<?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" dtd-version="3.0">
  <front>
    <journal-meta>
<journal-id journal-id-type="publisher">OS</journal-id>
<journal-title-group>
<journal-title>Ocean Science</journal-title>
<abbrev-journal-title abbrev-type="publisher">OS</abbrev-journal-title>
<abbrev-journal-title abbrev-type="nlm-ta">Ocean Sci.</abbrev-journal-title>
</journal-title-group>
<issn pub-type="epub">1812-0792</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/os-12-71-2016</article-id><title-group><article-title>Modeling the ocean and atmosphere during an extreme bora event
in northern Adriatic using one-way and two-way <?xmltex \hack{\break}?>atmosphere–ocean coupling</article-title>
      </title-group><?xmltex \runningtitle{Air--sea response to bora using atmosphere--ocean coupling}?><?xmltex \runningauthor{M.~Li\v{c}er et~al.}?>
      <contrib-group>
        <contrib contrib-type="author" corresp="yes" rid="aff1">
          <name><surname>Ličer</surname><given-names>M.</given-names></name>
          <email>matjaz.licer@mbss.org</email>
        <ext-link>https://orcid.org/0000-0003-2304-2505</ext-link></contrib>
        <contrib contrib-type="author" corresp="no" rid="aff1">
          <name><surname>Smerkol</surname><given-names>P.</given-names></name>
          
        </contrib>
        <contrib contrib-type="author" corresp="no" rid="aff1">
          <name><surname>Fettich</surname><given-names>A.</given-names></name>
          
        </contrib>
        <contrib contrib-type="author" corresp="no" rid="aff2">
          <name><surname>Ravdas</surname><given-names>M.</given-names></name>
          
        </contrib>
        <contrib contrib-type="author" corresp="no" rid="aff2">
          <name><surname>Papapostolou</surname><given-names>A.</given-names></name>
          
        </contrib>
        <contrib contrib-type="author" corresp="no" rid="aff2">
          <name><surname>Mantziafou</surname><given-names>A.</given-names></name>
          
        </contrib>
        <contrib contrib-type="author" corresp="no" rid="aff3">
          <name><surname>Strajnar</surname><given-names>B.</given-names></name>
          
        </contrib>
        <contrib contrib-type="author" corresp="no" rid="aff3">
          <name><surname>Cedilnik</surname><given-names>J.</given-names></name>
          
        </contrib>
        <contrib contrib-type="author" corresp="no" rid="aff3">
          <name><surname>Jeromel</surname><given-names>M.</given-names></name>
          
        </contrib>
        <contrib contrib-type="author" corresp="no" rid="aff3">
          <name><surname>Jerman</surname><given-names>J.</given-names></name>
          
        </contrib>
        <contrib contrib-type="author" corresp="no" rid="aff3">
          <name><surname>Petan</surname><given-names>S.</given-names></name>
          
        </contrib>
        <contrib contrib-type="author" corresp="no" rid="aff1">
          <name><surname>Malačič</surname><given-names>V.</given-names></name>
          
        </contrib>
        <contrib contrib-type="author" corresp="no" rid="aff2">
          <name><surname>Sofianos</surname><given-names>S.</given-names></name>
          
        </contrib>
        <aff id="aff1"><label>1</label><institution>NIB, National Institute of Biology, Marine Biology Station, Piran, Slovenia</institution>
        </aff>
        <aff id="aff2"><label>2</label><institution>UOA, Ocean Physics and Modeling Group, University of Athens,
Athens, Greece</institution>
        </aff>
        <aff id="aff3"><label>3</label><institution>ARSO, Slovenian Environment Agency, Ljubljana,
Slovenia</institution>
        </aff>
      </contrib-group>
      <author-notes><corresp id="corr1">M. Ličer (matjaz.licer@mbss.org)</corresp></author-notes><pub-date><day>15</day><month>January</month><year>2016</year></pub-date>
      
      <volume>12</volume>
      <issue>1</issue>
      <fpage>71</fpage><lpage>86</lpage>
      <history>
        <date date-type="received"><day>1</day><month>June</month><year>2015</year></date>
           <date date-type="rev-request"><day>13</day><month>July</month><year>2015</year></date>
           <date date-type="rev-recd"><day>5</day><month>November</month><year>2015</year></date>
           <date date-type="accepted"><day>11</day><month>December</month><year>2015</year></date>
      </history>
      <permissions>
<license license-type="open-access">
<license-p>This work is licensed under a Creative Commons Attribution 3.0 Unported License. To view a copy of this license, visit <ext-link ext-link-type="uri" xlink:href="http://creativecommons.org/licenses/by/3.0/">http://creativecommons.org/licenses/by/3.0/</ext-link></license-p>
</license>
</permissions><self-uri xlink:href="https://os.copernicus.org/articles/12/71/2016/os-12-71-2016.html">This article is available from https://os.copernicus.org/articles/12/71/2016/os-12-71-2016.html</self-uri>
<self-uri xlink:href="https://os.copernicus.org/articles/12/71/2016/os-12-71-2016.pdf">The full text article is available as a PDF file from https://os.copernicus.org/articles/12/71/2016/os-12-71-2016.pdf</self-uri>


      <abstract>
    <p>We have studied the performances of (a) a two-way coupled atmosphere–ocean
modeling system and (b) one-way coupled ocean model (forced by the atmosphere
model), as compared to the available in situ measurements during and after a
strong Adriatic bora wind event in February 2012, which led to extreme
air–sea interactions. The simulations span the period between January and
March 2012. The models used were ALADIN (Aire Limitée Adaptation
dynamique Développement InterNational) (4.4 <inline-formula><mml:math display="inline"><mml:mi mathvariant="normal">km</mml:mi></mml:math></inline-formula> resolution) on the
atmosphere side and an Adriatic setup of Princeton ocean model (POM)
(<inline-formula><mml:math display="inline"><mml:mrow><mml:msup><mml:mn mathvariant="normal">1</mml:mn><mml:mo>∘</mml:mo></mml:msup><mml:mo>/</mml:mo><mml:mn>30</mml:mn><mml:mo>×</mml:mo><mml:msup><mml:mn mathvariant="normal">1</mml:mn><mml:mo>∘</mml:mo></mml:msup><mml:mo>/</mml:mo><mml:mn>30</mml:mn></mml:mrow></mml:math></inline-formula> angular resolution) on the ocean side.
The atmosphere–ocean coupling was implemented using the OASIS3-MCT model
coupling toolkit. Two-way coupling ocean feedback to the atmosphere is
limited to sea surface temperature. We have compared modeled
atmosphere–ocean fluxes and sea temperatures from both setups to platform
and CTD (conductivity, temperature, and depth) measurements from three
locations in the northern Adriatic. We present objective verification of
2 <inline-formula><mml:math display="inline"><mml:mi mathvariant="normal">m</mml:mi></mml:math></inline-formula> atmosphere temperature forecasts using mean bias and standard
deviation of errors scores from 23 meteorological stations in the eastern
part of Italy. We show that turbulent fluxes from both setups differ up to
20 % during the bora but not significantly before and after the event.
When compared to observations, two-way coupling ocean temperatures exhibit a
4 times lower root mean square error (RMSE) than those from one-way coupled
system. Two-way coupling improves sensible heat fluxes at all stations but
does not improve latent heat loss. The spatial average of the two-way coupled
atmosphere component is up to 0.3 <inline-formula><mml:math display="inline"><mml:msup><mml:mi/><mml:mo>∘</mml:mo></mml:msup></mml:math></inline-formula>C colder than the one-way coupled
setup, which is an improvement for prognostic lead times up to 20 h. Daily
spatial average of the standard deviation of air temperature errors shows
0.15 <inline-formula><mml:math display="inline"><mml:msup><mml:mi/><mml:mo>∘</mml:mo></mml:msup></mml:math></inline-formula>C improvement in the case of coupled system compared to the
uncoupled. Coupled and uncoupled circulations in the northern Adriatic are
predominantly wind-driven and show no significant mesoscale differences.</p>
  </abstract>
    </article-meta>
  </front>
<body>
      

      <?xmltex \floatpos{t}?><fig id="Ch1.F1" specific-use="star"><caption><p>Study area: the Adriatic Sea and surrounding topography. Red arrows
depict northern Adriatic bora wind jet directions. Bora jet names are in
white letters. The inset (zoom from white dashed rectangle) depicts the
northern
Adriatic along with the Gulf of Trieste. Red <inline-formula><mml:math display="inline"><mml:mo>+</mml:mo></mml:math></inline-formula> signs in the inset mark the
position of Vida oceanographic buoy, Paloma platform and Acqua Alta platform.
JP and SAP denote Jabuka pit and south Adriatic pit, respectively. Light-blue
rectangles show POM river sources marked by the following abbreviations
(listed with yellow letters from the north in counter-clockwise direction):
S – Soča/Isonzo, St – Stella, T – Tagliamento, L – Livenza, P – Piave, Si – Sile,
Ad – Adige, PO – Po, R – Reno, M – Marecchia, V – Vibrata, Of – Ofanto,
Sm – Seman, Sk – Shkumbi, Dr – Drini, B – Bojana, Du – HPP (hydroelectric power plant) Dubrovnik, O – Ombla,
N – Neretva, C – Cetina, J – Jadro, K – Krka, Z – Zrmanja, Se – HPP
Senj, Cr – Crikvenica, Ra – Raša, Mr – Mirna, Ri – Rižana.</p></caption>
      <?xmltex \igopts{width=398.338583pt}?><graphic xlink:href="https://os.copernicus.org/articles/12/71/2016/os-12-71-2016-f01.png"/>

    </fig>

<sec id="Ch1.S1" sec-type="intro">
  <title>Introduction</title>
      <p>The Adriatic Sea is a semi-enclosed basin in the north of the central
Mediterranean, oriented in the SE–NW direction
(Fig. <xref ref-type="fig" rid="Ch1.F1"/>). The northern part of the Adriatic consists
of a shallow shelf with depths not exceeding 60 <inline-formula><mml:math display="inline"><mml:mi mathvariant="normal">m</mml:mi></mml:math></inline-formula>. Further southeast
the depths in the middle Adriatic reach 260 <inline-formula><mml:math display="inline"><mml:mi mathvariant="normal">m</mml:mi></mml:math></inline-formula> in two Jabuka pits,
while the southernmost part, the south Adriatic basin, is the deepest part of
the Adriatic with depths over 1200 <inline-formula><mml:math display="inline"><mml:mi mathvariant="normal">m</mml:mi></mml:math></inline-formula>. The eastern coast of the
Adriatic (Croatia) is mostly rocky, steep, with an abundance of rocky
islands, while the western Adriatic coast (Italy) is low and sandy with
several lagoons <xref ref-type="bibr" rid="bib1.bibx1" id="paren.1"/>.</p>
      <p>The northernmost part of the northern Adriatic is the shallow, semi-enclosed
Gulf of Trieste. The main source of freshwater in the Gulf of Trieste is the
Soča/Isonzo river, while other important riverine inputs to the northern
Adriatic water stem from the Tagliamento, Livenza, Piave, Adige and, most
importantly, Po, rivers. Buoyant spreading of their low salinity waters tends
to generate a classical coastal current, which, under the influence of
Earth's rotation, remains more or less confined to the western Adriatic
(Italian) coast where it flows south as a part of Western Adriatic Current.
Adriatic river climatological discharges, used ubiquitously in the Adriatic
modeling community, were compiled in <xref ref-type="bibr" rid="bib1.bibx28" id="text.2"/>, while another, more
recent, work investigated circulation, dense water formation and its
spreading along the Adriatic basin during the February 2012 bora, using new
Croatian river climatologies (based on 2009–2011 river–runoff observations)
as hydrological forcings in their circulation model <xref ref-type="bibr" rid="bib1.bibx14" id="paren.3"/>.</p>
      <p>Dominant wind forcings can be separated into two classes: the cold
northeasterly bora wind and the warm southeasterly scirocco. The bora denotes
a predominantly winter occurrence of strong, cold and dry continental air
flowing over the Dinaric orographic barriers over the Adriatic sea. The
generating mechanism of the bora was thought to be of katabatic origin
<xref ref-type="bibr" rid="bib1.bibx16" id="paren.4"/>, but several investigations now show that severe episodes of
bora are generated by orographic wave-breaking <xref ref-type="bibr" rid="bib1.bibx13 bib1.bibx33" id="paren.5"/>. The bora seaward airflow is channeled mainly over Trieste, Senj
and other gaps in the Dinaric Alps (see Fig. <xref ref-type="fig" rid="Ch1.F1"/> for
jet locations) on the eastern coast of the Adriatic <xref ref-type="bibr" rid="bib1.bibx5" id="paren.6"/>,
producing the well known wind jets over the ocean surface, which generate
surface offshore water removal in the east and downwelling on the west
Adriatic coast <xref ref-type="bibr" rid="bib1.bibx15" id="paren.7"/>. Bora wind episodes lead to intense
air–sea interactions, which increase the net upward heat fluxes through the
ocean surface and induce negative buoyancy flux on the northern Adriatic
shelf <xref ref-type="bibr" rid="bib1.bibx29" id="paren.8"/>, thus making it one of the most important Mediterranean
dense water formation sites.</p>
      <p>The Adriatic sea has often been investigated using one-way coupled
atmosphere–ocean modeling chains. A series of observations and simulations
have been used routinely to analyze the occurrence of the well-known double
gyre system, occurring north of the 45<inline-formula><mml:math display="inline"><mml:msup><mml:mi/><mml:mo>∘</mml:mo></mml:msup></mml:math></inline-formula> N parallel during intense
bora events <xref ref-type="bibr" rid="bib1.bibx16 bib1.bibx35" id="paren.9"/>. Measurements <xref ref-type="bibr" rid="bib1.bibx35" id="paren.10"/>
suggested a positive surface current curl to the north of and a negative
current curl south of Rovinj. This pattern was reproduced numerically by
several modeling groups <xref ref-type="bibr" rid="bib1.bibx16 bib1.bibx24" id="paren.11"/>. It was also reported
that the NW cyclonic gyre, the so-called Trieste gyre, has a barotropic
current field while the anticyclonic gyre south of Rovinj is baroclinic in
nature <xref ref-type="bibr" rid="bib1.bibx16" id="paren.12"/>. These findings indicate a controlling influence of the
shallower topography along the Italian coast. This was further explored to
quantify the extent of topographic control of wind-driven circulation during
bora and scirocco <xref ref-type="bibr" rid="bib1.bibx18" id="paren.13"/>. It was shown that the topographic
influence during bora episodes is substantial.</p>
      <p>Interannual simulations of the Adriatic deep-water formation indicate that it
is the atmospheric forcing on synoptic spatial and temporal scales that
determines the characteristics and variability of the Adriatic dense water
<xref ref-type="bibr" rid="bib1.bibx19" id="paren.14"/>. Consequently incorporation of an ocean feedback into
the atmospheric model, namely, the two-way atmosphere–ocean coupling, should
lead to improvements in modeled ocean and atmosphere response during synoptic
events and their consequent relaxation on longer timescales. Two-way
atmosphere–ocean coupling approach to modeling the Adriatic was to the best
of our knowledge first implemented to achieve realistic simulations of the
Adriatic baroclinic circulation during bora episodes in the winter and spring
2001 <xref ref-type="bibr" rid="bib1.bibx25 bib1.bibx26" id="paren.15"/> and to study marine atmospheric
conditions over the Adriatic during February 2003 <xref ref-type="bibr" rid="bib1.bibx8" id="paren.16"/>. A
coupled atmosphere–wave–ocean modeling system (without ocean feedback to
the atmosphere) over the Adriatic domain was recently used to simulate the
formation and relaxation of Adriatic dense water during and after
February 2012 bora event <xref ref-type="bibr" rid="bib1.bibx2" id="paren.17"/>. All these studies indicate that
two-way atmosphere–ocean coupling could be beneficial for modeling the
air–sea state during periods of intense atmosphere–ocean interactions.</p>
      <p>As already noted, the February 2012 bora was a landmark event <xref ref-type="bibr" rid="bib1.bibx29 bib1.bibx22" id="paren.18"/> due to both its strength and its duration. Consequently, there
were several routine (and ad hoc) field campaigns performed in the Gulf of
Trieste and in mid-Adriatic immediately before and immediately after the
strongest bora episode, and the event itself was subsequently widely reported
<xref ref-type="bibr" rid="bib1.bibx29 bib1.bibx22" id="paren.19"/> as well as modeled <xref ref-type="bibr" rid="bib1.bibx14 bib1.bibx2" id="paren.20"/>.
Atmospheric forcing caused exceptionally large surface heat fluxes, leading
to an almost unprecedented water cooling and dense water formation in the
Gulf of Trieste, where the sea temperatures dropped to 4 <inline-formula><mml:math display="inline"><mml:msup><mml:mi/><mml:mo>∘</mml:mo></mml:msup></mml:math></inline-formula>C and
density anomalies exceeded 30.55 <inline-formula><mml:math display="inline"><mml:mrow><mml:mi mathvariant="normal">kg</mml:mi><mml:mspace width="0.125em" linebreak="nobreak"/><mml:msup><mml:mi mathvariant="normal">m</mml:mi><mml:mrow><mml:mo>-</mml:mo><mml:mn mathvariant="normal">3</mml:mn></mml:mrow></mml:msup></mml:mrow></mml:math></inline-formula> <xref ref-type="bibr" rid="bib1.bibx29 bib1.bibx22" id="paren.21"/>. Intensive air–sea interactions and the substantial number of
observations make this event a suitable candidate for verifying two-way
atmosphere–ocean coupled simulations.</p>
      <p>In this paper we present two-way atmosphere–ocean coupling simulations of
the February 2012 bora, along with comparisons with an identically set up,
one-way coupled system. In Sect. <xref ref-type="sec" rid="Ch1.S2"/> we present the two numerical
systems in detail, focusing first on the one-way coupled configuration and
then on the two-way coupled configuration. In Sect. <xref ref-type="sec" rid="Ch1.S3"/> the
observational data are described, together with dates and locations of the CTD (conductivity, temperature, and depth)
casts in early 2012. In Sect. <xref ref-type="sec" rid="Ch1.S4"/> the results of both one-way and
two-way coupled numerical systems are presented and discussed, focusing on
air–sea fluxes and air and sea temperatures, and we compare these with
observations. Concluding remarks are presented in Sect. <xref ref-type="sec" rid="Ch1.S5"/>.</p>
</sec>
<sec id="Ch1.S2">
  <title>Modeling systems</title>
<sec id="Ch1.S2.SS1">
  <title>The ALADIN atmospheric model: a one-way coupled setup</title>
      <p>ALADIN (Aire Limitée Adaptation dynamique Développement
InterNational) <xref ref-type="bibr" rid="bib1.bibx11" id="paren.22"/> is a spectral limited area mesoscale
numerical weather prediction (NWP) model developed by a consortium of
national weather services led by Météo France. ALADIN has been used
for routine weather forecasting in Slovenia since 1997. It uses a two time
level semi-Lagrangian semi-implicit advection scheme with several choices of
physical packages. Its dynamical core is shared with the Integrated
Forecasting System of the European Centre for Medium-range Weather Forecasts (ECMWF)
and of AROME <xref ref-type="bibr" rid="bib1.bibx30" id="paren.23"><named-content content-type="pre">Application de la Recherche à
l'Opérationnel à Méso-Echelle;</named-content></xref>.</p>
      <p>The version of the model used for the experiments in this paper is that of
the currently operational version used at Slovenian weather service. It runs
on a <inline-formula><mml:math display="inline"><mml:mrow><mml:mn>432</mml:mn><mml:mo>×</mml:mo><mml:mn>432</mml:mn></mml:mrow></mml:math></inline-formula> horizontal Lambert conic conformal grid with
4.4 <inline-formula><mml:math display="inline"><mml:mi mathvariant="normal">km</mml:mi></mml:math></inline-formula> resolution and 87 vertical levels with the model top at
1 <inline-formula><mml:math display="inline"><mml:mi mathvariant="normal">hPa</mml:mi></mml:math></inline-formula> and model integration time step of 180 <inline-formula><mml:math display="inline"><mml:mi mathvariant="normal">s</mml:mi></mml:math></inline-formula>. The model
domain spans <inline-formula><mml:math display="inline"><mml:mrow><mml:msup><mml:mn>0.7</mml:mn><mml:mo>∘</mml:mo></mml:msup><mml:mspace width="0.125em" linebreak="nobreak"/><mml:mi mathvariant="normal">W</mml:mi><mml:mo>,</mml:mo><mml:msup><mml:mn>28.6</mml:mn><mml:mo>∘</mml:mo></mml:msup><mml:mspace width="0.125em" linebreak="nobreak"/><mml:mi mathvariant="normal">E</mml:mi></mml:mrow></mml:math></inline-formula> in longitude
and <inline-formula><mml:math display="inline"><mml:mrow><mml:msup><mml:mn>37.4</mml:mn><mml:mo>∘</mml:mo></mml:msup><mml:mspace linebreak="nobreak" width="0.125em"/><mml:mi mathvariant="normal">N</mml:mi><mml:mo>,</mml:mo><mml:msup><mml:mn>55.0</mml:mn><mml:mo>∘</mml:mo></mml:msup><mml:mspace width="0.125em" linebreak="nobreak"/><mml:mi mathvariant="normal">N</mml:mi></mml:mrow></mml:math></inline-formula> in latitude. The
physics package used in the model uses Modular, Multi-scale, Microphysics and
Transport (3MT) structure <xref ref-type="bibr" rid="bib1.bibx12" id="paren.24"/>. Initial conditions for the
model are provided by atmospheric analysis with 3-hourly three-dimensional
variational assimilation (3D-Var) <xref ref-type="bibr" rid="bib1.bibx11 bib1.bibx31" id="paren.25"/> and
optimal interpolation for surface and soil variables. Most of the
conventional and satellite observations are assimilated, including surface
stations (Synop) and ship reports, radiosondes, Automated Meteorological Data
Relay (AMDAR), atmospheric motion vectors (AMV), Advanced Microwave Sounding
Unit (AMSU)-A and -B, Metop-A and -B polar orbiters and water vapor channels
from the geostationary Meteosat Second Generation Spinning Enhanced Visible
and Infrared Imager (SEVIRI). Observations within <inline-formula><mml:math display="inline"><mml:mo>±</mml:mo></mml:math></inline-formula>1.5 h from the
analysis time are assimilated. Information at the domain edge is obtained
from the global model by applying Davies relaxation <xref ref-type="bibr" rid="bib1.bibx6" id="paren.26"/>.
Lateral boundary conditions (LBCs) are provided by the ECMWF boundary
conditions optional project. LBCs are applied with a 1 h period in the
assimilation cycle and a 3 h period during model forecasts. Boundary
condition information is interpolated linearly for time steps in-between
these times.<?xmltex \hack{\newpage}?></p>
      <p>Within one-way coupled ALADIN (from now on referred to as ALADIN1w, while the
two-way coupled version will be referred to as ALADIN2w) itself there is no
implementation of ocean component; ocean variables values are therefore kept
constant during the model run. In the operational (one-way coupled) ALADIN1w
in Slovenia, sea surface temperature (SST) is initialized from the most
recent host model analysis of the ECMWF model that uses Operational Sea Surface Temperature and Sea Ice Analysis <xref ref-type="bibr" rid="bib1.bibx7" id="paren.27"><named-content content-type="pre">OSTIA;</named-content></xref>,
supplied by the National Environmental Satellite, Data and Information
Service (NESDIS) of the American National Ocean and Atmospheric
Administration (NOAA). OSTIA combines satellite data with in situ
measurements by applying optimal interpolation. In ALADIN1w the SST remains
constant over each 24 h of the model run.</p>
      <p>In both ALADIN1w and ALADIN2w, turbulent fluxes over the ocean surface are parameterized in the usual fashion as follows.
Wind stress is calculated as

                <disp-formula id="Ch1.E1" content-type="numbered"><mml:math display="block"><mml:mrow><mml:mo>|</mml:mo><mml:mi mathvariant="bold-italic">τ</mml:mi><mml:mo>|</mml:mo><mml:mo>=</mml:mo><mml:msub><mml:mi mathvariant="italic">ρ</mml:mi><mml:mi mathvariant="normal">a</mml:mi></mml:msub><mml:msub><mml:mi>c</mml:mi><mml:mi mathvariant="normal">pa</mml:mi></mml:msub><mml:msub><mml:mi>C</mml:mi><mml:mi mathvariant="normal">D</mml:mi></mml:msub><mml:mo>|</mml:mo><mml:mi mathvariant="bold-italic">V</mml:mi><mml:msup><mml:mo>|</mml:mo><mml:mn mathvariant="normal">2</mml:mn></mml:msup><mml:mo>,</mml:mo></mml:mrow></mml:math></disp-formula>

          where <inline-formula><mml:math display="inline"><mml:mrow><mml:msub><mml:mi mathvariant="italic">ρ</mml:mi><mml:mi mathvariant="normal">a</mml:mi></mml:msub></mml:mrow></mml:math></inline-formula> is the moist air density, <inline-formula><mml:math display="inline"><mml:mrow><mml:msub><mml:mi>c</mml:mi><mml:mi mathvariant="normal">pa</mml:mi></mml:msub></mml:mrow></mml:math></inline-formula> is the specific heat of air
at constant pressure, <inline-formula><mml:math display="inline"><mml:mrow><mml:msub><mml:mi>C</mml:mi><mml:mi mathvariant="normal">D</mml:mi></mml:msub></mml:mrow></mml:math></inline-formula> is the drag coefficient and <inline-formula><mml:math display="inline"><mml:mi mathvariant="bold-italic">V</mml:mi></mml:math></inline-formula> is the
surface wind velocity. Sensible heat flux is parameterized as

                <disp-formula id="Ch1.E2" content-type="numbered"><mml:math display="block"><mml:mrow><mml:msub><mml:mi>Q</mml:mi><mml:mi mathvariant="normal">H</mml:mi></mml:msub><mml:mo>=</mml:mo><mml:msub><mml:mi mathvariant="italic">ρ</mml:mi><mml:mi mathvariant="normal">a</mml:mi></mml:msub><mml:msub><mml:mi>c</mml:mi><mml:mi mathvariant="normal">pa</mml:mi></mml:msub><mml:msub><mml:mi>C</mml:mi><mml:mi mathvariant="normal">H</mml:mi></mml:msub><mml:mo>|</mml:mo><mml:mi mathvariant="bold-italic">V</mml:mi><mml:mo>|</mml:mo><mml:mo>(</mml:mo><mml:msub><mml:mi>T</mml:mi><mml:mi mathvariant="normal">air</mml:mi></mml:msub><mml:mo>-</mml:mo><mml:msub><mml:mi>T</mml:mi><mml:mi mathvariant="normal">sea</mml:mi></mml:msub><mml:mo>)</mml:mo><mml:mo>,</mml:mo></mml:mrow></mml:math></disp-formula>

          where <inline-formula><mml:math display="inline"><mml:mrow><mml:msub><mml:mi>C</mml:mi><mml:mi mathvariant="normal">H</mml:mi></mml:msub></mml:mrow></mml:math></inline-formula> is the sensible heat transfer coefficient, <inline-formula><mml:math display="inline"><mml:mrow><mml:msub><mml:mi>T</mml:mi><mml:mi mathvariant="normal">air</mml:mi></mml:msub></mml:mrow></mml:math></inline-formula> is air
temperature at the lowest model level (10 <inline-formula><mml:math display="inline"><mml:mi mathvariant="normal">m</mml:mi></mml:math></inline-formula> approximately) and
<inline-formula><mml:math display="inline"><mml:mrow><mml:msub><mml:mi>T</mml:mi><mml:mi mathvariant="normal">sea</mml:mi></mml:msub></mml:mrow></mml:math></inline-formula> is sea surface temperature.</p>
      <p>Latent heat flux is calculated as
            <disp-formula id="Ch1.E3" content-type="numbered"><mml:math display="block"><mml:mrow><mml:msub><mml:mi>Q</mml:mi><mml:mi>E</mml:mi></mml:msub><mml:mo>=</mml:mo><mml:msub><mml:mi mathvariant="italic">ρ</mml:mi><mml:mi mathvariant="normal">a</mml:mi></mml:msub><mml:mi>L</mml:mi><mml:msub><mml:mi>C</mml:mi><mml:mi>E</mml:mi></mml:msub><mml:mo>|</mml:mo><mml:mi mathvariant="bold-italic">V</mml:mi><mml:mo>|</mml:mo><mml:mo>(</mml:mo><mml:msub><mml:mi>q</mml:mi><mml:mi mathvariant="normal">air</mml:mi></mml:msub><mml:mo>-</mml:mo><mml:msub><mml:mi>q</mml:mi><mml:mi mathvariant="normal">sat</mml:mi></mml:msub><mml:mo>(</mml:mo><mml:msub><mml:mi>T</mml:mi><mml:mi mathvariant="normal">sea</mml:mi></mml:msub><mml:mo>)</mml:mo><mml:mo>)</mml:mo><mml:mo>,</mml:mo></mml:mrow></mml:math></disp-formula>
          where <inline-formula><mml:math display="inline"><mml:mi>L</mml:mi></mml:math></inline-formula> is latent heat of vaporization, <inline-formula><mml:math display="inline"><mml:mrow><mml:msub><mml:mi>C</mml:mi><mml:mi>E</mml:mi></mml:msub></mml:mrow></mml:math></inline-formula> is the vapor transfer
coefficient, <inline-formula><mml:math display="inline"><mml:mrow><mml:msub><mml:mi>q</mml:mi><mml:mi mathvariant="normal">air</mml:mi></mml:msub></mml:mrow></mml:math></inline-formula> is the atmospheric specific humidity at the
lowest model level and <inline-formula><mml:math display="inline"><mml:mrow><mml:msub><mml:mi>q</mml:mi><mml:mi mathvariant="normal">sat</mml:mi></mml:msub><mml:mo>(</mml:mo><mml:msub><mml:mi>T</mml:mi><mml:mi mathvariant="normal">sea</mml:mi></mml:msub><mml:mo>)</mml:mo></mml:mrow></mml:math></inline-formula> is the saturated
specific humidity at the sea surface temperature.</p>
      <p>The turbulent transfer coefficients <inline-formula><mml:math display="inline"><mml:mrow><mml:msub><mml:mi>C</mml:mi><mml:mi mathvariant="normal">D</mml:mi></mml:msub></mml:mrow></mml:math></inline-formula>, <inline-formula><mml:math display="inline"><mml:mrow><mml:msub><mml:mi>C</mml:mi><mml:mi mathvariant="normal">H</mml:mi></mml:msub></mml:mrow></mml:math></inline-formula> and <inline-formula><mml:math display="inline"><mml:mrow><mml:msub><mml:mi>C</mml:mi><mml:mi>E</mml:mi></mml:msub></mml:mrow></mml:math></inline-formula> are computed using a
modified Louis scheme that takes into account different roughness lengths for
heat and momentum <xref ref-type="bibr" rid="bib1.bibx20" id="paren.28"/>.</p>
      <p>Net radiation at the surface, <inline-formula><mml:math display="inline"><mml:mrow><mml:msub><mml:mi>R</mml:mi><mml:mi mathvariant="normal">n</mml:mi></mml:msub></mml:mrow></mml:math></inline-formula>, is computed as

                <disp-formula id="Ch1.E4" content-type="numbered"><mml:math display="block"><mml:mrow><mml:msub><mml:mi>R</mml:mi><mml:mi mathvariant="normal">n</mml:mi></mml:msub><mml:mo>=</mml:mo><mml:msub><mml:mi>j</mml:mi><mml:mi mathvariant="normal">SW</mml:mi></mml:msub><mml:mo>+</mml:mo><mml:mi mathvariant="italic">ϵ</mml:mi><mml:mo>(</mml:mo><mml:msub><mml:mi>R</mml:mi><mml:mrow><mml:mi mathvariant="normal">atm</mml:mi><mml:mo>↓</mml:mo></mml:mrow></mml:msub><mml:mo>-</mml:mo><mml:mi mathvariant="italic">σ</mml:mi><mml:msubsup><mml:mi>T</mml:mi><mml:mi mathvariant="normal">sea</mml:mi><mml:mn mathvariant="normal">4</mml:mn></mml:msubsup><mml:mo>)</mml:mo><mml:mo>,</mml:mo></mml:mrow></mml:math></disp-formula>

          where <inline-formula><mml:math display="inline"><mml:mrow><mml:msub><mml:mi>j</mml:mi><mml:mi mathvariant="normal">SW</mml:mi></mml:msub></mml:mrow></mml:math></inline-formula> is the net shortwave radiation at the ocean surface,
<inline-formula><mml:math display="inline"><mml:mrow><mml:mi mathvariant="italic">ϵ</mml:mi><mml:mo>=</mml:mo><mml:mn>0.96</mml:mn></mml:mrow></mml:math></inline-formula> is the atmosphere and ocean long-wave radiation emissivity
factor, <inline-formula><mml:math display="inline"><mml:mrow><mml:msub><mml:mi>R</mml:mi><mml:mrow><mml:mi mathvariant="normal">atm</mml:mi><mml:mo>↓</mml:mo></mml:mrow></mml:msub></mml:mrow></mml:math></inline-formula> is the atmospheric downward long-wave radiation,
<inline-formula><mml:math display="inline"><mml:mi mathvariant="italic">σ</mml:mi></mml:math></inline-formula> is the Stefan–Boltzmann constant and <inline-formula><mml:math display="inline"><mml:mrow><mml:msub><mml:mi>T</mml:mi><mml:mi mathvariant="normal">sea</mml:mi></mml:msub></mml:mrow></mml:math></inline-formula> is the sea surface
temperature.</p>
      <p>The total heat flux <inline-formula><mml:math display="inline"><mml:mrow><mml:msub><mml:mi>Q</mml:mi><mml:mi mathvariant="normal">T</mml:mi></mml:msub></mml:mrow></mml:math></inline-formula> presented in this paper was obtained from

                <disp-formula id="Ch1.E5" content-type="numbered"><mml:math display="block"><mml:mrow><mml:msub><mml:mi>Q</mml:mi><mml:mi mathvariant="normal">T</mml:mi></mml:msub><mml:mo>=</mml:mo><mml:msub><mml:mi>Q</mml:mi><mml:mi>E</mml:mi></mml:msub><mml:mo>+</mml:mo><mml:msub><mml:mi>Q</mml:mi><mml:mi mathvariant="normal">H</mml:mi></mml:msub><mml:mo>+</mml:mo><mml:msub><mml:mi>R</mml:mi><mml:mi mathvariant="normal">n</mml:mi></mml:msub><mml:mo>.</mml:mo></mml:mrow></mml:math></disp-formula></p>
      <p>Throughout the paper, fluxes are depicted as positive when they represent an energy gain
for the ocean and as negative when they represent energy loss from the ocean.</p>
</sec>
<sec id="Ch1.S2.SS2">
  <title>POM: one-way coupled system setup</title>
      <p>The ocean model for one-way and two-way coupling is the Princeton ocean model (POM)
<xref ref-type="bibr" rid="bib1.bibx3 bib1.bibx21" id="paren.29"/>. Our configuration of POM was set
up for the Adriatic domain with 271 longitude cells spanning
12.0–21.0<inline-formula><mml:math display="inline"><mml:msup><mml:mi/><mml:mo>∘</mml:mo></mml:msup></mml:math></inline-formula> E and with 209 latitude cells spanning
39.0–45.9375<inline-formula><mml:math display="inline"><mml:msup><mml:mi/><mml:mo>∘</mml:mo></mml:msup></mml:math></inline-formula> <inline-formula><mml:math display="inline"><mml:mi mathvariant="normal">N</mml:mi></mml:math></inline-formula>. Grid horizontal angular resolution is
<inline-formula><mml:math display="inline"><mml:mrow><mml:msup><mml:mn mathvariant="normal">1</mml:mn><mml:mo>∘</mml:mo></mml:msup><mml:mo>/</mml:mo><mml:mn>30</mml:mn><mml:mo>×</mml:mo><mml:msup><mml:mn mathvariant="normal">1</mml:mn><mml:mo>∘</mml:mo></mml:msup><mml:mo>/</mml:mo><mml:mn>30</mml:mn></mml:mrow></mml:math></inline-formula> and its vertical discretization contains
25 <inline-formula><mml:math display="inline"><mml:mi mathvariant="italic">σ</mml:mi></mml:math></inline-formula> layers. We will refer to this one-way coupled Adriatic setup of
the model as POM1w and the two-way coupled system as POM2w. As noted below,
POM2w setup inherits most of its features from the one-way coupled setup.</p>
      <p>River outflows, depicted in (Fig. <xref ref-type="fig" rid="Ch1.F1"/>), were included in
both POM1w and POM2w using climatologies for the rivers Stella, Tagliamento,
Livenza, Piave, Sile, Adige, Reno, Marecchia, Vibrata, Ofanto, Bojana, Drini,
Shkumbi and Seman from <xref ref-type="bibr" rid="bib1.bibx28" id="paren.30"/>. Climatologies for rivers Mirna,
Raša, Crikvenica, HPP (hydroelectric power plant) Senj, Zrmanja, Krka,
Jadro, Cetina, Neretva, Ombla and HPP Dubrovnik were taken from
<xref ref-type="bibr" rid="bib1.bibx14" id="paren.31"/>. Po River discharges were included using daily means of
hourly discharge observations from Pontelagoscuro station. Soča/Isonzo
river discharge was implemented using hourly operational HFS (hydrological
forecasting system) forecasts from the hydrological forecasting system for
the Soča/Isonzo River (see subsection <xref ref-type="sec" rid="Ch1.S2.SS4"/> below). Rižana
climatology was created from measurements provided by the Slovenian
Environment Agency (ARSO). River temperatures were not defined and at the
river mouth river discharges attain the temperature of the surrounding ocean,
but with zero salinity. Po River discharge was split into five separate
neighboring cells, partly to mimic the Po River delta but also because high
Po River discharges (above 2000 <inline-formula><mml:math display="inline"><mml:mrow><mml:msup><mml:mi mathvariant="normal">m</mml:mi><mml:mn mathvariant="normal">3</mml:mn></mml:msup><mml:mspace linebreak="nobreak" width="0.125em"/><mml:msup><mml:mi mathvariant="normal">s</mml:mi><mml:mrow><mml:mo>-</mml:mo><mml:mn mathvariant="normal">1</mml:mn></mml:mrow></mml:msup></mml:mrow></mml:math></inline-formula>), when applied to a
single discharge cell, sometimes cause violations of the
Courant–Friedrichs–Lewy (CFL) stability criteria.</p>
      <p>Lateral boundary conditions at the open boundary are applied for temperature,
salinity, sea surface elevation and zonal and meridional velocity components.
All these are taken daily from MyOcean Mediterranean Forecasting System
<xref ref-type="bibr" rid="bib1.bibx32" id="paren.32"><named-content content-type="pre">MFS;</named-content></xref> and the model is forced with them for 24 h of
model runtime. Every 24 h of model runtime, the simulation run was
terminated and ocean models were hot started from their previous-step restart
files. Both coupled and uncoupled ALADIN setups are initialized from the same
3-hourly 3D-Var data assimilation cycle using OSTIA SST as the surface
boundary condition. Tidal forcing in both ocean models is provided by the
Oregon State University Tidal Prediction Software (OTPS), set up on the
POM1w/2w grid, but tidal effects are not considered in this paper.</p>
      <p>Both setups were initialized on 6 November 2011, 00:00 UTC, with temperature,
salinity, sea surface elevation and zonal and meridional velocity components
of the MyOcean MFS <xref ref-type="bibr" rid="bib1.bibx32" id="paren.33"/>.</p>
      <p>POM1w mode splitting is inherited from POM: the external time step, used for
the barotropic mode computation, is set to 1 <inline-formula><mml:math display="inline"><mml:mi mathvariant="normal">s</mml:mi></mml:math></inline-formula>; the internal time
step, used for the baroclinic mode calculations, is set to 90 <inline-formula><mml:math display="inline"><mml:mi mathvariant="normal">s</mml:mi></mml:math></inline-formula>.
These time steps and grid resolutions were set in order to satisfy the CFL
condition for numerical stability. The vertical turbulence closure scheme is
the usual 2.5 level Mellor–Yamada; horizontal diffusion is treated using the
Smagorinsky formula.</p>
      <p>Atmospheric input in POM1w is implemented through the input of hourly
ALADIN1w fields contained in ALADIN1w gridded binary files. POM1w bulk flux
computations were disabled and the POM1w model obtains all its fluxes from
the ALADIN1w model.</p>
</sec>
<sec id="Ch1.S2.SS3">
  <title>Two-way atmosphere–ocean coupled system setup</title>
      <p>Apart from the coupling interface, the two-way coupled setup of POM (referred
to as POM2w) is identical to the POM1w setup. The spatial oceanic grid and
vertical discretization of the POM2w model domain remain the same as with
POM1w, as do the river discharges and turbulence closures. The baroclinic
time step in POM2w was adjusted to match the ALADIN2w computational time step
of 180 <inline-formula><mml:math display="inline"><mml:mi mathvariant="normal">s</mml:mi></mml:math></inline-formula>. The exchange of coupling quantities between ALADIN2w and
POM2w was enforced at each computational time step, namely, every 180 <inline-formula><mml:math display="inline"><mml:mi mathvariant="normal">s</mml:mi></mml:math></inline-formula>
of coupled system runtime. As in POM1w, the bulk heat flux computation in
POM2w was disabled to be consistent with the POM1w setup and to obtain
a balanced heat flux budget in the atmosphere–ocean system. Heat fluxes are
computed only by ALADIN2w and are sent to POM2w via a coupling exchange
scheme.</p>
      <p>The atmosphere–ocean coupling itself is implemented using the OASIS3-MCT
model coupling toolkit <xref ref-type="bibr" rid="bib1.bibx34" id="paren.34"/> with the data flow as shown on
Fig. <xref ref-type="fig" rid="Ch1.F2"/>. In the Figure, the rounded rectangles represent
distinct OASIS objects, which are effectively treated by OASIS as independent
models, even though that is only true in the case of POM2w and ALADIN2w, and
is not true for pseudo-MFS and pseudo-MERGER, as will be explained
immediately below.</p>

      <?xmltex \floatpos{t}?><fig id="Ch1.F2"><caption><p>Two-way atmosphere–ocean coupling scheme for one coupling time step.
Rounded rectangles denote distinct OASIS models, effectively treated by OASIS
as independent. The arrows denote exchanged coupling quantities and direction
of the transfer. Time instant of any specific coupling exchanges within one
coupling time step grows from left (earliest) to right
(latest).</p></caption>
          <?xmltex \igopts{width=236.157874pt}?><graphic xlink:href="https://os.copernicus.org/articles/12/71/2016/os-12-71-2016-f02.png"/>

        </fig>

      <?xmltex \floatpos{h!}?><fig id="Ch1.F3"><caption><p>Domains of all OASIS models used in the atmosphere–ocean coupling
scheme. White rectangle: atmospheric model ALADIN domain. Orange rectangle:
ocean circulation model POM2w domain. Black rectangle: remapping pseudo model
pseudo-MERGER domain. Blue rectangle: domain defined by the MyOcean MFS
NetCDF files, read by the pseudo-MFS pseudo model.</p></caption>
          <?xmltex \igopts{width=199.169291pt}?><graphic xlink:href="https://os.copernicus.org/articles/12/71/2016/os-12-71-2016-f03.png"/>

        </fig>

      <p>The coupling scheme uses four OASIS objects with domains shown on Fig. <xref ref-type="fig" rid="Ch1.F3"/>.
A more detailed description of all OASIS objects is as follows:
<list list-type="order"><list-item>
      <p>ALADIN. Atmospheric model. Receives the SST field from the pseudo-MERGER pseudo-model and sends the
computed mean sea-level pressure, air temperature, precipitation,
wind speed (<inline-formula><mml:math display="inline"><mml:mi>u</mml:mi></mml:math></inline-formula> and <inline-formula><mml:math display="inline"><mml:mi>v</mml:mi></mml:math></inline-formula> directions), humidity, sensible and latent heat
fluxes, and shortwave and long-wave downward radiation fields to the POM
model.</p></list-item><list-item>
      <p>POM. Ocean circulation model. Receives mean sea-level pressure,
air temperature, precipitation, wind speed (<inline-formula><mml:math display="inline"><mml:mi>u</mml:mi></mml:math></inline-formula> and <inline-formula><mml:math display="inline"><mml:mi>v</mml:mi></mml:math></inline-formula> directions),
relative humidity, sensible and latent heat fluxes, and shortwave and
long-wave downward radiation fields from the ALADIN2w model. It sends
the computed SST field to the pseudo-MERGER pseudo-model.</p></list-item><list-item>
      <p>Pseudo-MFS. A pseudo-model that reads the daily mean SST field
from the MyOcean MFS model <xref ref-type="bibr" rid="bib1.bibx32" id="paren.35"/> NetCDF file, and sends
it to the pseudo-MERGER pseudo-model.</p></list-item><list-item>
      <p>Pseudo-MERGER. A pseudo-model that receives the SST fields
from POM2w and MFS models, merges them on a common ALADIN2w grid and
sends the merged SST field to the ALADIN2w model. Inclusion of
pseudo-MFS and pseudo-MERGER pseudo models into the OASIS scheme was
necessary because the ALADIN2w domain extends well beyond the POM2w
domain (see Fig. <xref ref-type="fig" rid="Ch1.F3"/>). We therefore needed to provide
an SST estimate to ALADIN2w in regions outside the POM2w domain and
this was achieved using MFS NetCDF files.</p></list-item></list>
The time step for exchanging fields is the same for all models –
180 <inline-formula><mml:math display="inline"><mml:mi mathvariant="normal">s</mml:mi></mml:math></inline-formula>. At the first time step, the models are initialized
independently. From the second time step on, the following sequence takes
place: (1) all coupling fields (see ALADIN2w-POM2w connection on
Fig. <xref ref-type="fig" rid="Ch1.F2"/>) are sent from ALADIN2w to the POM2w model. (2) POM2w
integrates the model differential equations for one time step (180 <inline-formula><mml:math display="inline"><mml:mi mathvariant="normal">s</mml:mi></mml:math></inline-formula>)
and sends the SST field (on the Adriatic domain) to the pseudo-MERGER pseudo
model. Simultaneously, pseudo-MFS sends its (daily mean) SST (on the
Mediterranean domain) to pseudo-MERGER. (3) pseudo-MERGER merges both SST
fields and remaps the merged field to the ALADIN2w model grid.</p>
      <p>In Aladin2w outside POM2w domain we opted for the use of MFS SST field
instead of OSTIA SST field. This was done because POM2w is forced by the MFS
and merging POM2w SST with OSTIA SST might introduce substantial SST
discontinuities at POM2w domain boundary. To further reduce sharp SST
gradients on the pseudo-MFS/POM2w boundary, a linear interpolation between
MFS and POM2w SST fields is performed within pseudo-MERGER during the merging
and remapping process. The interpolation of SST takes place in all cells
located up to five grid cells north of the POM2w open boundary (39<inline-formula><mml:math display="inline"><mml:msup><mml:mi/><mml:mo>∘</mml:mo></mml:msup></mml:math></inline-formula> <inline-formula><mml:math display="inline"><mml:mi mathvariant="normal">N</mml:mi></mml:math></inline-formula> parallel between Italy and Greece, see
Fig. <xref ref-type="fig" rid="Ch1.F3"/>). (4) This merged and remapped MFS-POM2w SST field is
sent from pseudo-MERGER to ALADIN2w, completing one coupling time step.</p>
      <p>The ALADIN analyses used to initialize 24 h two-way coupled runs are taken
from an independent data assimilation cycle, computed in advance. The same
analyses are used in one-way and two-way coupled setups. This means that the
initial conditions for ALADIN2w do not contain oceanographic information from
the previous POM2w run; only OSTIA analyses are used for SST in the ALADIN
assimilation cycle. The full two-way ocean–atmosphere assimilation system
would consist of a fully coupled ocean–atmosphere model for integration
during the data assimilation cycle of ALADIN as well: such a strategy was
successfully implemented in <xref ref-type="bibr" rid="bib1.bibx27" id="paren.36"/> and its implementation is left
for future work.</p>
</sec>
<sec id="Ch1.S2.SS4">
  <?xmltex \opttitle{The hydrological forecasting system on the So\v{c}a/Isonzo river basin}?><title>The hydrological forecasting system on the Soča/Isonzo river basin</title>
      <p>To the best of our knowledge, Soča/Isonzo river–runoff measurements at
the river mouth are not available for the February 2012 bora event or for the
preconditioning phase. We therefore decided to employ a HFS on the
Soča/Isonzo river basin to remedy this lack of measurements. HFS has been
operative at the Slovenian Environment Agency since early 2012 and provides
Soča/Isonzo river hourly discharges for ocean circulation models in the
numerical experiments presented in this paper. The most important input to
the HFS system is that of the meteorological forecasts for precipitation and
air temperature at ground level from the ALADIN and ECMWF atmospheric models,
which are updated four and two times a day, respectively. The HFS is based on
DHI MIKE11 engines – the hydrological model for rainfall–runoff simulations
and the 1-D hydrodynamic (HD) model for river routing simulations. The
rainfall–runoff model <xref ref-type="bibr" rid="bib1.bibx23" id="paren.37"/> is a deterministic, conceptual,
lumped model describing the land phase of the hydrological cycle in
a simplified form. The parameters of the physical and semi-empirical model
formulations for the snow, surface, root-zone and groundwater storages
constitute the average values for each of the sub-catchments. The
hydrological Soča/Isonzo river basin model is defined in 30
sub-catchments with an average area of 115 <inline-formula><mml:math display="inline"><mml:mrow><mml:msup><mml:mi mathvariant="normal">km</mml:mi><mml:mn mathvariant="normal">2</mml:mn></mml:msup></mml:mrow></mml:math></inline-formula>. In the model snow
module the sub-catchments are divided into 100 <inline-formula><mml:math display="inline"><mml:mi mathvariant="normal">m</mml:mi></mml:math></inline-formula> high elevation zones
giving the model setup a semi-distributed character. The river network of the
Soča/Isonzo river basin HD model is represented by 17 branches having a
calculation point density of less than 2000 <inline-formula><mml:math display="inline"><mml:mi mathvariant="normal">m</mml:mi></mml:math></inline-formula> in river length. The
upstream and lateral boundary conditions of the HD model are the runoff
hydrographs simulated by the model while the downstream boundary condition is
the constant water level of the Adriatic Sea.</p>
      <p>The rainfall–runoff model was calibrated on the measured precipitation and
discharge data in the period between 1998 and 2007. The precipitation data
set contained daily and hourly values from approximately 115 meteorological
stations from the entire river basin. The discharge data set was prepared for
19 Slovenian hydrological stations only as the discharge values for the
Italian part of the basin (approximately one third of the basin area) did not
exist. The Italian subcatchments were therefore assigned parameter values
similar to the parameters of the calibrated Slovenian subcatchments that have
common topographic and hydrological characteristics. The model calibration
was performed on low and high flow conditions as well as on long-term water
balance on the downstream-lying hydrological stations Soča/Isonzo Solkan
and Vipava Miren in southwestern mainland Slovenia (not shown on the map).
Model parameters were verified on the Slovenian hydrological stations with
the 2008–2010 data set.<?xmltex \hack{\vspace{-3mm}}?></p>
</sec>
</sec>
<sec id="Ch1.S3">
  <title>Observations</title>
      <p>The oceanographic buoy Vida is a coastal observation platform, operated by
the National Institute of Biology (NIB). It is located in the southern part
of the Gulf of Trieste at 13.55505<inline-formula><mml:math display="inline"><mml:msup><mml:mi/><mml:mo>∘</mml:mo></mml:msup></mml:math></inline-formula> E, 45.5488<inline-formula><mml:math display="inline"><mml:msup><mml:mi/><mml:mo>∘</mml:mo></mml:msup></mml:math></inline-formula> N, as
depicted in the inset of Fig. <xref ref-type="fig" rid="Ch1.F1"/> (marked with a red
cross). Data from the buoy are multifaceted (air temperature, air humidity,
currents, waves, sea temperature, salinity, dissolved oxygen, chlorophyll
concentration, etc.) and are transferred in real-time to the NIB server via
an Ethernet link and are publicly available
(<uri>http://www.nib.si/mbp/en/buoy/</uri>). A Nortek AWAC acoustic Doppler
current profiler is situated on the sea bed beneath the buoy to monitor
profiles of current (at 1 m intervals along the water column) and the sea
floor temperature at a depth of 22.5 <inline-formula><mml:math display="inline"><mml:mi mathvariant="normal">m</mml:mi></mml:math></inline-formula>. The sea temperature at
2.5 <inline-formula><mml:math display="inline"><mml:mi mathvariant="normal">m</mml:mi></mml:math></inline-formula> beneath the surface is measured using a SeaBird 16plus Seacat
salinity and temperature sensor with a sampling period of 300 <inline-formula><mml:math display="inline"><mml:mi mathvariant="normal">s</mml:mi></mml:math></inline-formula>.
Temperature measurements used for comparisons in this paper were downsampled
in time to 1 h temporal resolution using hourly instant values.</p>
      <p>An acoustic wind gauge (Gill instrument, WindMaster Pro Ultrasonic
Anemometer) is also installed on the oceanographic buoy at approximately 5 m above sea level, and is used to measure wind speed and direction with
a sampling frequency of 10 <inline-formula><mml:math display="inline"><mml:mi mathvariant="normal">Hz</mml:mi></mml:math></inline-formula>. Wind speed measurements presented in
this paper were downsampled in time to a 30 min temporal resolution using
half-hourly instant values.</p>
      <p>In 2012 the NIB made regular monthly CTD measurements in the Slovenian part
of the Gulf of Trieste. These campaigns took place in the vicinity of buoy
Vida (13.55505<inline-formula><mml:math display="inline"><mml:msup><mml:mi/><mml:mo>∘</mml:mo></mml:msup></mml:math></inline-formula> E, 45.5488<inline-formula><mml:math display="inline"><mml:msup><mml:mi/><mml:mo>∘</mml:mo></mml:msup></mml:math></inline-formula> N) on 26 January, 16 February,
27 February and on 12 March, providing temperature, salinity and density
anomaly vertical profiles at buoy Vida before and after the bora episode.</p>
      <p>The Paloma platform, located in the center of the Gulf of Trieste
(45.618<inline-formula><mml:math display="inline"><mml:msup><mml:mi/><mml:mo>∘</mml:mo></mml:msup></mml:math></inline-formula> <inline-formula><mml:math display="inline"><mml:mi mathvariant="normal">N</mml:mi></mml:math></inline-formula>, 13.565<inline-formula><mml:math display="inline"><mml:msup><mml:mi/><mml:mo>∘</mml:mo></mml:msup></mml:math></inline-formula> <inline-formula><mml:math display="inline"><mml:mi mathvariant="normal">E</mml:mi></mml:math></inline-formula>) is operated jointly by
the Regional Environmental Protection Agency (ARPA FVG) and National Research
Council, Institute of Marine Sciences (CNR-ISMAR) (for details see
<xref ref-type="bibr" rid="bib1.bibx29" id="altparen.38"/>). The observed quantities from Paloma used in this paper
were seawater temperature at 3 and 15 <inline-formula><mml:math display="inline"><mml:mi mathvariant="normal">m</mml:mi></mml:math></inline-formula> depths, air temperatures,
relative humidity, mean hourly wind velocity and hourly mean pressure.</p>
      <p>The Acqua Alta platform is located in the Gulf of Venice
(45.314<inline-formula><mml:math display="inline"><mml:msup><mml:mi/><mml:mo>∘</mml:mo></mml:msup></mml:math></inline-formula> <inline-formula><mml:math display="inline"><mml:mi mathvariant="normal">N</mml:mi></mml:math></inline-formula>, 12.519<inline-formula><mml:math display="inline"><mml:msup><mml:mi/><mml:mo>∘</mml:mo></mml:msup></mml:math></inline-formula> <inline-formula><mml:math display="inline"><mml:mi mathvariant="normal">E</mml:mi></mml:math></inline-formula>) and is operated by the
CNR-ISMAR. The
following observational quantities from Acqua Alta were used to estimate in
situ bulk fluxes during January and February 2012: air temperature, air
pressure, wind speed and seawater temperature at 2 m depth.</p>
      <p>To compute the bulk fluxes ALADIN uses an implicit scheme, which takes into
account several prognostic quantities (like wind gustiness and friction
velocity) that were not available within the existing observational data.
This flux scheme therefore could not be used on observational data in the
same manner as implemented in ALADIN itself. To compute the fluxes from
measurements, we therefore used a COARE 2.0 algorithm <xref ref-type="bibr" rid="bib1.bibx10" id="paren.39"/> that
needs no additional data apart from those available.<?xmltex \hack{\vspace{-3mm}}?></p>

      <?xmltex \floatpos{t}?><fig id="Ch1.F4"><caption><p>Hourly wind speeds measured at oceanographic platforms Vida, Paloma
and Acqua Alta during January and February 2012.</p></caption>
        <?xmltex \igopts{width=199.169291pt}?><graphic xlink:href="https://os.copernicus.org/articles/12/71/2016/os-12-71-2016-f04.png"/>

      </fig>

</sec>
<sec id="Ch1.S4">
  <title>Results and discussion</title>
      <p>In the February 2012 bora episode, there was a strong bora with wind speeds
above 20 <inline-formula><mml:math display="inline"><mml:mrow><mml:mi mathvariant="normal">m</mml:mi><mml:mspace width="0.125em" linebreak="nobreak"/><mml:msup><mml:mi mathvariant="normal">s</mml:mi><mml:mrow><mml:mo>-</mml:mo><mml:mn mathvariant="normal">1</mml:mn></mml:mrow></mml:msup></mml:mrow></mml:math></inline-formula> blowing from 5 to 9 February 2012, on 9
February the wind speed fell to about 7 <inline-formula><mml:math display="inline"><mml:mrow><mml:mi mathvariant="normal">m</mml:mi><mml:mspace linebreak="nobreak" width="0.125em"/><mml:msup><mml:mi mathvariant="normal">s</mml:mi><mml:mrow><mml:mo>-</mml:mo><mml:mn mathvariant="normal">1</mml:mn></mml:mrow></mml:msup></mml:mrow></mml:math></inline-formula> for about 10 h,
and rose again to reach wind speeds of 15–20 <inline-formula><mml:math display="inline"><mml:mrow><mml:mi mathvariant="normal">m</mml:mi><mml:mspace linebreak="nobreak" width="0.125em"/><mml:msup><mml:mi mathvariant="normal">s</mml:mi><mml:mrow><mml:mo>-</mml:mo><mml:mn mathvariant="normal">1</mml:mn></mml:mrow></mml:msup></mml:mrow></mml:math></inline-formula>, peaking
between 10 and 14 February. Between 14 and 19 February the
bora subsided and wind speeds generally stayed below 5 <inline-formula><mml:math display="inline"><mml:mrow><mml:mi mathvariant="normal">m</mml:mi><mml:mspace linebreak="nobreak" width="0.125em"/><mml:msup><mml:mi mathvariant="normal">s</mml:mi><mml:mrow><mml:mo>-</mml:mo><mml:mn mathvariant="normal">1</mml:mn></mml:mrow></mml:msup></mml:mrow></mml:math></inline-formula>, see
Fig. <xref ref-type="fig" rid="Ch1.F4"/>. Most images presented below were computed during these
time windows.</p>
<sec id="Ch1.S4.SS1">
  <title>Air–sea fluxes</title>
      <p>As noted above, both coupled and uncoupled system use the following ALADIN
fluxes at the air–sea interface: sensible heat flux <inline-formula><mml:math display="inline"><mml:mrow><mml:msub><mml:mi>Q</mml:mi><mml:mi mathvariant="normal">H</mml:mi></mml:msub></mml:mrow></mml:math></inline-formula>, latent heat flux
<inline-formula><mml:math display="inline"><mml:mrow><mml:msub><mml:mi>Q</mml:mi><mml:mi>E</mml:mi></mml:msub></mml:mrow></mml:math></inline-formula> and net radiation <inline-formula><mml:math display="inline"><mml:mrow><mml:msub><mml:mi>R</mml:mi><mml:mi mathvariant="normal">n</mml:mi></mml:msub></mml:mrow></mml:math></inline-formula> (composed from solar and infrared
contributions) at the ocean surface. The only difference between coupled and
uncoupled flux calculation in ALADIN stems from the sea surface temperature
boundary condition in ALADIN: ALADIN1w calculates its fluxes using sea
surface temperature from the OSTIA product while ALADIN2w computes these
fluxes using POM2w sea surface temperature. OSTIA SST in the northern
Adriatic was consistently overestimated throughout the event (see green lines
in all the panels of Fig. <xref ref-type="fig" rid="Ch1.F11"/>), which led to an
overestimation of fluxes in the one-way coupled setup, as shown below.</p>

      <?xmltex \floatpos{t}?><fig id="Ch1.F5"><caption><p>Mean fluxes in the Adriatic during 28 January–16 February. First
row left: mean latent heat flux <inline-formula><mml:math display="inline"><mml:mrow><mml:msub><mml:mi>Q</mml:mi><mml:mi>E</mml:mi></mml:msub><mml:mspace linebreak="nobreak" width="0.125em"/><mml:mo>[</mml:mo><mml:msup><mml:mi mathvariant="normal">Wm</mml:mi><mml:mrow><mml:mo>-</mml:mo><mml:mn mathvariant="normal">2</mml:mn></mml:mrow></mml:msup><mml:mo>]</mml:mo></mml:mrow></mml:math></inline-formula> from ALADIN1w. First
row middle: mean latent heat flux <inline-formula><mml:math display="inline"><mml:mrow><mml:msub><mml:mi>Q</mml:mi><mml:mi>E</mml:mi></mml:msub><mml:mspace linebreak="nobreak" width="0.125em"/><mml:mo>[</mml:mo><mml:msup><mml:mi mathvariant="normal">Wm</mml:mi><mml:mrow><mml:mo>-</mml:mo><mml:mn mathvariant="normal">2</mml:mn></mml:mrow></mml:msup><mml:mo>]</mml:mo></mml:mrow></mml:math></inline-formula> from ALADIN2w.
First row right: difference in latent heat flux between ALADIN2w and
ALADIN1w. Second row: the same as first row, but for sensible heat flux
<inline-formula><mml:math display="inline"><mml:mrow><mml:msub><mml:mi>Q</mml:mi><mml:mi mathvariant="normal">H</mml:mi></mml:msub></mml:mrow></mml:math></inline-formula>. Third row: the same as first row, but for total heat flux <inline-formula><mml:math display="inline"><mml:mrow><mml:msub><mml:mi>Q</mml:mi><mml:mi mathvariant="normal">T</mml:mi></mml:msub></mml:mrow></mml:math></inline-formula>. (Note
that the fluxes are generally negative during the event, implying ocean heat
loss. Positive difference <inline-formula><mml:math display="inline"><mml:mrow><mml:msub><mml:mi>Q</mml:mi><mml:mrow><mml:mn mathvariant="normal">2</mml:mn><mml:mi mathvariant="normal">w</mml:mi></mml:mrow></mml:msub><mml:mo>-</mml:mo><mml:msub><mml:mi>Q</mml:mi><mml:mrow><mml:mn mathvariant="normal">1</mml:mn><mml:mi mathvariant="normal">w</mml:mi></mml:mrow></mml:msub></mml:mrow></mml:math></inline-formula> thus means that the uncoupled
system is losing more heat, while negative difference implies that the
coupled system is losing more heat.)</p></caption>
          <?xmltex \igopts{width=236.157874pt}?><graphic xlink:href="https://os.copernicus.org/articles/12/71/2016/os-12-71-2016-f05.png"/>

        </fig>

      <p>Figure <xref ref-type="fig" rid="Ch1.F5"/> depicts latent, sensible and total fluxes, averaged
in time over the period of the bora event between 28 January–16 February.
Latent heat flux (top row in Fig. <xref ref-type="fig" rid="Ch1.F5"/>) reaches a time-averaged
value of approximately <inline-formula><mml:math display="inline"><mml:mo>-</mml:mo></mml:math></inline-formula>200 W m<inline-formula><mml:math display="inline"><mml:msup><mml:mi/><mml:mrow><mml:mo>-</mml:mo><mml:mn mathvariant="normal">2</mml:mn></mml:mrow></mml:msup></mml:math></inline-formula> in the Trieste jet, about
<inline-formula><mml:math display="inline"><mml:mo>-</mml:mo></mml:math></inline-formula>450 W m<inline-formula><mml:math display="inline"><mml:msup><mml:mi/><mml:mrow><mml:mo>-</mml:mo><mml:mn mathvariant="normal">2</mml:mn></mml:mrow></mml:msup></mml:math></inline-formula> in the Senj jet, <inline-formula><mml:math display="inline"><mml:mo>-</mml:mo></mml:math></inline-formula>400 W m<inline-formula><mml:math display="inline"><mml:msup><mml:mi/><mml:mrow><mml:mo>-</mml:mo><mml:mn mathvariant="normal">2</mml:mn></mml:mrow></mml:msup></mml:math></inline-formula> in the
Karlobag jet and <inline-formula><mml:math display="inline"><mml:mo>-</mml:mo></mml:math></inline-formula>230 W m<inline-formula><mml:math display="inline"><mml:msup><mml:mi/><mml:mrow><mml:mo>-</mml:mo><mml:mn mathvariant="normal">2</mml:mn></mml:mrow></mml:msup></mml:math></inline-formula> in the Sukošan jet (see Fig. <xref ref-type="fig" rid="Ch1.F1"/> for the locations of these jets). The coupled system
exhibits somewhat lower latent heat losses, leading to positive values of
<inline-formula><mml:math display="inline"><mml:mrow><mml:msub><mml:mi>Q</mml:mi><mml:mrow><mml:mi mathvariant="normal">E</mml:mi><mml:mn mathvariant="normal">2</mml:mn><mml:mi mathvariant="normal">w</mml:mi></mml:mrow></mml:msub><mml:mo>-</mml:mo><mml:msub><mml:mi>Q</mml:mi><mml:mrow><mml:mi mathvariant="normal">E</mml:mi><mml:mn mathvariant="normal">1</mml:mn><mml:mi mathvariant="normal">w</mml:mi></mml:mrow></mml:msub></mml:mrow></mml:math></inline-formula> (top row, right-hand panel in Fig. <xref ref-type="fig" rid="Ch1.F5"/>).
This happens especially in the Gulf of Trieste, where the coupled system
appears to lose about 40 W m<inline-formula><mml:math display="inline"><mml:msup><mml:mi/><mml:mrow><mml:mo>-</mml:mo><mml:mn mathvariant="normal">2</mml:mn></mml:mrow></mml:msup></mml:math></inline-formula> (amounting to approximately
20 %) less heat due to evaporation. In the Trieste jet wake, along the
latitudes of Istria, the coupled system loses about 20 W m<inline-formula><mml:math display="inline"><mml:msup><mml:mi/><mml:mrow><mml:mo>-</mml:mo><mml:mn mathvariant="normal">2</mml:mn></mml:mrow></mml:msup></mml:math></inline-formula>
(about 10 %) less than the uncoupled one.</p>
      <p>In the Senj jet there are notable differences between the coupled and
uncoupled setups in the bay of Kvarner and between the island of Cres and the
mainland (see Fig. <xref ref-type="fig" rid="Ch1.F1"/> for location). The coupled
setup exhibits up to 75 W m<inline-formula><mml:math display="inline"><mml:msup><mml:mi/><mml:mrow><mml:mo>-</mml:mo><mml:mn mathvariant="normal">2</mml:mn></mml:mrow></mml:msup></mml:math></inline-formula> (amounting to approximately 20 %)
less total heat loss in those regions.</p>
      <p>In the coupled–uncoupled heat flux differences (see right-hand column of
Fig. <xref ref-type="fig" rid="Ch1.F5"/>) a bipolar zonal pattern occurs, reaching all the way
to Italy at the latitudes 44.5–44.8<inline-formula><mml:math display="inline"><mml:msup><mml:mi/><mml:mo>∘</mml:mo></mml:msup></mml:math></inline-formula> <inline-formula><mml:math display="inline"><mml:mi mathvariant="normal">N</mml:mi></mml:math></inline-formula> and
starting just below the tip of the Istrian peninsula: there is a northern
zonal belt where the uncoupled system loses more heat than the coupled (the
blue region starting right at the tip of Istria) and a southern zonal belt
where the opposite is observed (the red region starting at Lošinj island,
see Fig. <xref ref-type="fig" rid="Ch1.F1"/> for location). This arises from the fact
that the direction of the Senj jet in the two systems differs slightly. When
comparing the two jets this leads to regions where the central flow of Senj
jet form one model overlaps with a weaker lateral flow from the same jet from the
other model. The same pattern is encountered when comparing sensible heat
losses.</p>
      <p>A significant difference between the latent fluxes is noticeable along the
Italian coast. There is a coastal belt where an uncoupled system loses up to
50 W m<inline-formula><mml:math display="inline"><mml:msup><mml:mi/><mml:mrow><mml:mo>-</mml:mo><mml:mn mathvariant="normal">2</mml:mn></mml:mrow></mml:msup></mml:math></inline-formula> more heat than the coupled system. This happens due to
OSTIA SST overestimation in the northern Adriatic during the bora outbreak as
will be discussed in more detail in Sect. <xref ref-type="sec" rid="Ch1.S4.SS3"/>.</p>
      <p>Sensible heat losses (second row of Fig. <xref ref-type="fig" rid="Ch1.F5"/>) exceed <inline-formula><mml:math display="inline"><mml:mo>-</mml:mo></mml:math></inline-formula>200 W m<inline-formula><mml:math display="inline"><mml:msup><mml:mi/><mml:mrow><mml:mo>-</mml:mo><mml:mn mathvariant="normal">2</mml:mn></mml:mrow></mml:msup></mml:math></inline-formula>
only in the bora jets, being below <inline-formula><mml:math display="inline"><mml:mo>-</mml:mo></mml:math></inline-formula>100 W m<inline-formula><mml:math display="inline"><mml:msup><mml:mi/><mml:mrow><mml:mo>-</mml:mo><mml:mn mathvariant="normal">2</mml:mn></mml:mrow></mml:msup></mml:math></inline-formula>
in the bora wakes and elsewhere across the Adriatic. Time-averaged cooling
losses amount to approximately <inline-formula><mml:math display="inline"><mml:mo>-</mml:mo></mml:math></inline-formula>130 W m<inline-formula><mml:math display="inline"><mml:msup><mml:mi/><mml:mrow><mml:mo>-</mml:mo><mml:mn mathvariant="normal">2</mml:mn></mml:mrow></mml:msup></mml:math></inline-formula> in the Gulf of
Trieste. In the Senj jet, sensible losses amount to <inline-formula><mml:math display="inline"><mml:mo>-</mml:mo></mml:math></inline-formula>300 W m<inline-formula><mml:math display="inline"><mml:msup><mml:mi/><mml:mrow><mml:mo>-</mml:mo><mml:mn mathvariant="normal">2</mml:mn></mml:mrow></mml:msup></mml:math></inline-formula>
offshore from Senj and <inline-formula><mml:math display="inline"><mml:mo>-</mml:mo></mml:math></inline-formula>200 W m<inline-formula><mml:math display="inline"><mml:msup><mml:mi/><mml:mrow><mml:mo>-</mml:mo><mml:mn mathvariant="normal">2</mml:mn></mml:mrow></mml:msup></mml:math></inline-formula> downwind westward of Cres.
The Karlobag jet losses surpass <inline-formula><mml:math display="inline"><mml:mo>-</mml:mo></mml:math></inline-formula>300 W m<inline-formula><mml:math display="inline"><mml:msup><mml:mi/><mml:mrow><mml:mo>-</mml:mo><mml:mn mathvariant="normal">2</mml:mn></mml:mrow></mml:msup></mml:math></inline-formula> only in a very
limited region close to the mainland and the jet zonally decays westward on
the scale of about 10 nautical miles. Zonal extension of pronounced sensible
losses is in general approximately 30 nautical miles shorter than the zonal
extension of evaporative losses.</p>

      <?xmltex \floatpos{t}?><fig id="Ch1.F6"><caption><p>Daily averages of bulk fluxes [W m<inline-formula><mml:math display="inline"><mml:msup><mml:mi/><mml:mrow><mml:mo>-</mml:mo><mml:mn mathvariant="normal">2</mml:mn></mml:mrow></mml:msup></mml:math></inline-formula>] at Vida location
from ALADIN1w (red line) and ALADIN2w (blue line). Sensible and latent heat
fluxes at Vida location were computed using COARE 2.0 algorithm from
measurements at Vida: daily averages of these fluxes are depicted with black
squares with vertical error bars (see text for details on error estimation).
Top left: <inline-formula><mml:math display="inline"><mml:mrow><mml:msub><mml:mi>Q</mml:mi><mml:mi>E</mml:mi></mml:msub></mml:mrow></mml:math></inline-formula> – latent heat flux. Middle left: <inline-formula><mml:math display="inline"><mml:mrow><mml:msub><mml:mi>Q</mml:mi><mml:mi mathvariant="normal">H</mml:mi></mml:msub></mml:mrow></mml:math></inline-formula> – sensible heat flux.
Bottom left: <inline-formula><mml:math display="inline"><mml:mrow><mml:msub><mml:mi>Q</mml:mi><mml:mi mathvariant="normal">T</mml:mi></mml:msub></mml:mrow></mml:math></inline-formula> – total heat flux. Top right: <inline-formula><mml:math display="inline"><mml:mrow><mml:msub><mml:mi>Q</mml:mi><mml:mi mathvariant="normal">B</mml:mi></mml:msub></mml:mrow></mml:math></inline-formula> – long-wave downward
radiation. Middle right: <inline-formula><mml:math display="inline"><mml:mrow><mml:msub><mml:mi>Q</mml:mi><mml:mi mathvariant="normal">S</mml:mi></mml:msub></mml:mrow></mml:math></inline-formula> – Solar shortwave radiation. Bottom right:
<inline-formula><mml:math display="inline"><mml:mrow><mml:msub><mml:mi>Q</mml:mi><mml:mrow><mml:mi mathvariant="normal">T</mml:mi><mml:mn mathvariant="normal">2</mml:mn><mml:mi mathvariant="normal">w</mml:mi></mml:mrow></mml:msub><mml:mo>-</mml:mo><mml:msub><mml:mi>Q</mml:mi><mml:mrow><mml:mi mathvariant="normal">T</mml:mi><mml:mn mathvariant="normal">1</mml:mn><mml:mi mathvariant="normal">w</mml:mi></mml:mrow></mml:msub></mml:mrow></mml:math></inline-formula> – total heat flux difference between ALADIN2w and
ALADIN1w. Fluxes are depicted as negative when the ocean is losing
heat.</p></caption>
          <?xmltex \igopts{width=236.157874pt}?><graphic xlink:href="https://os.copernicus.org/articles/12/71/2016/os-12-71-2016-f06.png"/>

        </fig>

      <p>These described patterns dominate the time-averaged total heat flux
<inline-formula><mml:math display="inline"><mml:mrow><mml:msub><mml:mi>Q</mml:mi><mml:mi mathvariant="normal">T</mml:mi></mml:msub></mml:mrow></mml:math></inline-formula>,
which reflects them to a large extent as seen from the bottom row of Fig. <xref ref-type="fig" rid="Ch1.F5"/>.
The time-averaged total flux can be multiplied with the
respective time window (28 January–16 February) to obtain quantitative
estimates of the differences of total energy lost in the two systems (not
shown in the paper). Over the entire period of the bora event, the Gulf of
Trieste, the Gulf of Kvarner and the western Adriatic (Italian) coastal belt
lost approximately <inline-formula><mml:math display="inline"><mml:mrow><mml:mn>100</mml:mn><mml:mspace width="0.125em" linebreak="nobreak"/><mml:msup><mml:mi mathvariant="normal">MJm</mml:mi><mml:mrow><mml:mo>-</mml:mo><mml:mn mathvariant="normal">2</mml:mn></mml:mrow></mml:msup></mml:mrow></mml:math></inline-formula> less energy per unit area in the
coupled system than in the uncoupled system. On the other hand, the zonal
belt extending from the tip of Istria to the Italian coast lost about <inline-formula><mml:math display="inline"><mml:mrow><mml:mn>20</mml:mn><mml:mspace linebreak="nobreak" width="0.125em"/><mml:msup><mml:mi mathvariant="normal">MJm</mml:mi><mml:mrow><mml:mo>-</mml:mo><mml:mn mathvariant="normal">2</mml:mn></mml:mrow></mml:msup></mml:mrow></mml:math></inline-formula> more energy per unit area in the coupled than in the
uncoupled system. The northern Adriatic west of Istria lost on average 30–50 MJ m<inline-formula><mml:math display="inline"><mml:msup><mml:mi/><mml:mrow><mml:mo>-</mml:mo><mml:mn mathvariant="normal">2</mml:mn></mml:mrow></mml:msup></mml:math></inline-formula> more energy per unit area in the uncoupled system than
in the coupled system.</p>
      <p>To estimate flux differences between the two setups in the time domain, we
compared the flux time series from both setups at three observation platform
locations in the northern Adriatic, the Vida buoy and the Paloma and Acqua
Alta platforms (see the inset of Fig. <xref ref-type="fig" rid="Ch1.F1"/> for their
locations). We further compared the modeled fluxes from the two setups with
bulk fluxes computed from observations at all three platforms. These results
are shown in Figs. <xref ref-type="fig" rid="Ch1.F6"/>–<xref ref-type="fig" rid="Ch1.F8"/>.</p>
      <p>At the Vida and Paloma platforms, the following observational parameters were
used to compute the heat fluxes: sea temperature at <inline-formula><mml:math display="inline"><mml:mrow><mml:mn mathvariant="normal">2</mml:mn><mml:mspace linebreak="nobreak" width="0.125em"/><mml:mi mathvariant="normal">m</mml:mi></mml:mrow></mml:math></inline-formula> depth at
Vida or <inline-formula><mml:math display="inline"><mml:mrow><mml:mn mathvariant="normal">3</mml:mn><mml:mspace linebreak="nobreak" width="0.125em"/><mml:mi mathvariant="normal">m</mml:mi></mml:mrow></mml:math></inline-formula> at Paloma (note that the water columns at platform
locations exhibit negligible stratification during strong bora events), air
pressure, wind speed, relative humidity and air temperature. At Acqua Alta
station, measurements were used for all parameters except relative humidity.
The latter, for flux computations at Acqua Alta location, was taken from the
ALADIN1w model. Error estimates for fluxes, computed from observations at the
platforms, were determined in the following way. Uncertainties in observed
air temperatures, sea temperatures, relative humidity, atmospheric pressure
and wind speed were adopted from <xref ref-type="bibr" rid="bib1.bibx29" id="text.40"/> (note that the very same
data sets are used at Vida and Paloma as in <xref ref-type="bibr" rid="bib1.bibx29" id="altparen.41"/>). The
uncertainties of air temperature and SST were thus set to be 0.3 and
<inline-formula><mml:math display="inline"><mml:mrow><mml:mn>0.1</mml:mn><mml:mspace linebreak="nobreak" width="0.125em"/><mml:msup><mml:mi/><mml:mo>∘</mml:mo></mml:msup><mml:mi mathvariant="normal">C</mml:mi></mml:mrow></mml:math></inline-formula>, respectively. Uncertainty in relative humidity was
set to <inline-formula><mml:math display="inline"><mml:mrow><mml:mn mathvariant="normal">3</mml:mn><mml:mspace linebreak="nobreak" width="0.125em"/><mml:mi mathvariant="italic">%</mml:mi></mml:mrow></mml:math></inline-formula> while that of sea level air pressure was set to 0.3 hPa. A
relative error of 12 % was set for wind speed. Sensible and latent heat
flux error estimates were then computed using the Coupled Ocean–Atmosphere
Response Experiment (COARE) algorithm on an ensemble of observed air
temperature, sea temperature, relative humidity, atmosphere pressure and wind
speed measurements that were perturbed by the values of their respective
uncertainties stated above.</p>

      <?xmltex \floatpos{t}?><fig id="Ch1.F7"><caption><p>Same as Fig. <xref ref-type="fig" rid="Ch1.F6"/> but for Paloma location.</p></caption>
          <?xmltex \igopts{width=236.157874pt}?><graphic xlink:href="https://os.copernicus.org/articles/12/71/2016/os-12-71-2016-f07.png"/>

        </fig>

      <?xmltex \floatpos{t}?><fig id="Ch1.F8"><caption><p>Same as Fig. <xref ref-type="fig" rid="Ch1.F6"/> but for Acqua Alta location.</p></caption>
          <?xmltex \igopts{width=236.157874pt}?><graphic xlink:href="https://os.copernicus.org/articles/12/71/2016/os-12-71-2016-f08.png"/>

        </fig>

      <p><?xmltex \hack{\newpage}?>When comparing sensible and latent heat fluxes from ALADIN setups to those
from COARE 2.0, it has to be remembered that flux comparisons are generally
highly sensitive to parameterization, but Louis and COARE schemes agree
reasonably well <xref ref-type="bibr" rid="bib1.bibx9" id="paren.42"/>. Such was also the outcome of our
comparisons, as shown in Figs. <xref ref-type="fig" rid="Ch1.F6"/>–<xref ref-type="fig" rid="Ch1.F8"/>. On the
other hand, the fluxes in this paper differ from those presented in
<xref ref-type="bibr" rid="bib1.bibx29" id="paren.43"/> in which the Kondo scheme was used for transfer coefficients,
by as much as 30 % during the peak periods of bora (compare Figs. <xref ref-type="fig" rid="Ch1.F6"/>–<xref ref-type="fig" rid="Ch1.F8"/> in this paper to Fig. 6 in
<xref ref-type="bibr" rid="bib1.bibx29" id="altparen.44"/>). As demonstrated below, however, ocean cooling in the
northern Adriatic is well reproduced in POM2w (and indeed even overestimated
in POM1w) using our values of fluxes.</p>
      <p>As judged from observations at Vida, Paloma and Acqua Alta, it appears that
air–sea coupling has a positive impact on the sensible heat flux <inline-formula><mml:math display="inline"><mml:mrow><mml:msub><mml:mi>Q</mml:mi><mml:mi mathvariant="normal">H</mml:mi></mml:msub></mml:mrow></mml:math></inline-formula> (see
middle left panes in Figs. <xref ref-type="fig" rid="Ch1.F6"/>–<xref ref-type="fig" rid="Ch1.F8"/>): at all three
stations the sensible heat fluxes, as computed from observations, are better
reproduced in ALADIN2w. The latent heat fluxes, computed from observations,
are on the other hand somewhat better reproduced by ALADIN1w. The root mean
square differences (RMSD) between ALADIN1w(2w) setups and fluxes, computed
from observations, are shown in Table <xref ref-type="table" rid="Ch1.T1"/>.</p>
      <p>It should be noted that the fluxes in these images are depicted as negative
when the ocean is losing heat. A positive total heat flux difference <inline-formula><mml:math display="inline"><mml:mrow><mml:msub><mml:mi>Q</mml:mi><mml:mrow><mml:mi mathvariant="normal">T</mml:mi><mml:mn mathvariant="normal">2</mml:mn><mml:mi mathvariant="normal">w</mml:mi></mml:mrow></mml:msub><mml:mo>-</mml:mo><mml:msub><mml:mi>Q</mml:mi><mml:mrow><mml:mi mathvariant="normal">T</mml:mi><mml:mn mathvariant="normal">1</mml:mn><mml:mi mathvariant="normal">w</mml:mi></mml:mrow></mml:msub><mml:mo>&gt;</mml:mo><mml:mn mathvariant="normal">0</mml:mn></mml:mrow></mml:math></inline-formula> between a coupled and an uncoupled setup therefore implies
that the uncoupled system is losing more heat than the coupled one. This was
the case at the Paloma and Acqua Alta stations during the bora event. Total heat
flux difference during the event reached 150 W m<inline-formula><mml:math display="inline"><mml:msup><mml:mi/><mml:mrow><mml:mo>-</mml:mo><mml:mn mathvariant="normal">2</mml:mn></mml:mrow></mml:msup></mml:math></inline-formula> at Paloma
(bottom right panel in Fig. <xref ref-type="fig" rid="Ch1.F7"/>) and 100 W m<inline-formula><mml:math display="inline"><mml:msup><mml:mi/><mml:mrow><mml:mo>-</mml:mo><mml:mn mathvariant="normal">2</mml:mn></mml:mrow></mml:msup></mml:math></inline-formula>
at Acqua Alta (bottom right panel in Fig. <xref ref-type="fig" rid="Ch1.F8"/>), amounting to as
much as 20 % of the total flux during the bora outbreak. Before and
after the bora event the air–sea coupling does not appear to have created a
significant impact on the fluxes.</p>

<?xmltex \floatpos{t}?><table-wrap id="Ch1.T1"><caption><p>Root mean square differences (RMSD) between modeled ALADIN fluxes (modified Louis scheme) and fluxes from in situ observations (COARE 2.0).</p></caption><oasis:table frame="topbot"><oasis:tgroup cols="4">
     <oasis:colspec colnum="1" colname="col1" align="left"/>
     <oasis:colspec colnum="2" colname="col2" align="right"/>
     <oasis:colspec colnum="3" colname="col3" align="right"/>
     <oasis:colspec colnum="4" colname="col4" align="right"/>
     <oasis:thead>
       <oasis:row rowsep="1">  
         <oasis:entry colname="col1">RMSD [Wm<inline-formula><mml:math display="inline"><mml:msup><mml:mi/><mml:mrow><mml:mo>-</mml:mo><mml:mn mathvariant="normal">2</mml:mn></mml:mrow></mml:msup></mml:math></inline-formula>] at:</oasis:entry>  
         <oasis:entry colname="col2">Vida</oasis:entry>  
         <oasis:entry colname="col3">Paloma</oasis:entry>  
         <oasis:entry colname="col4">Acqua Alta</oasis:entry>
       </oasis:row>
     </oasis:thead>
     <oasis:tbody>
       <oasis:row>  
         <oasis:entry colname="col1"><inline-formula><mml:math display="inline"><mml:mrow><mml:msub><mml:mi>Q</mml:mi><mml:mi mathvariant="normal">H</mml:mi></mml:msub><mml:mo>:</mml:mo></mml:mrow></mml:math></inline-formula> 1w-COARE</oasis:entry>  
         <oasis:entry colname="col2">13.84</oasis:entry>  
         <oasis:entry colname="col3">33.67</oasis:entry>  
         <oasis:entry colname="col4">18.12</oasis:entry>
       </oasis:row>
       <oasis:row rowsep="1">  
         <oasis:entry colname="col1"><inline-formula><mml:math display="inline"><mml:mrow><mml:msub><mml:mi>Q</mml:mi><mml:mi mathvariant="normal">H</mml:mi></mml:msub><mml:mo>:</mml:mo></mml:mrow></mml:math></inline-formula> 2w-COARE</oasis:entry>  
         <oasis:entry colname="col2">12.86</oasis:entry>  
         <oasis:entry colname="col3">17.69</oasis:entry>  
         <oasis:entry colname="col4">12.38</oasis:entry>
       </oasis:row>
       <oasis:row>  
         <oasis:entry colname="col1"><inline-formula><mml:math display="inline"><mml:mrow><mml:msub><mml:mi>Q</mml:mi><mml:mi>E</mml:mi></mml:msub><mml:mo>:</mml:mo></mml:mrow></mml:math></inline-formula> 1w-COARE</oasis:entry>  
         <oasis:entry colname="col2">25.76</oasis:entry>  
         <oasis:entry colname="col3">30.34</oasis:entry>  
         <oasis:entry colname="col4">17.04</oasis:entry>
       </oasis:row>
       <oasis:row>  
         <oasis:entry colname="col1"><inline-formula><mml:math display="inline"><mml:mrow><mml:msub><mml:mi>Q</mml:mi><mml:mi>E</mml:mi></mml:msub><mml:mo>:</mml:mo></mml:mrow></mml:math></inline-formula> 2w-COARE</oasis:entry>  
         <oasis:entry colname="col2">25.06</oasis:entry>  
         <oasis:entry colname="col3">53.00</oasis:entry>  
         <oasis:entry colname="col4">21.34</oasis:entry>
       </oasis:row>
     </oasis:tbody>
   </oasis:tgroup></oasis:table></table-wrap>

      <p>At the Vida location, fluxes from the two setups do not differ significantly,
even during the bora event (bottom right panel in Fig. <xref ref-type="fig" rid="Ch1.F6"/>).
This is most probably due to its location, which, unlike Paloma and Acqua
Alta, is very close (within one ALADIN or POM grid point) to the land, and
the impact of the ocean feedback is consequently less pronounced in the
atmospheric model.</p>
</sec>
<sec id="Ch1.S4.SS2">
  <title>Atmosphere temperatures</title>
      <p>The temperature at the lowest model level (approximately 10 m) fields over
the Adriatic part of the computational domain was determined for various
forecast ranges on 10 February 2012 (Fig. <xref ref-type="fig" rid="Ch1.F9"/>). The
differences between near-surface temperatures obtained from ALADIN1w and
ALADIN2w, shown in the right column of Fig. <xref ref-type="fig" rid="Ch1.F9"/>, increase with
time, which is an expected consequence of the fact that both ALADIN1w and
ALADIN2w are initialized from the same initial conditions at the beginning
(00:00 UTC) of each 24 <inline-formula><mml:math display="inline"><mml:mi mathvariant="normal">h</mml:mi></mml:math></inline-formula> forecast run. At a forecast range of 1 h (top
row of Fig. <xref ref-type="fig" rid="Ch1.F9"/>), the differences are limited mostly to the
shallow coastal regions (especially along the Italian Adriatic coast) where
the differences in SST between OSTIA and POW2w are largest. At this time the
near-surface air temperatures over land in the two cases are almost
identical. Later on during the forecast these differences increase, as shown
in the right-hand column of Fig. <xref ref-type="fig" rid="Ch1.F9"/>). The air masses from the
two setups are advected along with the dominant wind, i.e., towards the
southwest, and the pattern of their respective differences resembles the wind
field (e.g., pronounced jet structures perpendicular to the eastern Adriatic
coast in the right column of Fig. <xref ref-type="fig" rid="Ch1.F9"/>). These temperature
differences are also seen across the Italian Peninsula and over the
Tyrrhenian Sea, where the differences in near-surface temperature exhibit a
characteristic dipole-like pattern, which is related to displacement of the
precipitation features around the center of a low-pressure area (shown as a
“<bold>+</bold>” character in the middle and bottom panels of the right column of
Fig. <xref ref-type="fig" rid="Ch1.F9"/>). This difference, arising as a non-linear response
of the atmospheric models to the differences between OSTIA and POM2w, is
shifted eastward into the southern Adriatic at a forecast time of 23 h,
in accordance with the trajectory of the pressure low.</p>

      <?xmltex \floatpos{t}?><fig id="Ch1.F9"><caption><p>ALADIN 10 <inline-formula><mml:math display="inline"><mml:mi mathvariant="normal">m</mml:mi></mml:math></inline-formula> temperatures [<inline-formula><mml:math display="inline"><mml:msup><mml:mi/><mml:mo>∘</mml:mo></mml:msup></mml:math></inline-formula><inline-formula><mml:math display="inline"><mml:mi mathvariant="normal">C</mml:mi></mml:math></inline-formula>] in ALADIN1w (left
column) and ALADIN2w (middle column) on 10 February 2012 at: 01:00 UTC (top
row), 13:00 UTC (middle row) and 23:00 UTC (bottom row). Right column depicts
temperature differences [<inline-formula><mml:math display="inline"><mml:msup><mml:mi/><mml:mo>∘</mml:mo></mml:msup></mml:math></inline-formula>C] in both setups. Black <bold>+</bold>
signs in middle and bottom panels of the right column indicate the
approximate location of the center of the low pressure area at 13:00 UTC (middle
panel) and 23:00 UTC (bottom panel).</p></caption>
          <?xmltex \igopts{width=236.157874pt}?><graphic xlink:href="https://os.copernicus.org/articles/12/71/2016/os-12-71-2016-f09.jpg"/>

        </fig>

<?xmltex \floatpos{t}?><table-wrap id="Ch1.T2" specific-use="star"><caption><p>Biases and root mean square errors (RMSE) [<inline-formula><mml:math display="inline"><mml:msup><mml:mi/><mml:mo>∘</mml:mo></mml:msup></mml:math></inline-formula>C] between modeled POM1w,
POM2w ocean temperatures and in situ observations at Vida, Paloma and Acqua Alta observational platforms during February 2012.</p></caption><oasis:table frame="topbot"><oasis:tgroup cols="5">
     <oasis:colspec colnum="1" colname="col1" align="left"/>
     <oasis:colspec colnum="2" colname="col2" align="right"/>
     <oasis:colspec colnum="3" colname="col3" align="right"/>
     <oasis:colspec colnum="4" colname="col4" align="right"/>
     <oasis:colspec colnum="5" colname="col5" align="right"/>
     <oasis:thead>
       <oasis:row rowsep="1">  
         <oasis:entry colname="col1">Station (sensor depth):</oasis:entry>  
         <oasis:entry colname="col2">Vida (2 <inline-formula><mml:math display="inline"><mml:mi mathvariant="normal">m</mml:mi></mml:math></inline-formula>)</oasis:entry>  
         <oasis:entry colname="col3">Paloma (3 <inline-formula><mml:math display="inline"><mml:mi mathvariant="normal">m</mml:mi></mml:math></inline-formula>)</oasis:entry>  
         <oasis:entry colname="col4">Paloma (15 <inline-formula><mml:math display="inline"><mml:mi mathvariant="normal">m</mml:mi></mml:math></inline-formula>)</oasis:entry>  
         <oasis:entry colname="col5">Acqua Alta (2 <inline-formula><mml:math display="inline"><mml:mi mathvariant="normal">m</mml:mi></mml:math></inline-formula>)</oasis:entry>
       </oasis:row>
     </oasis:thead>
     <oasis:tbody>
       <oasis:row>  
         <oasis:entry colname="col1">T BIAS 1w [<inline-formula><mml:math display="inline"><mml:msup><mml:mi/><mml:mo>∘</mml:mo></mml:msup></mml:math></inline-formula>C]</oasis:entry>  
         <oasis:entry colname="col2"><inline-formula><mml:math display="inline"><mml:mo>-</mml:mo></mml:math></inline-formula>1.88</oasis:entry>  
         <oasis:entry colname="col3"><inline-formula><mml:math display="inline"><mml:mo>-</mml:mo></mml:math></inline-formula>2.53</oasis:entry>  
         <oasis:entry colname="col4"><inline-formula><mml:math display="inline"><mml:mo>-</mml:mo></mml:math></inline-formula>1.90</oasis:entry>  
         <oasis:entry colname="col5"><inline-formula><mml:math display="inline"><mml:mo>-</mml:mo></mml:math></inline-formula>1.16</oasis:entry>
       </oasis:row>
       <oasis:row rowsep="1">  
         <oasis:entry colname="col1">T BIAS 2w [<inline-formula><mml:math display="inline"><mml:msup><mml:mi/><mml:mo>∘</mml:mo></mml:msup></mml:math></inline-formula>C]</oasis:entry>  
         <oasis:entry colname="col2"><inline-formula><mml:math display="inline"><mml:mo>-</mml:mo></mml:math></inline-formula>0.33</oasis:entry>  
         <oasis:entry colname="col3"><inline-formula><mml:math display="inline"><mml:mo>-</mml:mo></mml:math></inline-formula>0.53</oasis:entry>  
         <oasis:entry colname="col4"><inline-formula><mml:math display="inline"><mml:mo>-</mml:mo></mml:math></inline-formula>0.35</oasis:entry>  
         <oasis:entry colname="col5"><inline-formula><mml:math display="inline"><mml:mo>-</mml:mo></mml:math></inline-formula>0.22</oasis:entry>
       </oasis:row>
       <oasis:row>  
         <oasis:entry colname="col1">T RMSE 1w [<inline-formula><mml:math display="inline"><mml:msup><mml:mi/><mml:mo>∘</mml:mo></mml:msup></mml:math></inline-formula>C]</oasis:entry>  
         <oasis:entry colname="col2">2.06</oasis:entry>  
         <oasis:entry colname="col3">2.68</oasis:entry>  
         <oasis:entry colname="col4">1.99</oasis:entry>  
         <oasis:entry colname="col5">1.24</oasis:entry>
       </oasis:row>
       <oasis:row>  
         <oasis:entry colname="col1">T RMSE 2w [<inline-formula><mml:math display="inline"><mml:msup><mml:mi/><mml:mo>∘</mml:mo></mml:msup></mml:math></inline-formula>C]</oasis:entry>  
         <oasis:entry colname="col2">0.59</oasis:entry>  
         <oasis:entry colname="col3">0.79</oasis:entry>  
         <oasis:entry colname="col4">0.48</oasis:entry>  
         <oasis:entry colname="col5">0.46</oasis:entry>
       </oasis:row>
     </oasis:tbody>
   </oasis:tgroup></oasis:table></table-wrap>

      <?xmltex \floatpos{t}?><fig id="Ch1.F10" specific-use="star"><caption><p>Left panel: bias (top) and standard deviation of errors (bottom)
scores for ALADIN1w (red curve) and ALADIN2w (blue curve), depending on
forecast length for temperature at 2 <inline-formula><mml:math display="inline"><mml:mi mathvariant="normal">m</mml:mi></mml:math></inline-formula> in February 2012. The green
curve, quantified by the right-hand axis, depicts the number of cases
(available number of measurements at all stations at each forecast hour
combined) used in the statistics. Right panel: red circles mark the
geographical locations of all 23 meteorological stations used for
verifications.</p></caption>
          <?xmltex \igopts{width=341.433071pt}?><graphic xlink:href="https://os.copernicus.org/articles/12/71/2016/os-12-71-2016-f10.png"/>

        </fig>

      <?xmltex \floatpos{t}?><fig id="Ch1.F11"><caption><p>Modeled and observed temperatures at Vida, Paloma and Acqua Alta
locations. Observed in situ temperatures are plotted with a black line, OSTIA
SST boundary condition for ALADIN1w with a green line, POM1w modeled
temperature with red line and POM2w modeled temperature with a blue line.
First row: temperatures at Vida location. Second row: temperatures for Paloma
location at <inline-formula><mml:math display="inline"><mml:mrow><mml:mn mathvariant="normal">3</mml:mn><mml:mspace width="0.125em" linebreak="nobreak"/><mml:mi mathvariant="normal">m</mml:mi></mml:mrow></mml:math></inline-formula> depth. Third row: temperatures at Paloma at
<inline-formula><mml:math display="inline"><mml:mrow><mml:mn>15</mml:mn><mml:mspace linebreak="nobreak" width="0.125em"/><mml:mi mathvariant="normal">m</mml:mi></mml:mrow></mml:math></inline-formula> depth. Fourth row: temperatures for Acqua Alta locations at
<inline-formula><mml:math display="inline"><mml:mrow><mml:mn mathvariant="normal">2</mml:mn><mml:mspace linebreak="nobreak" width="0.125em"/><mml:mi mathvariant="normal">m</mml:mi></mml:mrow></mml:math></inline-formula> depth.</p></caption>
          <?xmltex \igopts{width=199.169291pt}?><graphic xlink:href="https://os.copernicus.org/articles/12/71/2016/os-12-71-2016-f11.png"/>

        </fig>

      <?xmltex \floatpos{t}?><fig id="Ch1.F12"><caption><p>Differences, biases and RMSEs of modeled and observed temperatures
at Vida, Paloma and Acqua Alta locations. Top left: sea temperature
difference between observations and POM1w (red line) and POM2w (blue line) at
Vida location. Top right: the same for Acqua Alta location. Bottom left: the
same for Paloma location at <inline-formula><mml:math display="inline"><mml:mrow><mml:mn mathvariant="normal">3</mml:mn><mml:mspace width="0.125em" linebreak="nobreak"/><mml:mi mathvariant="normal">m</mml:mi></mml:mrow></mml:math></inline-formula> depth. Bottom right: the same for
Paloma location at <inline-formula><mml:math display="inline"><mml:mrow><mml:mn>15</mml:mn><mml:mspace linebreak="nobreak" width="0.125em"/><mml:mi mathvariant="normal">m</mml:mi></mml:mrow></mml:math></inline-formula> depth. Respective biases and RMSEs are
included below each panel.</p></caption>
          <?xmltex \igopts{width=236.157874pt}?><graphic xlink:href="https://os.copernicus.org/articles/12/71/2016/os-12-71-2016-f12.png"/>

        </fig>

      <?xmltex \floatpos{t}?><fig id="Ch1.F13"><caption><p>Sea surface temperature [<inline-formula><mml:math display="inline"><mml:msup><mml:mi/><mml:mo>∘</mml:mo></mml:msup></mml:math></inline-formula>C] in POM1w (top panel)
and POM2w (bottom panel) at the end of the bora event on 14 February 2012,
00:00 UTC. Stations locations are drawn in pink color as a square (Acqua Alta),
circle (Paloma) and triangle (Vida).</p></caption>
          <?xmltex \igopts{width=236.157874pt}?><graphic xlink:href="https://os.copernicus.org/articles/12/71/2016/os-12-71-2016-f13.png"/>

        </fig>

<table-wrap id="Ch1.T3" specific-use="star"><caption><p>Biases and root mean square errors (RMSE) [<inline-formula><mml:math display="inline"><mml:msup><mml:mi/><mml:mo>∘</mml:mo></mml:msup></mml:math></inline-formula>C] between modeled POM1w,
POM2w ocean temperatures and CTD observations at Vida location on four different dates before and after the 2012 bora outbreak.</p></caption><oasis:table frame="topbot"><oasis:tgroup cols="5">
     <oasis:colspec colnum="1" colname="col1" align="left"/>
     <oasis:colspec colnum="2" colname="col2" align="right"/>
     <oasis:colspec colnum="3" colname="col3" align="right"/>
     <oasis:colspec colnum="4" colname="col4" align="right"/>
     <oasis:colspec colnum="5" colname="col5" align="right"/>
     <oasis:thead>
       <oasis:row rowsep="1">  
         <oasis:entry colname="col1">Date of CTD measurement:</oasis:entry>  
         <oasis:entry colname="col2">26 January</oasis:entry>  
         <oasis:entry colname="col3">16 February</oasis:entry>  
         <oasis:entry colname="col4">27 February</oasis:entry>  
         <oasis:entry colname="col5">12 March</oasis:entry>
       </oasis:row>
     </oasis:thead>
     <oasis:tbody>
       <oasis:row>  
         <oasis:entry colname="col1">T BIAS 1w [<inline-formula><mml:math display="inline"><mml:msup><mml:mi/><mml:mo>∘</mml:mo></mml:msup></mml:math></inline-formula>C]</oasis:entry>  
         <oasis:entry colname="col2"><inline-formula><mml:math display="inline"><mml:mo>-</mml:mo></mml:math></inline-formula>0.24</oasis:entry>  
         <oasis:entry colname="col3"><inline-formula><mml:math display="inline"><mml:mo>-</mml:mo></mml:math></inline-formula>1.55</oasis:entry>  
         <oasis:entry colname="col4"><inline-formula><mml:math display="inline"><mml:mo>-</mml:mo></mml:math></inline-formula>2.48</oasis:entry>  
         <oasis:entry colname="col5"><inline-formula><mml:math display="inline"><mml:mo>-</mml:mo></mml:math></inline-formula>2.06</oasis:entry>
       </oasis:row>
       <oasis:row rowsep="1">  
         <oasis:entry colname="col1">T BIAS 2w [<inline-formula><mml:math display="inline"><mml:msup><mml:mi/><mml:mo>∘</mml:mo></mml:msup></mml:math></inline-formula>C]</oasis:entry>  
         <oasis:entry colname="col2"><inline-formula><mml:math display="inline"><mml:mo>-</mml:mo></mml:math></inline-formula>0.07</oasis:entry>  
         <oasis:entry colname="col3">0.24</oasis:entry>  
         <oasis:entry colname="col4"><inline-formula><mml:math display="inline"><mml:mo>-</mml:mo></mml:math></inline-formula>0.66</oasis:entry>  
         <oasis:entry colname="col5"><inline-formula><mml:math display="inline"><mml:mo>-</mml:mo></mml:math></inline-formula>0.16</oasis:entry>
       </oasis:row>
       <oasis:row>  
         <oasis:entry colname="col1">T RMSE 1w [<inline-formula><mml:math display="inline"><mml:msup><mml:mi/><mml:mo>∘</mml:mo></mml:msup></mml:math></inline-formula>C]</oasis:entry>  
         <oasis:entry colname="col2">0.25</oasis:entry>  
         <oasis:entry colname="col3">1.61</oasis:entry>  
         <oasis:entry colname="col4">2.50</oasis:entry>  
         <oasis:entry colname="col5">2.06</oasis:entry>
       </oasis:row>
       <oasis:row>  
         <oasis:entry colname="col1">T RMSE 2w [<inline-formula><mml:math display="inline"><mml:msup><mml:mi/><mml:mo>∘</mml:mo></mml:msup></mml:math></inline-formula>C]</oasis:entry>  
         <oasis:entry colname="col2">0.08</oasis:entry>  
         <oasis:entry colname="col3">0.42</oasis:entry>  
         <oasis:entry colname="col4">0.72</oasis:entry>  
         <oasis:entry colname="col5">0.18</oasis:entry>
       </oasis:row>
     </oasis:tbody>
   </oasis:tgroup></oasis:table></table-wrap>

      <p>This progressive impact during the forecast is further confirmed by objective
verification of 2 <inline-formula><mml:math display="inline"><mml:mi mathvariant="normal">m</mml:mi></mml:math></inline-formula> temperature forecasts. The mean bias and standard
deviation of errors scores for 2 <inline-formula><mml:math display="inline"><mml:mi mathvariant="normal">m</mml:mi></mml:math></inline-formula> temperatures, computed for 23
downwind meteorological stations (see right panel of Fig. <xref ref-type="fig" rid="Ch1.F10"/> for their locations) in the eastern part of Italy
for all days in February 2012 are shown in Fig. <xref ref-type="fig" rid="Ch1.F10"/>.
The two-way coupled experiment is considerably and systematically colder than
the one-way coupled setup, which is an improvement for prognostic lead times
up to 18 h. The difference in standard deviation of errors is rather
small after 24 h of forecast, but still shows a slight improvement in
case of ALADIN2w compared to ALADIN1w.</p>
      <p>The above suggests that the improvements due to two-way air–sea coupling are
not due to more detailed ALADIN2w model dynamics but mostly due to improved
forecast of the temperature of the air advected across Adriatic. The impact
on other meteorological variables and/or other regions is much smaller. This
is somewhat expected, since the ALADIN atmospheric component communicates
asymmetrically with the ocean model in the coupled system – only SST is fed
back from POM2w to ALADIN2w. The improvement of the temperature error
variance and the impact of ocean on other meteorological fields might be
greater during the warm part of the year, when small-scale atmospheric
processes are more important and their initialization could be a direct
consequence of ocean dynamics.</p>
</sec>
<sec id="Ch1.S4.SS3">
  <title>Ocean temperatures</title>
      <p>Figure <xref ref-type="fig" rid="Ch1.F11"/> depicts the ocean temperature time series, while
Fig. <xref ref-type="fig" rid="Ch1.F12"/> shows model-observation temperature differences
and their respective biases and root mean square errors (RMSEs) at Vida,
Paloma and Acqua Alta stations. Ocean temperatures are reproduced
significantly better in the POM2w, with RMSEs up to 4 times smaller than
those in POM1w, as shown in Fig. <xref ref-type="fig" rid="Ch1.F12"/>. These statistics
are also gathered in Table <xref ref-type="table" rid="Ch1.T2"/>.</p>
      <p>The latter serves as an illustration of the significance of the atmosphere
boundary condition over the sea surface, especially in setups where heat
fluxes from the atmospheric model are the only ones used in the ocean model.
The OSTIA sea surface temperature boundary condition in the northern Adriatic
was consistently warmer than observations by more than 2 <inline-formula><mml:math display="inline"><mml:msup><mml:mi/><mml:mo>∘</mml:mo></mml:msup></mml:math></inline-formula>C
in the Gulf of Trieste throughout the bora event (see the green line in the
first three rows in Fig. <xref ref-type="fig" rid="Ch1.F11"/>). On the other hand, OSTIA
seawater temperatures at Acqua Alta station reproduces the sea surface
temperature fairly well (see the green line in the bottom row in Fig. <xref ref-type="fig" rid="Ch1.F11"/>). OSTIA overestimation of SST led to a modeled
cumulative ocean heat loss and cooling in the Gulf of Trieste and along the
Venetian coast during the event (see Fig. <xref ref-type="fig" rid="Ch1.F12"/>) in the
POM1w. The uncoupled system at Paloma location was losing about 130 W m<inline-formula><mml:math display="inline"><mml:msup><mml:mi/><mml:mrow><mml:mo>-</mml:mo><mml:mn mathvariant="normal">2</mml:mn></mml:mrow></mml:msup></mml:math></inline-formula>
more heat per unit area than the coupled setup (see bottom
right panel in Fig. <xref ref-type="fig" rid="Ch1.F7"/>). At Acqua Alta, however, the amounts
of these losses are lower, of the order of 70 W m<inline-formula><mml:math display="inline"><mml:msup><mml:mi/><mml:mrow><mml:mo>-</mml:mo><mml:mn mathvariant="normal">2</mml:mn></mml:mrow></mml:msup></mml:math></inline-formula> (see bottom
right panel in Fig. <xref ref-type="fig" rid="Ch1.F8"/>), which could explain the absence of
negative ocean temperature drift during the bora outbreak at this location
(compare the red lines in all the panels of Fig. <xref ref-type="fig" rid="Ch1.F12"/>).</p>
      <p>A possible way to mitigate such drifts in uncoupled systems could be the introduction of flux corrections of the form

                <disp-formula id="Ch1.E6" content-type="numbered"><mml:math display="block"><mml:mrow><mml:mi mathvariant="italic">δ</mml:mi><mml:mi>Q</mml:mi><mml:mo>∼</mml:mo><mml:msub><mml:mfenced open="(" close=")"><mml:mstyle displaystyle="true"><mml:mfrac style="display"><mml:mrow><mml:mo>∂</mml:mo><mml:msub><mml:mi>Q</mml:mi><mml:mi mathvariant="normal">T</mml:mi></mml:msub></mml:mrow><mml:mrow><mml:mo>∂</mml:mo><mml:mi>T</mml:mi></mml:mrow></mml:mfrac></mml:mstyle></mml:mfenced><mml:mrow><mml:mi>T</mml:mi><mml:mo>=</mml:mo><mml:msup><mml:mi>T</mml:mi><mml:mtext>OSTIA</mml:mtext></mml:msup></mml:mrow></mml:msub><mml:mo>(</mml:mo><mml:msup><mml:mi>T</mml:mi><mml:mtext>CLIM</mml:mtext></mml:msup><mml:mo>-</mml:mo><mml:msup><mml:mi>T</mml:mi><mml:mtext>OSTIA</mml:mtext></mml:msup><mml:mo>)</mml:mo><mml:mo>,</mml:mo></mml:mrow></mml:math></disp-formula>

          where <inline-formula><mml:math display="inline"><mml:mrow><mml:mo>(</mml:mo><mml:mo>∂</mml:mo><mml:msub><mml:mi>Q</mml:mi><mml:mi mathvariant="normal">T</mml:mi></mml:msub><mml:mo>/</mml:mo><mml:mo>∂</mml:mo><mml:mi>T</mml:mi><mml:mo>)</mml:mo></mml:mrow></mml:math></inline-formula> is the temperature derivative of the
total flux from Eq. (<xref ref-type="disp-formula" rid="Ch1.E5"/>), <inline-formula><mml:math display="inline"><mml:mrow><mml:msup><mml:mi>T</mml:mi><mml:mtext>CLIM</mml:mtext></mml:msup></mml:mrow></mml:math></inline-formula> is climatological sea
surface temperature field and <inline-formula><mml:math display="inline"><mml:mrow><mml:msup><mml:mi>T</mml:mi><mml:mtext>OSTIA</mml:mtext></mml:msup></mml:mrow></mml:math></inline-formula> is the atmospheric model sea
surface temperature boundary condition used to compute bulk fluxes. Such flux
corrections would effectively nudge the uncoupled ocean model temperatures
towards its climatology and at least in principle prevent excessive cooling.
Of course in this case climatology needs to be computed in advance and is
itself burdened with errors. Our work shows that two-way atmosphere–ocean
coupling presents an alternative, and more consistent solution to temperature
drifts caused by air–sea flux misestimation.</p>
      <p>Sea surface temperatures at the end of the bora event on 14 February 2012,
00:00 UTC, from POM1w and POM2w are shown in Fig. <xref ref-type="fig" rid="Ch1.F13"/>. When
compared to POM2w (bottom panel of Fig. <xref ref-type="fig" rid="Ch1.F13"/>), POM1w exhibits
1–2 <inline-formula><mml:math display="inline"><mml:msup><mml:mi/><mml:mo>∘</mml:mo></mml:msup></mml:math></inline-formula>C cooler temperatures in the Gulf of Trieste and along
the coastal belt of the Venetian Gulf (top panel of Fig. <xref ref-type="fig" rid="Ch1.F13"/>).
Comparisons with ocean temperature CTD measurements at Vida location also
reflect this discrepancy between POM2w and POM1w, as shown in Table <xref ref-type="table" rid="Ch1.T3"/>. POM1w exhibits a persistent cold bias, which increased
substantially after the bora event. This POM1w behavior is proposed to be
stemming from the overestimation of sea surface temperatures in the
atmospheric model OSTIA boundary conditions (green lines in Fig. <xref ref-type="fig" rid="Ch1.F11"/>), which leads to excessive heat losses and excessive
cooling in the Gulf of Trieste and along the narrow Italian coastal belt. The
models and observations shown in Fig. <xref ref-type="fig" rid="Ch1.F11"/> for example
indicate that OSTIA SSTs were most overestimated at the Paloma station and that
POM1w sea temperatures at the Paloma station were consistently the coolest of all
three stations.</p>
</sec>
<sec id="Ch1.S4.SS4">
  <title>Ocean circulation</title>
      <p>Our numerical experiments are consistent with earlier findings
<xref ref-type="bibr" rid="bib1.bibx35 bib1.bibx24 bib1.bibx16 bib1.bibx27 bib1.bibx18" id="paren.45"/> in the sense
that a prolonged bora generates a double cyclonic–anticyclonic gyre system in
the northern Adriatic. A double gyre is related to the wind stress curl in the
northern Adriatic, arising from pronounced Trieste and Senj jets to the north
and south of Istria and moderate to low wind speeds in between. It was often
reported <xref ref-type="bibr" rid="bib1.bibx16" id="paren.46"><named-content content-type="pre">i.e.,</named-content></xref> that this wind distribution leads to
formation of a cyclonic gyre in the northwestern (Venetian) part of the
northern Adriatic, and of an anticyclonic gyre to the south of Rovinj,
without any notable offshore surface currents along the west coast of Istria.</p>
      <p>Circulation from POM2w at 1 and 12 <inline-formula><mml:math display="inline"><mml:mi mathvariant="normal">m</mml:mi></mml:math></inline-formula> depths is depicted in
Fig. <xref ref-type="fig" rid="Ch1.F14"/>. A cyclonic gyre is positioned at roughly
(45.15<inline-formula><mml:math display="inline"><mml:msup><mml:mi/><mml:mo>∘</mml:mo></mml:msup></mml:math></inline-formula> N, 12.9<inline-formula><mml:math display="inline"><mml:msup><mml:mi/><mml:mo>∘</mml:mo></mml:msup></mml:math></inline-formula> E), while its anticyclonic counterpart is
much smaller and located closer to Rovinj at about (<inline-formula><mml:math display="inline"><mml:mo>-</mml:mo></mml:math></inline-formula>0.1<inline-formula><mml:math display="inline"><mml:msup><mml:mi/><mml:mo>∘</mml:mo></mml:msup></mml:math></inline-formula> N,
0.5<inline-formula><mml:math display="inline"><mml:msup><mml:mi/><mml:mo>∘</mml:mo></mml:msup></mml:math></inline-formula> E) to the southeast with respect to the center of the cyclonic
gyre. In the northern arm of the cyclonic gyre along the Italian coast and
along the Venetian coast surface current velocities attain up to 0.5 m s<inline-formula><mml:math 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>. At 12 <inline-formula><mml:math display="inline"><mml:mi mathvariant="normal">m</mml:mi></mml:math></inline-formula> depth, the cyclonic gyre current
velocities are less than 0.25 m s<inline-formula><mml:math 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>. Currents at 12 <inline-formula><mml:math display="inline"><mml:mi mathvariant="normal">m</mml:mi></mml:math></inline-formula>
depth further indicate a compensating inflow of water into the Gulf of
Trieste along its southern coastline.</p>

      <?xmltex \floatpos{t}?><fig id="Ch1.F14"><caption><p>Time averaged ocean currents in POM2w at <inline-formula><mml:math display="inline"><mml:mrow><mml:mn mathvariant="normal">1</mml:mn><mml:mspace linebreak="nobreak" width="0.125em"/><mml:mi mathvariant="normal">m</mml:mi></mml:mrow></mml:math></inline-formula> depth (top
panel) and at <inline-formula><mml:math display="inline"><mml:mrow><mml:mn>12</mml:mn><mml:mspace linebreak="nobreak" width="0.125em"/><mml:mi mathvariant="normal">m</mml:mi></mml:mrow></mml:math></inline-formula> depth (bottom panel) during 9–14 February 2012,
00:00 UTC.</p></caption>
          <?xmltex \igopts{width=199.169291pt}?><graphic xlink:href="https://os.copernicus.org/articles/12/71/2016/os-12-71-2016-f14.png"/>

        </fig>

      <p>The northern Adriatic surface circulation patterns from POM1w are very
similar and they are not shown explicitly. This similarity in the POM1w and
POM2w circulations is due to the fact that circulation in the northern
Adriatic during this bora outbreak was completely wind dominated. This can be
shown by estimating the Ekman depth during the bora outbreak. Wind speeds of
bora in the Gulf of Trieste during the period 9–14 February 2012 were
measured to be well above <inline-formula><mml:math display="inline"><mml:mrow><mml:mn>10</mml:mn><mml:mspace width="0.125em" linebreak="nobreak"/><mml:mi mathvariant="normal">m</mml:mi><mml:mspace width="0.125em" linebreak="nobreak"/><mml:msup><mml:mi mathvariant="normal">s</mml:mi><mml:mrow><mml:mo>-</mml:mo><mml:mn mathvariant="normal">1</mml:mn></mml:mrow></mml:msup></mml:mrow></mml:math></inline-formula> for most of the time; see
Fig. <xref ref-type="fig" rid="Ch1.F4"/>. Following <xref ref-type="bibr" rid="bib1.bibx17" id="text.47"/> the Ekman depth in the
Gulf of Trieste can be expressed in terms of wind speed and they estimated its
value with the following:

                <disp-formula id="Ch1.E7" content-type="numbered"><mml:math display="block"><mml:mrow><mml:msub><mml:mi>D</mml:mi><mml:mi mathvariant="normal">E</mml:mi></mml:msub><mml:mo>=</mml:mo><mml:mstyle displaystyle="true"><mml:mfrac style="display"><mml:mrow><mml:msqrt><mml:mn mathvariant="normal">2</mml:mn></mml:msqrt><mml:mi mathvariant="italic">π</mml:mi></mml:mrow><mml:mrow><mml:mi>f</mml:mi><mml:mi>k</mml:mi></mml:mrow></mml:mfrac></mml:mstyle><mml:mfenced open="(" close=")"><mml:mstyle displaystyle="true"><mml:mfrac style="display"><mml:mrow><mml:msub><mml:mi mathvariant="italic">ρ</mml:mi><mml:mtext>air</mml:mtext></mml:msub></mml:mrow><mml:mrow><mml:msub><mml:mi mathvariant="italic">ρ</mml:mi><mml:mtext>sea</mml:mtext></mml:msub></mml:mrow></mml:mfrac></mml:mstyle></mml:mfenced><mml:msub><mml:mi>C</mml:mi><mml:mi mathvariant="normal">D</mml:mi></mml:msub><mml:mo>|</mml:mo><mml:mi mathvariant="bold-italic">V</mml:mi><mml:mo>|</mml:mo><mml:mo>≈</mml:mo><mml:mn>47</mml:mn><mml:mspace linebreak="nobreak" width="0.125em"/><mml:mi mathvariant="normal">m</mml:mi><mml:mo>,</mml:mo></mml:mrow></mml:math></disp-formula>

          where <inline-formula><mml:math display="inline"><mml:mrow><mml:msub><mml:mi mathvariant="italic">ρ</mml:mi><mml:mtext>air</mml:mtext></mml:msub><mml:mo>=</mml:mo><mml:mn>1.29</mml:mn><mml:mspace width="0.125em" linebreak="nobreak"/><mml:mi mathvariant="normal">kg</mml:mi><mml:mspace width="0.125em" linebreak="nobreak"/><mml:msup><mml:mi mathvariant="normal">m</mml:mi><mml:mrow><mml:mo>-</mml:mo><mml:mn mathvariant="normal">3</mml:mn></mml:mrow></mml:msup></mml:mrow></mml:math></inline-formula> and
<inline-formula><mml:math display="inline"><mml:mrow><mml:msub><mml:mi mathvariant="italic">ρ</mml:mi><mml:mtext>sea</mml:mtext></mml:msub><mml:mo>=</mml:mo><mml:mn>1029</mml:mn><mml:mspace linebreak="nobreak" width="0.125em"/><mml:mi mathvariant="normal">kg</mml:mi><mml:mspace linebreak="nobreak" width="0.125em"/><mml:msup><mml:mi mathvariant="normal">m</mml:mi><mml:mrow><mml:mo>-</mml:mo><mml:mn mathvariant="normal">3</mml:mn></mml:mrow></mml:msup></mml:mrow></mml:math></inline-formula> are air and seawater densities, <inline-formula><mml:math display="inline"><mml:mrow><mml:msub><mml:mi>C</mml:mi><mml:mi mathvariant="normal">D</mml:mi></mml:msub><mml:mo>=</mml:mo><mml:mn>2.6</mml:mn><mml:mo>×</mml:mo><mml:msup><mml:mn>10</mml:mn><mml:mrow><mml:mo>-</mml:mo><mml:mn mathvariant="normal">3</mml:mn></mml:mrow></mml:msup></mml:mrow></mml:math></inline-formula> is
the drag coefficient, <inline-formula><mml:math display="inline"><mml:mrow><mml:mi>f</mml:mi><mml:mo>=</mml:mo><mml:mn>1.03</mml:mn><mml:mo>×</mml:mo><mml:msup><mml:mn>10</mml:mn><mml:mrow><mml:mo>-</mml:mo><mml:mn mathvariant="normal">4</mml:mn></mml:mrow></mml:msup></mml:mrow></mml:math></inline-formula> is the Coriolis parameter, <inline-formula><mml:math display="inline"><mml:mrow><mml:mo>|</mml:mo><mml:mi mathvariant="bold-italic">V</mml:mi><mml:mo>|</mml:mo><mml:mo>≈</mml:mo><mml:mn>10</mml:mn><mml:mspace width="0.125em" linebreak="nobreak"/><mml:mi mathvariant="normal">m</mml:mi><mml:mspace linebreak="nobreak" width="0.125em"/><mml:msup><mml:mi mathvariant="normal">s</mml:mi><mml:mrow><mml:mo>-</mml:mo><mml:mn mathvariant="normal">1</mml:mn></mml:mrow></mml:msup></mml:mrow></mml:math></inline-formula> is
the wind speed during the bora episode and <inline-formula><mml:math display="inline"><mml:mrow><mml:mi>k</mml:mi><mml:mo>≈</mml:mo><mml:mn>0.03</mml:mn></mml:mrow></mml:math></inline-formula> is the ratio of
Ekman current velocity to wind speed. Ekman depth <inline-formula><mml:math display="inline"><mml:mrow><mml:msub><mml:mi>D</mml:mi><mml:mi mathvariant="normal">E</mml:mi></mml:msub></mml:mrow></mml:math></inline-formula> during this bora event was therefore larger
than the ocean depths along the Trieste jet (about <inline-formula><mml:math display="inline"><mml:mrow><mml:mo>∼</mml:mo><mml:mn>30</mml:mn><mml:mspace width="0.125em" linebreak="nobreak"/><mml:mi mathvariant="normal">m</mml:mi></mml:mrow></mml:math></inline-formula>, see Fig. <xref ref-type="fig" rid="Ch1.F1"/>),
which implies a wind-dominated circulation in the Trieste jet (and in the northern Adriatic at Istrian latitudes)
throughout the period of 9–14 February 2012. Ekman dynamics confines the outflow from the Gulf of Trieste to the
northern part of its entrance (see upper panel of Fig. <xref ref-type="fig" rid="Ch1.F14"/>). This circulation feature is well known
and is also confirmed by high-frequency radar measurements in the Gulf, see <xref ref-type="bibr" rid="bib1.bibx4" id="text.48"/>. Due to land-locked
nature of the Gulf, however, this Ekman layer outflow is modified by the compensating inflow, which needs to take place
at larger depths along the southern part of the Gulf to replenish it with water.</p>
      <p>These cool water masses from the Gulf of Trieste and Venetian coast then
recirculate in the northern Adriatic due to bora induced
cyclonic–anticyclonic circulation (see Fig. <xref ref-type="fig" rid="Ch1.F14"/>) and re-enter
the Gulf of Trieste some 10 days later. It appears therefore (see Figs. <xref ref-type="fig" rid="Ch1.F13"/> and <xref ref-type="fig" rid="Ch1.F14"/>) that higher POM2w temperatures at Vida
location are at least partly due to bora induced recirculation, which leads to
the intrusion of warmer water along the southern part the of the Gulf of
Trieste. Warm water intrusion to the Gulf is clearly visible in POM2w but
absent from the uncoupled setup. Judging from POM2w ocean temperatures
(bottom panel of Fig. <xref ref-type="fig" rid="Ch1.F13"/>), Vida is located well within the warm
water tongue while Paloma is on its lateral boundary (bottom panel of
Fig. <xref ref-type="fig" rid="Ch1.F13"/>).<?xmltex \hack{\newpage}?></p>
</sec>
</sec>
<sec id="Ch1.S5" sec-type="conclusions">
  <title>Conclusions</title>
      <p>In this paper we present one-way and two-way coupled modeling of ocean and
atmosphere during an extreme bora event in the northern Adriatic in February
2012. Ocean feedback to the atmosphere in the two-way coupled system was
limited to the sea surface temperature. The sea surface temperature boundary
condition for the atmospheric model in the one-way coupled system was taken
from the OSTIA SST product, provided by NOAA. The sea surface temperature
boundary condition for the atmospheric model in the two-way coupled system
was obtained from the ocean component of the two-way coupled system. Both
systems used fluxes computed in the ALADIN atmosphere component. OSTIA SST
was overestimated throughout the bora outbreak, leading to overestimation of
air–sea flux and excessive ocean cooling in the one-way coupled system.</p>
      <p>We used in situ observations from three marine observation platforms (Vida,
Paloma and Acqua Alta) in the northern Adriatic to compute in situ fluxes and
to compare them to modeled fluxes. When compared to observations sensible
heat fluxes were better modeled by the two-way coupled system, while latent
heat fluxes were somewhat better represented in the one-way coupled system.
An improvement of sensible heat fluxes in the two-way coupled system was not
unexpected, since SST was the parameter that was communicated from the ocean
to the atmosphere component of the two-way setup. At Paloma and Acqua Alta
sensible and latent fluxes differ significantly during the bora event while
their differences are less pronounced before and after the bora. At Vida
location flux differences between the two setups were not significant
throughout the simulations. This is probably connected to its location, which
is within 1 grid point from the land in both ocean and atmospheric model.
These results indicate that the two-way coupling impact will be largest
during periods of intense air–sea interactions and less pronounced otherwise.</p>
      <p>We performed an objective verification of near-surface air temperatures from
both setups, using measurements from 23 downwind stations in eastern Italy
during each calendar day in February 2012. The two-way coupled system shows a
lower temperature bias for prognostic lead times up to 18 h. Standard
deviation of modeled near-surface temperature errors is also somewhat lower
in the atmosphere component of the two-way coupled system. However, the
atmosphere in the two-way coupled system appears to benefit mostly from a
better representation of air temperatures over the ocean, which are later
advected over mainland Italy during the bora, improving the skill score.
Ocean temperature feedback in the two-way coupled setup did not induce any
new dynamics in the atmospheric model during this extreme winter case. We
expect this to happen during the summer when small-scale processes in the
atmosphere are more important. We leave this for further
investigations.<?xmltex \hack{\newpage}?></p>
      <p>Using in situ platform and CTD observations we show that the two-way coupled system
captures the observed ocean temperature response during the bora event significantly better
than the one-way coupled system. This is due to OSTIA SST overestimation during the bora,
leading to excessive heat losses from the ocean, as noted above. These heat losses induced
overcooling of  water in POM1w and this cold bias persisted in the one-way coupled ocean
model long after the cessation of bora. We have demonstrated that two-way atmosphere–ocean
coupling offers an efficient and consistent way to address this issue.</p>
      <p>Currently the initial conditions of the two ALADIN 24 h integration runs
are provided by an independent ALADIN assimilation cycle with 3-hourly
3D-Var,
which uses OSTIA SST analysis for the ocean temperature boundary condition. A
possible future improvement of the two-way coupled system would therefore be
the
inclusion of a two-way coupling in the ALADIN assimilation cycle also. On the
other hand, there would be a lack of SST information from measurements, which
now enter the two-way coupled system through the ALADIN assimilation cycle (apart
from information reaching the area of interest from the lateral boundary
conditions). Consequently, there would be obvious danger of a significant
model drift in longer timescales, especially over several seasons. To avoid
this and to ensure optimal use of SST measurements, a comprehensive data
assimilation system for the ocean component would be needed.</p>
</sec>

      
      </body>
    <back><ack><title>Acknowledgements</title><p>The manuscript benefited from suggestions by two anonymous referees. We are
indebted to the NIB staff that was performing the national monitoring program
(under the EU Water Framework Directive) in 2012, funded by the Slovenian
Agency of Environment, and to the research program funded by the Ministry of
Higher Education and Technology of Slovenia (contract P1-0237). We received
valuable suggestions and comments by Jeffrey W. Book and Sandro Carniel. Po
River discharges from Pontelagoscuro measuring station, operated by ARPA-SIMC
Emilia-Romagna, were kindly provided by Fabrizio Tonelli and Silvano Pecora.
Observations from Paloma station were kindly provided by Fabio Raicich from
CNR-ISMAR, Trieste, Italy. Acqua Alta observational data were kindly provided
by Alvise Benetazzo and Sandro Carniel from CNR-ISMAR, Venice, Italy. Neva
Pristov and Jasna Vehovar (ARSO) maintained computation operational
environment on ARSO and obtained all necessary ECMWF boundary conditions,
respectively. Rižana river climatology was computed from ARSO discharge
measurements by Martin Vodopivec (NIB). <?xmltex \hack{\newline}?><?xmltex \hack{\newline}?>
Edited by: L. Kantha</p></ack><ref-list>
    <title>References</title>

      <ref id="bib1.bibx1"><label>Artegiani et al.(1997)</label><mixed-citation>Artegiani, A., Bregant, D., Paschini, E., Pinardi, N., Raicich, F., and Russo, A.:
The Adriatic Sea General Circulation I: Air-Sea Interactions and Water Mass Structure, J. Phys. Oceanogr., 27, 1492–1514,
<ext-link xlink:href="http://dx.doi.org/10.1175/1520-0485(1997)027&lt;1492:TASGCP&gt;2.0.CO;2" ext-link-type="DOI">10.1175/1520-0485(1997)027&lt;1492:TASGCP&gt;2.0.CO;2</ext-link>,
1997.</mixed-citation></ref>
      <ref id="bib1.bibx2"><label>Benetazzo et al.(2014)</label><mixed-citation>Benetazzo, A., Bergamasco, A., Bonaldo, D., Falcieria, F. M., Sclavo, M.,
Langone, L., and Carniel, S.: Response of the Adriatic Sea to an intense cold air outbreak: dense water dynamics and wave-induced transport, Prog.
Oceanogr., 128, 115–138, <ext-link xlink:href="http://dx.doi.org/10.1016/j.pocean.2014.08.015" ext-link-type="DOI">10.1016/j.pocean.2014.08.015</ext-link>, 2014.</mixed-citation></ref>
      <ref id="bib1.bibx3"><label>Blumberg and Mellor(1987)</label><mixed-citation>
Blumberg, A. F. and Mellor, G. L.: A description of a three-dimensional coastal ocean circulation model, in: Three-Dimensional Coastal Ocean Models,
edited by: Heaps, N., American Geophysical Union, Washington, DC, USA, 1–16,
1987.</mixed-citation></ref>
      <ref id="bib1.bibx4"><label>Cosoli et al.(2013)</label><mixed-citation>Cosoli, S., Ličer, M., Vodopivec, M., and Malačič, V.: Surface circulation in the Gulf of Trieste (northern Adriatic Sea) from radar, model,
and ADCP comparisons, J. Geophys. Res.-Oceans, 118, 6183–6200,
<ext-link xlink:href="http://dx.doi.org/10.1002/2013JC009261" ext-link-type="DOI">10.1002/2013JC009261</ext-link>, 2013.</mixed-citation></ref>
      <ref id="bib1.bibx5"><label>Cushman-Roisin et al.(2001)</label><mixed-citation>
Cushman-Roisin, B., Gačić, M., Poulain, P.-M., and Artegiani, A.: Physical
Oceanography of the Adriatic Sea. Past, Present and Future,
Springer, London, UK, 47–48, 2001.</mixed-citation></ref>
      <ref id="bib1.bibx6"><label>Davies(1976)</label><mixed-citation>
Davies, H. C.: A lateral boundary formulation for multi-level prediction models, Q. J. Roy. Meteor. Soc., 102, 405–418, 1976.</mixed-citation></ref>
      <ref id="bib1.bibx7"><label>Donlon et al.(2012)</label><mixed-citation>Donlon, C. J., Martin, M., Stark, J. D., Roberts-Jones, J., Fiedler, E., and Wimmer, W.: The Operational Sea Surface Temperature and Sea Ice Analysis
(OSTIA), Remote Sens. Environ., 116, 140–158,
<ext-link xlink:href="http://dx.doi.org/10.1016/j.rse.2010.10.017" ext-link-type="DOI">10.1016/j.rse.2010.10.017</ext-link>, 2012.</mixed-citation></ref>
      <ref id="bib1.bibx8"><label>Dorman et al.(2006)</label><mixed-citation>Dorman, C. E., Carniel, S., Cavaleri, L., Sclavo, M., Chiggiato, J.,
Doyle, J., Haack, T., Pullen, J., Grbec, B., Vilibić, I.,
Janeković, I., Lee, C., Malačič, V., Orlić, M., Paschini, E.,
Russo, A., and Signell, R. P.: February 2003 marine atmospheric conditions and the bora over the northern Adriatic, J. Geophys. Res., 111, C03S03,
<ext-link xlink:href="http://dx.doi.org/10.1029/2005JC003134" ext-link-type="DOI">10.1029/2005JC003134</ext-link>, 2006.</mixed-citation></ref>
      <ref id="bib1.bibx9"><label>Eleuterio(1998)</label><mixed-citation>Eleuterio, D.: A Comparison of Bulk Aerodynamic Methods for Calculating Air-Sea Flux, Master Thesis, Naval Postgraduate School, Monterey, CA, USA, available at:
<uri>http://www.researchgate.net/publication/235119736_A_Comparison_of_Bulk_Aerodynamic_Methods_for_Calculating_Air-Sea_Flux</uri> (last access: 14 October 2015), 1998.</mixed-citation></ref>
      <ref id="bib1.bibx10"><label>Fairall et al.(1996)</label><mixed-citation>Fairall, C. W., Bradley, E. F., Rogers, D. P., Edson, J. B., and Young, G. S.:
Bulk parameterization of air-sea fluxes for Tropical Ocean Global Atmosphere
Coupled-Ocean Atmosphere Response Experiment, J. Geophys. Res., 101,
3747–3764, <ext-link xlink:href="http://dx.doi.org/10.1029/95JC03205" ext-link-type="DOI">10.1029/95JC03205</ext-link>, 1996.</mixed-citation></ref>
      <ref id="bib1.bibx11"><label>Fischer et al.(2005)</label><mixed-citation>Fischer, C., Montmerle, T., Berre, L., Auger, L., and
Ştefănescu, S. E.: An overview of the variational assimilation in the
ALADIN/France numerical weather-prediction system, Q. J. Roy. Meteor. Soc.,
131, 3477–3492, <ext-link xlink:href="http://dx.doi.org/10.1256/qj.05.115" ext-link-type="DOI">10.1256/qj.05.115</ext-link>, 2005.</mixed-citation></ref>
      <ref id="bib1.bibx12"><label>Gerard et al.(2009)</label><mixed-citation>Gerard, L., Piriou, J.-M., Brožková, R., Geleyn, J.-F., and
Banciu, D.: Cloud and precipitation parameterization in a meso-gamma-scale
operational weather prediction model, Mon. Weather Rev., 137, 3960–3977,
<ext-link xlink:href="http://dx.doi.org/10.1175/2009MWR2750.1" ext-link-type="DOI">10.1175/2009MWR2750.1</ext-link>, 2009.</mixed-citation></ref>
      <ref id="bib1.bibx13"><label>Grisogono and Belušić(2009)</label><mixed-citation>Grisogono, B. and Belušić, D.: A review of recent advances in understanding the meso- and microscale properties of severe Bora wind, Tellus, 61A, 1–16,
<ext-link xlink:href="http://dx.doi.org/10.1111/j.1600-0870.2008.00369.x" ext-link-type="DOI">10.1111/j.1600-0870.2008.00369.x</ext-link>, 2009.</mixed-citation></ref>
      <ref id="bib1.bibx14"><label>Janeković et al.(2014)</label><mixed-citation>Janeković, I., Mihanović, H., Vilibić, I., and Tudor, M.: Extreme
cooling and dense water formation estimates in open and coastal regions of
the Adriatic Sea during the winter of 2012, J. Geophys. Res.-Oceans, 119,
3200–3218, <ext-link xlink:href="http://dx.doi.org/10.1002/2014JC009865" ext-link-type="DOI">10.1002/2014JC009865</ext-link>, 2014.</mixed-citation></ref>
      <ref id="bib1.bibx15"><label>Kourafalou(2001)</label><mixed-citation>Kourafalou, V. H.: River plume development in semi-enclosed Mediterranean regions: north Adriatic Sea and northwestern Aegean Sea, J. Marine Syst., 30,
181–205, <ext-link xlink:href="http://dx.doi.org/10.1016/S0924-7963(01)00058-6" ext-link-type="DOI">10.1016/S0924-7963(01)00058-6</ext-link>, 2001.</mixed-citation></ref>
      <ref id="bib1.bibx16"><label>Kuzmić et al.(2007)</label><mixed-citation>Kuzmić, M., Janeković, I., Book, J. W., Martin, P. J., and
Doyle, J. D.: Modeling the northern Adriatic double-gyre response to intense
bora wind: a revisit, J. Geophys. Res., 111, C03S13,
<ext-link xlink:href="http://dx.doi.org/10.1029/2005JC003377" ext-link-type="DOI">10.1029/2005JC003377</ext-link>, 2007.</mixed-citation></ref>
      <ref id="bib1.bibx17"><label>Malačič and Petelin(2009)</label><mixed-citation>Malačič, V. and Petelin, B.: Climatic circulation in the Gulf of Trieste (northern Adriatic), J. Geophys. Res., 114, C07002,
<ext-link xlink:href="http://dx.doi.org/10.1029/2008JC004904" ext-link-type="DOI">10.1029/2008JC004904</ext-link>, 2009.</mixed-citation></ref>
      <ref id="bib1.bibx18"><label>Malačič et al.(2012)</label><mixed-citation>Malačič, V., Petelin, B., and Vodopivec, M.: Topographic control of wind-driven circulation in the northern Adriatic, J. Geophys. Res., 117,
C06032, <ext-link xlink:href="http://dx.doi.org/10.1029/2012JC008063" ext-link-type="DOI">10.1029/2012JC008063</ext-link>, 2012.</mixed-citation></ref>
      <ref id="bib1.bibx19"><label>Mantziafou and Lascaratos(2008)</label><mixed-citation>Mantziafou, A. and Lascaratos, A.: Deep-water formation in the Adriatic Sea:
interannual simulations for the years 1979–1999, Deep-Sea Res. Pt. I, 55,
1403–1427, <ext-link xlink:href="http://dx.doi.org/10.1016/j.dsr.2008.06.005" ext-link-type="DOI">10.1016/j.dsr.2008.06.005</ext-link>, 2008.</mixed-citation></ref>
      <ref id="bib1.bibx20"><label>Mascart et al.(1995)</label><mixed-citation>Mascart, P., Noilhan, J., and Giordani, H.: A modified parametrization of
flux-profile relationships in the surface layer using different roughness
length values for heat and momentum, Bound.-Lay. Meteorol., 72, 331–344,
<ext-link xlink:href="http://dx.doi.org/10.1007/BF00708998" ext-link-type="DOI">10.1007/BF00708998</ext-link>, 1995.</mixed-citation></ref>
      <ref id="bib1.bibx21"><label>Mellor(1998)</label><mixed-citation>
Mellor, G.: Users Guide for a Three-Dimensional, Primitive Equation,
Numerical Ocean Model, Tech. Rep., Princeton University, Princeton, NJ, USA, 1998.</mixed-citation></ref>
      <ref id="bib1.bibx22"><label>Mihanović et al.(2013)</label><mixed-citation>Mihanović, H., Vilibić, I., Carniel, S., Tudor, M., Russo, A.,
Bergamasco, A., Bubić, N., Ljubešić, Z., Viličić, D.,
Boldrin, A., Malačič, V., Celio, M., Comici, C., and Raicich, F.:
Exceptional dense water formation on the Adriatic shelf in the winter of 2012, Ocean Sci., 111, 561–572, <ext-link xlink:href="http://dx.doi.org/10.5194/os-9-561-2013" ext-link-type="DOI">10.5194/os-9-561-2013</ext-link>, 2013.</mixed-citation></ref>
      <ref id="bib1.bibx23"><label>Nielsen and Hansen(1973)</label><mixed-citation>
Nielsen, S. and Hansen, E.: Numerical simulation of the rainfall-runoff process on a daily basis, Nord. Hydrol., 4, 171–190, 1973.</mixed-citation></ref>
      <ref id="bib1.bibx24"><label>Paklar et al.(2001)</label><mixed-citation>Paklar, G. B., Isakov, V., Koračin, D., Kourafalou, V., and
Orlić, M.: A case study of bora-driven flow and density changes on the
Adriatic Shelf (January 1987), Cont. Shelf. Res., 21, 1751–1783,
<ext-link xlink:href="http://dx.doi.org/10.1016/S0278-4343(01)00029-2" ext-link-type="DOI">10.1016/S0278-4343(01)00029-2</ext-link>, 2001.</mixed-citation></ref>
      <ref id="bib1.bibx25"><label>Pullen et al.(2003)</label><mixed-citation>Pullen, J., Doyle, J., Hodur, R., Ogston, A., Book, J., Perkins, H., and
Signell, R.: Coupled ocean-atmosphere nested modeling of the Adriatic Sea
during winter and spring 2001, J. Geophys. Res., 108, 3320,
<ext-link xlink:href="http://dx.doi.org/10.1029/2003JC001780" ext-link-type="DOI">10.1029/2003JC001780</ext-link>, 2003.</mixed-citation></ref>
      <ref id="bib1.bibx26"><label>Pullen et al.(2006)</label><mixed-citation>Pullen, J., Doyle, J., and Signell, R.: Two-way air–sea coupling: a study of the Adriatic, Mon. Weather Rev., 134, 1465–1483, <ext-link xlink:href="http://dx.doi.org/10.1175/MWR3137.1" ext-link-type="DOI">10.1175/MWR3137.1</ext-link>,
2006.</mixed-citation></ref>
      <ref id="bib1.bibx27"><label>Pullen et al.(2007)</label><mixed-citation>Pullen, J., Doyle, J., Haack, T., Dorman, C., Signell, R., and Lee, C. M.:
Bora event variability and the role of air-sea feedback, J. Geophys. Res.,
112, 33407, <ext-link xlink:href="http://dx.doi.org/10.1029/2006JC003726" ext-link-type="DOI">10.1029/2006JC003726</ext-link>, 2007.</mixed-citation></ref>
      <ref id="bib1.bibx28"><label>Raicich(1994)</label><mixed-citation>
Raicich, F.: Notes on the Flow Rates of the Adriatic Rivers, tech. rep., CNR,
Ist. Sper. Talassografico, Trieste, Italy, 1994.</mixed-citation></ref>
      <ref id="bib1.bibx29"><label>Raicich et al.(2013)</label><mixed-citation>Raicich, F., Malačič, C., Celio, M., Giaiotti, D., Cantoni, C.,
Colucci, R. R., Čermelj, B., and Pucillo, A.: Extreme air-sea
interactions in the Gulf of Trieste (North Adriatic) during the strong Bora
event in winter 2012, J. Geophys. Res.-Oceans, 118, 5238–5250,
<ext-link xlink:href="http://dx.doi.org/10.1002/jgrc.20398" ext-link-type="DOI">10.1002/jgrc.20398</ext-link>, 2013.</mixed-citation></ref>
      <ref id="bib1.bibx30"><label>Seity et al.(2011)</label><mixed-citation>Seity, Y., Brousseau, P.,Malardel, S., Hello, G., Benard, P., Bouttier, F.,
Lac, C., and Masson, V.: The AROME-France convective-scale operational model, Mon.
Weather Rev., 139, 976–991, <ext-link xlink:href="http://dx.doi.org/10.1175/2010MWR3425.1" ext-link-type="DOI">10.1175/2010MWR3425.1</ext-link>,
2011.</mixed-citation></ref>
      <ref id="bib1.bibx31"><label>Strajnar et al.(2015)</label><mixed-citation>Strajnar, B., Žagar, N., and Berre, L.: Impact of new aircraft
observations Mode-S MRAR in a mesoscale NWP model, J. Geophys. Res.-Atmos.,
120, 3920–3938, <ext-link xlink:href="http://dx.doi.org/10.1002/2014JD022654" ext-link-type="DOI">10.1002/2014JD022654</ext-link>, 2015.</mixed-citation></ref>
      <ref id="bib1.bibx32"><label>Tonani et al.(2009)</label><mixed-citation>Tonani, M., Pinardi, N., Fratianni, C., Pistoia, J., Dobricic, S., Pensieri,
S., de Alfonso, M., and Nittis, K.: Mediterranean Forecasting System:
forecast and analysis assessment through skill scores, Ocean Sci., 5,
649–660, <ext-link xlink:href="http://dx.doi.org/10.5194/os-5-649-2009" ext-link-type="DOI">10.5194/os-5-649-2009</ext-link>, 2009.</mixed-citation></ref><?xmltex \hack{\newpage}?>
      <ref id="bib1.bibx33"><label>Tudor and Ivatek-Šahdan (2010)</label><mixed-citation>Tudor, M. and Ivatek-Šahdan, S.: The case study of bura of 1st and 3rd
February 2007, Meteorologische Zeitschrift, 19, 453–466,
<ext-link xlink:href="http://dx.doi.org/10.1127/0941-2948/2010/0475" ext-link-type="DOI">10.1127/0941-2948/2010/0475</ext-link>, 2010.</mixed-citation></ref>
      <ref id="bib1.bibx34"><label>Valcke(2013)</label><mixed-citation>Valcke, S.: The OASIS3 coupler: a European climate modelling community
software, Geosci. Model Dev., 6, 373–388, <ext-link xlink:href="http://dx.doi.org/10.5194/gmd-6-373-2013" ext-link-type="DOI">10.5194/gmd-6-373-2013</ext-link>, 2013.</mixed-citation></ref>
      <ref id="bib1.bibx35"><label>Zore-Armanda and Gačić(1987)</label><mixed-citation>
Zore-Armanda, M., and Gačić, M.: Effects of Bora on the circulation in the northern Adriatic, Ann. Geophys., 5, 93–102, 1987.</mixed-citation></ref>

  </ref-list><app-group content-type="float"><app><title/>

    </app></app-group></back>
    <!--<article-title-html>Modeling the ocean and atmosphere during an extreme bora event
in northern Adriatic using one-way and two-way atmosphere–ocean coupling</article-title-html>
<abstract-html><p class="p">We have studied the performances of (a) a two-way coupled atmosphere–ocean
modeling system and (b) one-way coupled ocean model (forced by the atmosphere
model), as compared to the available in situ measurements during and after a
strong Adriatic bora wind event in February 2012, which led to extreme
air–sea interactions. The simulations span the period between January and
March 2012. The models used were ALADIN (Aire Limitée Adaptation
dynamique Développement InterNational) (4.4 km resolution) on the
atmosphere side and an Adriatic setup of Princeton ocean model (POM)
(1°∕30 × 1°∕30 angular resolution) on the ocean side.
The atmosphere–ocean coupling was implemented using the OASIS3-MCT model
coupling toolkit. Two-way coupling ocean feedback to the atmosphere is
limited to sea surface temperature. We have compared modeled
atmosphere–ocean fluxes and sea temperatures from both setups to platform
and CTD (conductivity, temperature, and depth) measurements from three
locations in the northern Adriatic. We present objective verification of
2 m atmosphere temperature forecasts using mean bias and standard
deviation of errors scores from 23 meteorological stations in the eastern
part of Italy. We show that turbulent fluxes from both setups differ up to
20 % during the bora but not significantly before and after the event.
When compared to observations, two-way coupling ocean temperatures exhibit a
4 times lower root mean square error (RMSE) than those from one-way coupled
system. Two-way coupling improves sensible heat fluxes at all stations but
does not improve latent heat loss. The spatial average of the two-way coupled
atmosphere component is up to 0.3 °C colder than the one-way coupled
setup, which is an improvement for prognostic lead times up to 20 h. Daily
spatial average of the standard deviation of air temperature errors shows
0.15 °C improvement in the case of coupled system compared to the
uncoupled. Coupled and uncoupled circulations in the northern Adriatic are
predominantly wind-driven and show no significant mesoscale differences.</p></abstract-html>
<ref-html id="bib1.bib1"><label>Artegiani et al.(1997)</label><mixed-citation>
Artegiani, A., Bregant, D., Paschini, E., Pinardi, N., Raicich, F., and Russo, A.:
The Adriatic Sea General Circulation I: Air-Sea Interactions and Water Mass Structure, J. Phys. Oceanogr., 27, 1492–1514,
<a href="http://dx.doi.org/10.1175/1520-0485(1997)027&lt;1492:TASGCP&gt;2.0.CO;2" target="_blank">doi:10.1175/1520-0485(1997)027&lt;1492:TASGCP&gt;2.0.CO;2</a>,
1997.
</mixed-citation></ref-html>
<ref-html id="bib1.bib2"><label>Benetazzo et al.(2014)</label><mixed-citation>
Benetazzo, A., Bergamasco, A., Bonaldo, D., Falcieria, F. M., Sclavo, M.,
Langone, L., and Carniel, S.: Response of the Adriatic Sea to an intense cold air outbreak: dense water dynamics and wave-induced transport, Prog.
Oceanogr., 128, 115–138, <a href="http://dx.doi.org/10.1016/j.pocean.2014.08.015" target="_blank">doi:10.1016/j.pocean.2014.08.015</a>, 2014.
</mixed-citation></ref-html>
<ref-html id="bib1.bib3"><label>Blumberg and Mellor(1987)</label><mixed-citation>
Blumberg, A. F. and Mellor, G. L.: A description of a three-dimensional coastal ocean circulation model, in: Three-Dimensional Coastal Ocean Models,
edited by: Heaps, N., American Geophysical Union, Washington, DC, USA, 1–16,
1987.
</mixed-citation></ref-html>
<ref-html id="bib1.bib4"><label>Cosoli et al.(2013)</label><mixed-citation>
Cosoli, S., Ličer, M., Vodopivec, M., and Malačič, V.: Surface circulation in the Gulf of Trieste (northern Adriatic Sea) from radar, model,
and ADCP comparisons, J. Geophys. Res.-Oceans, 118, 6183–6200,
<a href="http://dx.doi.org/10.1002/2013JC009261" target="_blank">doi:10.1002/2013JC009261</a>, 2013.
</mixed-citation></ref-html>
<ref-html id="bib1.bib5"><label>Cushman-Roisin et al.(2001)</label><mixed-citation>
Cushman-Roisin, B., Gačić, M., Poulain, P.-M., and Artegiani, A.: Physical
Oceanography of the Adriatic Sea. Past, Present and Future,
Springer, London, UK, 47–48, 2001.
</mixed-citation></ref-html>
<ref-html id="bib1.bib6"><label>Davies(1976)</label><mixed-citation>
Davies, H. C.: A lateral boundary formulation for multi-level prediction models, Q. J. Roy. Meteor. Soc., 102, 405–418, 1976.
</mixed-citation></ref-html>
<ref-html id="bib1.bib7"><label>Donlon et al.(2012)</label><mixed-citation>
Donlon, C. J., Martin, M., Stark, J. D., Roberts-Jones, J., Fiedler, E., and Wimmer, W.: The Operational Sea Surface Temperature and Sea Ice Analysis
(OSTIA), Remote Sens. Environ., 116, 140–158,
<a href="http://dx.doi.org/10.1016/j.rse.2010.10.017" target="_blank">doi:10.1016/j.rse.2010.10.017</a>, 2012.
</mixed-citation></ref-html>
<ref-html id="bib1.bib8"><label>Dorman et al.(2006)</label><mixed-citation>
Dorman, C. E., Carniel, S., Cavaleri, L., Sclavo, M., Chiggiato, J.,
Doyle, J., Haack, T., Pullen, J., Grbec, B., Vilibić, I.,
Janeković, I., Lee, C., Malačič, V., Orlić, M., Paschini, E.,
Russo, A., and Signell, R. P.: February 2003 marine atmospheric conditions and the bora over the northern Adriatic, J. Geophys. Res., 111, C03S03,
<a href="http://dx.doi.org/10.1029/2005JC003134" target="_blank">doi:10.1029/2005JC003134</a>, 2006.
</mixed-citation></ref-html>
<ref-html id="bib1.bib9"><label>Eleuterio(1998)</label><mixed-citation>
Eleuterio, D.: A Comparison of Bulk Aerodynamic Methods for Calculating Air-Sea Flux, Master Thesis, Naval Postgraduate School, Monterey, CA, USA, available at:
<a href="http://www.researchgate.net/publication/235119736_A_Comparison_of_Bulk_Aerodynamic_Methods_for_Calculating_Air-Sea_Flux" target="_blank">doi:http://www.researchgate.net/publication/235119736_A_Comparison_of_Bulk_Aerodynamic_Methods_for_Calculating_Air-Sea_Flux</a> (last access: 14 October 2015), 1998.
</mixed-citation></ref-html>
<ref-html id="bib1.bib10"><label>Fairall et al.(1996)</label><mixed-citation>
Fairall, C. W., Bradley, E. F., Rogers, D. P., Edson, J. B., and Young, G. S.:
Bulk parameterization of air-sea fluxes for Tropical Ocean Global Atmosphere
Coupled-Ocean Atmosphere Response Experiment, J. Geophys. Res., 101,
3747–3764, <a href="http://dx.doi.org/10.1029/95JC03205" target="_blank">doi:10.1029/95JC03205</a>, 1996.
</mixed-citation></ref-html>
<ref-html id="bib1.bib11"><label>Fischer et al.(2005)</label><mixed-citation>
Fischer, C., Montmerle, T., Berre, L., Auger, L., and
Ştefănescu, S. E.: An overview of the variational assimilation in the
ALADIN/France numerical weather-prediction system, Q. J. Roy. Meteor. Soc.,
131, 3477–3492, <a href="http://dx.doi.org/10.1256/qj.05.115" target="_blank">doi:10.1256/qj.05.115</a>, 2005.
</mixed-citation></ref-html>
<ref-html id="bib1.bib12"><label>Gerard et al.(2009)</label><mixed-citation>
Gerard, L., Piriou, J.-M., Brožková, R., Geleyn, J.-F., and
Banciu, D.: Cloud and precipitation parameterization in a meso-gamma-scale
operational weather prediction model, Mon. Weather Rev., 137, 3960–3977,
<a href="http://dx.doi.org/10.1175/2009MWR2750.1" target="_blank">doi:10.1175/2009MWR2750.1</a>, 2009.
</mixed-citation></ref-html>
<ref-html id="bib1.bib13"><label>Grisogono and Belušić(2009)</label><mixed-citation>
Grisogono, B. and Belušić, D.: A review of recent advances in understanding the meso- and microscale properties of severe Bora wind, Tellus, 61A, 1–16,
<a href="http://dx.doi.org/10.1111/j.1600-0870.2008.00369.x" target="_blank">doi:10.1111/j.1600-0870.2008.00369.x</a>, 2009.
</mixed-citation></ref-html>
<ref-html id="bib1.bib14"><label>Janeković et al.(2014)</label><mixed-citation>
Janeković, I., Mihanović, H., Vilibić, I., and Tudor, M.: Extreme
cooling and dense water formation estimates in open and coastal regions of
the Adriatic Sea during the winter of 2012, J. Geophys. Res.-Oceans, 119,
3200–3218, <a href="http://dx.doi.org/10.1002/2014JC009865" target="_blank">doi:10.1002/2014JC009865</a>, 2014.
</mixed-citation></ref-html>
<ref-html id="bib1.bib15"><label>Kourafalou(2001)</label><mixed-citation>
Kourafalou, V. H.: River plume development in semi-enclosed Mediterranean regions: north Adriatic Sea and northwestern Aegean Sea, J. Marine Syst., 30,
181–205, <a href="http://dx.doi.org/10.1016/S0924-7963(01)00058-6" target="_blank">doi:10.1016/S0924-7963(01)00058-6</a>, 2001.
</mixed-citation></ref-html>
<ref-html id="bib1.bib16"><label>Kuzmić et al.(2007)</label><mixed-citation>
Kuzmić, M., Janeković, I., Book, J. W., Martin, P. J., and
Doyle, J. D.: Modeling the northern Adriatic double-gyre response to intense
bora wind: a revisit, J. Geophys. Res., 111, C03S13,
<a href="http://dx.doi.org/10.1029/2005JC003377" target="_blank">doi:10.1029/2005JC003377</a>, 2007.
</mixed-citation></ref-html>
<ref-html id="bib1.bib17"><label>Malačič and Petelin(2009)</label><mixed-citation>
Malačič, V. and Petelin, B.: Climatic circulation in the Gulf of Trieste (northern Adriatic), J. Geophys. Res., 114, C07002,
<a href="http://dx.doi.org/10.1029/2008JC004904" target="_blank">doi:10.1029/2008JC004904</a>, 2009.
</mixed-citation></ref-html>
<ref-html id="bib1.bib18"><label>Malačič et al.(2012)</label><mixed-citation>
Malačič, V., Petelin, B., and Vodopivec, M.: Topographic control of wind-driven circulation in the northern Adriatic, J. Geophys. Res., 117,
C06032, <a href="http://dx.doi.org/10.1029/2012JC008063" target="_blank">doi:10.1029/2012JC008063</a>, 2012.
</mixed-citation></ref-html>
<ref-html id="bib1.bib19"><label>Mantziafou and Lascaratos(2008)</label><mixed-citation>
Mantziafou, A. and Lascaratos, A.: Deep-water formation in the Adriatic Sea:
interannual simulations for the years 1979–1999, Deep-Sea Res. Pt. I, 55,
1403–1427, <a href="http://dx.doi.org/10.1016/j.dsr.2008.06.005" target="_blank">doi:10.1016/j.dsr.2008.06.005</a>, 2008.
</mixed-citation></ref-html>
<ref-html id="bib1.bib20"><label>Mascart et al.(1995)</label><mixed-citation>
Mascart, P., Noilhan, J., and Giordani, H.: A modified parametrization of
flux-profile relationships in the surface layer using different roughness
length values for heat and momentum, Bound.-Lay. Meteorol., 72, 331–344,
<a href="http://dx.doi.org/10.1007/BF00708998" target="_blank">doi:10.1007/BF00708998</a>, 1995.
</mixed-citation></ref-html>
<ref-html id="bib1.bib21"><label>Mellor(1998)</label><mixed-citation>
Mellor, G.: Users Guide for a Three-Dimensional, Primitive Equation,
Numerical Ocean Model, Tech. Rep., Princeton University, Princeton, NJ, USA, 1998.
</mixed-citation></ref-html>
<ref-html id="bib1.bib22"><label>Mihanović et al.(2013)</label><mixed-citation>
Mihanović, H., Vilibić, I., Carniel, S., Tudor, M., Russo, A.,
Bergamasco, A., Bubić, N., Ljubešić, Z., Viličić, D.,
Boldrin, A., Malačič, V., Celio, M., Comici, C., and Raicich, F.:
Exceptional dense water formation on the Adriatic shelf in the winter of 2012, Ocean Sci., 111, 561–572, <a href="http://dx.doi.org/10.5194/os-9-561-2013" target="_blank">doi:10.5194/os-9-561-2013</a>, 2013.
</mixed-citation></ref-html>
<ref-html id="bib1.bib23"><label>Nielsen and Hansen(1973)</label><mixed-citation>
Nielsen, S. and Hansen, E.: Numerical simulation of the rainfall-runoff process on a daily basis, Nord. Hydrol., 4, 171–190, 1973.
</mixed-citation></ref-html>
<ref-html id="bib1.bib24"><label>Paklar et al.(2001)</label><mixed-citation>
Paklar, G. B., Isakov, V., Koračin, D., Kourafalou, V., and
Orlić, M.: A case study of bora-driven flow and density changes on the
Adriatic Shelf (January 1987), Cont. Shelf. Res., 21, 1751–1783,
<a href="http://dx.doi.org/10.1016/S0278-4343(01)00029-2" target="_blank">doi:10.1016/S0278-4343(01)00029-2</a>, 2001.
</mixed-citation></ref-html>
<ref-html id="bib1.bib25"><label>Pullen et al.(2003)</label><mixed-citation>
Pullen, J., Doyle, J., Hodur, R., Ogston, A., Book, J., Perkins, H., and
Signell, R.: Coupled ocean-atmosphere nested modeling of the Adriatic Sea
during winter and spring 2001, J. Geophys. Res., 108, 3320,
<a href="http://dx.doi.org/10.1029/2003JC001780" target="_blank">doi:10.1029/2003JC001780</a>, 2003.
</mixed-citation></ref-html>
<ref-html id="bib1.bib26"><label>Pullen et al.(2006)</label><mixed-citation>
Pullen, J., Doyle, J., and Signell, R.: Two-way air–sea coupling: a study of the Adriatic, Mon. Weather Rev., 134, 1465–1483, <a href="http://dx.doi.org/10.1175/MWR3137.1" target="_blank">doi:10.1175/MWR3137.1</a>,
2006.
</mixed-citation></ref-html>
<ref-html id="bib1.bib27"><label>Pullen et al.(2007)</label><mixed-citation>
Pullen, J., Doyle, J., Haack, T., Dorman, C., Signell, R., and Lee, C. M.:
Bora event variability and the role of air-sea feedback, J. Geophys. Res.,
112, 33407, <a href="http://dx.doi.org/10.1029/2006JC003726" target="_blank">doi:10.1029/2006JC003726</a>, 2007.
</mixed-citation></ref-html>
<ref-html id="bib1.bib28"><label>Raicich(1994)</label><mixed-citation>
Raicich, F.: Notes on the Flow Rates of the Adriatic Rivers, tech. rep., CNR,
Ist. Sper. Talassografico, Trieste, Italy, 1994.
</mixed-citation></ref-html>
<ref-html id="bib1.bib29"><label>Raicich et al.(2013)</label><mixed-citation>
Raicich, F., Malačič, C., Celio, M., Giaiotti, D., Cantoni, C.,
Colucci, R. R., Čermelj, B., and Pucillo, A.: Extreme air-sea
interactions in the Gulf of Trieste (North Adriatic) during the strong Bora
event in winter 2012, J. Geophys. Res.-Oceans, 118, 5238–5250,
<a href="http://dx.doi.org/10.1002/jgrc.20398" target="_blank">doi:10.1002/jgrc.20398</a>, 2013.
</mixed-citation></ref-html>
<ref-html id="bib1.bib30"><label>Seity et al.(2011)</label><mixed-citation>
Seity, Y., Brousseau, P.,Malardel, S., Hello, G., Benard, P., Bouttier, F.,
Lac, C., and Masson, V.: The AROME-France convective-scale operational model, Mon.
Weather Rev., 139, 976–991, <a href="http://dx.doi.org/10.1175/2010MWR3425.1" target="_blank">doi:10.1175/2010MWR3425.1</a>,
2011.
</mixed-citation></ref-html>
<ref-html id="bib1.bib31"><label>Strajnar et al.(2015)</label><mixed-citation>
Strajnar, B., Žagar, N., and Berre, L.: Impact of new aircraft
observations Mode-S MRAR in a mesoscale NWP model, J. Geophys. Res.-Atmos.,
120, 3920–3938, <a href="http://dx.doi.org/10.1002/2014JD022654" target="_blank">doi:10.1002/2014JD022654</a>, 2015.
</mixed-citation></ref-html>
<ref-html id="bib1.bib32"><label>Tonani et al.(2009)</label><mixed-citation>
Tonani, M., Pinardi, N., Fratianni, C., Pistoia, J., Dobricic, S., Pensieri,
S., de Alfonso, M., and Nittis, K.: Mediterranean Forecasting System:
forecast and analysis assessment through skill scores, Ocean Sci., 5,
649–660, <a href="http://dx.doi.org/10.5194/os-5-649-2009" target="_blank">doi:10.5194/os-5-649-2009</a>, 2009.
</mixed-citation></ref-html>
<ref-html id="bib1.bib33"><label>Tudor and Ivatek-Šahdan (2010)</label><mixed-citation>
Tudor, M. and Ivatek-Šahdan, S.: The case study of bura of 1st and 3rd
February 2007, Meteorologische Zeitschrift, 19, 453–466,
<a href="http://dx.doi.org/10.1127/0941-2948/2010/0475" target="_blank">doi:10.1127/0941-2948/2010/0475</a>, 2010.
</mixed-citation></ref-html>
<ref-html id="bib1.bib34"><label>Valcke(2013)</label><mixed-citation>
Valcke, S.: The OASIS3 coupler: a European climate modelling community
software, Geosci. Model Dev., 6, 373–388, <a href="http://dx.doi.org/10.5194/gmd-6-373-2013" target="_blank">doi:10.5194/gmd-6-373-2013</a>, 2013.
</mixed-citation></ref-html>
<ref-html id="bib1.bib35"><label>Zore-Armanda and Gačić(1987)</label><mixed-citation>
Zore-Armanda, M., and Gačić, M.: Effects of Bora on the circulation in the northern Adriatic, Ann. Geophys., 5, 93–102, 1987.
</mixed-citation></ref-html>--></article>
