<?xml version="1.0" encoding="UTF-8"?>
<!DOCTYPE article PUBLIC "-//NLM//DTD Journal Publishing DTD v3.0 20080202//EN" "https://jats.nlm.nih.gov/nlm-dtd/publishing/3.0/journalpublishing3.dtd">
<article xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink" article-type="research-article" specific-use="SMUR" dtd-version="3.0" xml:lang="en">
<front>
<journal-meta>
<journal-id journal-id-type="publisher">EGUsphere</journal-id>
<journal-title-group>
<journal-title>EGUsphere</journal-title>
<abbrev-journal-title abbrev-type="publisher">EGUsphere</abbrev-journal-title>
<abbrev-journal-title abbrev-type="nlm-ta">EGUsphere</abbrev-journal-title>
</journal-title-group>
<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/egusphere-2026-6594</article-id>
<title-group>
<article-title>A physically constrained framework for reconstructing high-resolution solar irradiance from accumulated radiation in machine-learning weather forecasts</article-title>
</title-group>
<contrib-group><contrib contrib-type="author" xlink:type="simple"><name name-style="western"><surname>Dai</surname>
<given-names>Zekai</given-names>
</name>
<xref ref-type="aff" rid="aff1">
<sup>1</sup>
</xref>
</contrib>
<contrib contrib-type="author" xlink:type="simple"><name name-style="western"><surname>Wang</surname>
<given-names>Man</given-names>
</name>
<xref ref-type="aff" rid="aff2">
<sup>2</sup>
</xref>
</contrib>
<contrib contrib-type="author" xlink:type="simple"><name name-style="western"><surname>Wang</surname>
<given-names>Yongwei</given-names>
</name>
<xref ref-type="aff" rid="aff1">
<sup>1</sup>
</xref>
<xref ref-type="aff" rid="aff3">
<sup>3</sup>
</xref>
</contrib>
<contrib contrib-type="author" xlink:type="simple"><name name-style="western"><surname>Yang</surname>
<given-names>Yuanjian</given-names>
</name>
<xref ref-type="aff" rid="aff1">
<sup>1</sup>
</xref>
<xref ref-type="aff" rid="aff3">
<sup>3</sup>
</xref>
</contrib>
<contrib contrib-type="author" xlink:type="simple"><name name-style="western"><surname>Li</surname>
<given-names>Xindin</given-names>
</name>
<xref ref-type="aff" rid="aff1">
<sup>1</sup>
</xref>
</contrib>
<contrib contrib-type="author" xlink:type="simple"><name name-style="western"><surname>Jiang</surname>
<given-names>Xiaoying</given-names>
</name>
<xref ref-type="aff" rid="aff1">
<sup>1</sup>
</xref>
</contrib>
<contrib contrib-type="author" xlink:type="simple"><name name-style="western"><surname>Wang</surname>
<given-names>Yajun</given-names>
</name>
<xref ref-type="aff" rid="aff1">
<sup>1</sup>
</xref>
</contrib>
<contrib contrib-type="author" xlink:type="simple"><name name-style="western"><surname>Ji</surname>
<given-names>Jiangwei</given-names>
</name>
<xref ref-type="aff" rid="aff1">
<sup>1</sup>
</xref>
</contrib>
</contrib-group><aff id="aff1">
<label>1</label>
<addr-line>School of Atmospheric Physics, Nanjing University of Information Science and Technology, Nanjing, China</addr-line>
</aff>
<aff id="aff2">
<label>2</label>
<addr-line>Yunnan Institute of Meteorological Sciences, Kunming, Yunnan 650034, China</addr-line>
</aff>
<aff id="aff3">
<label>3</label>
<addr-line>China Meteorological Administration Xiong’an Atmospheric Boundary Layer Key Laboratory, Xiong’an New Area, Hebei 071700, China</addr-line>
</aff>
<pub-date pub-type="epub">
<day>16</day>
<month>11</month>
<year>2026</year>
</pub-date>
<volume>2026</volume>
<fpage>1</fpage>
<lpage>40</lpage>
<permissions>
<copyright-statement>Copyright: &#x000a9; 2026 Zekai Dai et al.</copyright-statement>
<copyright-year>2026</copyright-year>
<license license-type="open-access">
<license-p>This work is licensed under the Creative Commons Attribution 4.0 International License. To view a copy of this licence, visit <ext-link ext-link-type="uri"  xlink:href="https://creativecommons.org/licenses/by/4.0/">https://creativecommons.org/licenses/by/4.0/</ext-link></license-p>
</license>
</permissions>
<self-uri xlink:href="https://egusphere.copernicus.org/preprints/2026/egusphere-2026-6594/">This article is available from https://egusphere.copernicus.org/preprints/2026/egusphere-2026-6594/</self-uri>
<self-uri xlink:href="https://egusphere.copernicus.org/preprints/2026/egusphere-2026-6594/egusphere-2026-6594.pdf">The full text article is available as a PDF file from https://egusphere.copernicus.org/preprints/2026/egusphere-2026-6594/egusphere-2026-6594.pdf</self-uri>
<abstract>
<p>Machine-learning weather prediction (MLWP) supports global forecasting, but solar radiation is often provided on coarse grids as multihour accumulations, creating a scale mismatch with local applications. We present SolarFINE, a physically constrained hierarchical framework that converts 0.25&amp;deg; six-hour accumulated radiation and concurrent meteorological states to surface solar irradiance at 0.05&amp;deg; and 10-min resolution. It reconstructs hourly coarse-grid irradiance from solar geometry and meteorological states under a six-hour integral-consistency constraint, then constructs an hourly high-resolution spatial background and generates subhourly variability using conditional diffusion. The temporal stage was trained on ERA5, the high-resolution stage was supervised with a satellite radiation product, and SolarFINE-T was adapted to Artificial Intelligence Forecasting System (AIFS) inputs. Evaluation covered hourly reconstruction, high-resolution spatial structure, and AIFS-driven end-to-end station performance. In controlled ERA5 experiments, SolarFINE-T achieved a daytime hourly root-mean-square error (RMSE) of 40.37 W m&lt;sup&gt;&amp;minus;2&lt;/sup&gt;, 29.9 % lower than with solar-geometry allocation. At 0.05&amp;deg; and 10-min resolution, SolarFINE-J achieved an RMSE of 132.42 W m&lt;sup&gt;&amp;minus;2&lt;/sup&gt; and more closely reproduced the spatial power spectrum, spatial variability, and high-irradiance tail of the high-resolution reference product. In the independent station evaluation, RMSE skill scores relative to solar-geometry allocation were positive on forecast days 1&amp;ndash;15, rising from 0.084 to 0.174. Improvements were largest and most consistent at intermediate clear-sky index. Adapting to AIFS inputs improved skill, whereas forcing hourly fields to match the original six-hour accumulation reduced it. The results indicate that MLWP-driven solar-irradiance downscaling requires exploiting atmospheric information beyond solar geometry while accounting for upstream forecast errors.</p>
</abstract>
<counts><page-count count="40"/></counts>
</article-meta>
</front>
<body/>
<back>
</back>
</article>