvariant="normal">O</mml:mi><mml:mn mathvariant="normal">2</mml:mn></mml:msub></mml:mrow><mml:mo>→</mml:mo><mml:mrow class="chem"><mml:mi mathvariant="normal">NO</mml:mi></mml:mrow><mml:mo>+</mml:mo><mml:mrow class="chem"><mml:mi mathvariant="normal">O</mml:mi></mml:mrow><mml:mspace linebreak="nobreak" width="0.25em"/><mml:mo>.</mml:mo></mml:mrow></mml:math></disp-formula>

          The dissociation energy of <inline-formula><mml:math id="M22" display="inline"><mml:mrow class="chem"><mml:msub><mml:mi mathvariant="normal">N</mml:mi><mml:mn mathvariant="normal">2</mml:mn></mml:msub></mml:mrow></mml:math></inline-formula>
into ground state atoms <inline-formula><mml:math id="M23" display="inline"><mml:mrow class="chem"><mml:mi mathvariant="normal">N</mml:mi><mml:msup><mml:mo>(</mml:mo><mml:mn mathvariant="normal">4</mml:mn></mml:msup><mml:mi mathvariant="normal">S</mml:mi><mml:mo>)</mml:mo></mml:mrow></mml:math></inline-formula>
is about 9.8 eV
(<inline-formula><mml:math id="M24" display="inline"><mml:mrow><mml:mi mathvariant="italic">λ</mml:mi><mml:mo>≈</mml:mo><mml:mn mathvariant="normal">127</mml:mn></mml:mrow></mml:math></inline-formula> nm) <xref ref-type="bibr" rid="bib1.bibx39 bib1.bibx24 bib1.bibx35" id="paren.25"/>.
This energy together with the excitation energy to <inline-formula><mml:math id="M25" display="inline"><mml:mrow class="chem"><mml:mi mathvariant="normal">N</mml:mi><mml:msup><mml:mo>(</mml:mo><mml:mn mathvariant="normal">2</mml:mn></mml:msup><mml:mi mathvariant="normal">D</mml:mi><mml:mo>)</mml:mo></mml:mrow></mml:math></inline-formula>
is denoted by <inline-formula><mml:math id="M26" display="inline"><mml:mrow><mml:mi>h</mml:mi><mml:mi mathvariant="italic">ν</mml:mi></mml:mrow></mml:math></inline-formula> in Reaction (<xref ref-type="disp-formula" rid="Ch1.E1"/>) and
can be provided by a number of sources,
most notably by auroral or photoelectrons
as well as by soft solar X-rays.</p>
      <p id="d1e584">The <inline-formula><mml:math id="M27" display="inline"><mml:mrow class="chem"><mml:mi mathvariant="normal">NO</mml:mi></mml:mrow></mml:math></inline-formula> content is reduced by photodissociation,

                <disp-formula id="Ch1.E3" content-type="numbered reaction"><mml:math id="M28" display="block"><mml:mrow><mml:mstyle class="stylechange" displaystyle="true"/><mml:mrow class="chem"><mml:mi mathvariant="normal">NO</mml:mi></mml:mrow><mml:mo>+</mml:mo><mml:mi>h</mml:mi><mml:mi mathvariant="italic">ν</mml:mi><mml:mo>→</mml:mo><mml:mrow class="chem"><mml:mi mathvariant="normal">N</mml:mi></mml:mrow><mml:mo>+</mml:mo><mml:mrow class="chem"><mml:mi mathvariant="normal">O</mml:mi></mml:mrow><mml:mspace linebreak="nobreak" width="1em"/><mml:mo>(</mml:mo><mml:mi mathvariant="italic">λ</mml:mi><mml:mo>&lt;</mml:mo><mml:mn mathvariant="normal">191</mml:mn><mml:mspace linebreak="nobreak" width="0.125em"/><mml:mi mathvariant="normal">nm</mml:mi><mml:mo>)</mml:mo><mml:mspace linebreak="nobreak" width="0.25em"/><mml:mo>,</mml:mo></mml:mrow></mml:math></disp-formula>

          by photoionization,

                <disp-formula id="Ch1.E4" content-type="numbered reaction"><mml:math id="M29" display="block"><mml:mrow><mml:mstyle class="stylechange" displaystyle="true"/><mml:mrow class="chem"><mml:mi mathvariant="normal">NO</mml:mi></mml:mrow><mml:mo>+</mml:mo><mml:mi>h</mml:mi><mml:mi mathvariant="italic">ν</mml:mi><mml:mo>→</mml:mo><mml:msup><mml:mrow class="chem"><mml:mi mathvariant="normal">NO</mml:mi></mml:mrow><mml:mo>+</mml:mo></mml:msup><mml:mo>+</mml:mo><mml:msup><mml:mi>e</mml:mi><mml:mo>-</mml:mo></mml:msup><mml:mspace width="1em" linebreak="nobreak"/><mml:mo>(</mml:mo><mml:mi mathvariant="italic">λ</mml:mi><mml:mo>&lt;</mml:mo><mml:mn mathvariant="normal">134</mml:mn><mml:mspace width="0.125em" linebreak="nobreak"/><mml:mi mathvariant="normal">nm</mml:mi><mml:mo>)</mml:mo><mml:mspace linebreak="nobreak" width="0.25em"/><mml:mo>,</mml:mo></mml:mrow></mml:math></disp-formula>

          and by reacting with atomic nitrogen,

                <disp-formula id="Ch1.E5" content-type="numbered reaction"><mml:math id="M30" display="block"><mml:mrow><mml:mstyle class="stylechange" displaystyle="true"/><mml:mrow class="chem"><mml:mi mathvariant="normal">NO</mml:mi></mml:mrow><mml:mo>+</mml:mo><mml:mrow class="chem"><mml:mi mathvariant="normal">N</mml:mi></mml:mrow><mml:mo>→</mml:mo><mml:mrow class="chem"><mml:msub><mml:mi mathvariant="normal">N</mml:mi><mml:mn mathvariant="normal">2</mml:mn></mml:msub></mml:mrow><mml:mo>+</mml:mo><mml:mrow class="chem"><mml:mi mathvariant="normal">O</mml:mi></mml:mrow><mml:mspace width="0.25em" linebreak="nobreak"/><mml:mo>.</mml:mo></mml:mrow></mml:math></disp-formula>

          <inline-formula><mml:math id="M31" display="inline"><mml:mrow class="chem"><mml:msub><mml:mi mathvariant="normal">N</mml:mi><mml:mn mathvariant="normal">2</mml:mn></mml:msub><mml:mi mathvariant="normal">O</mml:mi></mml:mrow></mml:math></inline-formula> has been retrieved in the mesosphere and thermosphere
from MIPAS (see, e.g. <xref ref-type="bibr" rid="bib1.bibx29 bib1.bibx28" id="altparen.26"/>) and from
Scisat-1/ACE-FTS <xref ref-type="bibr" rid="bib1.bibx57" id="paren.27"/>.
Model–measurement studies by <xref ref-type="bibr" rid="bib1.bibx56" id="text.28"/> attributed
the source of this <inline-formula><mml:math id="M32" display="inline"><mml:mrow class="chem"><mml:msub><mml:mi mathvariant="normal">N</mml:mi><mml:mn mathvariant="normal">2</mml:mn></mml:msub><mml:mi mathvariant="normal">O</mml:mi></mml:mrow></mml:math></inline-formula>
to being most likely the reaction
between <inline-formula><mml:math id="M33" display="inline"><mml:mrow class="chem"><mml:msub><mml:mi mathvariant="normal">NO</mml:mi><mml:mn mathvariant="normal">2</mml:mn></mml:msub></mml:mrow></mml:math></inline-formula> and <inline-formula><mml:math id="M34" display="inline"><mml:mrow class="chem"><mml:mi mathvariant="normal">N</mml:mi></mml:mrow></mml:math></inline-formula> atoms produced by
particle precipitation:

                <disp-formula id="Ch1.E6" content-type="numbered reaction"><mml:math id="M35" display="block"><mml:mrow><mml:mstyle displaystyle="true" class="stylechange"/><mml:mrow class="chem"><mml:mi mathvariant="normal">N</mml:mi></mml:mrow><mml:mo>+</mml:mo><mml:mrow class="chem"><mml:msub><mml:mi mathvariant="normal">NO</mml:mi><mml:mn mathvariant="normal">2</mml:mn></mml:msub></mml:mrow><mml:mo>→</mml:mo><mml:mrow class="chem"><mml:msub><mml:mi mathvariant="normal">N</mml:mi><mml:mn mathvariant="normal">2</mml:mn></mml:msub><mml:mi mathvariant="normal">O</mml:mi></mml:mrow><mml:mo>+</mml:mo><mml:mrow class="chem"><mml:mi mathvariant="normal">O</mml:mi></mml:mrow><mml:mspace linebreak="nobreak" width="0.25em"/><mml:mo>.</mml:mo></mml:mrow></mml:math></disp-formula>

          We note that photo-excitation and photolysis at 185 nm (vacuum UV) of
<inline-formula><mml:math id="M36" display="inline"><mml:mrow class="chem"><mml:mi mathvariant="normal">NO</mml:mi></mml:mrow></mml:math></inline-formula> or <inline-formula><mml:math id="M37" display="inline"><mml:mrow class="chem"><mml:msub><mml:mi mathvariant="normal">NO</mml:mi><mml:mn mathvariant="normal">2</mml:mn></mml:msub></mml:mrow></mml:math></inline-formula> mixtures in nitrogen, <inline-formula><mml:math id="M38" display="inline"><mml:mrow class="chem"><mml:msub><mml:mi mathvariant="normal">N</mml:mi><mml:mn mathvariant="normal">2</mml:mn></mml:msub></mml:mrow></mml:math></inline-formula>, or helium mixtures at
1 atm leads to <inline-formula><mml:math id="M39" display="inline"><mml:mrow class="chem"><mml:msub><mml:mi mathvariant="normal">N</mml:mi><mml:mn mathvariant="normal">2</mml:mn></mml:msub><mml:mi mathvariant="normal">O</mml:mi></mml:mrow></mml:math></inline-formula> formation <xref ref-type="bibr" rid="bib1.bibx44" id="paren.29"/>.
Both mechanisms explaining the production of <inline-formula><mml:math id="M40" display="inline"><mml:mrow class="chem"><mml:msub><mml:mi mathvariant="normal">N</mml:mi><mml:mn mathvariant="normal">2</mml:mn></mml:msub><mml:mi mathvariant="normal">O</mml:mi></mml:mrow></mml:math></inline-formula> involve excited states
of <inline-formula><mml:math id="M41" display="inline"><mml:mrow class="chem"><mml:mi mathvariant="normal">NO</mml:mi></mml:mrow></mml:math></inline-formula>. Hence these pathways contribute to the loss of <inline-formula><mml:math id="M42" display="inline"><mml:mrow class="chem"><mml:mi mathvariant="normal">NO</mml:mi></mml:mrow></mml:math></inline-formula> and
potentially an additional daytime source of <inline-formula><mml:math id="M43" display="inline"><mml:mrow class="chem"><mml:msub><mml:mi mathvariant="normal">N</mml:mi><mml:mn mathvariant="normal">2</mml:mn></mml:msub><mml:mi mathvariant="normal">O</mml:mi></mml:mrow></mml:math></inline-formula> in the upper atmosphere.
<inline-formula><mml:math id="M44" display="inline"><mml:mrow class="chem"><mml:msub><mml:mi mathvariant="normal">N</mml:mi><mml:mn mathvariant="normal">2</mml:mn></mml:msub><mml:mi mathvariant="normal">O</mml:mi></mml:mrow></mml:math></inline-formula> acts as an intermediate reservoir at high altitudes
(<inline-formula><mml:math id="M45" display="inline"><mml:mrow><mml:mi mathvariant="italic">⪆</mml:mi><mml:mn mathvariant="normal">90</mml:mn></mml:mrow></mml:math></inline-formula> km; see <xref ref-type="bibr" rid="bib1.bibx57" id="altparen.30"/>),
reacting with <inline-formula><mml:math id="M46" display="inline"><mml:mrow class="chem"><mml:mi mathvariant="normal">O</mml:mi></mml:mrow></mml:math></inline-formula>(<inline-formula><mml:math id="M47" display="inline"><mml:msup><mml:mi/><mml:mn mathvariant="normal">1</mml:mn></mml:msup></mml:math></inline-formula>D) in two well-known channels
to <inline-formula><mml:math id="M48" display="inline"><mml:mrow class="chem"><mml:msub><mml:mi mathvariant="normal">N</mml:mi><mml:mn mathvariant="normal">2</mml:mn></mml:msub></mml:mrow></mml:math></inline-formula> and <inline-formula><mml:math id="M49" display="inline"><mml:mrow class="chem"><mml:msub><mml:mi mathvariant="normal">O</mml:mi><mml:mn mathvariant="normal">2</mml:mn></mml:msub></mml:mrow></mml:math></inline-formula> as well as to 2<inline-formula><mml:math id="M50" display="inline"><mml:mrow class="chem"><mml:mi mathvariant="normal">NO</mml:mi></mml:mrow></mml:math></inline-formula>.
However, the largest <inline-formula><mml:math id="M51" display="inline"><mml:mrow class="chem"><mml:msub><mml:mi mathvariant="normal">N</mml:mi><mml:mn mathvariant="normal">2</mml:mn></mml:msub><mml:mi mathvariant="normal">O</mml:mi></mml:mrow></mml:math></inline-formula> abundances are located below 60 km
and originate primarily from the transport of tropospheric <inline-formula><mml:math id="M52" display="inline"><mml:mrow class="chem"><mml:msub><mml:mi mathvariant="normal">N</mml:mi><mml:mn mathvariant="normal">2</mml:mn></mml:msub><mml:mi mathvariant="normal">O</mml:mi></mml:mrow></mml:math></inline-formula>
into the stratosphere through the Brewer–Dobson
circulation <xref ref-type="bibr" rid="bib1.bibx28 bib1.bibx29 bib1.bibx57" id="paren.31"/>
but can reach up
to 70 km in geomagnetic storm conditions <xref ref-type="bibr" rid="bib1.bibx28 bib1.bibx57" id="paren.32"/>.
Both source and sink reactions indicate that
<inline-formula><mml:math id="M53" display="inline"><mml:mrow class="chem"><mml:mi mathvariant="normal">NO</mml:mi></mml:mrow></mml:math></inline-formula> behaves differently in sunlit conditions than
in dark conditions.
<inline-formula><mml:math id="M54" display="inline"><mml:mrow class="chem"><mml:mi mathvariant="normal">NO</mml:mi></mml:mrow></mml:math></inline-formula> is produced by particle precipitation at auroral latitudes,
but in dark conditions (without photolysis) it is only depleted
by reacting with atomic nitrogen (Reaction <xref ref-type="disp-formula" rid="Ch1.E5"/>).
This asymmetry between production and depletion
in dark conditions results in different lifetimes
of <inline-formula><mml:math id="M55" display="inline"><mml:mrow class="chem"><mml:mi mathvariant="normal">NO</mml:mi></mml:mrow></mml:math></inline-formula>.</p>
      <?pagebreak page2137?><p id="d1e1034">Early work to parametrize <inline-formula><mml:math id="M56" display="inline"><mml:mrow class="chem"><mml:mi mathvariant="normal">NO</mml:mi></mml:mrow></mml:math></inline-formula> in the lower thermosphere
(100–150 km) used SNOE measurements
from March 1998 to September 2000 <xref ref-type="bibr" rid="bib1.bibx45" id="paren.33"/>.
With these 2.5 years of data and using empirical orthogonal functions,
the so-called NOEM (Nitric Oxide Empirical Model) estimates
<inline-formula><mml:math id="M57" display="inline"><mml:mrow class="chem"><mml:mi mathvariant="normal">NO</mml:mi></mml:mrow></mml:math></inline-formula> in the lower thermosphere as a function of the solar
<inline-formula><mml:math id="M58" display="inline"><mml:mrow><mml:msub><mml:mi>f</mml:mi><mml:mrow><mml:mn mathvariant="normal">10.7</mml:mn><mml:mspace width="0.125em" linebreak="nobreak"/><mml:mi mathvariant="normal">cm</mml:mi></mml:mrow></mml:msub></mml:mrow></mml:math></inline-formula> radio flux, the solar declination angle, and
the planetary Kp index.
NOEM is still used as prior input for
<inline-formula><mml:math id="M59" display="inline"><mml:mrow class="chem"><mml:mi mathvariant="normal">NO</mml:mi></mml:mrow></mml:math></inline-formula> retrieval, for example, from MIPAS <xref ref-type="bibr" rid="bib1.bibx16 bib1.bibx30" id="paren.34"/>
and SCIAMACHY <xref ref-type="bibr" rid="bib1.bibx15" id="paren.35"/> spectra.
However, 2.5 years is relatively short compared to the 11-year
solar cycle, and the years 1998 to 2000 encompass a period of
elevated solar activity.
To address this, a longer time series from AIM/SOFIE was used to determine the
important drivers of <inline-formula><mml:math id="M60" display="inline"><mml:mrow class="chem"><mml:mi mathvariant="normal">NO</mml:mi></mml:mrow></mml:math></inline-formula> in the lower thermosphere (90–140 km)
by <xref ref-type="bibr" rid="bib1.bibx37" id="text.36"/>.
Other recent work uses 10 years of <inline-formula><mml:math id="M61" display="inline"><mml:mrow class="chem"><mml:mi mathvariant="normal">NO</mml:mi></mml:mrow></mml:math></inline-formula> data from Odin/SMR
from 85 to 115 km <xref ref-type="bibr" rid="bib1.bibx40" id="paren.37"/>.
<xref ref-type="bibr" rid="bib1.bibx33" id="text.38"/> derived a semi-empirical model of <inline-formula><mml:math id="M62" display="inline"><mml:mrow class="chem"><mml:msub><mml:mi mathvariant="normal">NO</mml:mi><mml:mi>y</mml:mi></mml:msub></mml:mrow></mml:math></inline-formula>
in the stratosphere and mesosphere from MIPAS data.
Here we use Envisat/SCIAMACHY <inline-formula><mml:math id="M63" display="inline"><mml:mrow class="chem"><mml:mi mathvariant="normal">NO</mml:mi></mml:mrow></mml:math></inline-formula> data from the nominal limb
mode <xref ref-type="bibr" rid="bib1.bibx15 bib1.bibx14" id="paren.39"/>.
Apart from providing a similarly long time series of <inline-formula><mml:math id="M64" display="inline"><mml:mrow class="chem"><mml:mi mathvariant="normal">NO</mml:mi></mml:mrow></mml:math></inline-formula> data,
the nominal Envisat/SCIAMACHY <inline-formula><mml:math id="M65" display="inline"><mml:mrow class="chem"><mml:mi mathvariant="normal">NO</mml:mi></mml:mrow></mml:math></inline-formula> data cover the mesosphere
from 60 to 90 km <xref ref-type="bibr" rid="bib1.bibx15" id="paren.40"/>, bridging the gap between
the stratosphere and lower thermosphere models.</p>
      <p id="d1e1155">The paper is organized as follows: we present the data used
in this work in Sect. <xref ref-type="sec" rid="Ch1.S2"/>.
The two model variants, linear and non-linear, are described in
Sect. <xref ref-type="sec" rid="Ch1.S3"/>.
Details about the parameter and uncertainty estimation are explained
in Sect. <xref ref-type="sec" rid="Ch1.S4"/>, and we present the results
in Sect. <xref ref-type="sec" rid="Ch1.S5"/>.
Finally we conclude our findings in Sect. <xref ref-type="sec" rid="Ch1.S6"/>.</p>
</sec>
</sec>
<sec id="Ch1.S2">
  <title>Data</title>
<sec id="Ch1.S2.SS1">
  <?xmltex \opttitle{SCIAMACHY {$\protect\chem{NO}$}}?><title>SCIAMACHY <inline-formula><mml:math id="M66" display="inline"><mml:mrow class="chem"><mml:mi mathvariant="normal">NO</mml:mi></mml:mrow></mml:math></inline-formula></title>
      <p id="d1e1188">We use the SCIAMACHY nitric oxide data set version 6.2.1 <xref ref-type="bibr" rid="bib1.bibx14" id="paren.41"/>
retrieved
from the nominal limb scan mode (<inline-formula><mml:math id="M67" display="inline"><mml:mrow><mml:mo>≈</mml:mo><mml:mn mathvariant="normal">0</mml:mn></mml:mrow></mml:math></inline-formula>–93 km).
For a detailed instrument description, see <xref ref-type="bibr" rid="bib1.bibx18" id="text.42"/> and <xref ref-type="bibr" rid="bib1.bibx17" id="text.43"/>,
and for details of the retrieval algorithm, see <xref ref-type="bibr" rid="bib1.bibx13 bib1.bibx15" id="text.44"/>.</p>
      <p id="d1e1213">The data were retrieved for the whole Envisat period
(August 2002–April 2012).
This satellite was orbiting in a sun-synchronous orbit
at around 800 km altitude, with Equator crossing times
of 10:00 and 22:00 local time.
The <inline-formula><mml:math id="M68" display="inline"><mml:mrow class="chem"><mml:mi mathvariant="normal">NO</mml:mi></mml:mrow></mml:math></inline-formula> number densities from the SCIAMACHY nominal mode
were retrieved from the <inline-formula><mml:math id="M69" display="inline"><mml:mrow class="chem"><mml:mi mathvariant="normal">NO</mml:mi></mml:mrow></mml:math></inline-formula> gamma band emissions.
Since those emissions are fluorescent emissions excited by solar UV,
SCIAMACHY <inline-formula><mml:math id="M70" display="inline"><mml:mrow class="chem"><mml:mi mathvariant="normal">NO</mml:mi></mml:mrow></mml:math></inline-formula> data are only available for the 10:00
dayside (downleg) part of the orbit.
Furthermore, the retrieval was carried out for altitudes from 60
to 160 km, but above approximately 90 km, the data reflect the
scaled a priori densities from NOEM <xref ref-type="bibr" rid="bib1.bibx15" id="paren.45"/>.
We therefore restrict the modelling to the mesosphere below 90 km.</p>
      <p id="d1e1243">We averaged the individual orbital data longitudinally on a daily basis
according to their geomagnetic latitude within 10<inline-formula><mml:math id="M71" display="inline"><mml:msup><mml:mi/><mml:mo>∘</mml:mo></mml:msup></mml:math></inline-formula> bins.
The geomagnetic latitude was determined according to the
eccentric dipole approximation of the
12th generation of the International Geomagnetic Reference Field
(IGRF12) <xref ref-type="bibr" rid="bib1.bibx65" id="paren.46"/>.
In the vertical direction the original retrieval
grid altitudes (2 km bins) were used.
Note that mesospheric <inline-formula><mml:math id="M72" display="inline"><mml:mrow class="chem"><mml:mi mathvariant="normal">NO</mml:mi></mml:mrow></mml:math></inline-formula> concentrations are related to geomagnetically
as well as geographically based processes, but disentangling them is beyond the
scope of the paper.
Follow-up studies can build on the method presented here and study,
for example, longitudinally resolved time series.</p>
      <p id="d1e1266">The measurement sensitivity is taken into account via the
averaging kernel diagonal elements, and days where its
binned average was below 0.002 were excluded from the time series.
Considering this criterion, each bin
(geomagnetic latitude and altitude)
contains about 3400 data points.</p>
</sec>
<sec id="Ch1.S2.SS2">
  <title>Proxies</title>
      <p id="d1e1275">We use two proxies to model the <inline-formula><mml:math id="M73" display="inline"><mml:mrow class="chem"><mml:mi mathvariant="normal">NO</mml:mi></mml:mrow></mml:math></inline-formula> number densities,
one accounting for the solar irradiance variations and
one accounting for the geomagnetic activity.
Various proxies have been used or proposed to account for the
solar-irradiance-induced variations in mesospheric–thermospheric <inline-formula><mml:math id="M74" display="inline"><mml:mrow class="chem"><mml:mi mathvariant="normal">NO</mml:mi></mml:mrow></mml:math></inline-formula>,
which are in particular related to the 11-year solar cycle.
The NOEM (Nitric Oxide Empirical Model; <xref ref-type="bibr" rid="bib1.bibx45" id="altparen.47"/>) uses
the natural logarithm of the solar 10.7 cm radio flux <inline-formula><mml:math id="M75" display="inline"><mml:mrow><mml:msub><mml:mi>f</mml:mi><mml:mn mathvariant="normal">10.7</mml:mn></mml:msub></mml:mrow></mml:math></inline-formula>.
More recent work on AIM/SOFIE <inline-formula><mml:math id="M76" display="inline"><mml:mrow class="chem"><mml:mi mathvariant="normal">NO</mml:mi></mml:mrow></mml:math></inline-formula> <xref ref-type="bibr" rid="bib1.bibx37" id="paren.48"/>
uses the solar Lyman-<inline-formula><mml:math id="M77" display="inline"><mml:mi mathvariant="italic">α</mml:mi></mml:math></inline-formula> index
because some of the main production and loss processes are driven by UV photons.
Besides accounting for the long-term variation of <inline-formula><mml:math id="M78" display="inline"><mml:mrow class="chem"><mml:mi mathvariant="normal">NO</mml:mi></mml:mrow></mml:math></inline-formula> with
solar activity, the Lyman-<inline-formula><mml:math id="M79" display="inline"><mml:mi mathvariant="italic">α</mml:mi></mml:math></inline-formula> index also includes short-term
UV variations and the associated <inline-formula><mml:math id="M80" display="inline"><mml:mrow class="chem"><mml:mi mathvariant="normal">NO</mml:mi></mml:mrow></mml:math></inline-formula> production,
for example, caused by solar flares.
<xref ref-type="bibr" rid="bib1.bibx8" id="text.49"/> have shown that the Lyman-<inline-formula><mml:math id="M81" display="inline"><mml:mi mathvariant="italic">α</mml:mi></mml:math></inline-formula> index
directly relates to the observed <inline-formula><mml:math id="M82" display="inline"><mml:mrow class="chem"><mml:mi mathvariant="normal">NO</mml:mi></mml:mrow></mml:math></inline-formula>
at low latitudes (30<inline-formula><mml:math id="M83" display="inline"><mml:msup><mml:mi/><mml:mo>∘</mml:mo></mml:msup></mml:math></inline-formula> S–30<inline-formula><mml:math id="M84" display="inline"><mml:msup><mml:mi/><mml:mo>∘</mml:mo></mml:msup></mml:math></inline-formula> N).
Thus we use it in this work as a proxy for <inline-formula><mml:math id="M85" display="inline"><mml:mrow class="chem"><mml:mi mathvariant="normal">NO</mml:mi></mml:mrow></mml:math></inline-formula>.</p>
      <p id="d1e1395">In the same manner as for the irradiance variations, the “right”
geomagnetic index to model particle-induced variations of <inline-formula><mml:math id="M86" display="inline"><mml:mrow class="chem"><mml:mi mathvariant="normal">NO</mml:mi></mml:mrow></mml:math></inline-formula>
is a matter of opinion.
Kp is the oldest and most commonly used geomagnetic index;
it was, for example, used in earlier work by <xref ref-type="bibr" rid="bib1.bibx45" id="text.50"/>
for modelling <inline-formula><mml:math id="M87" display="inline"><mml:mrow class="chem"><mml:mi mathvariant="normal">NO</mml:mi></mml:mrow></mml:math></inline-formula> in the mesosphere and lower thermosphere.
Kp is derived from magnetometer stations distributed at
different latitudes and mostly in the Northern Hemisphere (NH).
However, <xref ref-type="bibr" rid="bib1.bibx36" id="text.51"/> found that the auroral
electrojet index (AE) <xref ref-type="bibr" rid="bib1.bibx20" id="paren.52"/>
correlated better with SOFIE-derived <inline-formula><mml:math id="M88" display="inline"><mml:mrow class="chem"><mml:mi mathvariant="normal">NO</mml:mi></mml:mrow></mml:math></inline-formula>
concentrations (<xref ref-type="bibr" rid="bib1.bibx36 bib1.bibx37" id="altparen.53"/>; see also <xref ref-type="bibr" rid="bib1.bibx60" id="altparen.54"/>).
The AE index is derived from stations distributed almost evenly
within the auroral latitude band.
This distribution enables the AE index to be more closely
related to the energy input into the atmosphere at these latitudes.
Therefore, we use the auroral electrojet index (AE)
as a proxy for geomagnetically induced <inline-formula><mml:math id="M89" display="inline"><mml:mrow class="chem"><mml:mi mathvariant="normal">NO</mml:mi></mml:mrow></mml:math></inline-formula>.
To account for the 10:00 satellite sampling,
we average the hourly AE index from noon
the day before to noon on the measurement day.</p>
      <?pagebreak page2138?><p id="d1e1446">It should be noted that tests using Kp (or its linear equivalent, Ap) instead of AE
and using <inline-formula><mml:math id="M90" display="inline"><mml:mrow><mml:msub><mml:mi>f</mml:mi><mml:mn mathvariant="normal">10.7</mml:mn></mml:msub></mml:mrow></mml:math></inline-formula> instead of Lyman-<inline-formula><mml:math id="M91" display="inline"><mml:mi mathvariant="italic">α</mml:mi></mml:math></inline-formula>
suggested that the particular
choice of index did not lead to significantly different results.
Our choice of AE rather than Kp
and Lyman-<inline-formula><mml:math id="M92" display="inline"><mml:mi mathvariant="italic">α</mml:mi></mml:math></inline-formula> over <inline-formula><mml:math id="M93" display="inline"><mml:mrow><mml:msub><mml:mi>f</mml:mi><mml:mn mathvariant="normal">10.7</mml:mn></mml:msub></mml:mrow></mml:math></inline-formula>
is physically based and motivated
as described above.</p>
</sec>
</sec>
<sec id="Ch1.S3">
  <title>Regression model</title>
      <p id="d1e1492">We denote the number density by <inline-formula><mml:math id="M94" display="inline"><mml:mrow><mml:msub><mml:mi>x</mml:mi><mml:mi mathvariant="normal">NO</mml:mi></mml:msub></mml:mrow></mml:math></inline-formula>
as a function of
the (geomagnetic; see Sect. <xref ref-type="sec" rid="Ch1.S2.SS1"/>) latitude <inline-formula><mml:math id="M95" display="inline"><mml:mi mathvariant="italic">ϕ</mml:mi></mml:math></inline-formula>,
the altitude <inline-formula><mml:math id="M96" display="inline"><mml:mi>z</mml:mi></mml:math></inline-formula>,
and the time (measurement day) <inline-formula><mml:math id="M97" display="inline"><mml:mi>t</mml:mi></mml:math></inline-formula>:
<inline-formula><mml:math id="M98" display="inline"><mml:mrow><mml:msub><mml:mi>x</mml:mi><mml:mi mathvariant="normal">NO</mml:mi></mml:msub><mml:mo>(</mml:mo><mml:mi mathvariant="italic">ϕ</mml:mi><mml:mo>,</mml:mo><mml:mi>z</mml:mi><mml:mo>,</mml:mo><mml:mi>t</mml:mi><mml:mo>)</mml:mo></mml:mrow></mml:math></inline-formula>.
In the following we often drop the
subscript <inline-formula><mml:math id="M99" display="inline"><mml:mrow class="chem"><mml:mi mathvariant="normal">NO</mml:mi></mml:mrow></mml:math></inline-formula> and combine the time direction
into a vector <inline-formula><mml:math id="M100" display="inline"><mml:mi mathvariant="bold-italic">x</mml:mi></mml:math></inline-formula>, with the <inline-formula><mml:math id="M101" display="inline"><mml:mi>i</mml:mi></mml:math></inline-formula>th entry
denoting the density at time <inline-formula><mml:math id="M102" display="inline"><mml:mrow><mml:msub><mml:mi>t</mml:mi><mml:mi>i</mml:mi></mml:msub></mml:mrow></mml:math></inline-formula>, such that
<inline-formula><mml:math id="M103" display="inline"><mml:mrow><mml:msub><mml:mi>x</mml:mi><mml:mi>i</mml:mi></mml:msub><mml:mo>(</mml:mo><mml:mi mathvariant="italic">ϕ</mml:mi><mml:mo>,</mml:mo><mml:mi>z</mml:mi><mml:mo>)</mml:mo><mml:mo>=</mml:mo><mml:mi>x</mml:mi><mml:mo>(</mml:mo><mml:mi mathvariant="italic">ϕ</mml:mi><mml:mo>,</mml:mo><mml:mi>z</mml:mi><mml:mo>,</mml:mo><mml:msub><mml:mi>t</mml:mi><mml:mi>i</mml:mi></mml:msub><mml:mo>)</mml:mo></mml:mrow></mml:math></inline-formula>.</p>
<sec id="Ch1.S3.SS1">
  <title>Linear model</title>
      <p id="d1e1636">In the (multi-)linear case, we relate the nitric oxide
number densities <inline-formula><mml:math id="M104" display="inline"><mml:mrow><mml:msub><mml:mi>x</mml:mi><mml:mi mathvariant="normal">NO</mml:mi></mml:msub><mml:mo>(</mml:mo><mml:mi mathvariant="italic">ϕ</mml:mi><mml:mo>,</mml:mo><mml:mi>z</mml:mi><mml:mo>,</mml:mo><mml:mi>t</mml:mi><mml:mo>)</mml:mo></mml:mrow></mml:math></inline-formula> to the
two proxies, the solar Lyman-<inline-formula><mml:math id="M105" display="inline"><mml:mi mathvariant="italic">α</mml:mi></mml:math></inline-formula> index (Ly<inline-formula><mml:math id="M106" display="inline"><mml:mi mathvariant="italic">α</mml:mi></mml:math></inline-formula>(<inline-formula><mml:math id="M107" display="inline"><mml:mi>t</mml:mi></mml:math></inline-formula>)) and
the geomagnetic AE index (AE(<inline-formula><mml:math id="M108" display="inline"><mml:mi>t</mml:mi></mml:math></inline-formula>)).
Harmonic terms with
<inline-formula><mml:math id="M109" display="inline"><mml:mrow><mml:mi mathvariant="italic">ω</mml:mi><mml:mo>=</mml:mo><mml:mn mathvariant="normal">2</mml:mn><mml:mi mathvariant="italic">π</mml:mi><mml:mspace width="0.125em" linebreak="nobreak"/><mml:msup><mml:mi mathvariant="normal">a</mml:mi><mml:mrow><mml:mo>-</mml:mo><mml:mn mathvariant="normal">1</mml:mn></mml:mrow></mml:msup><mml:mo>=</mml:mo><mml:mn mathvariant="normal">2</mml:mn><mml:mi mathvariant="italic">π</mml:mi><mml:mo>(</mml:mo><mml:mn mathvariant="normal">365.25</mml:mn><mml:mspace width="0.125em" linebreak="nobreak"/><mml:mi mathvariant="normal">d</mml:mi><mml:msup><mml:mo>)</mml:mo><mml:mrow><mml:mo>-</mml:mo><mml:mn mathvariant="normal">1</mml:mn></mml:mrow></mml:msup></mml:mrow></mml:math></inline-formula>
account for annual and semi-annual variations.
The linear model, including a constant offset for the background density,
describes the <inline-formula><mml:math id="M110" display="inline"><mml:mrow class="chem"><mml:mi mathvariant="normal">NO</mml:mi></mml:mrow></mml:math></inline-formula> density according to Eq. (<xref ref-type="disp-formula" rid="Ch1.E7"/>):


                <disp-formula specific-use="align" content-type="numbered"><mml:math id="M111" display="block"><mml:mtable displaystyle="true"><mml:mtr><mml:mtd><mml:mrow><mml:mstyle displaystyle="true" class="stylechange"/><mml:msub><mml:mi>x</mml:mi><mml:mi mathvariant="normal">NO</mml:mi></mml:msub><mml:mo>(</mml:mo><mml:mi mathvariant="italic">ϕ</mml:mi><mml:mo>,</mml:mo><mml:mi>z</mml:mi><mml:mo>,</mml:mo><mml:mi>t</mml:mi><mml:mo>)</mml:mo></mml:mrow></mml:mtd><mml:mtd><mml:mrow><mml:mstyle class="stylechange" displaystyle="true"/><mml:mo>=</mml:mo><mml:mi>a</mml:mi><mml:mo>(</mml:mo><mml:mi mathvariant="italic">ϕ</mml:mi><mml:mo>,</mml:mo><mml:mi>z</mml:mi><mml:mo>)</mml:mo><mml:mo>+</mml:mo><mml:mi>b</mml:mi><mml:mo>(</mml:mo><mml:mi mathvariant="italic">ϕ</mml:mi><mml:mo>,</mml:mo><mml:mi>z</mml:mi><mml:mo>)</mml:mo><mml:mo>⋅</mml:mo><mml:mi mathvariant="normal">Ly</mml:mi><mml:mi mathvariant="italic">α</mml:mi><mml:mo>(</mml:mo><mml:mi>t</mml:mi><mml:mo>)</mml:mo><mml:mo>+</mml:mo><mml:mi>c</mml:mi><mml:mo>(</mml:mo><mml:mi mathvariant="italic">ϕ</mml:mi><mml:mo>,</mml:mo><mml:mi>z</mml:mi><mml:mo>)</mml:mo><mml:mo>⋅</mml:mo><mml:mi mathvariant="normal">AE</mml:mi><mml:mo>(</mml:mo><mml:mi>t</mml:mi><mml:mo>)</mml:mo></mml:mrow></mml:mtd></mml:mtr><mml:mtr><mml:mtd><mml:mstyle class="stylechange" displaystyle="true"/></mml:mtd><mml:mtd><mml:mrow><mml:mstyle class="stylechange" displaystyle="true"/><?xmltex \hack{\quad}?><mml:mo>+</mml:mo><mml:munderover><mml:mo movablelimits="false">∑</mml:mo><mml:mrow><mml:mi>n</mml:mi><mml:mo>=</mml:mo><mml:mn mathvariant="normal">1</mml:mn></mml:mrow><mml:mn mathvariant="normal">2</mml:mn></mml:munderover><mml:mfenced close="" open="["><mml:mrow><mml:msub><mml:mi>d</mml:mi><mml:mi>n</mml:mi></mml:msub><mml:mo>(</mml:mo><mml:mi mathvariant="italic">ϕ</mml:mi><mml:mo>,</mml:mo><mml:mi>z</mml:mi><mml:mo>)</mml:mo><mml:mi>cos⁡</mml:mi><mml:mo>(</mml:mo><mml:mi>n</mml:mi><mml:mi mathvariant="italic">ω</mml:mi><mml:mi>t</mml:mi><mml:mo>)</mml:mo></mml:mrow></mml:mfenced></mml:mrow></mml:mtd></mml:mtr><mml:mlabeledtr id="Ch1.E7"><mml:mtd/><mml:mtd><mml:mstyle class="stylechange" displaystyle="true"/></mml:mtd><mml:mtd><mml:mrow><mml:mstyle displaystyle="true" class="stylechange"/><?xmltex \hack{\qquad}?><mml:mo>+</mml:mo><mml:msub><mml:mi>e</mml:mi><mml:mi>n</mml:mi></mml:msub><mml:mo>(</mml:mo><mml:mi mathvariant="italic">ϕ</mml:mi><mml:mo>,</mml:mo><mml:mi>z</mml:mi><mml:mo>)</mml:mo><mml:mfenced open="" close="]"><mml:mrow><mml:mi>sin⁡</mml:mi><mml:mo>(</mml:mo><mml:mi>n</mml:mi><mml:mi mathvariant="italic">ω</mml:mi><mml:mi>t</mml:mi><mml:mo>)</mml:mo></mml:mrow></mml:mfenced><mml:mspace width="0.25em" linebreak="nobreak"/><mml:mo>.</mml:mo></mml:mrow></mml:mtd></mml:mlabeledtr></mml:mtable></mml:math></disp-formula>

            The linear model can be written in matrix form for
the <inline-formula><mml:math id="M112" display="inline"><mml:mi>n</mml:mi></mml:math></inline-formula> measurement times <inline-formula><mml:math id="M113" display="inline"><mml:mrow><mml:msub><mml:mi>t</mml:mi><mml:mn mathvariant="normal">1</mml:mn></mml:msub><mml:mo>,</mml:mo><mml:mi mathvariant="normal">…</mml:mi><mml:mo>,</mml:mo><mml:msub><mml:mi>t</mml:mi><mml:mi>n</mml:mi></mml:msub></mml:mrow></mml:math></inline-formula> as Eq. (<xref ref-type="disp-formula" rid="Ch1.E8"/>),
with the parameter vector <inline-formula><mml:math id="M114" display="inline"><mml:mi mathvariant="bold-italic">β</mml:mi></mml:math></inline-formula> given by
<inline-formula><mml:math id="M115" display="inline"><mml:mrow><mml:msub><mml:mi mathvariant="bold-italic">β</mml:mi><mml:mi mathvariant="normal">lin</mml:mi></mml:msub><mml:mo>=</mml:mo><mml:mo>(</mml:mo><mml:mi>a</mml:mi><mml:mo>,</mml:mo><mml:mi>b</mml:mi><mml:mo>,</mml:mo><mml:mi>c</mml:mi><mml:mo>,</mml:mo><mml:msub><mml:mi>d</mml:mi><mml:mn mathvariant="normal">1</mml:mn></mml:msub><mml:mo>,</mml:mo><mml:msub><mml:mi>e</mml:mi><mml:mn mathvariant="normal">1</mml:mn></mml:msub><mml:mo>,</mml:mo><mml:msub><mml:mi>d</mml:mi><mml:mn mathvariant="normal">2</mml:mn></mml:msub><mml:mo>,</mml:mo><mml:msub><mml:mi>e</mml:mi><mml:mn mathvariant="normal">2</mml:mn></mml:msub><mml:msup><mml:mo>)</mml:mo><mml:mo>⊤</mml:mo></mml:msup><mml:mo>∈</mml:mo><mml:msup><mml:mi mathvariant="double-struck">R</mml:mi><mml:mn mathvariant="normal">7</mml:mn></mml:msup></mml:mrow></mml:math></inline-formula>
and the model matrix <inline-formula><mml:math id="M116" display="inline"><mml:mrow><mml:mi mathvariant="bold">X</mml:mi><mml:mo>∈</mml:mo><mml:msup><mml:mi mathvariant="double-struck">R</mml:mi><mml:mrow><mml:mi>n</mml:mi><mml:mo>×</mml:mo><mml:mn mathvariant="normal">7</mml:mn></mml:mrow></mml:msup></mml:mrow></mml:math></inline-formula>.</p>
      <p id="d1e2067"><disp-formula id="Ch1.E8" content-type="numbered"><mml:math id="M117" display="block"><mml:mtable columnspacing="1em" class="aligned" rowspacing="0.2ex" displaystyle="true" columnalign="right left"><mml:mtr><mml:mtd><mml:mrow><mml:mstyle class="stylechange" displaystyle="true"/><mml:msub><mml:mi mathvariant="bold-italic">x</mml:mi><mml:mi mathvariant="normal">NO</mml:mi></mml:msub><mml:mo>(</mml:mo><mml:mi mathvariant="italic">ϕ</mml:mi><mml:mo>,</mml:mo><mml:mi>z</mml:mi><mml:mo>)</mml:mo></mml:mrow></mml:mtd><mml:mtd><mml:mrow><mml:mstyle displaystyle="true" class="stylechange"/><mml:mo>=</mml:mo><mml:mfenced close="" open="("><mml:mtable class="matrix" columnalign="center center center center center" framespacing="0em"><mml:mtr><mml:mtd><mml:mn mathvariant="normal">1</mml:mn></mml:mtd><mml:mtd><mml:mrow><mml:mi mathvariant="normal">Ly</mml:mi><mml:mi mathvariant="italic">α</mml:mi><mml:mo>(</mml:mo><mml:msub><mml:mi>t</mml:mi><mml:mn mathvariant="normal">1</mml:mn></mml:msub><mml:mo>)</mml:mo></mml:mrow></mml:mtd><mml:mtd><mml:mrow><mml:mi mathvariant="normal">AE</mml:mi><mml:mo>(</mml:mo><mml:msub><mml:mi>t</mml:mi><mml:mn mathvariant="normal">1</mml:mn></mml:msub><mml:mo>)</mml:mo></mml:mrow></mml:mtd><mml:mtd><mml:mrow><mml:mi>cos⁡</mml:mi><mml:mo>(</mml:mo><mml:mi mathvariant="italic">ω</mml:mi><mml:msub><mml:mi>t</mml:mi><mml:mn mathvariant="normal">1</mml:mn></mml:msub><mml:mo>)</mml:mo></mml:mrow></mml:mtd><mml:mtd><mml:mrow><mml:mi>sin⁡</mml:mi><mml:mo>(</mml:mo><mml:mi mathvariant="italic">ω</mml:mi><mml:msub><mml:mi>t</mml:mi><mml:mn mathvariant="normal">1</mml:mn></mml:msub><mml:mo>)</mml:mo></mml:mrow></mml:mtd></mml:mtr><mml:mtr><mml:mtd><mml:mi mathvariant="normal">⋮</mml:mi></mml:mtd></mml:mtr><mml:mtr><mml:mtd><mml:mn mathvariant="normal">1</mml:mn></mml:mtd><mml:mtd><mml:mrow><mml:mi mathvariant="normal">Ly</mml:mi><mml:mi mathvariant="italic">α</mml:mi><mml:mo>(</mml:mo><mml:msub><mml:mi>t</mml:mi><mml:mi>n</mml:mi></mml:msub><mml:mo>)</mml:mo></mml:mrow></mml:mtd><mml:mtd><mml:mrow><mml:mi mathvariant="normal">AE</mml:mi><mml:mo>(</mml:mo><mml:msub><mml:mi>t</mml:mi><mml:mi>n</mml:mi></mml:msub><mml:mo>)</mml:mo></mml:mrow></mml:mtd><mml:mtd><mml:mrow><mml:mi>cos⁡</mml:mi><mml:mo>(</mml:mo><mml:mi mathvariant="italic">ω</mml:mi><mml:msub><mml:mi>t</mml:mi><mml:mi>n</mml:mi></mml:msub><mml:mo>)</mml:mo></mml:mrow></mml:mtd><mml:mtd><mml:mrow><mml:mi>sin⁡</mml:mi><mml:mo>(</mml:mo><mml:mi mathvariant="italic">ω</mml:mi><mml:msub><mml:mi>t</mml:mi><mml:mi>n</mml:mi></mml:msub><mml:mo>)</mml:mo></mml:mrow></mml:mtd></mml:mtr></mml:mtable></mml:mfenced></mml:mrow></mml:mtd></mml:mtr><mml:mtr><mml:mtd><mml:mstyle displaystyle="true" class="stylechange"/></mml:mtd><mml:mtd><mml:mrow><mml:mstyle displaystyle="true" class="stylechange"/><?xmltex \hack{\qquad}?><mml:mfenced close=")" open=""><mml:mtable class="matrix" columnalign="center center" framespacing="0em"><mml:mtr><mml:mtd><mml:mrow><mml:mi>cos⁡</mml:mi><mml:mo>(</mml:mo><mml:mn mathvariant="normal">2</mml:mn><mml:mi mathvariant="italic">ω</mml:mi><mml:msub><mml:mi>t</mml:mi><mml:mn mathvariant="normal">1</mml:mn></mml:msub><mml:mo>)</mml:mo></mml:mrow></mml:mtd><mml:mtd><mml:mrow><mml:mi>sin⁡</mml:mi><mml:mo>(</mml:mo><mml:mn mathvariant="normal">2</mml:mn><mml:mi mathvariant="italic">ω</mml:mi><mml:msub><mml:mi>t</mml:mi><mml:mn mathvariant="normal">1</mml:mn></mml:msub><mml:mo>)</mml:mo></mml:mrow></mml:mtd></mml:mtr><mml:mtr><mml:mtd/><mml:mtd><mml:mi mathvariant="normal">⋮</mml:mi></mml:mtd></mml:mtr><mml:mtr><mml:mtd><mml:mrow><mml:mi>cos⁡</mml:mi><mml:mo>(</mml:mo><mml:mn mathvariant="normal">2</mml:mn><mml:mi mathvariant="italic">ω</mml:mi><mml:msub><mml:mi>t</mml:mi><mml:mi>n</mml:mi></mml:msub><mml:mo>)</mml:mo></mml:mrow></mml:mtd><mml:mtd><mml:mrow><mml:mi>sin⁡</mml:mi><mml:mo>(</mml:mo><mml:mn mathvariant="normal">2</mml:mn><mml:mi mathvariant="italic">ω</mml:mi><mml:msub><mml:mi>t</mml:mi><mml:mi>n</mml:mi></mml:msub><mml:mo>)</mml:mo></mml:mrow></mml:mtd></mml:mtr></mml:mtable></mml:mfenced><mml:mo>⋅</mml:mo><mml:mfenced close=")" open="("><mml:mtable class="matrix" columnalign="center" framespacing="0em"><mml:mtr><mml:mtd><mml:mi>a</mml:mi></mml:mtd></mml:mtr><mml:mtr><mml:mtd><mml:mi>b</mml:mi></mml:mtd></mml:mtr><mml:mtr><mml:mtd><mml:mi>c</mml:mi></mml:mtd></mml:mtr><mml:mtr><mml:mtd><mml:mrow><mml:msub><mml:mi>d</mml:mi><mml:mn mathvariant="normal">1</mml:mn></mml:msub></mml:mrow></mml:mtd></mml:mtr><mml:mtr><mml:mtd><mml:mrow><mml:msub><mml:mi>e</mml:mi><mml:mn mathvariant="normal">1</mml:mn></mml:msub></mml:mrow></mml:mtd></mml:mtr><mml:mtr><mml:mtd><mml:mrow><mml:msub><mml:mi>d</mml:mi><mml:mn mathvariant="normal">2</mml:mn></mml:msub></mml:mrow></mml:mtd></mml:mtr><mml:mtr><mml:mtd><mml:mrow><mml:msub><mml:mi>e</mml:mi><mml:mn mathvariant="normal">2</mml:mn></mml:msub></mml:mrow></mml:mtd></mml:mtr></mml:mtable></mml:mfenced></mml:mrow></mml:mtd></mml:mtr><mml:mtr><mml:mtd><mml:mstyle displaystyle="true" class="stylechange"/></mml:mtd><mml:mtd><mml:mrow><mml:mstyle displaystyle="true" class="stylechange"/><mml:mo>=</mml:mo><mml:mi mathvariant="bold">X</mml:mi><mml:mo>⋅</mml:mo><mml:mi mathvariant="bold-italic">β</mml:mi></mml:mrow></mml:mtd></mml:mtr></mml:mtable></mml:math></disp-formula></p>
      <p id="d1e2374">We determine the coefficients via least squares, minimizing the
squared differences of the modelled number densities to the measured ones.</p>
</sec>
<sec id="Ch1.S3.SS2">
  <title>Non-linear model</title>
      <p id="d1e2383">In contrast to the linear model above,
we modify the AE index by a finite lifetime <inline-formula><mml:math id="M118" display="inline"><mml:mi mathvariant="italic">τ</mml:mi></mml:math></inline-formula>, which
varies according to season; we denote this modified version by
<inline-formula><mml:math id="M119" display="inline"><mml:mover accent="true"><mml:mi mathvariant="normal">AE</mml:mi><mml:mo mathvariant="normal">̃</mml:mo></mml:mover></mml:math></inline-formula>.
We then omit the harmonic parts in the model, and the non-linear
model is given by Eq. (<xref ref-type="disp-formula" rid="Ch1.E9"/>):
            <disp-formula id="Ch1.E9" content-type="numbered"><mml:math id="M120" display="block"><mml:mrow><mml:msub><mml:mi>x</mml:mi><mml:mi mathvariant="normal">NO</mml:mi></mml:msub><mml:mo>(</mml:mo><mml:mi mathvariant="italic">ϕ</mml:mi><mml:mo>,</mml:mo><mml:mi>z</mml:mi><mml:mo>,</mml:mo><mml:mi>t</mml:mi><mml:mo>)</mml:mo><mml:mo>=</mml:mo><mml:mi>a</mml:mi><mml:mo>(</mml:mo><mml:mi mathvariant="italic">ϕ</mml:mi><mml:mo>,</mml:mo><mml:mi>z</mml:mi><mml:mo>)</mml:mo><mml:mo>+</mml:mo><mml:mi>b</mml:mi><mml:mo>(</mml:mo><mml:mi mathvariant="italic">ϕ</mml:mi><mml:mo>,</mml:mo><mml:mi>z</mml:mi><mml:mo>)</mml:mo><mml:mo>⋅</mml:mo><mml:mi mathvariant="normal">Ly</mml:mi><mml:mi mathvariant="italic">α</mml:mi><mml:mo>(</mml:mo><mml:mi>t</mml:mi><mml:mo>)</mml:mo><mml:mo>+</mml:mo><mml:mi>c</mml:mi><mml:mo>(</mml:mo><mml:mi mathvariant="italic">ϕ</mml:mi><mml:mo>,</mml:mo><mml:mi>z</mml:mi><mml:mo>)</mml:mo><mml:mo>⋅</mml:mo><mml:mover accent="true"><mml:mi mathvariant="normal">AE</mml:mi><mml:mo mathvariant="normal">̃</mml:mo></mml:mover><mml:mo>(</mml:mo><mml:mi>t</mml:mi><mml:mo>)</mml:mo><mml:mspace linebreak="nobreak" width="0.25em"/><mml:mo>.</mml:mo></mml:mrow></mml:math></disp-formula>
          Although this approach shifts all seasonal variations to the
AE index and thus attributes them to particle-induced effects,
we found that the residual traces of particle-unrelated seasonal
effects were minor compared to the overall improvement of the fit.
Additional harmonic terms only increase the number of free
parameters without substantially improving the fit further.</p>
      <p id="d1e2503">The lifetime-corrected <inline-formula><mml:math id="M121" display="inline"><mml:mover accent="true"><mml:mi mathvariant="normal">AE</mml:mi><mml:mo mathvariant="normal">̃</mml:mo></mml:mover></mml:math></inline-formula> is given
by the sum of the previous 60 days' AE values,
each multiplied by an exponential decay factor:
            <disp-formula id="Ch1.E10" content-type="numbered"><mml:math id="M122" display="block"><mml:mrow><mml:mover accent="true"><mml:mi mathvariant="normal">AE</mml:mi><mml:mo mathvariant="normal">̃</mml:mo></mml:mover><mml:mo>(</mml:mo><mml:mi>t</mml:mi><mml:mo>)</mml:mo><mml:mo>=</mml:mo><mml:munderover><mml:mo movablelimits="false">∑</mml:mo><mml:mrow><mml:msub><mml:mi>t</mml:mi><mml:mi>i</mml:mi></mml:msub><mml:mo>=</mml:mo><mml:mn mathvariant="normal">0</mml:mn></mml:mrow><mml:mrow><mml:mn mathvariant="normal">60</mml:mn><mml:mspace width="0.125em" linebreak="nobreak"/><mml:mi mathvariant="normal">d</mml:mi></mml:mrow></mml:munderover><mml:mi mathvariant="normal">AE</mml:mi><mml:mo>(</mml:mo><mml:mi>t</mml:mi><mml:mo>-</mml:mo><mml:msub><mml:mi>t</mml:mi><mml:mi>i</mml:mi></mml:msub><mml:mo>)</mml:mo><mml:mo>⋅</mml:mo><mml:mi>exp⁡</mml:mi><mml:mfenced close="}" open="{"><mml:mrow><mml:mo>-</mml:mo><mml:mstyle displaystyle="true"><mml:mfrac style="display"><mml:mrow><mml:msub><mml:mi>t</mml:mi><mml:mi>i</mml:mi></mml:msub></mml:mrow><mml:mi mathvariant="italic">τ</mml:mi></mml:mfrac></mml:mstyle></mml:mrow></mml:mfenced><mml:mspace linebreak="nobreak" width="0.25em"/><mml:mo>.</mml:mo></mml:mrow></mml:math></disp-formula>
          The total lifetime <inline-formula><mml:math id="M123" display="inline"><mml:mi mathvariant="italic">τ</mml:mi></mml:math></inline-formula> is given by a constant part <inline-formula><mml:math id="M124" display="inline"><mml:mrow><mml:msub><mml:mi mathvariant="italic">τ</mml:mi><mml:mn mathvariant="normal">0</mml:mn></mml:msub></mml:mrow></mml:math></inline-formula> plus
the non-negative fraction of a seasonally
varying part <inline-formula><mml:math id="M125" display="inline"><mml:mrow><mml:msub><mml:mi mathvariant="italic">τ</mml:mi><mml:mi>t</mml:mi></mml:msub></mml:mrow></mml:math></inline-formula>:

                <disp-formula specific-use="align" content-type="numbered"><mml:math id="M126" display="block"><mml:mtable displaystyle="true"><mml:mlabeledtr id="Ch1.E11"><mml:mtd/><mml:mtd><mml:mrow><mml:mstyle displaystyle="true" class="stylechange"/><mml:mi mathvariant="italic">τ</mml:mi></mml:mrow></mml:mtd><mml:mtd><mml:mrow><mml:mstyle displaystyle="true" class="stylechange"/><mml:mo>=</mml:mo><mml:msub><mml:mi mathvariant="italic">τ</mml:mi><mml:mn mathvariant="normal">0</mml:mn></mml:msub><mml:mo>+</mml:mo><mml:mfenced open="{" close=""><mml:mtable rowspacing="0.2ex" class="cases" columnspacing="1em" columnalign="left left" framespacing="0em"><mml:mtr><mml:mtd><mml:mrow><mml:msub><mml:mi mathvariant="italic">τ</mml:mi><mml:mi>t</mml:mi></mml:msub><mml:mspace width="0.25em" linebreak="nobreak"/><mml:mo>,</mml:mo></mml:mrow></mml:mtd><mml:mtd><mml:mrow><mml:msub><mml:mi mathvariant="italic">τ</mml:mi><mml:mi>t</mml:mi></mml:msub><mml:mo>≥</mml:mo><mml:mn mathvariant="normal">0</mml:mn></mml:mrow></mml:mtd></mml:mtr><mml:mtr><mml:mtd><mml:mrow><mml:mn mathvariant="normal">0</mml:mn><mml:mspace linebreak="nobreak" width="0.25em"/><mml:mo>,</mml:mo></mml:mrow></mml:mtd><mml:mtd><mml:mrow><mml:msub><mml:mi mathvariant="italic">τ</mml:mi><mml:mi>t</mml:mi></mml:msub><mml:mo>&lt;</mml:mo><mml:mn mathvariant="normal">0</mml:mn></mml:mrow></mml:mtd></mml:mtr></mml:mtable></mml:mfenced><mml:mspace width="0.25em" linebreak="nobreak"/><mml:mo>,</mml:mo></mml:mrow></mml:mtd></mml:mlabeledtr><mml:mlabeledtr id="Ch1.E12"><mml:mtd/><mml:mtd><mml:mrow><mml:mstyle displaystyle="true" class="stylechange"/><mml:msub><mml:mi mathvariant="italic">τ</mml:mi><mml:mi>t</mml:mi></mml:msub></mml:mrow></mml:mtd><mml:mtd><mml:mrow><mml:mstyle class="stylechange" displaystyle="true"/><mml:mo>=</mml:mo><mml:mi>d</mml:mi><mml:mi>cos⁡</mml:mi><mml:mo>(</mml:mo><mml:mi mathvariant="italic">ω</mml:mi><mml:mi>t</mml:mi><mml:mo>)</mml:mo><mml:mo>+</mml:mo><mml:mi>e</mml:mi><mml:mi>sin⁡</mml:mi><mml:mo>(</mml:mo><mml:mi mathvariant="italic">ω</mml:mi><mml:mi>t</mml:mi><mml:mo>)</mml:mo><mml:mspace linebreak="nobreak" width="0.25em"/><mml:mo>,</mml:mo></mml:mrow></mml:mtd></mml:mlabeledtr></mml:mtable></mml:math></disp-formula>

            where <inline-formula><mml:math id="M127" display="inline"><mml:mrow><mml:msub><mml:mi mathvariant="italic">τ</mml:mi><mml:mi>t</mml:mi></mml:msub></mml:mrow></mml:math></inline-formula> accounts for
the different lifetime during winter and summer.
The parameter vector for this model is given by
<inline-formula><mml:math id="M128" display="inline"><mml:mrow><mml:msub><mml:mi mathvariant="bold-italic">β</mml:mi><mml:mi mathvariant="normal">nonlin</mml:mi></mml:msub><mml:mo>=</mml:mo><mml:mo>(</mml:mo><mml:mi>a</mml:mi><mml:mo>,</mml:mo><mml:mi>b</mml:mi><mml:mo>,</mml:mo><mml:mi>c</mml:mi><mml:mo>,</mml:mo><mml:msub><mml:mi mathvariant="italic">τ</mml:mi><mml:mn mathvariant="normal">0</mml:mn></mml:msub><mml:mo>,</mml:mo><mml:mi>d</mml:mi><mml:mo>,</mml:mo><mml:mi>e</mml:mi><mml:msup><mml:mo>)</mml:mo><mml:mo>⊤</mml:mo></mml:msup><mml:mo>∈</mml:mo><mml:msup><mml:mi mathvariant="double-struck">R</mml:mi><mml:mn mathvariant="normal">6</mml:mn></mml:msup></mml:mrow></mml:math></inline-formula>,
and we describe how we determine these coefficients and their
uncertainties in the next section.</p>
</sec>
</sec>
<sec id="Ch1.S4">
  <title>Parameter and uncertainty estimation</title>
      <p id="d1e2808">The parameters are usually estimated by maximizing the likelihood,
or, in the case of additional prior constraints, by maximizing
the posterior probability.
In the linear case and in the case of independently
identically distributed Gaussian measurement uncertainties,
the maximum likelihood solutions
are given by the usual
linear least squares solutions.
Estimating the parameters in the non-linear case is more involved.
Various methods exist, for example, conjugate gradient,
random (Monte Carlo) sampling, or exhaustive search methods.
The assessment and selection of the method to estimate the
parameters in the non-linear case are given below.</p>
<sec id="Ch1.S4.SS1">
  <title>Maximum posterior probability</title>
      <?pagebreak page2139?><p id="d1e2816">Because of the complicated structure of the model function in Eq. (<xref ref-type="disp-formula" rid="Ch1.E9"/>), in particular the lifetime parts in
Eqs. (<xref ref-type="disp-formula" rid="Ch1.E11"/>) and (<xref ref-type="disp-formula" rid="Ch1.E12"/>),
the usual gradient methods converge slowly, if at all.
Therefore,
we fit the parameters and assess their uncertainty ranges using Markov chain
Monte Carlo (MCMC) sampling <xref ref-type="bibr" rid="bib1.bibx21" id="paren.55"/>.
This method samples probability distributions, and we apply it to
sample the parameter space, putting emphasis on parameter values
with a high posterior probability.
The posterior distribution is given in the Bayesian sense
as the product of the likelihood and the prior distribution:
            <disp-formula id="Ch1.E13" content-type="numbered"><mml:math id="M129" display="block"><mml:mrow><mml:mi>p</mml:mi><mml:mo>(</mml:mo><mml:msub><mml:mi mathvariant="bold-italic">x</mml:mi><mml:mi mathvariant="normal">mod</mml:mi></mml:msub><mml:mo>|</mml:mo><mml:mi mathvariant="bold-italic">y</mml:mi><mml:mo>)</mml:mo><mml:mo>∝</mml:mo><mml:mi>p</mml:mi><mml:mo>(</mml:mo><mml:msub><mml:mi mathvariant="bold-italic">x</mml:mi><mml:mi mathvariant="normal">mod</mml:mi></mml:msub><mml:mo>|</mml:mo><mml:mi mathvariant="bold-italic">y</mml:mi><mml:mo>,</mml:mo><mml:mi mathvariant="bold-italic">β</mml:mi><mml:mo>)</mml:mo><mml:mi>p</mml:mi><mml:mo>(</mml:mo><mml:mi mathvariant="bold-italic">β</mml:mi><mml:mo>)</mml:mo><mml:mspace width="0.25em" linebreak="nobreak"/><mml:mo>.</mml:mo></mml:mrow></mml:math></disp-formula>
          We denote the vector of the measured densities by <inline-formula><mml:math id="M130" display="inline"><mml:mi mathvariant="bold-italic">y</mml:mi></mml:math></inline-formula>
and the modelled densities by <inline-formula><mml:math id="M131" display="inline"><mml:mrow><mml:msub><mml:mi mathvariant="bold-italic">x</mml:mi><mml:mi mathvariant="normal">mod</mml:mi></mml:msub></mml:mrow></mml:math></inline-formula>, similar
to Eqs. (<xref ref-type="disp-formula" rid="Ch1.E7"/>) and (<xref ref-type="disp-formula" rid="Ch1.E9"/>).
To find the best parameters <inline-formula><mml:math id="M132" display="inline"><mml:mi mathvariant="bold-italic">β</mml:mi></mml:math></inline-formula> for the model,
we maximize <inline-formula><mml:math id="M133" display="inline"><mml:mrow><mml:mi>log⁡</mml:mi><mml:mi>p</mml:mi><mml:mo>(</mml:mo><mml:msub><mml:mi mathvariant="bold-italic">x</mml:mi><mml:mi mathvariant="normal">mod</mml:mi></mml:msub><mml:mo>|</mml:mo><mml:mi mathvariant="bold-italic">y</mml:mi><mml:mo>)</mml:mo></mml:mrow></mml:math></inline-formula>.</p>
      <p id="d1e2935">The likelihood <inline-formula><mml:math id="M134" display="inline"><mml:mrow><mml:mi>p</mml:mi><mml:mo>(</mml:mo><mml:msub><mml:mi mathvariant="bold-italic">x</mml:mi><mml:mi mathvariant="normal">mod</mml:mi></mml:msub><mml:mo>|</mml:mo><mml:mi mathvariant="bold-italic">y</mml:mi><mml:mo>,</mml:mo><mml:mi mathvariant="bold-italic">β</mml:mi><mml:mo>)</mml:mo></mml:mrow></mml:math></inline-formula>
is in our case given by a Gaussian distribution of the residuals,
the difference of the model to the data, given in Eq. (<xref ref-type="disp-formula" rid="Ch1.E14"/>):</p>
      <p id="d1e2965"><disp-formula specific-use="align" content-type="numbered"><mml:math id="M135" display="block"><mml:mtable displaystyle="true"><mml:mtr><mml:mtd><mml:mrow><mml:mstyle displaystyle="true" class="stylechange"/><mml:mi>p</mml:mi><mml:mo>(</mml:mo><mml:msub><mml:mi mathvariant="bold-italic">x</mml:mi><mml:mi mathvariant="normal">mod</mml:mi></mml:msub><mml:mo>|</mml:mo><mml:mi mathvariant="bold-italic">y</mml:mi><mml:mo>,</mml:mo><mml:mi mathvariant="bold-italic">β</mml:mi><mml:mo>)</mml:mo></mml:mrow></mml:mtd><mml:mtd><mml:mrow><mml:mstyle displaystyle="true" class="stylechange"/><mml:mo>=</mml:mo><mml:mi mathvariant="script">N</mml:mi><mml:mfenced close=")" open="("><mml:mrow><mml:mi mathvariant="bold-italic">y</mml:mi><mml:mo>,</mml:mo><mml:msub><mml:mi mathvariant="bold">S</mml:mi><mml:mi>y</mml:mi></mml:msub></mml:mrow></mml:mfenced></mml:mrow></mml:mtd></mml:mtr><mml:mtr><mml:mtd><mml:mstyle displaystyle="true" class="stylechange"/></mml:mtd><mml:mtd><mml:mrow><mml:mstyle class="stylechange" displaystyle="true"/><mml:mo>=</mml:mo><mml:mi>C</mml:mi><mml:mi>exp⁡</mml:mi><mml:mfenced close="" open="{"><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:msup><mml:mfenced open="(" close=")"><mml:mrow><mml:mi mathvariant="bold-italic">y</mml:mi><mml:mo>-</mml:mo><mml:msub><mml:mi mathvariant="bold-italic">x</mml:mi><mml:mi mathvariant="normal">mod</mml:mi></mml:msub><mml:mo>(</mml:mo><mml:mi mathvariant="bold-italic">β</mml:mi><mml:mo>)</mml:mo></mml:mrow></mml:mfenced><mml:mo>⊤</mml:mo></mml:msup></mml:mrow></mml:mfenced></mml:mrow></mml:mtd></mml:mtr><mml:mlabeledtr id="Ch1.E14"><mml:mtd/><mml:mtd><mml:mstyle class="stylechange" displaystyle="true"/></mml:mtd><mml:mtd><mml:mrow><mml:mstyle class="stylechange" displaystyle="true"/><?xmltex \hack{\quad}?><mml:mfenced close="}" open=""><mml:mrow><mml:msubsup><mml:mi mathvariant="bold">S</mml:mi><mml:mi>y</mml:mi><mml:mrow><mml:mo>-</mml:mo><mml:mn mathvariant="normal">1</mml:mn></mml:mrow></mml:msubsup><mml:mfenced close=")" open="("><mml:mrow><mml:mi mathvariant="bold-italic">y</mml:mi><mml:mo>-</mml:mo><mml:msub><mml:mi mathvariant="bold-italic">x</mml:mi><mml:mi mathvariant="normal">mod</mml:mi></mml:msub><mml:mo>(</mml:mo><mml:mi mathvariant="bold-italic">β</mml:mi><mml:mo>)</mml:mo></mml:mrow></mml:mfenced></mml:mrow></mml:mfenced><mml:mo>.</mml:mo></mml:mrow></mml:mtd></mml:mlabeledtr></mml:mtable></mml:math></disp-formula></p>
      <p id="d1e3094">Note that the normalization constant <inline-formula><mml:math id="M136" display="inline"><mml:mi>C</mml:mi></mml:math></inline-formula> in Eq. (<xref ref-type="disp-formula" rid="Ch1.E14"/>)
does not influence the value of the maximal
likelihood.
The covariance matrix <inline-formula><mml:math id="M137" display="inline"><mml:mrow><mml:msub><mml:mi mathvariant="bold">S</mml:mi><mml:mi>y</mml:mi></mml:msub></mml:mrow></mml:math></inline-formula> contains the squared standard errors
of the daily zonal means on the diagonal,
<inline-formula><mml:math id="M138" display="inline"><mml:mrow><mml:msub><mml:mi mathvariant="bold">S</mml:mi><mml:mi>y</mml:mi></mml:msub><mml:mo>=</mml:mo><mml:mi mathvariant="normal">diag</mml:mi><mml:mo>(</mml:mo><mml:msubsup><mml:mi mathvariant="italic">σ</mml:mi><mml:mi>y</mml:mi><mml:mn mathvariant="normal">2</mml:mn></mml:msubsup><mml:mo>)</mml:mo></mml:mrow></mml:math></inline-formula>.</p>
      <p id="d1e3144">The prior distribution <inline-formula><mml:math id="M139" display="inline"><mml:mrow><mml:mi>p</mml:mi><mml:mo>(</mml:mo><mml:mi mathvariant="bold-italic">β</mml:mi><mml:mo>)</mml:mo></mml:mrow></mml:math></inline-formula> restricts the parameters
to lie within certain ranges, and
the bounds we used for the sampling are listed in
Table <xref ref-type="table" rid="Ch1.T1"/>.
Within those bounds
we assume uniform (flat) prior distributions
for the offset, the geomagnetic and solar amplitudes, and in the
linear case also for the annual and semi-annual harmonics.
We penalize large lifetimes using an
exponential distribution <inline-formula><mml:math id="M140" display="inline"><mml:mrow><mml:mi>p</mml:mi><mml:mo>(</mml:mo><mml:mi mathvariant="italic">τ</mml:mi><mml:mo>)</mml:mo><mml:mo>∝</mml:mo><mml:mi>exp⁡</mml:mi><mml:mo mathvariant="italic">{</mml:mo><mml:mo>-</mml:mo><mml:mi mathvariant="italic">τ</mml:mi><mml:mo>/</mml:mo><mml:msub><mml:mi mathvariant="italic">σ</mml:mi><mml:mi mathvariant="italic">τ</mml:mi></mml:msub><mml:mo mathvariant="italic">}</mml:mo></mml:mrow></mml:math></inline-formula>
for each lifetime parameter, i.e. for <inline-formula><mml:math id="M141" display="inline"><mml:mrow><mml:msub><mml:mi mathvariant="italic">τ</mml:mi><mml:mn mathvariant="normal">0</mml:mn></mml:msub></mml:mrow></mml:math></inline-formula>, <inline-formula><mml:math id="M142" display="inline"><mml:mi>d</mml:mi></mml:math></inline-formula>, and <inline-formula><mml:math id="M143" display="inline"><mml:mi>e</mml:mi></mml:math></inline-formula> in
Eqs. (<xref ref-type="disp-formula" rid="Ch1.E11"/>) and (<xref ref-type="disp-formula" rid="Ch1.E12"/>).
The scale width <inline-formula><mml:math id="M144" display="inline"><mml:mrow><mml:msub><mml:mi mathvariant="italic">σ</mml:mi><mml:mi mathvariant="italic">τ</mml:mi></mml:msub></mml:mrow></mml:math></inline-formula> of this exponential distribution
is fixed to 1 day.
This choice of prior distributions for the lifetime parameters
prevents sampling of the edges of the
parameter space
at places with small geomagnetic coefficients.
In those regions the lifetime may be ambiguous and less meaningful.</p>

<?xmltex \floatpos{t}?><table-wrap id="Ch1.T1"><caption><p id="d1e3241">Parameter search space for the non-linear model and
uncertainty estimation.</p></caption><oasis:table frame="topbot"><?xmltex \begin{scaleboxenv}{.98}[.98]?><oasis:tgroup cols="4">
     <oasis:colspec colnum="1" colname="col1" align="left"/>
     <oasis:colspec colnum="2" colname="col2" align="left"/>
     <oasis:colspec colnum="3" colname="col3" align="left"/>
     <oasis:colspec colnum="4" colname="col4" align="left"/>
     <oasis:thead>
       <oasis:row>
         <oasis:entry colname="col1">Parameter</oasis:entry>
         <oasis:entry colname="col2">Lower bound</oasis:entry>
         <oasis:entry colname="col3">Upper bound</oasis:entry>
         <oasis:entry colname="col4">Prior</oasis:entry>
       </oasis:row>
       <oasis:row rowsep="1">
         <oasis:entry colname="col1"/>
         <oasis:entry colname="col2"/>
         <oasis:entry colname="col3"/>
         <oasis:entry colname="col4">form</oasis:entry>
       </oasis:row>
     </oasis:thead>
     <oasis:tbody>
       <oasis:row>
         <oasis:entry colname="col1">Offset (<inline-formula><mml:math id="M145" display="inline"><mml:mi>a</mml:mi></mml:math></inline-formula>)</oasis:entry>
         <oasis:entry colname="col2"><inline-formula><mml:math id="M146" display="inline"><mml:mrow><mml:mo>-</mml:mo><mml:msup><mml:mn mathvariant="normal">10</mml:mn><mml:mn mathvariant="normal">10</mml:mn></mml:msup><mml:mspace linebreak="nobreak" width="0.125em"/></mml:mrow></mml:math></inline-formula>cm<inline-formula><mml:math id="M147" 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="col3"><inline-formula><mml:math id="M148" display="inline"><mml:mrow><mml:msup><mml:mn mathvariant="normal">10</mml:mn><mml:mn mathvariant="normal">10</mml:mn></mml:msup><mml:mspace linebreak="nobreak" width="0.125em"/></mml:mrow></mml:math></inline-formula>cm<inline-formula><mml:math id="M149" 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">flat</oasis:entry>
       </oasis:row>
       <oasis:row>
         <oasis:entry colname="col1">Lyman-<inline-formula><mml:math id="M150" display="inline"><mml:mi mathvariant="italic">α</mml:mi></mml:math></inline-formula> amplitude (<inline-formula><mml:math id="M151" display="inline"><mml:mi>b</mml:mi></mml:math></inline-formula>)</oasis:entry>
         <oasis:entry colname="col2"><inline-formula><mml:math id="M152" display="inline"><mml:mrow><mml:mo>-</mml:mo><mml:msup><mml:mn mathvariant="normal">10</mml:mn><mml:mn mathvariant="normal">10</mml:mn></mml:msup><mml:mspace width="0.125em" linebreak="nobreak"/></mml:mrow></mml:math></inline-formula>cm<inline-formula><mml:math id="M153" 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="col3"><inline-formula><mml:math id="M154" display="inline"><mml:mrow><mml:msup><mml:mn mathvariant="normal">10</mml:mn><mml:mn mathvariant="normal">10</mml:mn></mml:msup><mml:mspace width="0.125em" linebreak="nobreak"/></mml:mrow></mml:math></inline-formula>cm<inline-formula><mml:math id="M155" 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">flat</oasis:entry>
       </oasis:row>
       <oasis:row>
         <oasis:entry colname="col1">AE amplitude (<inline-formula><mml:math id="M156" display="inline"><mml:mi>c</mml:mi></mml:math></inline-formula>)</oasis:entry>
         <oasis:entry colname="col2"><inline-formula><mml:math id="M157" display="inline"><mml:mrow><mml:mn mathvariant="normal">0</mml:mn><mml:mspace width="0.125em" linebreak="nobreak"/></mml:mrow></mml:math></inline-formula>cm<inline-formula><mml:math id="M158" 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="col3"><inline-formula><mml:math id="M159" display="inline"><mml:mrow><mml:msup><mml:mn mathvariant="normal">10</mml:mn><mml:mn mathvariant="normal">10</mml:mn></mml:msup><mml:mspace width="0.125em" linebreak="nobreak"/></mml:mrow></mml:math></inline-formula>cm<inline-formula><mml:math id="M160" 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">flat</oasis:entry>
       </oasis:row>
       <oasis:row>
         <oasis:entry colname="col1"><inline-formula><mml:math id="M161" display="inline"><mml:mrow><mml:msub><mml:mi mathvariant="italic">τ</mml:mi><mml:mn mathvariant="normal">0</mml:mn></mml:msub></mml:mrow></mml:math></inline-formula></oasis:entry>
         <oasis:entry colname="col2">0 days</oasis:entry>
         <oasis:entry colname="col3">100 days</oasis:entry>
         <oasis:entry colname="col4">exp</oasis:entry>
       </oasis:row>
       <oasis:row>
         <oasis:entry colname="col1"><inline-formula><mml:math id="M162" display="inline"><mml:mi mathvariant="italic">τ</mml:mi></mml:math></inline-formula> cosine amplitude (<inline-formula><mml:math id="M163" display="inline"><mml:mi>d</mml:mi></mml:math></inline-formula>)</oasis:entry>
         <oasis:entry colname="col2"><inline-formula><mml:math id="M164" display="inline"><mml:mo>-</mml:mo></mml:math></inline-formula>100 days</oasis:entry>
         <oasis:entry colname="col3">100 days</oasis:entry>
         <oasis:entry colname="col4">exp</oasis:entry>
       </oasis:row>
       <oasis:row>
         <oasis:entry colname="col1"><inline-formula><mml:math id="M165" display="inline"><mml:mi mathvariant="italic">τ</mml:mi></mml:math></inline-formula> sine amplitude (<inline-formula><mml:math id="M166" display="inline"><mml:mi>e</mml:mi></mml:math></inline-formula>)</oasis:entry>
         <oasis:entry colname="col2"><inline-formula><mml:math id="M167" display="inline"><mml:mo>-</mml:mo></mml:math></inline-formula>100 days</oasis:entry>
         <oasis:entry colname="col3">100 days</oasis:entry>
         <oasis:entry colname="col4">exp</oasis:entry>
       </oasis:row>
     </oasis:tbody>
   </oasis:tgroup><?xmltex \end{scaleboxenv}?></oasis:table></table-wrap>

</sec>
<sec id="Ch1.S4.SS2">
  <title>Correlations</title>
      <p id="d1e3596">In the simple case, the measurement covariance matrix <inline-formula><mml:math id="M168" display="inline"><mml:mrow><mml:msub><mml:mi mathvariant="bold">S</mml:mi><mml:mi>y</mml:mi></mml:msub></mml:mrow></mml:math></inline-formula>
contains the measurement uncertainties on the diagonal,
in our case the (squared) standard error of the zonal means
denoted by <inline-formula><mml:math id="M169" display="inline"><mml:mrow><mml:msub><mml:mi mathvariant="italic">σ</mml:mi><mml:mi>y</mml:mi></mml:msub></mml:mrow></mml:math></inline-formula>,
<inline-formula><mml:math id="M170" display="inline"><mml:mrow><mml:msub><mml:mi mathvariant="bold">S</mml:mi><mml:mi>y</mml:mi></mml:msub><mml:mo>=</mml:mo><mml:mi mathvariant="normal">diag</mml:mi><mml:mo>(</mml:mo><mml:msubsup><mml:mi mathvariant="italic">σ</mml:mi><mml:mi>y</mml:mi><mml:mn mathvariant="normal">2</mml:mn></mml:msubsup><mml:mo>)</mml:mo></mml:mrow></mml:math></inline-formula>.
However, the standard error of the mean
might underestimate the true uncertainties.
In addition, possible correlations may occur which
are not accounted for using a diagonal <inline-formula><mml:math id="M171" display="inline"><mml:mrow><mml:msub><mml:mi mathvariant="bold">S</mml:mi><mml:mi>y</mml:mi></mml:msub></mml:mrow></mml:math></inline-formula>.</p>
      <p id="d1e3658">Both problems can be addressed by adding a
covariance kernel <inline-formula><mml:math id="M172" display="inline"><mml:mi mathvariant="bold">K</mml:mi></mml:math></inline-formula> to <inline-formula><mml:math id="M173" display="inline"><mml:mrow><mml:msub><mml:mi mathvariant="bold">S</mml:mi><mml:mi>y</mml:mi></mml:msub></mml:mrow></mml:math></inline-formula>.
Various forms of covariance kernels can be
used <xref ref-type="bibr" rid="bib1.bibx53" id="paren.56"/>,
depending on the underlying process leading to the
measurement or residual uncertainties.
Since we have no prior knowledge about the true correlations,
we use a commonly chosen
kernel of the Matérn 3/2
type <xref ref-type="bibr" rid="bib1.bibx46 bib1.bibx42 bib1.bibx53" id="paren.57"/>.
This kernel only depends on the (time) distance between
the measurements <inline-formula><mml:math id="M174" display="inline"><mml:mrow><mml:msub><mml:mi>t</mml:mi><mml:mrow><mml:mi>i</mml:mi><mml:mi>j</mml:mi></mml:mrow></mml:msub><mml:mo>=</mml:mo><mml:mo fence="true">|</mml:mo><mml:msub><mml:mi>t</mml:mi><mml:mi>i</mml:mi></mml:msub><mml:mo>-</mml:mo><mml:msub><mml:mi>t</mml:mi><mml:mi>j</mml:mi></mml:msub><mml:mo fence="true">|</mml:mo></mml:mrow></mml:math></inline-formula>
and has two parameters, the “strength” <inline-formula><mml:math id="M175" display="inline"><mml:mi mathvariant="italic">σ</mml:mi></mml:math></inline-formula>
and correlation length <inline-formula><mml:math id="M176" display="inline"><mml:mi mathvariant="italic">ρ</mml:mi></mml:math></inline-formula>:
            <disp-formula id="Ch1.E15" content-type="numbered"><mml:math id="M177" display="block"><mml:mrow><mml:msub><mml:mi>K</mml:mi><mml:mrow><mml:mi>i</mml:mi><mml:mi>j</mml:mi></mml:mrow></mml:msub><mml:mo>=</mml:mo><mml:msup><mml:mi mathvariant="italic">σ</mml:mi><mml:mn mathvariant="normal">2</mml:mn></mml:msup><mml:mfenced open="(" close=")"><mml:mrow><mml:mn mathvariant="normal">1</mml:mn><mml:mo>+</mml:mo><mml:mstyle displaystyle="true"><mml:mfrac style="display"><mml:mrow><mml:msqrt><mml:mn mathvariant="normal">3</mml:mn></mml:msqrt><mml:mspace linebreak="nobreak" width="0.125em"/><mml:msub><mml:mi>t</mml:mi><mml:mrow><mml:mi>i</mml:mi><mml:mi>j</mml:mi></mml:mrow></mml:msub></mml:mrow><mml:mi mathvariant="italic">ρ</mml:mi></mml:mfrac></mml:mstyle></mml:mrow></mml:mfenced><mml:mi>exp⁡</mml:mi><mml:mfenced open="{" close="}"><mml:mrow><mml:mo>-</mml:mo><mml:mstyle displaystyle="true"><mml:mfrac style="display"><mml:mrow><mml:msqrt><mml:mn mathvariant="normal">3</mml:mn></mml:msqrt><mml:mspace width="0.125em" linebreak="nobreak"/><mml:msub><mml:mi>t</mml:mi><mml:mrow><mml:mi>i</mml:mi><mml:mi>j</mml:mi></mml:mrow></mml:msub></mml:mrow><mml:mi mathvariant="italic">ρ</mml:mi></mml:mfrac></mml:mstyle></mml:mrow></mml:mfenced><mml:mspace width="0.25em" linebreak="nobreak"/><mml:mo>.</mml:mo></mml:mrow></mml:math></disp-formula>
          Both parameters are estimated together with the model
parameter vector <inline-formula><mml:math id="M178" display="inline"><mml:mi mathvariant="bold-italic">β</mml:mi></mml:math></inline-formula>.
We found that using the kernel (<xref ref-type="disp-formula" rid="Ch1.E15"/>)
in a covariance matrix <inline-formula><mml:math id="M179" display="inline"><mml:mrow><mml:msub><mml:mi mathvariant="bold">S</mml:mi><mml:mi>y</mml:mi></mml:msub></mml:mrow></mml:math></inline-formula>
with the entries
            <disp-formula id="Ch1.E16" content-type="numbered"><mml:math id="M180" display="block"><mml:mrow><mml:msub><mml:msub><mml:mi>S</mml:mi><mml:mi>y</mml:mi></mml:msub><mml:mrow><mml:mi>i</mml:mi><mml:mi>j</mml:mi></mml:mrow></mml:msub><mml:mo>=</mml:mo><mml:msub><mml:mi>K</mml:mi><mml:mrow><mml:mi>i</mml:mi><mml:mi>j</mml:mi></mml:mrow></mml:msub><mml:mo>+</mml:mo><mml:msub><mml:mi mathvariant="italic">δ</mml:mi><mml:mrow><mml:mi>i</mml:mi><mml:mi>j</mml:mi></mml:mrow></mml:msub><mml:msubsup><mml:msub><mml:mi mathvariant="italic">σ</mml:mi><mml:mi>y</mml:mi></mml:msub><mml:mi>i</mml:mi><mml:mn mathvariant="normal">2</mml:mn></mml:msubsup><mml:mspace width="0.25em" linebreak="nobreak"/></mml:mrow></mml:math></disp-formula>
          worked best and led to stable and reliable parameter sampling.
Note that an additional “white noise” term
<inline-formula><mml:math id="M181" display="inline"><mml:mrow><mml:msup><mml:mi mathvariant="italic">σ</mml:mi><mml:mn mathvariant="normal">2</mml:mn></mml:msup><mml:mi mathvariant="bold">I</mml:mi></mml:mrow></mml:math></inline-formula> could be added to
the covariance matrix to account for data uncertainties that are still
underestimated.
However, this additional white noise term did not
improve the convergence, nor did it influence the fitted
parameters significantly.</p>
      <p id="d1e3884">The approximately <inline-formula><mml:math id="M182" display="inline"><mml:mrow><mml:mn mathvariant="normal">3000</mml:mn><mml:mo>×</mml:mo><mml:mn mathvariant="normal">3000</mml:mn></mml:mrow></mml:math></inline-formula> covariance matrix
of the Gaussian process model for the residuals was
evaluated using the <xref ref-type="bibr" rid="bib1.bibx22" id="text.58"/> approximation
and the provided Python code <xref ref-type="bibr" rid="bib1.bibx23" id="paren.59"/>.
For one-dimensional data sets, this approach is computationally
faster than the full Cholesky decomposition, which is usually used to invert the covariance matrix <inline-formula><mml:math id="M183" display="inline"><mml:mrow><mml:msub><mml:mi mathvariant="bold">S</mml:mi><mml:mi>y</mml:mi></mml:msub></mml:mrow></mml:math></inline-formula>.
With this approximation, we achieved sensible Monte Carlo
sampling times to facilitate evaluating all <inline-formula><mml:math id="M184" display="inline"><mml:mrow><mml:mn mathvariant="normal">18</mml:mn><mml:mo>×</mml:mo><mml:mn mathvariant="normal">16</mml:mn></mml:mrow></mml:math></inline-formula>
latitude <inline-formula><mml:math id="M185" display="inline"><mml:mo>×</mml:mo></mml:math></inline-formula> altitude bins on a small cluster in about 1 day.
We used the <monospace>emcee</monospace> package <xref ref-type="bibr" rid="bib1.bibx21" id="paren.60"/> for the Monte Carlo sampling,
set up to use
112 walkers and 800 samples for the initial fit of the parameters,
followed by another 800 so-called burn-in samples and 1400 production
samples.
The full code can be found in <xref ref-type="bibr" rid="bib1.bibx11" id="text.61"/>.</p>
</sec>
</sec>
<?pagebreak page2140?><sec id="Ch1.S5">
  <title>Results</title>
      <p id="d1e3953">We demonstrate the parameter estimates using example time series
<inline-formula><mml:math id="M186" display="inline"><mml:mrow><mml:msub><mml:mi mathvariant="bold-italic">x</mml:mi><mml:mi mathvariant="normal">NO</mml:mi></mml:msub></mml:mrow></mml:math></inline-formula> at 70 km at 65<inline-formula><mml:math id="M187" display="inline"><mml:msup><mml:mi/><mml:mo>∘</mml:mo></mml:msup></mml:math></inline-formula> S, 5<inline-formula><mml:math id="M188" display="inline"><mml:msup><mml:mi/><mml:mo>∘</mml:mo></mml:msup></mml:math></inline-formula> N, and 65<inline-formula><mml:math id="M189" display="inline"><mml:msup><mml:mi/><mml:mo>∘</mml:mo></mml:msup></mml:math></inline-formula> N.
<inline-formula><mml:math id="M190" display="inline"><mml:mrow class="chem"><mml:mi mathvariant="normal">NO</mml:mi></mml:mrow></mml:math></inline-formula> shows different behaviour in these regions,
showing the most variation with respect to
the solar cycle and geomagnetic activity at high latitudes.
In contrast, at low latitudes the geomagnetic influence should be
reduced <xref ref-type="bibr" rid="bib1.bibx8 bib1.bibx37 bib1.bibx40" id="paren.62"/>.
We briefly only show the results for the linear model and point out
some of its shortcomings.
Thereafter we show the results from the non-linear model and
continue to use that for further analysis of the coefficients.</p>
<sec id="Ch1.S5.SS1">
  <title>Time series fits</title>
      <p id="d1e4011">The fitted densities of the linear model Eq. (<xref ref-type="disp-formula" rid="Ch1.E7"/>)
compared to the data are shown in Fig. <xref ref-type="fig" rid="Ch1.F1"/>a, b, and c
for the three example latitude bins (65<inline-formula><mml:math id="M191" display="inline"><mml:msup><mml:mi/><mml:mo>∘</mml:mo></mml:msup></mml:math></inline-formula> S, 5<inline-formula><mml:math id="M192" display="inline"><mml:msup><mml:mi/><mml:mo>∘</mml:mo></mml:msup></mml:math></inline-formula> N, 65<inline-formula><mml:math id="M193" display="inline"><mml:msup><mml:mi/><mml:mo>∘</mml:mo></mml:msup></mml:math></inline-formula> N) at 70 km.
The linear model works well at high southern and low latitudes.
At high northern latitudes and to a lesser extent at high southern latitudes,
the linear model captures the summer <inline-formula><mml:math id="M194" display="inline"><mml:mrow class="chem"><mml:mi mathvariant="normal">NO</mml:mi></mml:mrow></mml:math></inline-formula> variations well.
However, the model underestimates the high values in the polar winter
at active times (2004–2007) and overestimates the low winter values
at quiet times (2009–2011).</p>

      <?xmltex \floatpos{t}?><fig id="Ch1.F1" specific-use="star"><caption><p id="d1e4056">Time series data and linear model values and residuals
at 70 km for 65<inline-formula><mml:math id="M195" display="inline"><mml:msup><mml:mi/><mml:mo>∘</mml:mo></mml:msup></mml:math></inline-formula> S <bold>(a, d)</bold>, 5<inline-formula><mml:math id="M196" display="inline"><mml:msup><mml:mi/><mml:mo>∘</mml:mo></mml:msup></mml:math></inline-formula> N <bold>(b, e)</bold>, and 65<inline-formula><mml:math id="M197" display="inline"><mml:msup><mml:mi/><mml:mo>∘</mml:mo></mml:msup></mml:math></inline-formula> N <bold>(c, f)</bold>.
Panels <bold>(a)</bold>–<bold>(c)</bold> show the data (black dots with <inline-formula><mml:math id="M198" display="inline"><mml:mrow><mml:mn mathvariant="normal">2</mml:mn><mml:mi mathvariant="italic">σ</mml:mi></mml:mrow></mml:math></inline-formula> error bars) and the
model values (blue line).
Panels <bold>(d)</bold>–<bold>(f)</bold> show the residuals as black dots with <inline-formula><mml:math id="M199" display="inline"><mml:mrow><mml:mn mathvariant="normal">2</mml:mn><mml:mi mathvariant="italic">σ</mml:mi></mml:mrow></mml:math></inline-formula> error bars.</p></caption>
          <?xmltex \igopts{width=455.244094pt}?><graphic xlink:href="https://acp.copernicus.org/articles/19/2135/2019/acp-19-2135-2019-f01.png"/>

        </fig>

      <p id="d1e4135">For the sample time series (65<inline-formula><mml:math id="M200" display="inline"><mml:msup><mml:mi/><mml:mo>∘</mml:mo></mml:msup></mml:math></inline-formula> S, 5<inline-formula><mml:math id="M201" display="inline"><mml:msup><mml:mi/><mml:mo>∘</mml:mo></mml:msup></mml:math></inline-formula> N, 65<inline-formula><mml:math id="M202" display="inline"><mml:msup><mml:mi/><mml:mo>∘</mml:mo></mml:msup></mml:math></inline-formula> N at 70 km),
the fits using the non-linear model Eq. (<xref ref-type="disp-formula" rid="Ch1.E9"/>)
are shown in Fig. <xref ref-type="fig" rid="Ch1.F2"/>a, b, and c.
The non-linear model better captures both the summer <inline-formula><mml:math id="M203" display="inline"><mml:mrow class="chem"><mml:mi mathvariant="normal">NO</mml:mi></mml:mrow></mml:math></inline-formula> variations as
well as the high values in the winter, especially at high northern latitudes.
However, at times of high solar activity (2003–2006) and
in particular at times of a strongly disturbed mesosphere (2004, 2006, 2012),
the residuals are still significant.
At high southern and low latitudes, the improvement over the linear model is
less evident.
At low latitudes, the <inline-formula><mml:math id="M204" display="inline"><mml:mrow class="chem"><mml:mi mathvariant="normal">NO</mml:mi></mml:mrow></mml:math></inline-formula> content is apparently mostly related
to the 11-year solar cycle, and the particle influence is
suppressed.
Since this cycle is covered by the Lyman-<inline-formula><mml:math id="M205" display="inline"><mml:mi mathvariant="italic">α</mml:mi></mml:math></inline-formula> index,
both models perform similarly, but the non-linear version
has one less parameter.
In both regions the residuals show traces of seasonal variations
that are not related to particle effects.
The linear model appears to capture these variations better than the non-linear model.
However, by objective measures including the number of model parameters<fn id="Ch1.Footn1"><p id="d1e4193">Past and recent research in model selection provides a number of
choices on how to compare models objectively.
The results are so-called information criteria which aim to provide a
consistent way of how to compare models,
most notably the “Akaike information criterion” (AIC; <xref ref-type="bibr" rid="bib1.bibx1" id="altparen.63"/>),
the “Bayesian information criterion” or “Schwarz criterion” (BIC or SIC; <xref ref-type="bibr" rid="bib1.bibx55" id="altparen.64"/>),
the “deviance information criterion” (DIC; <xref ref-type="bibr" rid="bib1.bibx64 bib1.bibx2" id="altparen.65"/>),
or the “widely applicable information criterion” (WAIC; <xref ref-type="bibr" rid="bib1.bibx69 bib1.bibx67" id="altparen.66"/>).
Alternatively, the “standardized mean squared error” (SMSE)
or the “mean standardized log loss” (MSLL) <xref ref-type="bibr" rid="bib1.bibx53" id="paren.67"><named-content content-type="post">chap. 2</named-content></xref>
gives an impression of the quality of regression models with respect to each other.</p></fn>,
the non-linear version fits the data better in all bins (not shown here).
At high southern latitudes, the SCIAMACHY data are less
densely sampled compared to high northern latitudes
(see <xref ref-type="bibr" rid="bib1.bibx15" id="altparen.68"/>).
In addition to the sampling differences, geomagnetic latitudes encompass
a wider geographic range in the Southern Hemisphere (SH)
than in the Northern Hemisphere (NH), and the AE index is derived from stations in the NH.
Both effects can lower the <inline-formula><mml:math id="M206" display="inline"><mml:mrow class="chem"><mml:mi mathvariant="normal">NO</mml:mi></mml:mrow></mml:math></inline-formula> concentrations that SCIAMACHY
observes in the SH, particularly at the winter maxima.
The lifetime variation that improves the fit in the NH
is thus less effective in the SH.</p>

      <?xmltex \floatpos{t}?><fig id="Ch1.F2" specific-use="star"><caption><p id="d1e4228">Same as Fig. <xref ref-type="fig" rid="Ch1.F1"/> but for the non-linear model.</p></caption>
          <?xmltex \igopts{width=455.244094pt}?><graphic xlink:href="https://acp.copernicus.org/articles/19/2135/2019/acp-19-2135-2019-f02.png"/>

        </fig>

</sec>
<sec id="Ch1.S5.SS2">
  <title>Parameter morphologies</title>
      <p id="d1e4245">Using the non-linear model, we show the latitude–altitude
distributions of the medians of the sampled
Lyman-<inline-formula><mml:math id="M207" display="inline"><mml:mi mathvariant="italic">α</mml:mi></mml:math></inline-formula> and geomagnetic
index coefficients in Fig. <xref ref-type="fig" rid="Ch1.F3"/>.
The white regions indicate
values outside of the 95 % confidence region
or whose sampled distribution has a skewness larger than 0.33.
The MCMC method samples the parameter probability
distributions. Since we require the geomagnetic index
and constant lifetime parameters to be larger than zero (see Table <xref ref-type="table" rid="Ch1.T1"/>),
these sampled distributions are sometimes skewed towards zero, even though
the 95 % credible region is still larger than zero.
Excluding heavily skewed distributions avoids those cases
because the
“true” parameter is apparently zero.</p>

      <?xmltex \floatpos{t}?><fig id="Ch1.F3" specific-use="star"><caption><p id="d1e4261">Latitude–altitude distributions of the fitted solar index parameter
(Lyman-<inline-formula><mml:math id="M208" display="inline"><mml:mi mathvariant="italic">α</mml:mi></mml:math></inline-formula>, <bold>a</bold>) and the geomagnetic index parameter (AE, <bold>b</bold>) from
the non-linear model. </p></caption>
          <?xmltex \igopts{width=384.112205pt}?><graphic xlink:href="https://acp.copernicus.org/articles/19/2135/2019/acp-19-2135-2019-f03.png"/>

        </fig>

      <p id="d1e4283">The Lyman-<inline-formula><mml:math id="M209" display="inline"><mml:mi mathvariant="italic">α</mml:mi></mml:math></inline-formula> parameter distribution shows that its
largest influence is at middle and low latitudes between
65 and 80 km.
Another increase of the Lyman-<inline-formula><mml:math id="M210" display="inline"><mml:mi mathvariant="italic">α</mml:mi></mml:math></inline-formula>
coefficient is indicated at higher altitudes above 90 km.
The penetration of Lyman-<inline-formula><mml:math id="M211" display="inline"><mml:mi mathvariant="italic">α</mml:mi></mml:math></inline-formula> radiation decreases with
decreasing altitude as a result of scattering and absorption
by air molecules.
On the other hand, the concentration of air decreases with altitude.
At this stage we do not have an unambiguous explanation of this behaviour,
but it may be related to reaction pathways as laid out
by <xref ref-type="bibr" rid="bib1.bibx49" id="text.69"/>, which would relate the <inline-formula><mml:math id="M212" display="inline"><mml:mrow class="chem"><mml:mi mathvariant="normal">NO</mml:mi></mml:mrow></mml:math></inline-formula> concentrations to the <inline-formula><mml:math id="M213" display="inline"><mml:mrow class="chem"><mml:msub><mml:mi mathvariant="normal">CO</mml:mi><mml:mn mathvariant="normal">2</mml:mn></mml:msub></mml:mrow></mml:math></inline-formula>
and <inline-formula><mml:math id="M214" display="inline"><mml:mrow class="chem"><mml:msub><mml:mi mathvariant="normal">H</mml:mi><mml:mn mathvariant="normal">2</mml:mn></mml:msub><mml:mi mathvariant="normal">O</mml:mi></mml:mrow></mml:math></inline-formula> (or <inline-formula><mml:math id="M215" display="inline"><mml:mrow class="chem"><mml:mi mathvariant="normal">OH</mml:mi></mml:mrow></mml:math></inline-formula>, respectively) profiles.
The Lyman-<inline-formula><mml:math id="M216" display="inline"><mml:mi mathvariant="italic">α</mml:mi></mml:math></inline-formula> coefficients are all negative below 65 km.
We also observe negative values at high northern
latitude at all altitudes and at high southern latitudes
above 85 km.
These negative coefficients indicate that
<inline-formula><mml:math id="M217" display="inline"><mml:mrow class="chem"><mml:mi mathvariant="normal">NO</mml:mi></mml:mrow></mml:math></inline-formula> photodissociation or conversion to other
species outweighs its production via UV radiation in those places.
The north–south asymmetry may be related to sampling
and the difference in illumination with respect to
geomagnetic latitudes; see Sect. <xref ref-type="sec" rid="Ch1.S5.SS1"/>.</p>
      <p id="d1e4368">The geomagnetic influence is largest at high latitudes
between 50 and 75<inline-formula><mml:math id="M218" display="inline"><mml:msup><mml:mi/><mml:mo>∘</mml:mo></mml:msup></mml:math></inline-formula> above about 65 km.
The AE coefficients peak at around 72 km and indicate
a further increase above 90 km.
This pattern of the geomagnetic influence matches
the one found in <xref ref-type="bibr" rid="bib1.bibx60" id="text.70"/>.
Unfortunately both increased influences above 90 km
in Lyman-<inline-formula><mml:math id="M219" display="inline"><mml:mi mathvariant="italic">α</mml:mi></mml:math></inline-formula> and AE
cannot be studied at higher latitudes due to a large a priori contribution
to the data.</p>
      <?pagebreak page2141?><p id="d1e4391">The latitude–altitude distributions of the lifetime parameters
are shown in Fig. <xref ref-type="fig" rid="Ch1.F4"/>.
All values shown are within the
95 % confidence region.
As for the coefficients above, we also exclude regions where
the skewness was larger than 0.33.</p>

      <?xmltex \floatpos{t}?><fig id="Ch1.F4" specific-use="star"><caption><p id="d1e4398">Latitude–altitude distributions of the fitted
base lifetime <inline-formula><mml:math id="M220" display="inline"><mml:mrow><mml:msub><mml:mi mathvariant="italic">τ</mml:mi><mml:mn mathvariant="normal">0</mml:mn></mml:msub></mml:mrow></mml:math></inline-formula> <bold>(a)</bold> and the amplitude of the
annual variation <inline-formula><mml:math id="M221" display="inline"><mml:mrow><mml:mo fence="true">|</mml:mo><mml:msub><mml:mi mathvariant="italic">τ</mml:mi><mml:mi>t</mml:mi></mml:msub><mml:mo fence="true">|</mml:mo></mml:mrow></mml:math></inline-formula> <bold>(b)</bold>
from the non-linear model.
</p></caption>
          <?xmltex \igopts{width=384.112205pt}?><graphic xlink:href="https://acp.copernicus.org/articles/19/2135/2019/acp-19-2135-2019-f04.png"/>

        </fig>

      <p id="d1e4439">The constant part of the lifetime, <inline-formula><mml:math id="M222" display="inline"><mml:mrow><mml:msub><mml:mi mathvariant="italic">τ</mml:mi><mml:mn mathvariant="normal">0</mml:mn></mml:msub></mml:mrow></mml:math></inline-formula>, is below
2 days in most bins, except for exceptionally large
values (<inline-formula><mml:math id="M223" display="inline"><mml:mrow><mml:mo>&gt;</mml:mo><mml:mn mathvariant="normal">10</mml:mn></mml:mrow></mml:math></inline-formula> days) at low latitudes (0–20<inline-formula><mml:math id="M224" display="inline"><mml:msup><mml:mi/><mml:mo>∘</mml:mo></mml:msup></mml:math></inline-formula> N) between 68 and 74 km.
Although we constrained the lifetime with an exponential
prior distribution, these large values apparently resulted
in a better fit to the data.
One explanation could be that because of the small geomagnetic
influence (the AE coefficient is small in this region),
the lifetime is more or less irrelevant.
The amplitude of the annual variation
(<inline-formula><mml:math id="M225" display="inline"><mml:mrow><mml:mo fence="true">|</mml:mo><mml:msub><mml:mi mathvariant="italic">τ</mml:mi><mml:mi mathvariant="normal">t</mml:mi></mml:msub><mml:mo fence="true">|</mml:mo><mml:mo>=</mml:mo><mml:msqrt><mml:mrow><mml:msubsup><mml:mi mathvariant="italic">τ</mml:mi><mml:mi mathvariant="normal">cos</mml:mi><mml:mn mathvariant="normal">2</mml:mn></mml:msubsup><mml:mo>+</mml:mo><mml:msubsup><mml:mi mathvariant="italic">τ</mml:mi><mml:mi mathvariant="normal">sin</mml:mi><mml:mn mathvariant="normal">2</mml:mn></mml:msubsup></mml:mrow></mml:msqrt><mml:mo>=</mml:mo><mml:msqrt><mml:mrow><mml:msup><mml:mi>d</mml:mi><mml:mn mathvariant="normal">2</mml:mn></mml:msup><mml:mo>+</mml:mo><mml:msup><mml:mi>e</mml:mi><mml:mn mathvariant="normal">2</mml:mn></mml:msup></mml:mrow></mml:msqrt></mml:mrow></mml:math></inline-formula>; see Eq. <xref ref-type="disp-formula" rid="Ch1.E12"/>)
is largest at high latitudes in the Northern Hemisphere
and at middle latitudes in the Southern Hemisphere.
This difference could be linked to the geomagnetic latitudes
which include a wider range of geographic latitudes in the Southern
Hemisphere compared to the Northern Hemisphere.
Therefore, the annual variation is less apparent in the Southern Hemisphere.
The amplitude also increases with decreasing altitude below 75 km
at middle and high latitudes
and with increasing altitude above that.
The increasing annual variation at low altitudes
can be the result of transport processes that
are not explicitly treated in our approach.
Note that the term <italic>lifetime</italic> is not
a pure (photo)chemical lifetime; rather it indicates
how long the AE signal persists in the <inline-formula><mml:math id="M226" display="inline"><mml:mrow class="chem"><mml:mi mathvariant="normal">NO</mml:mi></mml:mrow></mml:math></inline-formula> densities.
In that sense it combines the
(photo)chemical lifetime with transport effects
as discussed in <xref ref-type="bibr" rid="bib1.bibx60" id="text.71"/>.</p>
</sec>
<sec id="Ch1.S5.SS3">
  <title>Parameter profiles</title>
      <p id="d1e4546">For three selected latitude bins in the Northern Hemisphere
(5, 35, and 65<inline-formula><mml:math id="M227" display="inline"><mml:msup><mml:mi/><mml:mo>∘</mml:mo></mml:msup></mml:math></inline-formula> N)
we present profiles of the fitted parameters in Fig. <xref ref-type="fig" rid="Ch1.F5"/>.
The solid line indicates the median, and
the error bars indicate the 95 % confidence region.
As indicated in Fig. <xref ref-type="fig" rid="Ch1.F3"/>, the solar radiation
influence is largest between 65 and 80 km.
Its influence is also up to a factor of 2
larger at low and middle latitudes compared to high latitudes,
where the coefficient only differs significantly from zero
below 65 and above 82 km.
Similarly, the geomagnetic impact decreases with decreasing latitude by
1 order of magnitude from high to middle latitudes and at least
a further factor of 5 to lower latitudes.
The largest impact is around 70–72 km and possibly above 90 km
at high latitudes and is approximately constant between 66 and
76 km at middle and low latitudes.
Note that the scale in Fig. <xref ref-type="fig" rid="Ch1.F5"/>b
is logarithmic.
The lifetime variation shows that at high latitudes,
geomagnetically affected<?pagebreak page2142?> <inline-formula><mml:math id="M228" display="inline"><mml:mrow class="chem"><mml:mi mathvariant="normal">NO</mml:mi></mml:mrow></mml:math></inline-formula> persists longer during winter (the phase is close to zero for all altitudes at 65<inline-formula><mml:math id="M229" display="inline"><mml:msup><mml:mi/><mml:mo>∘</mml:mo></mml:msup></mml:math></inline-formula> N,
not shown here).
It persists up to 10 days longer between 85 and 70 km
and increasingly longer below, reaching 28 days at 60 km.</p>

      <?xmltex \floatpos{t}?><fig id="Ch1.F5" specific-use="star"><caption><p id="d1e4584">Coefficient profiles of the solar index parameter (Lyman-<inline-formula><mml:math id="M230" display="inline"><mml:mi mathvariant="italic">α</mml:mi></mml:math></inline-formula>,
left, <bold>a</bold>), the geomagnetic index parameter (AE, middle, <bold>b</bold>), and the
amplitude of the annual variation of the <inline-formula><mml:math id="M231" display="inline"><mml:mrow class="chem"><mml:mi mathvariant="normal">NO</mml:mi></mml:mrow></mml:math></inline-formula> lifetime (right, <bold>c</bold>) at
5 (green), 35 (orange), and 65<inline-formula><mml:math id="M232" display="inline"><mml:msup><mml:mi/><mml:mo>∘</mml:mo></mml:msup></mml:math></inline-formula> N (blue). The solid
line indicates the median, and the error bars indicate the 95 % confidence
region.</p></caption>
          <?xmltex \igopts{width=369.885827pt}?><graphic xlink:href="https://acp.copernicus.org/articles/19/2135/2019/acp-19-2135-2019-f05.png"/>

        </fig>

      <p id="d1e4627">For the same latitude bins in the Southern Hemisphere (5, 35, and 65<inline-formula><mml:math id="M233" display="inline"><mml:msup><mml:mi/><mml:mo>∘</mml:mo></mml:msup></mml:math></inline-formula> S) we present profiles of the fitted parameters in
Fig. <xref ref-type="fig" rid="Ch1.F6"/>. Similar to the coefficients in the Northern
Hemisphere (see Fig. <xref ref-type="fig" rid="Ch1.F5"/>), the solar radiation influence is
largest between 65 and 80 km and also up to a factor of 2 larger at
low and middle latitudes compared to high latitudes. However, the
Lyman-<inline-formula><mml:math id="M234" display="inline"><mml:mi mathvariant="italic">α</mml:mi></mml:math></inline-formula> coefficients at 65<inline-formula><mml:math id="M235" display="inline"><mml:msup><mml:mi/><mml:mo>∘</mml:mo></mml:msup></mml:math></inline-formula> S are significant below 82 km. Also
the geomagnetic AE coefficients show a similar pattern in the Southern
Hemisphere compared to the Northern Hemisphere, decreasing by orders of
magnitude from high to low latitudes. Note that the AE coefficients at high
latitudes are slightly lower than in the Northern Hemisphere, whereas the
coefficients at middle and low latitudes are slightly larger. This slight
asymmetry was also found in the study by <xref ref-type="bibr" rid="bib1.bibx60" id="text.72"/> and may be
related to AE being derived solely from stations in the Northern
Hemisphere <xref ref-type="bibr" rid="bib1.bibx43" id="paren.73"/>. With respect to latitude, the annual variation
of the lifetime seems to be reversed compared to the Northern Hemisphere,
with almost no variation at high latitudes and longer persisting <inline-formula><mml:math id="M236" display="inline"><mml:mrow class="chem"><mml:mi mathvariant="normal">NO</mml:mi></mml:mrow></mml:math></inline-formula> at
low latitudes. A faster descent in the southern polar vortex may be
responsible for the short lifetime at high southern latitudes. Another reason
may be the mixture of air from the inside and outside of the polar vertex when
averaging along geomagnetic latitudes since the 65<inline-formula><mml:math id="M237" display="inline"><mml:msup><mml:mi/><mml:mo>∘</mml:mo></mml:msup></mml:math></inline-formula> S geomagnetic
latitude band includes geographic locations from about 45 to
85<inline-formula><mml:math id="M238" display="inline"><mml:msup><mml:mi/><mml:mo>∘</mml:mo></mml:msup></mml:math></inline-formula> S. A third possibility may be the exclusion of the South
Atlantic Anomaly from the retrieval <xref ref-type="bibr" rid="bib1.bibx13 bib1.bibx15" id="paren.74"/>, for which presumably the particle-induced impact on <inline-formula><mml:math id="M239" display="inline"><mml:mrow class="chem"><mml:mi mathvariant="normal">NO</mml:mi></mml:mrow></mml:math></inline-formula> is largest.</p>

      <?xmltex \floatpos{t}?><fig id="Ch1.F6" specific-use="star"><caption><p id="d1e4707">Coefficient profiles of the solar index parameter (Lyman-<inline-formula><mml:math id="M240" display="inline"><mml:mi mathvariant="italic">α</mml:mi></mml:math></inline-formula>,
left, <bold>a</bold>), the geomagnetic index parameter (AE, middle, <bold>b</bold>), and the
amplitude of the annual variation of the <inline-formula><mml:math id="M241" display="inline"><mml:mrow class="chem"><mml:mi mathvariant="normal">NO</mml:mi></mml:mrow></mml:math></inline-formula> lifetime (right, <bold>c</bold>) at
5 (green), 35 (orange), and 65<inline-formula><mml:math id="M242" display="inline"><mml:msup><mml:mi/><mml:mo>∘</mml:mo></mml:msup></mml:math></inline-formula> S (blue). The solid
line indicates the median, and the error bars indicate the 95 % confidence
region.</p></caption>
          <?xmltex \igopts{width=369.885827pt}?><graphic xlink:href="https://acp.copernicus.org/articles/19/2135/2019/acp-19-2135-2019-f06.png"/>

        </fig>

</sec>
<sec id="Ch1.S5.SS4">
  <title>Discussion</title>
      <p id="d1e4757">The distribution of the parameters confirms our understanding
of the processes producing <inline-formula><mml:math id="M243" display="inline"><mml:mrow class="chem"><mml:mi mathvariant="normal">NO</mml:mi></mml:mrow></mml:math></inline-formula> in the mesosphere to a large extent. The Lyman-<inline-formula><mml:math id="M244" display="inline"><mml:mi mathvariant="italic">α</mml:mi></mml:math></inline-formula> coefficients are related to
radiative processes such as production by
UV or soft X-rays, either directly or via intermediary
of photoelectrons. The photons are not influenced by Earth's
magnetic field, and the influence of these processes is largest
at low latitudes and decreases towards higher latitudes.
We observe negative Lyman-<inline-formula><mml:math id="M245" display="inline"><mml:mi mathvariant="italic">α</mml:mi></mml:math></inline-formula> coefficients
below 65 km at all latitudes and at high
northern latitudes above 80 km.
These negative Lyman-<inline-formula><mml:math id="M246" display="inline"><mml:mi mathvariant="italic">α</mml:mi></mml:math></inline-formula> coefficients
indicate that at high solar activity,
photodissociation by <inline-formula><mml:math id="M247" display="inline"><mml:mrow><mml:mi mathvariant="italic">λ</mml:mi><mml:mo>&lt;</mml:mo><mml:mn mathvariant="normal">191</mml:mn><mml:mspace width="0.125em" linebreak="nobreak"/></mml:mrow></mml:math></inline-formula>nm photons,
photoionization by <inline-formula><mml:math id="M248" display="inline"><mml:mrow><mml:mi mathvariant="italic">λ</mml:mi><mml:mo>&lt;</mml:mo><mml:mn mathvariant="normal">134</mml:mn><mml:mspace width="0.125em" linebreak="nobreak"/></mml:mrow></mml:math></inline-formula>nm photons,
or collisional loss and conversion to other species
outweigh the production from higher energy photons (<inline-formula><mml:math id="M249" display="inline"><mml:mrow><mml:mo>&lt;</mml:mo><mml:mn mathvariant="normal">40</mml:mn></mml:mrow></mml:math></inline-formula> nm).
At high southern latitudes these negative Lyman-<inline-formula><mml:math id="M250" display="inline"><mml:mi mathvariant="italic">α</mml:mi></mml:math></inline-formula>
coefficients are not as pronounced as at high northern latitudes.
As mentioned in Sect. <xref ref-type="sec" rid="Ch1.S5.SS2"/>, this north–south
asymmetry<?pagebreak page2143?> may be related to sampling
and the difference in illumination with respect to
geomagnetic latitudes; see also Sect. <xref ref-type="sec" rid="Ch1.S5.SS1"/>.</p>
      <p id="d1e4837">The AE coefficients are largest at auroral latitudes, as
expected for the particle nature of the associated <inline-formula><mml:math id="M251" display="inline"><mml:mrow class="chem"><mml:mi mathvariant="normal">NO</mml:mi></mml:mrow></mml:math></inline-formula> production.
The AE coefficient can be considered an effective production rate
modulated by all short-term (<inline-formula><mml:math id="M252" display="inline"><mml:mrow><mml:mo>≪</mml:mo><mml:mn mathvariant="normal">1</mml:mn></mml:mrow></mml:math></inline-formula> day) processes.
To roughly estimate this production rate, we divided the
coefficient of the (daily) AE by 86 400 s, which follows
the approach in <xref ref-type="bibr" rid="bib1.bibx60" id="text.75"/>.
We find a maximum production rate of about
1 cm<inline-formula><mml:math id="M253" 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> nT<inline-formula><mml:math id="M254" display="inline"><mml:msup><mml:mi/><mml:mrow><mml:mo>-</mml:mo><mml:mn mathvariant="normal">1</mml:mn></mml:mrow></mml:msup></mml:math></inline-formula> s<inline-formula><mml:math id="M255" display="inline"><mml:msup><mml:mi/><mml:mrow><mml:mo>-</mml:mo><mml:mn mathvariant="normal">1</mml:mn></mml:mrow></mml:msup></mml:math></inline-formula> around 70–72 km.
This production rate also agrees with the one estimated
by <xref ref-type="bibr" rid="bib1.bibx60" id="text.76"/> through a superposed epoch analysis
of summertime <inline-formula><mml:math id="M256" display="inline"><mml:mrow class="chem"><mml:mi mathvariant="normal">NO</mml:mi></mml:mrow></mml:math></inline-formula>.
Comparing the <inline-formula><mml:math id="M257" display="inline"><mml:mrow class="chem"><mml:mi mathvariant="normal">NO</mml:mi></mml:mrow></mml:math></inline-formula> production to the ionization rates
from <xref ref-type="bibr" rid="bib1.bibx68" id="text.77"/> from 1 to 3 January 2005
(assuming approximately 1 <inline-formula><mml:math id="M258" display="inline"><mml:mrow class="chem"><mml:mi mathvariant="normal">NO</mml:mi></mml:mrow></mml:math></inline-formula> molecule per ion pair), our
model overestimates the ionization derived from AE on these days.
The AE values of 105, 355, and 435 nT translate
to 105, 355, and 435 <inline-formula><mml:math id="M259" display="inline"><mml:mrow class="chem"><mml:mi mathvariant="normal">NO</mml:mi></mml:mrow></mml:math></inline-formula> molecules cm<inline-formula><mml:math id="M260" 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> s<inline-formula><mml:math id="M261" display="inline"><mml:msup><mml:mi/><mml:mrow><mml:mo>-</mml:mo><mml:mn mathvariant="normal">1</mml:mn></mml:mrow></mml:msup></mml:math></inline-formula>, about 4 times
larger than would be estimated from the ionization rates in <xref ref-type="bibr" rid="bib1.bibx68" id="text.78"/>, but
agreeing with <xref ref-type="bibr" rid="bib1.bibx60" id="text.79"/>.
The factor of 4 may be related to the slightly different locations,
around 60<inline-formula><mml:math id="M262" display="inline"><mml:msup><mml:mi/><mml:mo>∘</mml:mo></mml:msup></mml:math></inline-formula> N <xref ref-type="bibr" rid="bib1.bibx68" id="paren.80"/> compared to around 65<inline-formula><mml:math id="M263" display="inline"><mml:msup><mml:mi/><mml:mo>∘</mml:mo></mml:msup></mml:math></inline-formula> N
here and in <xref ref-type="bibr" rid="bib1.bibx60" id="text.81"/>, in which the ionization rates may be higher.</p>
      <p id="d1e4992">The associated constant part <inline-formula><mml:math id="M264" display="inline"><mml:mrow><mml:msub><mml:mi mathvariant="italic">τ</mml:mi><mml:mn mathvariant="normal">0</mml:mn></mml:msub></mml:mrow></mml:math></inline-formula> of the lifetime ranges
from around 1 to around 4 days, except for large <inline-formula><mml:math id="M265" display="inline"><mml:mrow><mml:msub><mml:mi mathvariant="italic">τ</mml:mi><mml:mn mathvariant="normal">0</mml:mn></mml:msub></mml:mrow></mml:math></inline-formula> at
low latitudes around 70 km.
As already discussed in Sect. <xref ref-type="sec" rid="Ch1.S5.SS2"/>,
these large lifetimes may be a
side effect of the small geomagnetic coefficients and more
or less arbitrary.
The magnitude is similar to what was found in the
study by <xref ref-type="bibr" rid="bib1.bibx60" id="text.82"/> using only the summer data.</p>
      <?pagebreak page2144?><p id="d1e5022">The annual variation of the lifetime is largest at high
northern latitudes with a nearly constant amplitude of
10 days between 70 and 85 km.
An empirical lifetime of 10 days in winter was used by <xref ref-type="bibr" rid="bib1.bibx60" id="text.83"/>
to extend the <inline-formula><mml:math id="M266" display="inline"><mml:mrow class="chem"><mml:mi mathvariant="normal">NO</mml:mi></mml:mrow></mml:math></inline-formula> predicted by the summer analysis
to the larger values in winter.
Here we could confirm that 10 days is a good approximation
of the <inline-formula><mml:math id="M267" display="inline"><mml:mrow class="chem"><mml:mi mathvariant="normal">NO</mml:mi></mml:mrow></mml:math></inline-formula> lifetime in winter, but it varies with altitude.
The altitude distribution agrees with the increasing
photochemical lifetime at large solar zenith angles <xref ref-type="bibr" rid="bib1.bibx60" id="paren.84"><named-content content-type="post">Fig. 7b</named-content></xref>.
The larger values in our study are similarly related to
transport and mixing effects which alter the observed lifetime.
The small variation of the lifetime at high southern
latitudes could be a sampling issue because
SCIAMACHY only observes small variations there in winter
(see Figs. <xref ref-type="fig" rid="Ch1.F1"/> and <xref ref-type="fig" rid="Ch1.F2"/>).
Note that the results (in particular the large annual variation)
in the northernmost latitude bin should be taken with caution
because this bin is sparsely sampled by SCIAMACHY, and
the large winter <inline-formula><mml:math id="M268" display="inline"><mml:mrow class="chem"><mml:mi mathvariant="normal">NO</mml:mi></mml:mrow></mml:math></inline-formula> concentrations are
actually absent from the data.</p>
</sec>
</sec>
<sec id="Ch1.S6" sec-type="conclusions">
  <title>Conclusions</title>
      <p id="d1e5069">We propose an empirical model to estimate the <inline-formula><mml:math id="M269" display="inline"><mml:mrow class="chem"><mml:mi mathvariant="normal">NO</mml:mi></mml:mrow></mml:math></inline-formula> density in the
mesosphere (60–90 km) derived from measurements from SCIAMACHY nominal-mode limb scans. Our model calculates <inline-formula><mml:math id="M270" display="inline"><mml:mrow class="chem"><mml:mi mathvariant="normal">NO</mml:mi></mml:mrow></mml:math></inline-formula> number densities for
geomagnetic latitudes using the solar Lyman-<inline-formula><mml:math id="M271" display="inline"><mml:mi mathvariant="italic">α</mml:mi></mml:math></inline-formula> index and the
geomagnetic AE index. Two approaches were tested, a linear approach
containing annual and semi-annual harmonics and a non-linear version using a
finite and variable lifetime for the geomagnetically induced variations. From
our proposed models, the linear variant only describes part of the <inline-formula><mml:math id="M272" display="inline"><mml:mrow class="chem"><mml:mi mathvariant="normal">NO</mml:mi></mml:mrow></mml:math></inline-formula>
variations. It can describe the summer variations but underestimates the
large number densities during winter. The non-linear version derived from
the SCIAMACHY <inline-formula><mml:math id="M273" display="inline"><mml:mrow class="chem"><mml:mi mathvariant="normal">NO</mml:mi></mml:mrow></mml:math></inline-formula> data describes both variations using an annually
varying finite lifetime for the particle-induced <inline-formula><mml:math id="M274" display="inline"><mml:mrow class="chem"><mml:mi mathvariant="normal">NO</mml:mi></mml:mrow></mml:math></inline-formula>. However, in cases
of dynamic disturbances of the mesosphere, for example, in early 2004 or in
early 2006, the indirectly enhanced <inline-formula><mml:math id="M275" display="inline"><mml:mrow class="chem"><mml:mi mathvariant="normal">NO</mml:mi></mml:mrow></mml:math></inline-formula> <xref ref-type="bibr" rid="bib1.bibx51" id="paren.85"><named-content content-type="pre">see, for
example,</named-content></xref> is not captured by the model. These remaining
variations are treated as statistical noise.</p>
      <p id="d1e5133"><xref ref-type="bibr" rid="bib1.bibx60" id="text.86"/> use a superposed epoch analysis limited to
the polar summer to model the <inline-formula><mml:math id="M276" display="inline"><mml:mrow class="chem"><mml:mi mathvariant="normal">NO</mml:mi></mml:mrow></mml:math></inline-formula> data.
Here we extend that analysis to use all available SCIAMACHY
nominal-mode <inline-formula><mml:math id="M277" display="inline"><mml:mrow class="chem"><mml:mi mathvariant="normal">NO</mml:mi></mml:mrow></mml:math></inline-formula> data for all seasons.
However, during summer the present results show comparable
<inline-formula><mml:math id="M278" display="inline"><mml:mrow class="chem"><mml:mi mathvariant="normal">NO</mml:mi></mml:mrow></mml:math></inline-formula> production per AE unit and similar lifetimes to
the <xref ref-type="bibr" rid="bib1.bibx60" id="text.87"/> study.</p>
      <p id="d1e5165">The parameter distributions indicate in which regions the different processes
are significant.
We find that these distributions match our current understanding of
the processes producing and depleting <inline-formula><mml:math id="M279" display="inline"><mml:mrow class="chem"><mml:mi mathvariant="normal">NO</mml:mi></mml:mrow></mml:math></inline-formula> in the
mesosphere <xref ref-type="bibr" rid="bib1.bibx31 bib1.bibx32 bib1.bibx33 bib1.bibx60 bib1.bibx37 bib1.bibx40" id="paren.88"/>.
In particular, the influence of Lyman-<inline-formula><mml:math id="M280" display="inline"><mml:mi mathvariant="italic">α</mml:mi></mml:math></inline-formula>
(or solar UV radiation in general) is largest at low and middle latitudes,
which is explained by the direct production of <inline-formula><mml:math id="M281" display="inline"><mml:mrow class="chem"><mml:mi mathvariant="normal">NO</mml:mi></mml:mrow></mml:math></inline-formula> via solar UV
or soft X-ray radiation <xref ref-type="bibr" rid="bib1.bibx8 bib1.bibx9" id="paren.89"/>.
The geomagnetic influence is largest at high latitudes and is best explained by the
production from charged particles that enter the atmosphere
in the polar regions along the magnetic field.</p>
      <p id="d1e5197">A potential improvement would be to use actual measurements of precipitating particles
instead of the AE index.
Using measured fluxes could help to confirm our current understanding
of how those fluxes relate to ionization <xref ref-type="bibr" rid="bib1.bibx66 bib1.bibx68" id="paren.90"/>
and subsequent <inline-formula><mml:math id="M282" display="inline"><mml:mrow class="chem"><mml:mi mathvariant="normal">NO</mml:mi></mml:mrow></mml:math></inline-formula> production <xref ref-type="bibr" rid="bib1.bibx60" id="paren.91"/>.
Furthermore, including dynamical transport, as for example
in <xref ref-type="bibr" rid="bib1.bibx33" id="text.92"/>, could improve our knowledge of the
combined direct and indirect <inline-formula><mml:math id="M283" display="inline"><mml:mrow class="chem"><mml:mi mathvariant="normal">NO</mml:mi></mml:mrow></mml:math></inline-formula> production in the mesosphere.</p>
</sec>

      
      </body>
    <back><notes notes-type="codedataavailability">

      <p id="d1e5229">The SCIAMACHY <inline-formula><mml:math id="M284" display="inline"><mml:mrow class="chem"><mml:mi mathvariant="normal">NO</mml:mi></mml:mrow></mml:math></inline-formula> data set used in this study
is available at <ext-link xlink:href="https://doi.org/10.5281/zenodo.804371" ext-link-type="DOI">10.5281/zenodo.804371</ext-link>
<xref ref-type="bibr" rid="bib1.bibx14" id="paren.93"/>. The Python code to prepare the data (daily zonal
averaging) and to perform the analysis is available at
<ext-link xlink:href="https://doi.org/10.5281/zenodo.1401370" ext-link-type="DOI">10.5281/zenodo.1401370</ext-link> <xref ref-type="bibr" rid="bib1.bibx11" id="paren.94"/>. The daily zonal
mean <inline-formula><mml:math id="M285" display="inline"><mml:mrow class="chem"><mml:mi mathvariant="normal">NO</mml:mi></mml:mrow></mml:math></inline-formula> data and the sampled parameter distributions
are available at
<ext-link xlink:href="https://doi.org/10.5281/zenodo.1342701" ext-link-type="DOI">10.5281/zenodo.1342701</ext-link> <xref ref-type="bibr" rid="bib1.bibx12" id="paren.95"/>.
The solar Lyman-<inline-formula><mml:math id="M286" display="inline"><mml:mi mathvariant="italic">α</mml:mi></mml:math></inline-formula> index data were downloaded from
<uri>http://lasp.colorado.edu/lisird/data/composite_lyman_alpha/</uri> (last access: 19 December 2017),
the AE index data were downloaded from <uri>http://wdc.kugi.kyoto-u.ac.jp/aedir/</uri> (last access: 13 February 2019),
and the daily mean values used in this study are available within the
aforementioned data set.</p>
  </notes><notes notes-type="authorcontribution">

      <p id="d1e5283">SB developed the model, prepared and performed the
data analysis, and set up the paper. MS provided input on the model
and the idea of a variable <inline-formula><mml:math id="M287" display="inline"><mml:mrow class="chem"><mml:mi mathvariant="normal">NO</mml:mi></mml:mrow></mml:math></inline-formula> lifetime.
JPB and PJE contributed to the discussion and use of language.
All authors contributed to
the interpretation and discussion of the method and the results as well as
to the writing of the paper.</p>
  </notes><notes notes-type="competinginterests">

      <p id="d1e5297">The authors declare that they have no conflict of
interest.</p>
  </notes><ack><title>Acknowledgements</title><p id="d1e5303">Stefan Bender and Miriam Sinnhuber thank the Helmholtz Association for funding part of this project
under grant number VH-NG-624.
Stefan Bender and Patrick J. Espy acknowledge support from the Birkeland Center
for Space Sciences (BCSS), supported by the Research Council of
Norway under grant number 223252/F50.
The SCIAMACHY project was a national contribution to ESA Envisat,
funded by German Aerospace (DLR),
the Dutch Space Agency, SNO, and the Belgium Federal Science Policy Office, BELSPO.
The University of Bremen as principal investigator has led the scientific
support and development of the SCIAMACHY instrument as well as the scientific exploitation of its
data products.
This work was performed on the Abel<?pagebreak page2145?> Cluster,
owned by the University of Oslo and Uninett/Sigma2,
and operated by the Department for Research Computing at USIT,
the University of Oslo IT department (<uri>https://www.hpc.uio.no/</uri>;
last access: 13 February 2019).
<?xmltex \hack{\newline}?><?xmltex \hack{\newline}?>
Edited by: Franz-Josef Lübken<?xmltex \hack{\newline}?>
Reviewed by: two anonymous referees</p></ack><ref-list>
    <title>References</title>

      <ref id="bib1.bibx1"><label>Akaike(1974)</label><mixed-citation>Akaike, H.: A new look at the statistical model identification, IEEE T. Automat. Contr., 19, 716–723,
<ext-link xlink:href="https://doi.org/10.1109/tac.1974.1100705" ext-link-type="DOI">10.1109/tac.1974.1100705</ext-link>, 1974.</mixed-citation></ref>
      <ref id="bib1.bibx2"><label>Ando(2011)</label><mixed-citation>Ando, T.: Predictive Bayesian Model Selection, Am. J. Math.-S, 31, 13–38, <ext-link xlink:href="https://doi.org/10.1080/01966324.2011.10737798" ext-link-type="DOI">10.1080/01966324.2011.10737798</ext-link>,
2011.</mixed-citation></ref>
      <ref id="bib1.bibx3"><label>Bailey et al.(2002)Bailey, Barth, and Solomon</label><mixed-citation>Bailey, S. M., Barth, C. A., and Solomon, S. C.: A model of nitric oxide in the
lower thermosphere, J. Geophys. Res., 107, 1205, <ext-link xlink:href="https://doi.org/10.1029/2001JA000258" ext-link-type="DOI">10.1029/2001JA000258</ext-link>,
2002.</mixed-citation></ref>
      <ref id="bib1.bibx4"><label>Baldwin and Dunkerton(2001)</label><mixed-citation>Baldwin, M. P. and Dunkerton, T. J.: Stratospheric Harbingers of Anomalous
Weather Regimes, Science, 294, 581–584, <ext-link xlink:href="https://doi.org/10.1126/science.1063315" ext-link-type="DOI">10.1126/science.1063315</ext-link>, 2001.</mixed-citation></ref>
      <ref id="bib1.bibx5"><label>Barth(1992)</label><mixed-citation>Barth, C. A.: Nitric oxide in the lower thermosphere, Planet. Space Sci., 40,
315–336, <ext-link xlink:href="https://doi.org/10.1016/0032-0633(92)90067-x" ext-link-type="DOI">10.1016/0032-0633(92)90067-x</ext-link>, 1992.</mixed-citation></ref>
      <ref id="bib1.bibx6"><label>Barth(1995)</label><mixed-citation>Barth, C. A.: Nitric Oxide in the Lower Thermosphere, in: The Upper Mesosphere
and Lower Thermosphere: A Review of Experiment and Theory, edited by: Johnson,
R. M. and Killeen, T. L., American Geophysical Union,
225–233,  <ext-link xlink:href="https://doi.org/10.1029/gm087p0225" ext-link-type="DOI">10.1029/gm087p0225</ext-link>, 1995.</mixed-citation></ref>
      <ref id="bib1.bibx7"><label>Barth(2010)</label><mixed-citation>Barth, C. A.: Joule heating and nitric oxide in the thermosphere, 2, J.
Geophys. Res.-Space, 115, A10305,
<ext-link xlink:href="https://doi.org/10.1029/2010ja015565" ext-link-type="DOI">10.1029/2010ja015565</ext-link>, 2010.</mixed-citation></ref>
      <ref id="bib1.bibx8"><label>Barth et al.(1988)Barth, Tobiska, Siskind, and Cleary</label><mixed-citation>Barth, C. A., Tobiska, W. K., Siskind, D. E., and Cleary, D. D.:
Solar-terrestrial coupling: Low-latitude thermospheric nitric oxide, Geophys.
Res. Lett., 15, 92–94, <ext-link xlink:href="https://doi.org/10.1029/gl015i001p00092" ext-link-type="DOI">10.1029/gl015i001p00092</ext-link>, 1988.</mixed-citation></ref>
      <ref id="bib1.bibx9"><label>Barth et al.(2003)Barth, Mankoff, Bailey, and Solomon</label><mixed-citation>Barth, C. A., Mankoff, K. D., Bailey, S. M., and Solomon, S. C.: Global
observations of nitric oxide in the thermosphere, J. Geophys. Res., 108,
1027, <ext-link xlink:href="https://doi.org/10.1029/2002JA009458" ext-link-type="DOI">10.1029/2002JA009458</ext-link>, 2003.</mixed-citation></ref>
      <ref id="bib1.bibx10"><label>Barth et al.(2009)Barth, Lu, and Roble</label><mixed-citation>Barth, C. A., Lu, G., and Roble, R. G.: Joule heating and nitric oxide in the
thermosphere, J. Geophys. Res.-Space, 114, A05301,
<ext-link xlink:href="https://doi.org/10.1029/2008ja013765" ext-link-type="DOI">10.1029/2008ja013765</ext-link>, 2009.</mixed-citation></ref>
      <ref id="bib1.bibx11"><?xmltex \def\ref@label{{Bender(2018{\natexlab{a}})}}?><label>Bender(2018a)</label><mixed-citation>Bender, S.: st-bender/sciapy: Version 0.0.5 [source code],
<ext-link xlink:href="https://doi.org/10.5281/zenodo.1401370" ext-link-type="DOI">10.5281/zenodo.1401370</ext-link>,
2018a.</mixed-citation></ref>
      <ref id="bib1.bibx12"><?xmltex \def\ref@label{{Bender(2018{\natexlab{b}})}}?><label>Bender(2018b)</label><mixed-citation>Bender, S.: SCIAMACHY NO regression fit MCMC samples [data set],
<ext-link xlink:href="https://doi.org/10.5281/zenodo.1342701" ext-link-type="DOI">10.5281/zenodo.1342701</ext-link>,
2018b.</mixed-citation></ref>
      <ref id="bib1.bibx13"><?xmltex \def\ref@label{{Bender et~al.(2013)Bender, Sinnhuber, Burrows, Langowski, Funke, and
L\'{o}pez-Puertas}}?><label>Bender et al.(2013)Bender, Sinnhuber, Burrows, Langowski, Funke, and
López-Puertas</label><mixed-citation>Bender, S., Sinnhuber, M., Burrows, J. P., Langowski, M., Funke, B., and López-Puertas, M.: Retrieval of nitric oxide in the
mesosphere and lower thermosphere from SCIAMACHY limb spectra, Atmos. Meas. Tech., 6, 2521–2531, <ext-link xlink:href="https://doi.org/10.5194/amt-6-2521-2013" ext-link-type="DOI">10.5194/amt-6-2521-2013</ext-link>, 2013.</mixed-citation></ref>
      <ref id="bib1.bibx14"><?xmltex \def\ref@label{{Bender et~al.(2017{\natexlab{a}})Bender, Sinnhuber, Burrows, and
Langowski}}?><label>Bender et al.(2017a)Bender, Sinnhuber, Burrows, and
Langowski</label><mixed-citation>Bender, S., Sinnhuber, M., Burrows, J. P., and Langowski, M.: Nitric oxide (NO) data set (60–160 km) from SCIAMACHY nominal limb scans,
<ext-link xlink:href="https://doi.org/10.5281/zenodo.804371" ext-link-type="DOI">10.5281/zenodo.804371</ext-link>,
2017a.</mixed-citation></ref>
      <ref id="bib1.bibx15"><?xmltex \def\ref@label{{Bender et~al.(2017{\natexlab{b}})Bender, Sinnhuber, Langowski, and
Burrows}}?><label>Bender et al.(2017b)Bender, Sinnhuber, Langowski, and
Burrows</label><mixed-citation>Bender, S., Sinnhuber, M., Langowski, M., and Burrows, J. P.: Retrieval of nitric oxide in the mesosphere from SCIAMACHY nominal limb spectra, Atmos. Meas. Tech., 10, 209–220, <ext-link xlink:href="https://doi.org/10.5194/amt-10-209-2017" ext-link-type="DOI">10.5194/amt-10-209-2017</ext-link>,
2017b.</mixed-citation></ref>
      <ref id="bib1.bibx16"><?xmltex \def\ref@label{{Bermejo-Pantale\'{o}n et~al.(2011)Bermejo-Pantale\'{o}n, Funke,
L\'{o}pez-Puertas, Garc\'{i}a-Comas, Stiller, von Clarmann, Linden,
Grabowski, H\"{o}pfner, Kiefer, Glatthor, Kellmann, and
Lu}}?><label>Bermejo-Pantaleón et al.(2011)Bermejo-Pantaleón, Funke,
López-Puertas, García-Comas, Stiller, von Clarmann, Linden,
Grabowski, Höpfner, Kiefer, Glatthor, Kellmann, and
Lu</label><mixed-citation>Bermejo-Pantaleón, D., Funke, B., López-Puertas, M., García-Comas,
M., Stiller, G. P., von Clarmann, T., Linden, A., Grabowski, U., Höpfner,
M., Kiefer, M., Glatthor, N., Kellmann, S., and Lu, G.: Global observations
of thermospheric temperature and nitric oxide from MIPAS spectra at 5.3 <inline-formula><mml:math id="M288" display="inline"><mml:mrow class="unit"><mml:mi mathvariant="normal">µ</mml:mi></mml:mrow></mml:math></inline-formula>m, J. Geophys. Res., 116, A10313, <ext-link xlink:href="https://doi.org/10.1029/2011JA016752" ext-link-type="DOI">10.1029/2011JA016752</ext-link>,
2011.</mixed-citation></ref>
      <ref id="bib1.bibx17"><?xmltex \def\ref@label{{Bovensmann et~al.(1999)Bovensmann, Burrows, Buchwitz, Frerick,
No{\"{e}}l, Rozanov, Chance, and Goede}}?><label>Bovensmann et al.(1999)Bovensmann, Burrows, Buchwitz, Frerick,
Noël, Rozanov, Chance, and Goede</label><mixed-citation>Bovensmann, H., Burrows, J. P., Buchwitz, M., Frerick, J., Noël, S.,
Rozanov, V. V., Chance, K. V., and Goede, A. P. H.: SCIAMACHY: Mission
Objectives and Measurement Modes, J. Atmos. Sci., 56, 127–150,
<ext-link xlink:href="https://doi.org/10.1175/1520-0469(1999)056&lt;0127:SMOAMM&gt;2.0.CO;2" ext-link-type="DOI">10.1175/1520-0469(1999)056&lt;0127:SMOAMM&gt;2.0.CO;2</ext-link>, 1999.</mixed-citation></ref>
      <ref id="bib1.bibx18"><?xmltex \def\ref@label{{Burrows et~al.(1995)Burrows, H\"{o}lzle, Goede, Visser, and
Fricke}}?><label>Burrows et al.(1995)Burrows, Hölzle, Goede, Visser, and
Fricke</label><mixed-citation>Burrows, J. P., Hölzle, E., Goede, A. P. H., Visser, H., and Fricke, W.:
SCIAMACHY – scanning imaging absorption spectrometer for atmospheric
chartography, Acta Astronaut., 35, 445–451,
<ext-link xlink:href="https://doi.org/10.1016/0094-5765(94)00278-T" ext-link-type="DOI">10.1016/0094-5765(94)00278-T</ext-link>,
1995.</mixed-citation></ref>
      <ref id="bib1.bibx19"><label>Crutzen(1970)</label><mixed-citation>Crutzen, P. J.: The influence of nitrogen oxides on the atmospheric ozone
content, Q. J. Roy. Meteor. Soc., 96, 320–325,
<ext-link xlink:href="https://doi.org/10.1002/qj.49709640815" ext-link-type="DOI">10.1002/qj.49709640815</ext-link>, 1970.</mixed-citation></ref>
      <ref id="bib1.bibx20"><label>Davis and Sugiura(1966)</label><mixed-citation>Davis, T. N. and Sugiura, M.: Auroral electrojet activity index AE and its
universal time variations, J. Geophys. Res., 71, 785–801,
<ext-link xlink:href="https://doi.org/10.1029/jz071i003p00785" ext-link-type="DOI">10.1029/jz071i003p00785</ext-link>, 1966.</mixed-citation></ref>
      <ref id="bib1.bibx21"><label>Foreman-Mackey et al.(2013)Foreman-Mackey, Hogg, Lang, and
Goodman</label><mixed-citation>Foreman-Mackey, D., Hogg, D. W., Lang, D., and Goodman, J.: emcee: The
MCMC Hammer, Publ. Astron. Soc. Pac., 125, 306–313, <ext-link xlink:href="https://doi.org/10.1086/670067" ext-link-type="DOI">10.1086/670067</ext-link>, 2013.</mixed-citation></ref>
      <ref id="bib1.bibx22"><label>Foreman-Mackey et al.(2017a)Foreman-Mackey, Agol, Angus, and
Ambikasaran</label><mixed-citation>Foreman-Mackey, D., Agol, E., Angus, R., and Ambikasaran, S.: Fast and
scalable Gaussian process modeling with applications to astronomical time
series, Astronomical Journal, 154,   220, <ext-link xlink:href="https://doi.org/10.3847/1538-3881/aa9332" ext-link-type="DOI">10.3847/1538-3881/aa9332</ext-link>,
2017a.</mixed-citation></ref>
      <ref id="bib1.bibx23"><label>Foreman-Mackey et al.(2017b)Foreman-Mackey, Agol, Angus, Brewer,
Austin, Farr, Guillochon, Czekala, and Casey</label><mixed-citation>Foreman-Mackey, D., Agol, E., Angus, R., Brewer, B. J., Austin, P., Farr,
W. M., Guillochon, J., Czekala, I., and Casey, A.: dfm/celerite: celerite
v0.3.0, <ext-link xlink:href="https://doi.org/10.5281/zenodo.1048287" ext-link-type="DOI">10.5281/zenodo.1048287</ext-link>, 2017b.</mixed-citation></ref>
      <ref id="bib1.bibx24"><label>Frost et al.(1956)Frost, McDowell, and Bawn</label><mixed-citation>Frost, D. C., McDowell, C. A., and Bawn, C. E. H.: The dissociation energy of
the nitrogen molecule, P. Roy. Soc. Lond. A, 236, 278–284,
<ext-link xlink:href="https://doi.org/10.1098/rspa.1956.0135" ext-link-type="DOI">10.1098/rspa.1956.0135</ext-link>, 1956.</mixed-citation></ref>
      <ref id="bib1.bibx25"><?xmltex \def\ref@label{{Funke et~al.(2001)Funke, L{\'{o}}pez-Puertas, Stiller, v.~Clarmann,
and H{\"{o}}pfner}}?><label>Funke et al.(2001)Funke, López-Puertas, Stiller, v. Clarmann,
and Höpfner</label><mixed-citation>Funke, B., López-Puertas, M., Stiller, G., v. Clarmann, T., and
Höpfner, M.: A new non-LTE retrieval method for atmospheric
parameters from mipas-envisat emission spectra, Adv. Space Res., 27,
1099–1104, <ext-link xlink:href="https://doi.org/10.1016/S0273-1177(01)00169-7" ext-link-type="DOI">10.1016/S0273-1177(01)00169-7</ext-link>,
2001.</mixed-citation></ref>
      <ref id="bib1.bibx26"><?xmltex \def\ref@label{{Funke et~al.(2005{\natexlab{a}})Funke, L\'{o}pez-Puertas,
Gil-L\'{o}pez, von Clarmann, Stiller, Fischer, and Kellmann}}?><label>Funke et al.(2005a)Funke, López-Puertas,
Gil-López, von Clarmann, Stiller, Fischer, and Kellmann</label><mixed-citation>Funke, B., López-Puertas, M., Gil-López, S., von Clarmann, T., Stiller,
G. P., Fischer, H., and Kellmann, S.: Downward transport of upper atmospheric
NOx into the polar stratosphere and lower mesosphere during the Antarctic
2003 and Arctic 2002/2003 winters, J. Geophys. Res., 110, D24308,
<ext-link xlink:href="https://doi.org/10.1029/2005JD006463" ext-link-type="DOI">10.1029/2005JD006463</ext-link>, 2005a.</mixed-citation></ref>
      <ref id="bib1.bibx27"><?xmltex \def\ref@label{{Funke et~al.(2005{\natexlab{b}})Funke, L\'{o}pez-Puertas,
{von~Clarmann}, Stiller, Fischer, Glatthor, Grabowski, H\"{o}pfner, Kellmann,
Kiefer, Linden, Mengistu~Tsidu, Milz, Steck, and Wang}}?><label>Funke et al.(2005b)Funke, López-Puertas,
von Clarmann, Stiller, Fischer, Glatthor, Grabowski, Höpfner, Kellmann,
Kiefer, Linden, Mengistu Tsidu, Milz, Steck, and Wang</label><mixed-citation>Funke, B., López-Puertas, M., von Clarmann, T., Stiller, G. P., Fischer,
H., Glatthor, N., Grabowski, U., Höpfner, M., Kellmann, S., Kiefer, M.,
Linden, A., Mengistu Tsidu, G., Milz, M., Steck, T., and Wang, D. Y.:
Retrieval of stratospheric NOx from 5.3 and 6.2 <inline-formula><mml:math id="M289" display="inline"><mml:mrow class="unit"><mml:mi mathvariant="normal">µ</mml:mi></mml:mrow></mml:math></inline-formula>m nonlocal
thermodynamic equilibrium emissions measured by Michelson Interferometer for
Passive Atmospheric Sounding (MIPAS) on Envisat, J. Geophys. Res., 110,
D09302, <ext-link xlink:href="https://doi.org/10.1029/2004JD005225" ext-link-type="DOI">10.1029/2004JD005225</ext-link>, 2005b.</mixed-citation></ref>
      <?pagebreak page2146?><ref id="bib1.bibx28"><?xmltex \def\ref@label{{Funke et~al.(2008{\natexlab{a}})Funke, Garc{\'{\i}}a-Comas,
L{\'{o}}pez-Puertas, Glatthor, Stiller, von Clarmann, Semeniuk, and
McConnell}}?><label>Funke et al.(2008a)Funke, García-Comas,
López-Puertas, Glatthor, Stiller, von Clarmann, Semeniuk, and
McConnell</label><mixed-citation>Funke, B., García-Comas, M., López-Puertas, M., Glatthor, N., Stiller, G. P., von Clarmann, T., Semeniuk, K., and McConnell, J. C.:
Enhancement of <inline-formula><mml:math id="M290" display="inline"><mml:mrow class="chem"><mml:msub><mml:mi mathvariant="normal">N</mml:mi><mml:mn mathvariant="normal">2</mml:mn></mml:msub><mml:mi mathvariant="normal">O</mml:mi></mml:mrow></mml:math></inline-formula> during the October–November 2003 solar proton events, Atmos. Chem. Phys., 8, 3805–3815, <ext-link xlink:href="https://doi.org/10.5194/acp-8-3805-2008" ext-link-type="DOI">10.5194/acp-8-3805-2008</ext-link>, 2008.  2008a.</mixed-citation></ref>
      <ref id="bib1.bibx29"><?xmltex \def\ref@label{{Funke et~al.(2008{\natexlab{b}})Funke, L{\'{o}}pez-Puertas,
Garcia-Comas, Stiller, von Clarmann, and Glatthor}}?><label>Funke et al.(2008b)Funke, López-Puertas,
Garcia-Comas, Stiller, von Clarmann, and Glatthor</label><mixed-citation>Funke, B., López-Puertas, M., Garcia-Comas, M., Stiller, G. P., von Clarmann, T., and Glatthor, N.: Mesospheric <inline-formula><mml:math id="M291" display="inline"><mml:mrow class="chem"><mml:msub><mml:mi mathvariant="normal">N</mml:mi><mml:mn mathvariant="normal">2</mml:mn></mml:msub><mml:mi mathvariant="normal">O</mml:mi></mml:mrow></mml:math></inline-formula>
enhancements as observed by MIPAS on Envisat during the polar winters in 2002–2004, Atmos. Chem. Phys., 8, 5787–5800, <ext-link xlink:href="https://doi.org/10.5194/acp-8-5787-2008" ext-link-type="DOI">10.5194/acp-8-5787-2008</ext-link>, 2008.
2008b.</mixed-citation></ref>
      <ref id="bib1.bibx30"><?xmltex \def\ref@label{{Funke et~al.(2012)Funke, L{\'{o}}pez-Puertas, Garc{\'{\i}}a-Comas,
Kaufmann, H{\"{o}}pfner, and Stiller}}?><label>Funke et al.(2012)Funke, López-Puertas, García-Comas,
Kaufmann, Höpfner, and Stiller</label><mixed-citation>Funke, B., López-Puertas, M., García-Comas, M., Kaufmann, M.,
Höpfner, M., and Stiller, G.: GRANADA: A Generic RAdiative traNsfer
AnD non-LTE population algorithm, J. Quant. Spectrosc. Ra.,
113, 1771–1817, <ext-link xlink:href="https://doi.org/10.1016/j.jqsrt.2012.05.001" ext-link-type="DOI">10.1016/j.jqsrt.2012.05.001</ext-link>,
2012.</mixed-citation></ref>
      <ref id="bib1.bibx31"><?xmltex \def\ref@label{{Funke et~al.(2014{\natexlab{a}})Funke, L{\'{o}}pez-Puertas, Holt,
Randall, Stiller, and von Clarmann}}?><label>Funke et al.(2014a)Funke, López-Puertas, Holt,
Randall, Stiller, and von Clarmann</label><mixed-citation>Funke, B., López-Puertas, M., Holt, L., Randall, C. E., Stiller, G. P.,
and von Clarmann, T.: Hemispheric distributions and interannual variability
of NO<inline-formula><mml:math id="M292" display="inline"><mml:msub><mml:mi/><mml:mi>y</mml:mi></mml:msub></mml:math></inline-formula> produced by energetic particle precipitation in 2002–2012,
J. Geophys. Res.-Atmos., 119, 13565–13582,
<ext-link xlink:href="https://doi.org/10.1002/2014jd022423" ext-link-type="DOI">10.1002/2014jd022423</ext-link>, 2014a.</mixed-citation></ref>
      <ref id="bib1.bibx32"><?xmltex \def\ref@label{{Funke et~al.(2014{\natexlab{b}})Funke, L{\'{o}}pez-Puertas, Stiller,
and von Clarmann}}?><label>Funke et al.(2014b)Funke, López-Puertas, Stiller,
and von Clarmann</label><mixed-citation>Funke, B., López-Puertas, M., Stiller, G. P., and von Clarmann, T.:
Mesospheric and stratospheric NOy produced by energetic particle
precipitation during 2002–2012, J. Geophys. Res., 119, 4429–4446,
<ext-link xlink:href="https://doi.org/10.1002/2013JD021404" ext-link-type="DOI">10.1002/2013JD021404</ext-link>, 2014b.</mixed-citation></ref>
      <ref id="bib1.bibx33"><label>Funke et al.(2016)</label><mixed-citation>Funke, B., López-Puertas, M., Stiller, G. P., Versick, S., and von Clarmann, T.: A semi-empirical model for mesospheric
and stratospheric <inline-formula><mml:math id="M293" display="inline"><mml:mrow class="chem"><mml:msub><mml:mi mathvariant="normal">NO</mml:mi><mml:mi>y</mml:mi></mml:msub></mml:mrow></mml:math></inline-formula> produced by energetic particle precipitation, Atmos. Chem. Phys., 16, 8667–8693, <ext-link xlink:href="https://doi.org/10.5194/acp-16-8667-2016" ext-link-type="DOI">10.5194/acp-16-8667-2016</ext-link>, 2016.</mixed-citation></ref>
      <ref id="bib1.bibx34"><label>Funke et al.(2017)</label><mixed-citation>Funke, B., Ball, W., Bender, S., Gardini, A., Harvey, V. L., Lambert, A., López-Puertas, M., Marsh, D. R., Meraner, K., Nieder, H.,
Päivärinta, S.-M., Pérot, K., Randall, C. E., Reddmann, T., Rozanov, E., Schmidt, H., Seppälä, A., Sinnhuber, M.,
Sukhodolov, T., Stiller, G. P., Tsvetkova, N. D., Verronen, P. T., Versick, S., von Clarmann, T., Walker, K. A., and Yushkov, V.:
HEPPA-II model-measurement intercomparison project: EPP indirect effects during the dynamically perturbed NH winter 2008–2009, Atmos. Chem. Phys., 17, 3573–3604, <ext-link xlink:href="https://doi.org/10.5194/acp-17-3573-2017" ext-link-type="DOI">10.5194/acp-17-3573-2017</ext-link>, 2017.</mixed-citation></ref>
      <ref id="bib1.bibx35"><label>Heays et al.(2017)Heays, Bosman, and van Dishoeck</label><mixed-citation>Heays, A. N., Bosman, A. D., and van Dishoeck, E. F.: Photodissociation and
photoionisation of atoms and molecules of astrophysical interest, Astron. Astrophys., 602, A105, <ext-link xlink:href="https://doi.org/10.1051/0004-6361/201628742" ext-link-type="DOI">10.1051/0004-6361/201628742</ext-link>, 2017.</mixed-citation></ref>
      <ref id="bib1.bibx36"><?xmltex \def\ref@label{{Hendrickx et~al.(2015)Hendrickx, Megner, Gumbel, Siskind, Orsolini,
Tyss{\o}y, and Hervig}}?><label>Hendrickx et al.(2015)Hendrickx, Megner, Gumbel, Siskind, Orsolini,
Tyssøy, and Hervig</label><mixed-citation>Hendrickx, K., Megner, L., Gumbel, J., Siskind, D. E., Orsolini, Y. J.,
Tyssøy, H. N., and Hervig, M.: Observation of 27 day solar cycles in the
production and mesospheric descent of EPP-produced NO, J. Geophys. Res.,
120, 8978–8988, <ext-link xlink:href="https://doi.org/10.1002/2015JA021441" ext-link-type="DOI">10.1002/2015JA021441</ext-link>, 2015JA021441, 2015.</mixed-citation></ref>
      <ref id="bib1.bibx37"><label>Hendrickx et al.(2017)Hendrickx, Megner, Marsh, Gumbel, Strandberg,
and Martinsson</label><mixed-citation>Hendrickx, K., Megner, L., Marsh, D. R., Gumbel, J., Strandberg, R., and
Martinsson, F.: Relative Importance of Nitric Oxide Physical Drivers in the
Lower Thermosphere, Geophys. Res. Lett., 44, 10,081–10,087,
<ext-link xlink:href="https://doi.org/10.1002/2017gl074786" ext-link-type="DOI">10.1002/2017gl074786</ext-link>,
2017.</mixed-citation></ref>
      <ref id="bib1.bibx38"><label>Hendrickx et al.(2018)Hendrickx, Megner, Marsh, and
Smith-Johnsen</label><mixed-citation>Hendrickx, K., Megner, L., Marsh, D. R., and Smith-Johnsen, C.: Production and transport mechanisms of NO in the polar upper mesosphere and
lower thermosphere in observations and models, Atmos. Chem. Phys., 18, 9075–9089, <ext-link xlink:href="https://doi.org/10.5194/acp-18-9075-2018" ext-link-type="DOI">10.5194/acp-18-9075-2018</ext-link>, 2018.</mixed-citation></ref>
      <ref id="bib1.bibx39"><label>Hendrie(1954)</label><mixed-citation>Hendrie, J. M.: Dissociation Energy of N<inline-formula><mml:math id="M294" display="inline"><mml:msub><mml:mi/><mml:mn mathvariant="normal">2</mml:mn></mml:msub></mml:math></inline-formula>, J. Chem. Phys., 22, 1503–1507, <ext-link xlink:href="https://doi.org/10.1063/1.1740449" ext-link-type="DOI">10.1063/1.1740449</ext-link>, 1954.</mixed-citation></ref>
      <ref id="bib1.bibx40"><?xmltex \def\ref@label{{Kiviranta et~al.(2018)Kiviranta, P{\'{e}}rot, Eriksson, and
Murtagh}}?><label>Kiviranta et al.(2018)Kiviranta, Pérot, Eriksson, and
Murtagh</label><mixed-citation>Kiviranta, J., Pérot, K., Eriksson, P., and Murtagh, D.: An empirical model of nitric oxide in the upper mesosphere and lower thermosphere
based on 12 years of Odin SMR measurements, Atmos. Chem. Phys., 18, 13393–13410, <ext-link xlink:href="https://doi.org/10.5194/acp-18-13393-2018" ext-link-type="DOI">10.5194/acp-18-13393-2018</ext-link>, 2018.</mixed-citation></ref>
      <ref id="bib1.bibx41"><label>Kockarts(1980)</label><mixed-citation>Kockarts, G.: Nitric oxide cooling in the terrestrial thermosphere, Geophys.
Res. Lett., 7, 137–140, <ext-link xlink:href="https://doi.org/10.1029/gl007i002p00137" ext-link-type="DOI">10.1029/gl007i002p00137</ext-link>, 1980.</mixed-citation></ref>
      <ref id="bib1.bibx42"><label>MacKay(2003)</label><mixed-citation>MacKay, D.: Information Theory, Inference and Learning Algorithms, Cambridge
University Press,
available at: <uri>http://www.inference.org.uk/mackay/itila/book.html</uri> (last access: 17 November 2017), 2003.</mixed-citation></ref>
      <ref id="bib1.bibx43"><label>Mandea and Korte(2011)</label><mixed-citation>Mandea, M. and Korte, M. (Eds.): Geomagnetic Observations and Models, vol. 5 of
IAGA Special Sopron Book Series, <ext-link xlink:href="https://doi.org/10.1007/978-90-481-9858-0" ext-link-type="DOI">10.1007/978-90-481-9858-0</ext-link>,
2011.</mixed-citation></ref>
      <ref id="bib1.bibx44"><label>Maric and Burrows(1992)</label><mixed-citation>Maric, D. and Burrows, J.: Formation of N<inline-formula><mml:math id="M295" display="inline"><mml:msub><mml:mi/><mml:mn mathvariant="normal">2</mml:mn></mml:msub></mml:math></inline-formula>O in the
photolysis/photoexcitation of NO, NO<inline-formula><mml:math id="M296" display="inline"><mml:msub><mml:mi/><mml:mn mathvariant="normal">2</mml:mn></mml:msub></mml:math></inline-formula> and air, J. Photochem.
Photobiol., A, 66, 291–312, <ext-link xlink:href="https://doi.org/10.1016/1010-6030(92)80002-d" ext-link-type="DOI">10.1016/1010-6030(92)80002-d</ext-link>, 1992.</mixed-citation></ref>
      <ref id="bib1.bibx45"><label>Marsh et al.(2004)Marsh, Solomon, and Reynolds</label><mixed-citation>Marsh, D. R., Solomon, S. C., and Reynolds, A. E.: Empirical model of nitric
oxide in the lower thermosphere, J. Geophys. Res., 109, A07301,
<ext-link xlink:href="https://doi.org/10.1029/2003JA010199" ext-link-type="DOI">10.1029/2003JA010199</ext-link>, 2004.</mixed-citation></ref>
      <ref id="bib1.bibx46"><?xmltex \def\ref@label{{Mat{\'{e}}rn(1960)}}?><label>Matérn(1960)</label><mixed-citation>Matérn, B.: Spatial Variation: Stachastic Models and Their Application to
Some Problems in Forest Surveys and Other Sampling Investigations, vol. 36 of
Lecture Notes in Statistics, Statens skogsforskningsinstitut,
Springer-Verlag New York, 2 edn., <ext-link xlink:href="https://doi.org/10.1007/978-1-4615-7892-5" ext-link-type="DOI">10.1007/978-1-4615-7892-5</ext-link>, originally
published by the Swedish National Institute for Forestry Research, 1960,
1960.</mixed-citation></ref>
      <ref id="bib1.bibx47"><label>Matthes et al.(2017)</label><mixed-citation>Matthes, K., Funke, B., Andersson, M. E., Barnard, L., Beer, J., Charbonneau, P., Clilverd, M. A., Dudok de Wit, T., Haberreiter, M.,
Hendry, A., Jackman, C. H., Kretzschmar, M., Kruschke, T., Kunze, M., Langematz, U., Marsh, D. R., Maycock, A. C., Misios, S., Rodger, C. J.,
Scaife, A. A., Seppälä, A., Shangguan, M., Sinnhuber, M., Tourpali, K., Usoskin, I., van de Kamp, M., Verronen, P. T., and Versick, S.:
Solar forcing for CMIP6 (v3.2), Geosci. Model Dev., 10, 2247–2302, <ext-link xlink:href="https://doi.org/10.5194/gmd-10-2247-2017" ext-link-type="DOI">10.5194/gmd-10-2247-2017</ext-link>, 2017.</mixed-citation></ref>
      <ref id="bib1.bibx48"><?xmltex \def\ref@label{{Orsolini et~al.(2017)Orsolini, Limpasuvan, P{\'{e}}rot, Espy, Hibbins,
Lossow, Raaholt~Larsson, and Murtagh}}?><label>Orsolini et al.(2017)Orsolini, Limpasuvan, Pérot, Espy, Hibbins,
Lossow, Raaholt Larsson, and Murtagh</label><mixed-citation>Orsolini, Y. J., Limpasuvan, V., Pérot, K., Espy, P., Hibbins, R., Lossow,
S., Raaholt Larsson, K., and Murtagh, D.: Modelling the descent of nitric
oxide during the elevated stratopause event of January 2013, J. Atmos. Sol.-Terr. Phy., 155, 50–61, <ext-link xlink:href="https://doi.org/10.1016/j.jastp.2017.01.006" ext-link-type="DOI">10.1016/j.jastp.2017.01.006</ext-link>,
2017.</mixed-citation></ref>
      <ref id="bib1.bibx49"><label>Pendleton et al.(1983)Pendleton, Erman, Larsson, and
Witt</label><mixed-citation>Pendleton, W., Erman, P., Larsson, M., and Witt, G.: Observation of Strong NO
Gamma-Band Radiation Induced in Thin N2-CO2and N2-H2O Targets by Electron
Impact and Its Possible Relation to the Auroral Chemistry of NO, Phys. Scripta, 28, 532–538, <ext-link xlink:href="https://doi.org/10.1088/0031-8949/28/5/005" ext-link-type="DOI">10.1088/0031-8949/28/5/005</ext-link>, 1983.</mixed-citation></ref>
      <ref id="bib1.bibx50"><?xmltex \def\ref@label{{P{\'{e}}rot et~al.(2014)P{\'{e}}rot, Urban, and Murtagh}}?><label>Pérot et al.(2014)Pérot, Urban, and Murtagh</label><mixed-citation>Pérot, K., Urban, J., and Murtagh, D. P.: Unusually strong nitric oxide descent in the Arctic middle atmosphere in early 2013 as
observed by Odin/SMR, Atmos. Chem. Phys., 14, 8009–8015, <ext-link xlink:href="https://doi.org/10.5194/acp-14-8009-2014" ext-link-type="DOI">10.5194/acp-14-8009-2014</ext-link>, 2014.</mixed-citation></ref>
      <ref id="bib1.bibx51"><label>Randall et al.(2007)Randall, Harvey, Singleton, Bailey, Bernath,
Codrescu, Nakajima, and Russell</label><mixed-citation>Randall, C. E., Harvey, V. L., Singleton, C. S., Bailey, S. M., Bernath, P. F.,
Codrescu, M., Nakajima, H., and Russell, J. M.: Energetic particle
precipitation effects on the Southern Hemisphere stratosphere in 1992–2005,
J. Geophys. Res., 112,
D08308, <ext-link xlink:href="https://doi.org/10.1029/2006JD007696" ext-link-type="DOI">10.1029/2006JD007696</ext-link>, 2007.</mixed-citation></ref>
      <ref id="bib1.bibx52"><label>Randall et al.(2015)Randall, Harvey, Holt, Marsh, Kinnison, Funke,
and Bernath</label><mixed-citation>Randall, C. E., Harvey, V. L., Holt, L. A., Marsh, D. R., Kinnison, D., Funke,
B., and Bernath, P. F.: Simulation of energetic particle precipitation
effects during the<?pagebreak page2147?> 2003–2004 Arctic winter, J. Geophys. Res.-Space, 120, 5035–5048, <ext-link xlink:href="https://doi.org/10.1002/2015ja021196" ext-link-type="DOI">10.1002/2015ja021196</ext-link>, 2015.</mixed-citation></ref>
      <ref id="bib1.bibx53"><label>Rasmussen and Williams(2006)</label><mixed-citation>Rasmussen, C. E. and Williams, C. K. I.: Gaussian processes for machine
learning, Adaptive Computation and Machine Learning, MIT Press, Cambridge,
MA, available at: <uri>http://www.gaussianprocess.org/gpml/chapters/</uri> (last access: 2 November 2017), 2006.</mixed-citation></ref>
      <ref id="bib1.bibx54"><label>Roble(1995)</label><mixed-citation>Roble, R. G.: Energetics of the Mesosphere and Thermosphere, in: The Upper
Mesosphere and Lower Thermosphere: A Review of Experiment and Theory, edited
by: Johnson, R. M. and Killeen, T. L.,  American Geophysical Union,
1–21,  <ext-link xlink:href="https://doi.org/10.1029/gm087p0001" ext-link-type="DOI">10.1029/gm087p0001</ext-link>, 1995.</mixed-citation></ref>
      <ref id="bib1.bibx55"><label>Schwarz(1978)</label><mixed-citation>Schwarz, G.: Estimating the Dimension of a Model, Ann. Stat., 6,
461–464, <ext-link xlink:href="https://doi.org/10.1214/aos/1176344136" ext-link-type="DOI">10.1214/aos/1176344136</ext-link>, 1978.</mixed-citation></ref>
      <ref id="bib1.bibx56"><label>Semeniuk et al.(2008)Semeniuk, McConnell, Jin, Jarosz, Boone, and
Bernath</label><mixed-citation>Semeniuk, K., McConnell, J. C., Jin, J. J., Jarosz, J. R., Boone, C. D., and
Bernath, P. F.: <inline-formula><mml:math id="M297" display="inline"><mml:mrow class="chem"><mml:msub><mml:mi mathvariant="normal">N</mml:mi><mml:mn mathvariant="normal">2</mml:mn></mml:msub><mml:mi mathvariant="normal">O</mml:mi></mml:mrow></mml:math></inline-formula> production by high energy auroral electron
precipitation, J. Geophys. Res., 113, D16302, <ext-link xlink:href="https://doi.org/10.1029/2007jd009690" ext-link-type="DOI">10.1029/2007jd009690</ext-link>, 2008.</mixed-citation></ref>
      <ref id="bib1.bibx57"><label>Sheese et al.(2016)Sheese, Walker, Boone, Bernath, and
Funke</label><mixed-citation>Sheese, P. E., Walker, K. A., Boone, C. D., Bernath, P. F., and Funke, B.:
Nitrous oxide in the atmosphere: First measurements of a lower thermospheric
source, Geophys. Res. Lett., 43, 2866–2872, <ext-link xlink:href="https://doi.org/10.1002/2015gl067353" ext-link-type="DOI">10.1002/2015gl067353</ext-link>,
2016.</mixed-citation></ref>
      <ref id="bib1.bibx58"><label>Sinnhuber et al.(2011)Sinnhuber, Kazeminejad, and
Wissing</label><mixed-citation>Sinnhuber, M., Kazeminejad, S., and Wissing, J. M.: Interannual variation of
NOx from the lower thermosphere to the upper stratosphere in the years
1991–2005, J. Geophys. Res., 116, A02312, <ext-link xlink:href="https://doi.org/10.1029/2010JA015825" ext-link-type="DOI">10.1029/2010JA015825</ext-link>,
2011.</mixed-citation></ref>
      <ref id="bib1.bibx59"><label>Sinnhuber et al.(2012)Sinnhuber, Nieder, and
Wieters</label><mixed-citation>Sinnhuber, M., Nieder, H., and Wieters, N.: Energetic Particle Precipitation
and the Chemistry of the Mesosphere/Lower Thermosphere, Surv. Geophys., 33,
1281–1334, <ext-link xlink:href="https://doi.org/10.1007/s10712-012-9201-3" ext-link-type="DOI">10.1007/s10712-012-9201-3</ext-link>, 2012.</mixed-citation></ref>
      <ref id="bib1.bibx60"><label>Sinnhuber et al.(2016)Sinnhuber, Friederich, Bender, and
Burrows</label><mixed-citation>Sinnhuber, M., Friederich, F., Bender, S., and Burrows, J. P.: The response of
mesospheric NO to geomagnetic forcing in 2002–2012 as seen by SCIAMACHY,
J. Geophys. Res., 121, 3603–3620, <ext-link xlink:href="https://doi.org/10.1002/2015JA022284" ext-link-type="DOI">10.1002/2015JA022284</ext-link>,
2015JA022284, 2016.</mixed-citation></ref>
      <ref id="bib1.bibx61"><label>Sinnhuber et al.(2018)Sinnhuber, Berger, Funke, Nieder, Reddmann,
Stiller, Versick, von Clarmann, and Wissing</label><mixed-citation>Sinnhuber, M., Berger, U., Funke, B., Nieder, H., Reddmann, T., Stiller, G., Versick, S., von Clarmann, T., and Wissing, J. M.:
<inline-formula><mml:math id="M298" display="inline"><mml:mrow class="chem"><mml:msub><mml:mi mathvariant="normal">NO</mml:mi><mml:mi>y</mml:mi></mml:msub></mml:mrow></mml:math></inline-formula> production, ozone loss and changes in net radiative heating due to energetic particle precipitation in 2002–2010, Atmos. Chem. Phys.,
18, 1115–1147, <ext-link xlink:href="https://doi.org/10.5194/acp-18-1115-2018" ext-link-type="DOI">10.5194/acp-18-1115-2018</ext-link>, 2018.</mixed-citation></ref>
      <ref id="bib1.bibx62"><label>Siskind et al.(2000)Siskind, Nedoluha, Randall, Fromm, and
Russell</label><mixed-citation>Siskind, D. E., Nedoluha, G. E., Randall, C. E., Fromm, M., and Russell, J. M.:
An assessment of southern hemisphere stratospheric NOx enhancements due to
transport from the upper atmosphere, Geophys. Res. Lett., 27, 329–332,
<ext-link xlink:href="https://doi.org/10.1029/1999GL010940" ext-link-type="DOI">10.1029/1999GL010940</ext-link>, 2000.
</mixed-citation></ref><?xmltex \hack{\newpage}?>
      <ref id="bib1.bibx63"><label>Solomon et al.(1982)Solomon, Crutzen, and Roble</label><mixed-citation>Solomon, S., Crutzen, P. J., and Roble, R. G.: Photochemical coupling between
the thermosphere and the lower atmosphere: 1. Odd nitrogen from 50 to 120 km,
J. Geophys. Res., 87, 7206, <ext-link xlink:href="https://doi.org/10.1029/jc087ic09p07206" ext-link-type="DOI">10.1029/jc087ic09p07206</ext-link>,
1982.</mixed-citation></ref>
      <ref id="bib1.bibx64"><label>Spiegelhalter et al.(2002)Spiegelhalter, Best, Carlin, and van der
Linde</label><mixed-citation>Spiegelhalter, D. J., Best, N. G., Carlin, B. P., and van der Linde, A.:
Bayesian measures of model complexity and fit, J. Roy. Stat. Soc. B, 64, 583–639,
<ext-link xlink:href="https://doi.org/10.1111/1467-9868.00353" ext-link-type="DOI">10.1111/1467-9868.00353</ext-link>, 2002.</mixed-citation></ref>
      <ref id="bib1.bibx65"><?xmltex \def\ref@label{{Th{\'{e}}bault et~al.(2015)Th{\'{e}}bault, Finlay, Beggan, Alken, Aubert,
Barrois, Bertrand, Bondar, Boness, Brocco, Canet, Chambodut, Chulliat,
Co{\"{i}}sson, Civet, Du, Fournier, Fratter, Gillet, Hamilton, Hamoudi, Hulot,
Jager, Korte, Kuang, Lalanne, Langlais, L{\'{e}}ger, Lesur, Lowes, Macmillan,
Mandea, Manoj, Maus, Olsen, Petrov, Ridley, Rother, Sabaka, Saturnino,
Schachtschneider, Sirol, Tangborn, Thomson, T{\o}ffner-Clausen, Vigneron,
Wardinski, and Zvereva}}?><label>Thébault et al.(2015)Thébault, Finlay, Beggan, Alken, Aubert,
Barrois, Bertrand, Bondar, Boness, Brocco, Canet, Chambodut, Chulliat,
Coïsson, Civet, Du, Fournier, Fratter, Gillet, Hamilton, Hamoudi, Hulot,
Jager, Korte, Kuang, Lalanne, Langlais, Léger, Lesur, Lowes, Macmillan,
Mandea, Manoj, Maus, Olsen, Petrov, Ridley, Rother, Sabaka, Saturnino,
Schachtschneider, Sirol, Tangborn, Thomson, Tøffner-Clausen, Vigneron,
Wardinski, and Zvereva</label><mixed-citation>Thébault, E., Finlay, C. C., Beggan, C. D., Alken, P., Aubert, J., Barrois,
O., Bertrand, F., Bondar, T., Boness, A., Brocco, L., Canet, E., Chambodut,
A., Chulliat, A., Coïsson, P., Civet, F., Du, A., Fournier, A., Fratter,
I., Gillet, N., Hamilton, B., Hamoudi, M., Hulot, G., Jager, T., Korte, M.,
Kuang, W., Lalanne, X., Langlais, B., Léger, J.-M., Lesur, V., Lowes,
F. J., Macmillan, S., Mandea, M., Manoj, C., Maus, S., Olsen, N., Petrov, V.,
Ridley, V., Rother, M., Sabaka, T. J., Saturnino, D., Schachtschneider, R.,
Sirol, O., Tangborn, A., Thomson, A., Tøffner-Clausen, L., Vigneron, P.,
Wardinski, I., and Zvereva, T.: International Geomagnetic Reference Field:
the 12th generation, Earth Planet. Space, 67, 79–98,
<ext-link xlink:href="https://doi.org/10.1186/s40623-015-0228-9" ext-link-type="DOI">10.1186/s40623-015-0228-9</ext-link>, 2015.</mixed-citation></ref>
      <ref id="bib1.bibx66"><label>Turunen et al.(2009)</label><mixed-citation>Turunen, E., Verronen, P. T., Seppälä, A., Rodger, C. J., Clilverd, M. A.,
Tamminen, J., Enell, C.-F., and Ulich, T.: Impact of different energies of
precipitating particles on NOx generation in the middle and upper
atmosphere during geomagnetic storms, J. Atmos. Sol.-Terr. Phy., 71,
1176–1189, <ext-link xlink:href="https://doi.org/10.1016/j.jastp.2008.07.005" ext-link-type="DOI">10.1016/j.jastp.2008.07.005</ext-link>, 2009.</mixed-citation></ref>
      <ref id="bib1.bibx67"><label>Vehtari et al.(2016)Vehtari, Gelman, and Gabry</label><mixed-citation>Vehtari, A., Gelman, A., and Gabry, J.: Practical Bayesian model evaluation
using leave-one-out cross-validation and WAIC, Stat. Comput.,
27, 1413–1432, <ext-link xlink:href="https://doi.org/10.1007/s11222-016-9696-4" ext-link-type="DOI">10.1007/s11222-016-9696-4</ext-link>, 2016.</mixed-citation></ref>
      <ref id="bib1.bibx68"><label>Verronen et al.(2013)Verronen, Andersson, Rodger, Clilverd, Wang, and
Turunen</label><mixed-citation>Verronen, P. T., Andersson, M. E., Rodger, C. J., Clilverd, M. A., Wang, S.,
and Turunen, E.: Comparison of modeled and observed effects of radiation belt
electron precipitation on mesospheric hydroxyl and ozone, J.
Geophys. Res.-Atmos., 118, 11419–11428,
<ext-link xlink:href="https://doi.org/10.1002/jgrd.50845" ext-link-type="DOI">10.1002/jgrd.50845</ext-link>, 2013.</mixed-citation></ref>
      <ref id="bib1.bibx69"><label>Watanabe(2010)</label><mixed-citation>
Watanabe, S.: Asymptotic equivalence of Bayes cross validation and widely
applicable information criterion in singular learning theory, J.
Mach. Learn. Res., 11, 3571–3594,
2010.</mixed-citation></ref>

  </ref-list></back>
    <!--<article-title-html>Mesospheric nitric oxide model from SCIAMACHY data</article-title-html>
<abstract-html><p>We present an empirical model for nitric oxide (NO) in the mesosphere
( ≈ 60–90&thinsp;km) derived from SCIAMACHY (SCanning Imaging Absorption spectroMeter for Atmospheric CHartoghraphY) limb scan data.
This work complements and extends the NOEM
(Nitric Oxide Empirical Model; Marsh et al., 2004) and
SANOMA (SMR Acquired Nitric Oxide Model Atmosphere; Kiviranta et al., 2018)
empirical models in the lower thermosphere.
The regression ansatz builds on the heritage of studies
by Hendrickx et al. (2017) and the superposed epoch
analysis by Sinnhuber et al. (2016) which
estimate NO production from particle precipitation.</p><p>Our model relates the daily (longitudinally) averaged NO number densities from
SCIAMACHY (Bender et al., 2017b,a) as a function of geomagnetic latitude
to the solar Lyman-<i>α</i> and the geomagnetic
AE (auroral electrojet) indices.
We use a non-linear regression model, incorporating a finite and seasonally
varying lifetime for the geomagnetically induced NO.
We estimate the parameters by finding the maximum posterior probability
and calculate the parameter uncertainties using
Markov chain Monte Carlo sampling.
In addition to providing an estimate of the NO content in the mesosphere,
the regression coefficients indicate regions where certain processes dominate.</p></abstract-html>
<ref-html id="bib1.bib1"><label>Akaike(1974)</label><mixed-citation>
Akaike, H.: A new look at the statistical model identification, IEEE T. Automat. Contr., 19, 716–723,
<a href="https://doi.org/10.1109/tac.1974.1100705" target="_blank">https://doi.org/10.1109/tac.1974.1100705</a>, 1974.
</mixed-citation></ref-html>
<ref-html id="bib1.bib2"><label>Ando(2011)</label><mixed-citation>
Ando, T.: Predictive Bayesian Model Selection, Am. J. Math.-S, 31, 13–38, <a href="https://doi.org/10.1080/01966324.2011.10737798" target="_blank">https://doi.org/10.1080/01966324.2011.10737798</a>,
2011.
</mixed-citation></ref-html>
<ref-html id="bib1.bib3"><label>Bailey et al.(2002)Bailey, Barth, and Solomon</label><mixed-citation>
Bailey, S. M., Barth, C. A., and Solomon, S. C.: A model of nitric oxide in the
lower thermosphere, J. Geophys. Res., 107, 1205, <a href="https://doi.org/10.1029/2001JA000258" target="_blank">https://doi.org/10.1029/2001JA000258</a>,
2002.
</mixed-citation></ref-html>
<ref-html id="bib1.bib4"><label>Baldwin and Dunkerton(2001)</label><mixed-citation>
Baldwin, M. P. and Dunkerton, T. J.: Stratospheric Harbingers of Anomalous
Weather Regimes, Science, 294, 581–584, <a href="https://doi.org/10.1126/science.1063315" target="_blank">https://doi.org/10.1126/science.1063315</a>, 2001.
</mixed-citation></ref-html>
<ref-html id="bib1.bib5"><label>Barth(1992)</label><mixed-citation>
Barth, C. A.: Nitric oxide in the lower thermosphere, Planet. Space Sci., 40,
315–336, <a href="https://doi.org/10.1016/0032-0633(92)90067-x" target="_blank">https://doi.org/10.1016/0032-0633(92)90067-x</a>, 1992.
</mixed-citation></ref-html>
<ref-html id="bib1.bib6"><label>Barth(1995)</label><mixed-citation>
Barth, C. A.: Nitric Oxide in the Lower Thermosphere, in: The Upper Mesosphere
and Lower Thermosphere: A Review of Experiment and Theory, edited by: Johnson,
R. M. and Killeen, T. L., American Geophysical Union,
225–233,  <a href="https://doi.org/10.1029/gm087p0225" target="_blank">https://doi.org/10.1029/gm087p0225</a>, 1995.
</mixed-citation></ref-html>
<ref-html id="bib1.bib7"><label>Barth(2010)</label><mixed-citation>
Barth, C. A.: Joule heating and nitric oxide in the thermosphere, 2, J.
Geophys. Res.-Space, 115, A10305,
<a href="https://doi.org/10.1029/2010ja015565" target="_blank">https://doi.org/10.1029/2010ja015565</a>, 2010.
</mixed-citation></ref-html>
<ref-html id="bib1.bib8"><label>Barth et al.(1988)Barth, Tobiska, Siskind, and Cleary</label><mixed-citation>
Barth, C. A., Tobiska, W. K., Siskind, D. E., and Cleary, D. D.:
Solar-terrestrial coupling: Low-latitude thermospheric nitric oxide, Geophys.
Res. Lett., 15, 92–94, <a href="https://doi.org/10.1029/gl015i001p00092" target="_blank">https://doi.org/10.1029/gl015i001p00092</a>, 1988.
</mixed-citation></ref-html>
<ref-html id="bib1.bib9"><label>Barth et al.(2003)Barth, Mankoff, Bailey, and Solomon</label><mixed-citation>
Barth, C. A., Mankoff, K. D., Bailey, S. M., and Solomon, S. C.: Global
observations of nitric oxide in the thermosphere, J. Geophys. Res., 108,
1027, <a href="https://doi.org/10.1029/2002JA009458" target="_blank">https://doi.org/10.1029/2002JA009458</a>, 2003.
</mixed-citation></ref-html>
<ref-html id="bib1.bib10"><label>Barth et al.(2009)Barth, Lu, and Roble</label><mixed-citation>
Barth, C. A., Lu, G., and Roble, R. G.: Joule heating and nitric oxide in the
thermosphere, J. Geophys. Res.-Space, 114, A05301,
<a href="https://doi.org/10.1029/2008ja013765" target="_blank">https://doi.org/10.1029/2008ja013765</a>, 2009.
</mixed-citation></ref-html>
<ref-html id="bib1.bib11"><label>Bender(2018a)</label><mixed-citation>
Bender, S.: st-bender/sciapy: Version 0.0.5 [source code],
<a href="https://doi.org/10.5281/zenodo.1401370" target="_blank">https://doi.org/10.5281/zenodo.1401370</a>,
2018a.
</mixed-citation></ref-html>
<ref-html id="bib1.bib12"><label>Bender(2018b)</label><mixed-citation>
Bender, S.: SCIAMACHY NO regression fit MCMC samples [data set],
<a href="https://doi.org/10.5281/zenodo.1342701" target="_blank">https://doi.org/10.5281/zenodo.1342701</a>,
2018b.
</mixed-citation></ref-html>
<ref-html id="bib1.bib13"><label>Bender et al.(2013)Bender, Sinnhuber, Burrows, Langowski, Funke, and
López-Puertas</label><mixed-citation>
Bender, S., Sinnhuber, M., Burrows, J. P., Langowski, M., Funke, B., and López-Puertas, M.: Retrieval of nitric oxide in the
mesosphere and lower thermosphere from SCIAMACHY limb spectra, Atmos. Meas. Tech., 6, 2521–2531, <a href="https://doi.org/10.5194/amt-6-2521-2013" target="_blank">https://doi.org/10.5194/amt-6-2521-2013</a>, 2013.
</mixed-citation></ref-html>
<ref-html id="bib1.bib14"><label>Bender et al.(2017a)Bender, Sinnhuber, Burrows, and
Langowski</label><mixed-citation>
Bender, S., Sinnhuber, M., Burrows, J. P., and Langowski, M.: Nitric oxide (NO) data set (60–160&thinsp;km) from SCIAMACHY nominal limb scans,
<a href="https://doi.org/10.5281/zenodo.804371" target="_blank">https://doi.org/10.5281/zenodo.804371</a>,
2017a.
</mixed-citation></ref-html>
<ref-html id="bib1.bib15"><label>Bender et al.(2017b)Bender, Sinnhuber, Langowski, and
Burrows</label><mixed-citation>
Bender, S., Sinnhuber, M., Langowski, M., and Burrows, J. P.: Retrieval of nitric oxide in the mesosphere from SCIAMACHY nominal limb spectra, Atmos. Meas. Tech., 10, 209–220, <a href="https://doi.org/10.5194/amt-10-209-2017" target="_blank">https://doi.org/10.5194/amt-10-209-2017</a>,
2017b.
</mixed-citation></ref-html>
<ref-html id="bib1.bib16"><label>Bermejo-Pantaleón et al.(2011)Bermejo-Pantaleón, Funke,
López-Puertas, García-Comas, Stiller, von Clarmann, Linden,
Grabowski, Höpfner, Kiefer, Glatthor, Kellmann, and
Lu</label><mixed-citation>
Bermejo-Pantaleón, D., Funke, B., López-Puertas, M., García-Comas,
M., Stiller, G. P., von Clarmann, T., Linden, A., Grabowski, U., Höpfner,
M., Kiefer, M., Glatthor, N., Kellmann, S., and Lu, G.: Global observations
of thermospheric temperature and nitric oxide from MIPAS spectra at 5.3&thinsp;µm, J. Geophys. Res., 116, A10313, <a href="https://doi.org/10.1029/2011JA016752" target="_blank">https://doi.org/10.1029/2011JA016752</a>,
2011.
</mixed-citation></ref-html>
<ref-html id="bib1.bib17"><label>Bovensmann et al.(1999)Bovensmann, Burrows, Buchwitz, Frerick,
Noël, Rozanov, Chance, and Goede</label><mixed-citation>
Bovensmann, H., Burrows, J. P., Buchwitz, M., Frerick, J., Noël, S.,
Rozanov, V. V., Chance, K. V., and Goede, A. P. H.: SCIAMACHY: Mission
Objectives and Measurement Modes, J. Atmos. Sci., 56, 127–150,
<a href="https://doi.org/10.1175/1520-0469(1999)056&lt;0127:SMOAMM&gt;2.0.CO;2" target="_blank">https://doi.org/10.1175/1520-0469(1999)056&lt;0127:SMOAMM&gt;2.0.CO;2</a>, 1999.
</mixed-citation></ref-html>
<ref-html id="bib1.bib18"><label>Burrows et al.(1995)Burrows, Hölzle, Goede, Visser, and
Fricke</label><mixed-citation>
Burrows, J. P., Hölzle, E., Goede, A. P. H., Visser, H., and Fricke, W.:
SCIAMACHY – scanning imaging absorption spectrometer for atmospheric
chartography, Acta Astronaut., 35, 445–451,
<a href="https://doi.org/10.1016/0094-5765(94)00278-T" target="_blank">https://doi.org/10.1016/0094-5765(94)00278-T</a>,
1995.
</mixed-citation></ref-html>
<ref-html id="bib1.bib19"><label>Crutzen(1970)</label><mixed-citation>
Crutzen, P. J.: The influence of nitrogen oxides on the atmospheric ozone
content, Q. J. Roy. Meteor. Soc., 96, 320–325,
<a href="https://doi.org/10.1002/qj.49709640815" target="_blank">https://doi.org/10.1002/qj.49709640815</a>, 1970.
</mixed-citation></ref-html>
<ref-html id="bib1.bib20"><label>Davis and Sugiura(1966)</label><mixed-citation>
Davis, T. N. and Sugiura, M.: Auroral electrojet activity index AE and its
universal time variations, J. Geophys. Res., 71, 785–801,
<a href="https://doi.org/10.1029/jz071i003p00785" target="_blank">https://doi.org/10.1029/jz071i003p00785</a>, 1966.
</mixed-citation></ref-html>
<ref-html id="bib1.bib21"><label>Foreman-Mackey et al.(2013)Foreman-Mackey, Hogg, Lang, and
Goodman</label><mixed-citation>
Foreman-Mackey, D., Hogg, D. W., Lang, D., and Goodman, J.: emcee: The
MCMC Hammer, Publ. Astron. Soc. Pac., 125, 306–313, <a href="https://doi.org/10.1086/670067" target="_blank">https://doi.org/10.1086/670067</a>, 2013.
</mixed-citation></ref-html>
<ref-html id="bib1.bib22"><label>Foreman-Mackey et al.(2017a)Foreman-Mackey, Agol, Angus, and
Ambikasaran</label><mixed-citation>
Foreman-Mackey, D., Agol, E., Angus, R., and Ambikasaran, S.: Fast and
scalable Gaussian process modeling with applications to astronomical time
series, Astronomical Journal, 154,   220, <a href="https://doi.org/10.3847/1538-3881/aa9332" target="_blank">https://doi.org/10.3847/1538-3881/aa9332</a>,
2017a.
</mixed-citation></ref-html>
<ref-html id="bib1.bib23"><label>Foreman-Mackey et al.(2017b)Foreman-Mackey, Agol, Angus, Brewer,
Austin, Farr, Guillochon, Czekala, and Casey</label><mixed-citation>
Foreman-Mackey, D., Agol, E., Angus, R., Brewer, B. J., Austin, P., Farr,
W. M., Guillochon, J., Czekala, I., and Casey, A.: dfm/celerite: celerite
v0.3.0, <a href="https://doi.org/10.5281/zenodo.1048287" target="_blank">https://doi.org/10.5281/zenodo.1048287</a>, 2017b.
</mixed-citation></ref-html>
<ref-html id="bib1.bib24"><label>Frost et al.(1956)Frost, McDowell, and Bawn</label><mixed-citation>
Frost, D. C., McDowell, C. A., and Bawn, C. E. H.: The dissociation energy of
the nitrogen molecule, P. Roy. Soc. Lond. A, 236, 278–284,
<a href="https://doi.org/10.1098/rspa.1956.0135" target="_blank">https://doi.org/10.1098/rspa.1956.0135</a>, 1956.
</mixed-citation></ref-html>
<ref-html id="bib1.bib25"><label>Funke et al.(2001)Funke, López-Puertas, Stiller, v. Clarmann,
and Höpfner</label><mixed-citation>
Funke, B., López-Puertas, M., Stiller, G., v. Clarmann, T., and
Höpfner, M.: A new non-LTE retrieval method for atmospheric
parameters from mipas-envisat emission spectra, Adv. Space Res., 27,
1099–1104, <a href="https://doi.org/10.1016/S0273-1177(01)00169-7" target="_blank">https://doi.org/10.1016/S0273-1177(01)00169-7</a>,
2001.
</mixed-citation></ref-html>
<ref-html id="bib1.bib26"><label>Funke et al.(2005a)Funke, López-Puertas,
Gil-López, von Clarmann, Stiller, Fischer, and Kellmann</label><mixed-citation>
Funke, B., López-Puertas, M., Gil-López, S., von Clarmann, T., Stiller,
G. P., Fischer, H., and Kellmann, S.: Downward transport of upper atmospheric
NOx into the polar stratosphere and lower mesosphere during the Antarctic
2003 and Arctic 2002/2003 winters, J. Geophys. Res., 110, D24308,
<a href="https://doi.org/10.1029/2005JD006463" target="_blank">https://doi.org/10.1029/2005JD006463</a>, 2005a.
</mixed-citation></ref-html>
<ref-html id="bib1.bib27"><label>Funke et al.(2005b)Funke, López-Puertas,
von Clarmann, Stiller, Fischer, Glatthor, Grabowski, Höpfner, Kellmann,
Kiefer, Linden, Mengistu Tsidu, Milz, Steck, and Wang</label><mixed-citation>
Funke, B., López-Puertas, M., von Clarmann, T., Stiller, G. P., Fischer,
H., Glatthor, N., Grabowski, U., Höpfner, M., Kellmann, S., Kiefer, M.,
Linden, A., Mengistu Tsidu, G., Milz, M., Steck, T., and Wang, D. Y.:
Retrieval of stratospheric NOx from 5.3 and 6.2 µm nonlocal
thermodynamic equilibrium emissions measured by Michelson Interferometer for
Passive Atmospheric Sounding (MIPAS) on Envisat, J. Geophys. Res., 110,
D09302, <a href="https://doi.org/10.1029/2004JD005225" target="_blank">https://doi.org/10.1029/2004JD005225</a>, 2005b.
</mixed-citation></ref-html>
<ref-html id="bib1.bib28"><label>Funke et al.(2008a)Funke, García-Comas,
López-Puertas, Glatthor, Stiller, von Clarmann, Semeniuk, and
McConnell</label><mixed-citation>
Funke, B., García-Comas, M., López-Puertas, M., Glatthor, N., Stiller, G. P., von Clarmann, T., Semeniuk, K., and McConnell, J. C.:
Enhancement of N<sub>2</sub>O during the October–November 2003 solar proton events, Atmos. Chem. Phys., 8, 3805–3815, <a href="https://doi.org/10.5194/acp-8-3805-2008" target="_blank">https://doi.org/10.5194/acp-8-3805-2008</a>, 2008.  2008a.
</mixed-citation></ref-html>
<ref-html id="bib1.bib29"><label>Funke et al.(2008b)Funke, López-Puertas,
Garcia-Comas, Stiller, von Clarmann, and Glatthor</label><mixed-citation>
Funke, B., López-Puertas, M., Garcia-Comas, M., Stiller, G. P., von Clarmann, T., and Glatthor, N.: Mesospheric N<sub>2</sub>O
enhancements as observed by MIPAS on Envisat during the polar winters in 2002–2004, Atmos. Chem. Phys., 8, 5787–5800, <a href="https://doi.org/10.5194/acp-8-5787-2008" target="_blank">https://doi.org/10.5194/acp-8-5787-2008</a>, 2008.
2008b.
</mixed-citation></ref-html>
<ref-html id="bib1.bib30"><label>Funke et al.(2012)Funke, López-Puertas, García-Comas,
Kaufmann, Höpfner, and Stiller</label><mixed-citation>
Funke, B., López-Puertas, M., García-Comas, M., Kaufmann, M.,
Höpfner, M., and Stiller, G.: GRANADA: A Generic RAdiative traNsfer
AnD non-LTE population algorithm, J. Quant. Spectrosc. Ra.,
113, 1771–1817, <a href="https://doi.org/10.1016/j.jqsrt.2012.05.001" target="_blank">https://doi.org/10.1016/j.jqsrt.2012.05.001</a>,
2012.
</mixed-citation></ref-html>
<ref-html id="bib1.bib31"><label>Funke et al.(2014a)Funke, López-Puertas, Holt,
Randall, Stiller, and von Clarmann</label><mixed-citation>
Funke, B., López-Puertas, M., Holt, L., Randall, C. E., Stiller, G. P.,
and von Clarmann, T.: Hemispheric distributions and interannual variability
of NO<sub><i>y</i></sub> produced by energetic particle precipitation in 2002–2012,
J. Geophys. Res.-Atmos., 119, 13565–13582,
<a href="https://doi.org/10.1002/2014jd022423" target="_blank">https://doi.org/10.1002/2014jd022423</a>, 2014a.
</mixed-citation></ref-html>
<ref-html id="bib1.bib32"><label>Funke et al.(2014b)Funke, López-Puertas, Stiller,
and von Clarmann</label><mixed-citation>
Funke, B., López-Puertas, M., Stiller, G. P., and von Clarmann, T.:
Mesospheric and stratospheric NOy produced by energetic particle
precipitation during 2002–2012, J. Geophys. Res., 119, 4429–4446,
<a href="https://doi.org/10.1002/2013JD021404" target="_blank">https://doi.org/10.1002/2013JD021404</a>, 2014b.
</mixed-citation></ref-html>
<ref-html id="bib1.bib33"><label>Funke et al.(2016)</label><mixed-citation>
Funke, B., López-Puertas, M., Stiller, G. P., Versick, S., and von Clarmann, T.: A semi-empirical model for mesospheric
and stratospheric NO<sub><i>y</i></sub> produced by energetic particle precipitation, Atmos. Chem. Phys., 16, 8667–8693, <a href="https://doi.org/10.5194/acp-16-8667-2016" target="_blank">https://doi.org/10.5194/acp-16-8667-2016</a>, 2016.
</mixed-citation></ref-html>
<ref-html id="bib1.bib34"><label>Funke et al.(2017)</label><mixed-citation>
Funke, B., Ball, W., Bender, S., Gardini, A., Harvey, V. L., Lambert, A., López-Puertas, M., Marsh, D. R., Meraner, K., Nieder, H.,
Päivärinta, S.-M., Pérot, K., Randall, C. E., Reddmann, T., Rozanov, E., Schmidt, H., Seppälä, A., Sinnhuber, M.,
Sukhodolov, T., Stiller, G. P., Tsvetkova, N. D., Verronen, P. T., Versick, S., von Clarmann, T., Walker, K. A., and Yushkov, V.:
HEPPA-II model-measurement intercomparison project: EPP indirect effects during the dynamically perturbed NH winter 2008–2009, Atmos. Chem. Phys., 17, 3573–3604, <a href="https://doi.org/10.5194/acp-17-3573-2017" target="_blank">https://doi.org/10.5194/acp-17-3573-2017</a>, 2017.
</mixed-citation></ref-html>
<ref-html id="bib1.bib35"><label>Heays et al.(2017)Heays, Bosman, and van Dishoeck</label><mixed-citation>
Heays, A. N., Bosman, A. D., and van Dishoeck, E. F.: Photodissociation and
photoionisation of atoms and molecules of astrophysical interest, Astron. Astrophys., 602, A105, <a href="https://doi.org/10.1051/0004-6361/201628742" target="_blank">https://doi.org/10.1051/0004-6361/201628742</a>, 2017.
</mixed-citation></ref-html>
<ref-html id="bib1.bib36"><label>Hendrickx et al.(2015)Hendrickx, Megner, Gumbel, Siskind, Orsolini,
Tyssøy, and Hervig</label><mixed-citation>
Hendrickx, K., Megner, L., Gumbel, J., Siskind, D. E., Orsolini, Y. J.,
Tyssøy, H. N., and Hervig, M.: Observation of 27 day solar cycles in the
production and mesospheric descent of EPP-produced NO, J. Geophys. Res.,
120, 8978–8988, <a href="https://doi.org/10.1002/2015JA021441" target="_blank">https://doi.org/10.1002/2015JA021441</a>, 2015JA021441, 2015.
</mixed-citation></ref-html>
<ref-html id="bib1.bib37"><label>Hendrickx et al.(2017)Hendrickx, Megner, Marsh, Gumbel, Strandberg,
and Martinsson</label><mixed-citation>
Hendrickx, K., Megner, L., Marsh, D. R., Gumbel, J., Strandberg, R., and
Martinsson, F.: Relative Importance of Nitric Oxide Physical Drivers in the
Lower Thermosphere, Geophys. Res. Lett., 44, 10,081–10,087,
<a href="https://doi.org/10.1002/2017gl074786" target="_blank">https://doi.org/10.1002/2017gl074786</a>,
2017.
</mixed-citation></ref-html>
<ref-html id="bib1.bib38"><label>Hendrickx et al.(2018)Hendrickx, Megner, Marsh, and
Smith-Johnsen</label><mixed-citation>
Hendrickx, K., Megner, L., Marsh, D. R., and Smith-Johnsen, C.: Production and transport mechanisms of NO in the polar upper mesosphere and
lower thermosphere in observations and models, Atmos. Chem. Phys., 18, 9075–9089, <a href="https://doi.org/10.5194/acp-18-9075-2018" target="_blank">https://doi.org/10.5194/acp-18-9075-2018</a>, 2018.
</mixed-citation></ref-html>
<ref-html id="bib1.bib39"><label>Hendrie(1954)</label><mixed-citation>
Hendrie, J. M.: Dissociation Energy of N<sub>2</sub>, J. Chem. Phys., 22, 1503–1507, <a href="https://doi.org/10.1063/1.1740449" target="_blank">https://doi.org/10.1063/1.1740449</a>, 1954.
</mixed-citation></ref-html>
<ref-html id="bib1.bib40"><label>Kiviranta et al.(2018)Kiviranta, Pérot, Eriksson, and
Murtagh</label><mixed-citation>
Kiviranta, J., Pérot, K., Eriksson, P., and Murtagh, D.: An empirical model of nitric oxide in the upper mesosphere and lower thermosphere
based on 12 years of Odin SMR measurements, Atmos. Chem. Phys., 18, 13393–13410, <a href="https://doi.org/10.5194/acp-18-13393-2018" target="_blank">https://doi.org/10.5194/acp-18-13393-2018</a>, 2018.
</mixed-citation></ref-html>
<ref-html id="bib1.bib41"><label>Kockarts(1980)</label><mixed-citation>
Kockarts, G.: Nitric oxide cooling in the terrestrial thermosphere, Geophys.
Res. Lett., 7, 137–140, <a href="https://doi.org/10.1029/gl007i002p00137" target="_blank">https://doi.org/10.1029/gl007i002p00137</a>, 1980.
</mixed-citation></ref-html>
<ref-html id="bib1.bib42"><label>MacKay(2003)</label><mixed-citation>
MacKay, D.: Information Theory, Inference and Learning Algorithms, Cambridge
University Press,
available at: <a href="http://www.inference.org.uk/mackay/itila/book.html" target="_blank">http://www.inference.org.uk/mackay/itila/book.html</a> (last access: 17 November 2017), 2003.
</mixed-citation></ref-html>
<ref-html id="bib1.bib43"><label>Mandea and Korte(2011)</label><mixed-citation>
Mandea, M. and Korte, M. (Eds.): Geomagnetic Observations and Models, vol. 5 of
IAGA Special Sopron Book Series, <a href="https://doi.org/10.1007/978-90-481-9858-0" target="_blank">https://doi.org/10.1007/978-90-481-9858-0</a>,
2011.
</mixed-citation></ref-html>
<ref-html id="bib1.bib44"><label>Maric and Burrows(1992)</label><mixed-citation>
Maric, D. and Burrows, J.: Formation of N<sub>2</sub>O in the
photolysis/photoexcitation of NO, NO<sub>2</sub> and air, J. Photochem.
Photobiol., A, 66, 291–312, <a href="https://doi.org/10.1016/1010-6030(92)80002-d" target="_blank">https://doi.org/10.1016/1010-6030(92)80002-d</a>, 1992.
</mixed-citation></ref-html>
<ref-html id="bib1.bib45"><label>Marsh et al.(2004)Marsh, Solomon, and Reynolds</label><mixed-citation>
Marsh, D. R., Solomon, S. C., and Reynolds, A. E.: Empirical model of nitric
oxide in the lower thermosphere, J. Geophys. Res., 109, A07301,
<a href="https://doi.org/10.1029/2003JA010199" target="_blank">https://doi.org/10.1029/2003JA010199</a>, 2004.
</mixed-citation></ref-html>
<ref-html id="bib1.bib46"><label>Matérn(1960)</label><mixed-citation>
Matérn, B.: Spatial Variation: Stachastic Models and Their Application to
Some Problems in Forest Surveys and Other Sampling Investigations, vol. 36 of
Lecture Notes in Statistics, Statens skogsforskningsinstitut,
Springer-Verlag New York, 2 edn., <a href="https://doi.org/10.1007/978-1-4615-7892-5" target="_blank">https://doi.org/10.1007/978-1-4615-7892-5</a>, originally
published by the Swedish National Institute for Forestry Research, 1960,
1960.
</mixed-citation></ref-html>
<ref-html id="bib1.bib47"><label>Matthes et al.(2017)</label><mixed-citation>
Matthes, K., Funke, B., Andersson, M. E., Barnard, L., Beer, J., Charbonneau, P., Clilverd, M. A., Dudok de Wit, T., Haberreiter, M.,
Hendry, A., Jackman, C. H., Kretzschmar, M., Kruschke, T., Kunze, M., Langematz, U., Marsh, D. R., Maycock, A. C., Misios, S., Rodger, C. J.,
Scaife, A. A., Seppälä, A., Shangguan, M., Sinnhuber, M., Tourpali, K., Usoskin, I., van de Kamp, M., Verronen, P. T., and Versick, S.:
Solar forcing for CMIP6 (v3.2), Geosci. Model Dev., 10, 2247–2302, <a href="https://doi.org/10.5194/gmd-10-2247-2017" target="_blank">https://doi.org/10.5194/gmd-10-2247-2017</a>, 2017.
</mixed-citation></ref-html>
<ref-html id="bib1.bib48"><label>Orsolini et al.(2017)Orsolini, Limpasuvan, Pérot, Espy, Hibbins,
Lossow, Raaholt Larsson, and Murtagh</label><mixed-citation>
Orsolini, Y. J., Limpasuvan, V., Pérot, K., Espy, P., Hibbins, R., Lossow,
S., Raaholt Larsson, K., and Murtagh, D.: Modelling the descent of nitric
oxide during the elevated stratopause event of January 2013, J. Atmos. Sol.-Terr. Phy., 155, 50–61, <a href="https://doi.org/10.1016/j.jastp.2017.01.006" target="_blank">https://doi.org/10.1016/j.jastp.2017.01.006</a>,
2017.
</mixed-citation></ref-html>
<ref-html id="bib1.bib49"><label>Pendleton et al.(1983)Pendleton, Erman, Larsson, and
Witt</label><mixed-citation>
Pendleton, W., Erman, P., Larsson, M., and Witt, G.: Observation of Strong NO
Gamma-Band Radiation Induced in Thin N2-CO2and N2-H2O Targets by Electron
Impact and Its Possible Relation to the Auroral Chemistry of NO, Phys. Scripta, 28, 532–538, <a href="https://doi.org/10.1088/0031-8949/28/5/005" target="_blank">https://doi.org/10.1088/0031-8949/28/5/005</a>, 1983.
</mixed-citation></ref-html>
<ref-html id="bib1.bib50"><label>Pérot et al.(2014)Pérot, Urban, and Murtagh</label><mixed-citation>
Pérot, K., Urban, J., and Murtagh, D. P.: Unusually strong nitric oxide descent in the Arctic middle atmosphere in early 2013 as
observed by Odin/SMR, Atmos. Chem. Phys., 14, 8009–8015, <a href="https://doi.org/10.5194/acp-14-8009-2014" target="_blank">https://doi.org/10.5194/acp-14-8009-2014</a>, 2014.
</mixed-citation></ref-html>
<ref-html id="bib1.bib51"><label>Randall et al.(2007)Randall, Harvey, Singleton, Bailey, Bernath,
Codrescu, Nakajima, and Russell</label><mixed-citation>
Randall, C. E., Harvey, V. L., Singleton, C. S., Bailey, S. M., Bernath, P. F.,
Codrescu, M., Nakajima, H., and Russell, J. M.: Energetic particle
precipitation effects on the Southern Hemisphere stratosphere in 1992–2005,
J. Geophys. Res., 112,
D08308, <a href="https://doi.org/10.1029/2006JD007696" target="_blank">https://doi.org/10.1029/2006JD007696</a>, 2007.
</mixed-citation></ref-html>
<ref-html id="bib1.bib52"><label>Randall et al.(2015)Randall, Harvey, Holt, Marsh, Kinnison, Funke,
and Bernath</label><mixed-citation>
Randall, C. E., Harvey, V. L., Holt, L. A., Marsh, D. R., Kinnison, D., Funke,
B., and Bernath, P. F.: Simulation of energetic particle precipitation
effects during the 2003–2004 Arctic winter, J. Geophys. Res.-Space, 120, 5035–5048, <a href="https://doi.org/10.1002/2015ja021196" target="_blank">https://doi.org/10.1002/2015ja021196</a>, 2015.
</mixed-citation></ref-html>
<ref-html id="bib1.bib53"><label>Rasmussen and Williams(2006)</label><mixed-citation>
Rasmussen, C. E. and Williams, C. K. I.: Gaussian processes for machine
learning, Adaptive Computation and Machine Learning, MIT Press, Cambridge,
MA, available at: <a href="http://www.gaussianprocess.org/gpml/chapters/" target="_blank">http://www.gaussianprocess.org/gpml/chapters/</a> (last access: 2 November 2017), 2006.
</mixed-citation></ref-html>
<ref-html id="bib1.bib54"><label>Roble(1995)</label><mixed-citation>
Roble, R. G.: Energetics of the Mesosphere and Thermosphere, in: The Upper
Mesosphere and Lower Thermosphere: A Review of Experiment and Theory, edited
by: Johnson, R. M. and Killeen, T. L.,  American Geophysical Union,
1–21,  <a href="https://doi.org/10.1029/gm087p0001" target="_blank">https://doi.org/10.1029/gm087p0001</a>, 1995.
</mixed-citation></ref-html>
<ref-html id="bib1.bib55"><label>Schwarz(1978)</label><mixed-citation>
Schwarz, G.: Estimating the Dimension of a Model, Ann. Stat., 6,
461–464, <a href="https://doi.org/10.1214/aos/1176344136" target="_blank">https://doi.org/10.1214/aos/1176344136</a>, 1978.
</mixed-citation></ref-html>
<ref-html id="bib1.bib56"><label>Semeniuk et al.(2008)Semeniuk, McConnell, Jin, Jarosz, Boone, and
Bernath</label><mixed-citation>
Semeniuk, K., McConnell, J. C., Jin, J. J., Jarosz, J. R., Boone, C. D., and
Bernath, P. F.: N<sub>2</sub>O production by high energy auroral electron
precipitation, J. Geophys. Res., 113, D16302, <a href="https://doi.org/10.1029/2007jd009690" target="_blank">https://doi.org/10.1029/2007jd009690</a>, 2008.
</mixed-citation></ref-html>
<ref-html id="bib1.bib57"><label>Sheese et al.(2016)Sheese, Walker, Boone, Bernath, and
Funke</label><mixed-citation>
Sheese, P. E., Walker, K. A., Boone, C. D., Bernath, P. F., and Funke, B.:
Nitrous oxide in the atmosphere: First measurements of a lower thermospheric
source, Geophys. Res. Lett., 43, 2866–2872, <a href="https://doi.org/10.1002/2015gl067353" target="_blank">https://doi.org/10.1002/2015gl067353</a>,
2016.
</mixed-citation></ref-html>
<ref-html id="bib1.bib58"><label>Sinnhuber et al.(2011)Sinnhuber, Kazeminejad, and
Wissing</label><mixed-citation>
Sinnhuber, M., Kazeminejad, S., and Wissing, J. M.: Interannual variation of
NOx from the lower thermosphere to the upper stratosphere in the years
1991–2005, J. Geophys. Res., 116, A02312, <a href="https://doi.org/10.1029/2010JA015825" target="_blank">https://doi.org/10.1029/2010JA015825</a>,
2011.
</mixed-citation></ref-html>
<ref-html id="bib1.bib59"><label>Sinnhuber et al.(2012)Sinnhuber, Nieder, and
Wieters</label><mixed-citation>
Sinnhuber, M., Nieder, H., and Wieters, N.: Energetic Particle Precipitation
and the Chemistry of the Mesosphere/Lower Thermosphere, Surv. Geophys., 33,
1281–1334, <a href="https://doi.org/10.1007/s10712-012-9201-3" target="_blank">https://doi.org/10.1007/s10712-012-9201-3</a>, 2012.
</mixed-citation></ref-html>
<ref-html id="bib1.bib60"><label>Sinnhuber et al.(2016)Sinnhuber, Friederich, Bender, and
Burrows</label><mixed-citation>
Sinnhuber, M., Friederich, F., Bender, S., and Burrows, J. P.: The response of
mesospheric NO to geomagnetic forcing in 2002–2012 as seen by SCIAMACHY,
J. Geophys. Res., 121, 3603–3620, <a href="https://doi.org/10.1002/2015JA022284" target="_blank">https://doi.org/10.1002/2015JA022284</a>,
2015JA022284, 2016.
</mixed-citation></ref-html>
<ref-html id="bib1.bib61"><label>Sinnhuber et al.(2018)Sinnhuber, Berger, Funke, Nieder, Reddmann,
Stiller, Versick, von Clarmann, and Wissing</label><mixed-citation>
Sinnhuber, M., Berger, U., Funke, B., Nieder, H., Reddmann, T., Stiller, G., Versick, S., von Clarmann, T., and Wissing, J. M.:
NO<sub><i>y</i></sub> production, ozone loss and changes in net radiative heating due to energetic particle precipitation in 2002–2010, Atmos. Chem. Phys.,
18, 1115–1147, <a href="https://doi.org/10.5194/acp-18-1115-2018" target="_blank">https://doi.org/10.5194/acp-18-1115-2018</a>, 2018.
</mixed-citation></ref-html>
<ref-html id="bib1.bib62"><label>Siskind et al.(2000)Siskind, Nedoluha, Randall, Fromm, and
Russell</label><mixed-citation>
Siskind, D. E., Nedoluha, G. E., Randall, C. E., Fromm, M., and Russell, J. M.:
An assessment of southern hemisphere stratospheric NOx enhancements due to
transport from the upper atmosphere, Geophys. Res. Lett., 27, 329–332,
<a href="https://doi.org/10.1029/1999GL010940" target="_blank">https://doi.org/10.1029/1999GL010940</a>, 2000.

</mixed-citation></ref-html>
<ref-html id="bib1.bib63"><label>Solomon et al.(1982)Solomon, Crutzen, and Roble</label><mixed-citation>
Solomon, S., Crutzen, P. J., and Roble, R. G.: Photochemical coupling between
the thermosphere and the lower atmosphere: 1. Odd nitrogen from 50 to 120 km,
J. Geophys. Res., 87, 7206, <a href="https://doi.org/10.1029/jc087ic09p07206" target="_blank">https://doi.org/10.1029/jc087ic09p07206</a>,
1982.
</mixed-citation></ref-html>
<ref-html id="bib1.bib64"><label>Spiegelhalter et al.(2002)Spiegelhalter, Best, Carlin, and van der
Linde</label><mixed-citation>
Spiegelhalter, D. J., Best, N. G., Carlin, B. P., and van der Linde, A.:
Bayesian measures of model complexity and fit, J. Roy. Stat. Soc. B, 64, 583–639,
<a href="https://doi.org/10.1111/1467-9868.00353" target="_blank">https://doi.org/10.1111/1467-9868.00353</a>, 2002.
</mixed-citation></ref-html>
<ref-html id="bib1.bib65"><label>Thébault et al.(2015)Thébault, Finlay, Beggan, Alken, Aubert,
Barrois, Bertrand, Bondar, Boness, Brocco, Canet, Chambodut, Chulliat,
Coïsson, Civet, Du, Fournier, Fratter, Gillet, Hamilton, Hamoudi, Hulot,
Jager, Korte, Kuang, Lalanne, Langlais, Léger, Lesur, Lowes, Macmillan,
Mandea, Manoj, Maus, Olsen, Petrov, Ridley, Rother, Sabaka, Saturnino,
Schachtschneider, Sirol, Tangborn, Thomson, Tøffner-Clausen, Vigneron,
Wardinski, and Zvereva</label><mixed-citation>
Thébault, E., Finlay, C. C., Beggan, C. D., Alken, P., Aubert, J., Barrois,
O., Bertrand, F., Bondar, T., Boness, A., Brocco, L., Canet, E., Chambodut,
A., Chulliat, A., Coïsson, P., Civet, F., Du, A., Fournier, A., Fratter,
I., Gillet, N., Hamilton, B., Hamoudi, M., Hulot, G., Jager, T., Korte, M.,
Kuang, W., Lalanne, X., Langlais, B., Léger, J.-M., Lesur, V., Lowes,
F. J., Macmillan, S., Mandea, M., Manoj, C., Maus, S., Olsen, N., Petrov, V.,
Ridley, V., Rother, M., Sabaka, T. J., Saturnino, D., Schachtschneider, R.,
Sirol, O., Tangborn, A., Thomson, A., Tøffner-Clausen, L., Vigneron, P.,
Wardinski, I., and Zvereva, T.: International Geomagnetic Reference Field:
the 12th generation, Earth Planet. Space, 67, 79–98,
<a href="https://doi.org/10.1186/s40623-015-0228-9" target="_blank">https://doi.org/10.1186/s40623-015-0228-9</a>, 2015.
</mixed-citation></ref-html>
<ref-html id="bib1.bib66"><label>Turunen et al.(2009)</label><mixed-citation>
Turunen, E., Verronen, P. T., Seppälä, A., Rodger, C. J., Clilverd, M. A.,
Tamminen, J., Enell, C.-F., and Ulich, T.: Impact of different energies of
precipitating particles on NOx generation in the middle and upper
atmosphere during geomagnetic storms, J. Atmos. Sol.-Terr. Phy., 71,
1176–1189, <a href="https://doi.org/10.1016/j.jastp.2008.07.005" target="_blank">https://doi.org/10.1016/j.jastp.2008.07.005</a>, 2009.
</mixed-citation></ref-html>
<ref-html id="bib1.bib67"><label>Vehtari et al.(2016)Vehtari, Gelman, and Gabry</label><mixed-citation>
Vehtari, A., Gelman, A., and Gabry, J.: Practical Bayesian model evaluation
using leave-one-out cross-validation and WAIC, Stat. Comput.,
27, 1413–1432, <a href="https://doi.org/10.1007/s11222-016-9696-4" target="_blank">https://doi.org/10.1007/s11222-016-9696-4</a>, 2016.
</mixed-citation></ref-html>
<ref-html id="bib1.bib68"><label>Verronen et al.(2013)Verronen, Andersson, Rodger, Clilverd, Wang, and
Turunen</label><mixed-citation>
Verronen, P. T., Andersson, M. E., Rodger, C. J., Clilverd, M. A., Wang, S.,
and Turunen, E.: Comparison of modeled and observed effects of radiation belt
electron precipitation on mesospheric hydroxyl and ozone, J.
Geophys. Res.-Atmos., 118, 11419–11428,
<a href="https://doi.org/10.1002/jgrd.50845" target="_blank">https://doi.org/10.1002/jgrd.50845</a>, 2013.
</mixed-citation></ref-html>
<ref-html id="bib1.bib69"><label>Watanabe(2010)</label><mixed-citation>
Watanabe, S.: Asymptotic equivalence of Bayes cross validation and widely
applicable information criterion in singular learning theory, J.
Mach. Learn. Res., 11, 3571–3594,
2010.
</mixed-citation></ref-html>--></article>
