Skip to main content


Sunspots and aurora

Last weekend's polar lights have been witnessed by millions around the world. While many drove to dark places to watch this celestial spectacle, others grabbed a lawn chair and watched the baffling show from their own backyard. Children were woken up by their parents to watch the colours in the sky. The 10-11 May aurora certainly captured the attention and awe of the public and the media.

SILSO change of management

Due to a reshuffling of responsibilities, SILSO will from now on run under the single Directorship of Laure Lefèvre, and therefore any future questions related to the SILSO management or data should be addressed to  We  take this opportunity to thank both Frédéric Clette and Laure Lefèvre for their dedication to the SILSO operations during many years.

New extended hemispheric sunspot numbers

On July 19, 2021, we added a new time series to our data products: extended hemispheric sunspot numbers starting in 1874. Those reconstructed numbers are based on the sunspot areas listed in the Greenwich photographic catalogue, and are calibrated on the SILSO reference hemispheric series, which starts only in 1992. This newly published series results from a collaboration between SILSO, the University of Graz (Austria) and the Skolkovo Institute of Science and Technology (Moscow, Russia).

New prediction method

Starting from a collaboration with the NCEI (NOAA, Boulder USA), we implemented new 12-month ahead predictions based on the McNish and Lincoln method. This rather simple method is based on a single mean cycle profile and is thus of "climatology" type. It was used as a standard for many years at NOAA, and we now add it to our other more advanced Standard Curves and Combined methods, allowing direct comparisons. Likewise, we now also provide a Kalman-filter optimized version of these new ML predictions.

New 'Spotless days' page

Regular solar observers have noticed that since mid-2016, the Sun has occasionally been devoid of sunspots. As the current solar cycle 24 will gradually give way to the new solar cycle 25, several consecutive days and even weeks without sunspots will become the norm. In order to have an idea on the number of spotless days, and how these numbers compare to past solar cycles, SILSO has created a “Spotless Days page”.

Transition to the new Sunspot Number successfully completed

Today marked a triple transition for us: - Uploading the new Sunspot Number archive files containing the daily, monthly and yearly re-calibrated sunspot numbers and the new Group Number series - In our Web site, switching to the new "Data" pages giving access to the new files, to updated graphics and also to the past version of the Sunspot Number - Adapting and running the entire monthly procedure to produce the provisional Sunspot Numbers for June 2015 and the associated 12-months forecast and EISN. Thus a lot of work in a single day for our small team.

New Data pages with archive of past versions

After completing the new data files themselves, we are now finalizing the design of the new sections in our data pages. Indeed, although the new Sunspot Number will bring large improvements to the most prominent defects of the original series, it still contains more subtle or local inhomogeneities that will require more work for years to come. Therefore, in order to maintain a full history of present and future changes, we will now attach a new version number to each modification of the Sunspot Number time series.

Advancing towards the July 1st transition: no change for our observers

The preparation of the July 1st transition is progressing at high pace. Hundreds of lines of codes are adapted for the new sunspot number series and for the new array of data files. The corrections can be subtle changes in hard-coded formulae for our two 12-month forecast methods, as well as large-scale modifications in our data pipeline controlling the flux of new data to route them to the proper archive file and database. There are so many diverse changes that we cannot guarantee that everything will work perfectly on the first try. Our team is too small to make full prior simulations.