<?xml version="1.0" encoding="UTF-8"?><rss xmlns:dc="http://purl.org/dc/elements/1.1/" xmlns:content="http://purl.org/rss/1.0/modules/content/" xmlns:atom="http://www.w3.org/2005/Atom" version="2.0"><channel><title><![CDATA[Radar nowcasting model]]></title><description><![CDATA[<p dir="auto">Re: <a href="/topic/31500/radar-nowcasting">Radar Nowcasting</a><br />
Hello,<br />
I wonder which method is used for the radar nowcasting. Is it an open-source model ?<br />
Does it rely solely on "black-box" machine learning, or does it contain an physical equation such as a kind of "advection equation" ?<br />
Thank you,<br />
Mathieu</p>
]]></description><link>https://community.windy.com/topic/43559/radar-nowcasting-model</link><generator>RSS for Node</generator><lastBuildDate>Sun, 19 Apr 2026 18:52:59 GMT</lastBuildDate><atom:link href="https://community.windy.com/topic/43559.rss" rel="self" type="application/rss+xml"/><pubDate>Wed, 11 Feb 2026 11:18:29 GMT</pubDate><ttl>60</ttl><item><title><![CDATA[Reply to Radar nowcasting model on Thu, 26 Feb 2026 07:50:44 GMT]]></title><description><![CDATA[<p dir="auto">Thanks for the clarification, that makes sense and it is interesting that you keep radar to simple extrapolation. Have you tested any ML based approaches for short term convective development, or do you find optical flow sufficient for your typical forecast horizons?</p>
]]></description><link>https://community.windy.com/post/225175</link><guid isPermaLink="true">https://community.windy.com/post/225175</guid><dc:creator><![CDATA[grayon99]]></dc:creator><pubDate>Thu, 26 Feb 2026 07:50:44 GMT</pubDate></item><item><title><![CDATA[Reply to Radar nowcasting model on Thu, 12 Feb 2026 08:23:27 GMT]]></title><description><![CDATA[<p dir="auto"><a class="plugin-mentions-user plugin-mentions-a" href="/user/mathieu-landreau" aria-label="Profile: Mathieu-Landreau">@<bdi>Mathieu-Landreau</bdi></a> Hi, we use  <a href="https://en.wikipedia.org/wiki/Lucas%E2%80%93Kanade_method" rel="nofollow ugc">sparse Lucas Kanade optical flow</a> with <a href="https://en.wikipedia.org/wiki/Inverse_distance_weighting" rel="nofollow ugc">Inverse distance weighting</a> interpolation, similarly but not the same as <a href="https://pysteps.readthedocs.io/en/stable/generated/pysteps.motion.lucaskanade.dense_lucaskanade.html#pysteps.motion.lucaskanade.dense_lucaskanade" rel="nofollow ugc">pySTEPS dense_lucaskanade() method</a> or <a href="https://github.com/hydrogo/rainymotion/blob/master/rainymotion/models.py#L310" rel="nofollow ugc">rainymotion SparseSD model</a></p>
<p dir="auto">For satellite advection we use <a href="https://en.wikipedia.org/wiki/Semi-Lagrangian_scheme" rel="nofollow ugc">Semi Lagrangian scheme</a>, for radar we use just simple extrapolation</p>
]]></description><link>https://community.windy.com/post/224782</link><guid isPermaLink="true">https://community.windy.com/post/224782</guid><dc:creator><![CDATA[Suty]]></dc:creator><pubDate>Thu, 12 Feb 2026 08:23:27 GMT</pubDate></item></channel></rss>