The Theia workshop for Sentinel-2 L2A MAJA products was held in Toulouse on the 13th and 14th of June 2018.  
Attendance on the 13th of June
 About 80 people participated either on the 13th or 14th, and nearly 70 participants attended each day of  this workshop,  whose object was to collect feedback and share experiences on the quality, use and applications of the L2A surface reflectance products delivered by Theia from Sentinel-2 data.

Slides

Here are the links to read and download the presentations. The organisers are very grateful for the hard work presented in all these documents.

SESSION 1 – DATA / PRODUCTS

SESSION 2 -PLATFORMS

SESSION 3 –AGROSYSTEMS

SESSION 4 – LAND COVER

SESSION 5 – FOREST

SESSION 6 – WATER

Posters

(We have not received yet all the posters, we will add them as soon as they arrive)

Conclusions

Regarding Sentinels

  • All the community is very grateful to the Copernicus program of the European Union and ESA for providing such great and useful data as those provided by the sentinels, and especially, of course, Sentinel-2.

Regarding MAJA

  • MAJA L2A products are used for very diverse applications, although MAJA is more adapted to processing over land than over water. Implementing a specific atmospheric correction above water is therefore relevant (J.M.Martinez).
  • Using MAJA L2A products generally performs better than L1C and Sen2cor L2A, as shown in the slides from C.Desjardins, I.Sola, R.Van Hoolst, J.Inglada or R.Gaetano
  • MAJA still needs improvements: misses small clouds (R. Van Hoolst), has negative values (O.Hagolle, J.M.Martinez), and shows insufficiently accurate results above very uniform landscapes such as equatorial forests (J.B.Feret)
  • Given the advanced results obtained by MAJA regarding cloud detection, some users would be happy to use the cloud mask independently from the atmospheric correction. This will be implemented end of 2018 within MAJA NG, which will be split into modules, each one formatted as an OTB application, and wrapped with a python main program.
  • The L3 composite product (monthly synthesis of cloud free reflectance, poster from P.Kettig) prepared by Theia, will be a useful product to monitor targets with a slow evolution, such as forests.

Regarding Theia

  • Many users requested to extend the zones processed by Theia with MAJA
  • Several users asked for an improvement of the real time production by Theia, which is not fully robust and reliable yet.
  • Several users would like to access to Theia data in cloud platforms, such as DIAS or other platforms
  • Distribution of products under a Cloud Optimized Geotiff (COGEO) format would be appreciated, to access to part of images or specific bands
  • Theia initiative to distribute high quality products from Sentinel data is important and useful but lacks funding. The processed zones are still small and scattered, and the operational implementation of the various products developed under the CES initiative is too slow, because of the lack of financial support to help the research teams transform their breadboard and demonstrations into operational processors.

Attendance on the 13th of June afternoon (Gérard was probably taking notes about the presentation going on)

A tired but happy organising committee

One thought on “Slides and conclusions of the MAJA/Theia worskhop

Laisser un commentaire

Votre adresse e-mail ne sera pas publiée. Les champs obligatoires sont indiqués avec *

Ce site utilise Akismet pour réduire les indésirables. En savoir plus sur comment les données de vos commentaires sont utilisées.