ECMWF moves to open data

<div> <p><span>ECMWF is committed to move to an open data policy gradually over the next few years. ECMWF has already released </span><span>hundreds </span><span>of </span><span>web forecast charts and made archived data available with a Creative Commons </span><span>(</span><span>CC</span><span> </span><span>BY</span><span> </span><span>4.0</span><span>)</span><span> open licence</span><span> in 2020</span><span>.</span><span> </span><span>The potential uses and benefits these products bring for a range of users and sectors is vast and particularly key in less economically developed countries</span><span> and</span><span> has the potential to supercharge research efforts </span><span>leading to </span><span>improv</span><span>ement in</span><span> weather </span><span>predictions</span><span> </span><span>and </span><span>delivering important </span><span>socio-economic benefits</span><span>.</span><span> </span><span> </span></p> </div><div> <p><span>Making these products </span><span>available </span><span>with CC</span><span> </span><span>BY</span><span> </span><span>4.0 </span><span>licence</span><span>s</span><span> </span><span>means that users can </span><span>now </span><span>share, redistribute and adapt the information as they </span><span>need</span><span>, </span><span>including</span><span> for commercial applications, </span><span>as long as</span><span> they acknowledge </span><span>ECMWF as </span><span>the source</span><span>.</span><span> </span><span>Archived </span><span>d</span><span>ata from all past ECMWF forecasts offer</span><span> immense opportunities for machine learning, where a computer uses observations or other data, to ‘learn’ relationships between different variables</span><span> and will support ECMWF Machine Learning Roadmap</span><span> to 2030</span><span>. </span><span> </span></p> </div><div> <p><span>The next steps </span><span>will be presented and </span><span>will </span><span>involve an expansion of the free and open datasets already available to increase their use </span><span>by targeted audiences </span><span>in </span><span>real-time. </span><span>ECMWF </span><span>recognizes</span><span> that giving free access to all data holdings in real-time will be </span><span>technically </span><span>challenging</span><span> </span><span>in the short and medium term. </span><span>The plan is to maintain paid-for delivery services for demanding applications </span><span>that require access to large volumes of data</span><span> with </span><span>a service level agreement</span><span>. </span><span>This phased move towards free and open data aims to support creativity and innovation in the field of scientific research as well as weather applications. </span><span> </span></p> </div><div> <p><span>Furthermore, to facilitate the efficient access and processing of such large volumes of data, </span><span>ECMWF is piloting c</span><span>loud solutions for</span><span> organisations of</span><span> </span><span>Member States in collaboration with EUMETSAT and is investigating</span><span> fast connections with commercial public clouds.</span></p> </div>