Sentinel-HR, a metric resolution optical mission with global systematic and frequent observation, and free and open data

=>

Every four years, CNES brings together its scientific advising committees (and I have never been invited :( )  for its scientific prospective seminar. This seminar is where the scientific priorities for CNES's next space missions are defined. A few months before, a call for ideas for new missions is issued, and it was the opportunity to submit the idea of a mission called SENTINEL-HR (following Gérard Dedieu who had already proposed it two years ago).

 

Sentinel-HR mission

This mission would make it possible to observe all the lands at a metric resolution once a season. Taking into account the clouds, it would require a systematic revisit frequency of 20 to 30 days. As the mission name suggests, access to data would be free and open. The minimum configuration would have four bands (blue, green, red, near-infra-red), with optional bands in the short wave infra-red and the possibility of making stereoscopic observations.

2018 Image cover of France at a high resolution (6m), acquired by SPOT-6/7, for Theia/GeoSud project (1). Data were acquired between the 6 of april and the 11th of October .

 

As I already presented numerous tiles and for many years,  the benefit of Sentinel-2 data, compared to the SPOT missions of the past, is the ability to make frequent and systematic acquisitions, coupled with the free and open availability of data. This clearly explains the success of the SENTINEL-2 mission, which brought the number of users each year from a few thousands for SPOT to tens of thousands for SENTINEL-2.

Feasibility

Even if the feasibility studies did not start yet, the minimal mission really looks within reach: for instance, the SPOT6/7 satellites, launched in 2013, are able to acquire 2m resolution images for a field of view of 60 km.  SPOT1 had been designed to acquire all lands in 26 days, at Nadir, with a 120 km field of view (two fields of view of 60km). In 26 days, a satellite in SPOT's orbit makes 369 revolutions, which gives a total field of view of 369*120=44280 km,  10% more than the length of equator. .

 

Two satellites with a 60 km field of view would do the same.

Observation of the thinning of Mt Cook (Kerguelen) Ice cap (Favier, nature scientific reports, 2009, (2) thanks to SRTM et stereoscopic images from SPOT5, in m/yr, between 2000 and 2009. Systematic stereoscopic observations will make it possible to regularly monitor all the planet's glaciers.

Of course, there are probably other smart solutions which will need to be studied, either with more satellites, or with a larger field of view.

 

The most complicated point for Sentinel-HR would probably be data throughput. Currently, SPOT6/7 only acquire a few minutes of orbital shooting, whereas it would take up to 20 minutes for Sentinel-2. Setting up a stereoscopic shot with an angle difference of about 10 degrees would not be easy either, but our instrument designers will probably have good ideas.

Applications

Hedgerow detection (in red) in Britanny, from Kermap company (3), using high resolution optical images.

As for Sentinel-2, the applications would be very diverse.

 

 

Here are a few of them :

  • monitoring of artificialization and urban density, water masses, glaciers, morphology of rivers, forests, management of natural parks.
  • support for agricultural declarations or their control

    Monitoring of coastline and sandbanks at the Arcachon basin exit, by bathymetric inversion of SPOT5 images (Dulac, 2013 (4) Thanks to a special concentartion of SPOT5 image acquisitions on that zone (thanks to Kalideos project), it was possible to get an accurate evaluation of the movement amplitude in meters over two years. With the Sentinel-HR mission, it would be possible to generalize these studies on all the world's coasts.

  • monitoring of the tree elements (green frames, hedges ...) of the landscape
  • contribution of texture in the identification of land cover has been demonstrated
  • providing texture information to estimate biophysical parameters of forests such as biomass, age, or height
  • coastline and shallow bathymetry
  • for post-disaster damage assessments, this mission would provide assurance of having a high-resolution, recent reference image

 

If you are interested by the Sentinel-HR mission, please feel free to tell us, by commenting this post. We are interested in your reasons, and we will need application examples to promote this idea. And maybe, if you do it well, i'll get invited to CNES's prospective seminar ;) .

 

Of course it will be a long time until we get such a mission,  probably too expensive for CNES alone. But submitting this proposal will maybe allow to start studying it and make the idea progress.

 

Co-proposers

The proposal was submitted with several co-proposers :

  • Gérard Dedieu (CESBIO/CNES)
  • Anne Puissant (LIVE/CNRS)
  • Pierre Morel (TETIS/IRSTEA)
  • Etienne Berthier (LEGOS/CNRS)

 

 

Image references

[1] Geosud: <a href="http://ids.equipex-geosud.fr/">http://ids.equipex-geosud.fr/</a>

[2] <a href="https://www.nature.com/articles/srep32396">Favier, Vincent, D. Verfaillie, E. Berthier, M. Menegoz, V. Jomelli, J. E. Kay, L. Ducret et al. "Atmospheric drying as the main driver of dramatic glacier wastage in the southern Indian Ocean." <em>Scientific reports</em> 6 (2016): 32396.</a>

[Extraction of hedgerows by Kermap company: <a href="http://www.kermap.com/bocage/">http://www.kermap.com/bocage/</a>

[4] <a href="http://www.gis-cooc.org/images/stories/coocimage/presentations/Atelier_2013/gis_cooc_atelier%202013_bertrand_lubac.pdf">coastline monitoring </a>

 

 

 

 

 

 

 

 

 

 

 

 

Posted under: In English

2 comments

  • Clement ATZBERGER on 31/10/2018 at 20:17 said:

    Dear colleagues. Such a HR (metric) mission would indeed be great. Perhaps coupled with a low resolution (< 100m) thermal sensor in addition ?

    • Olivier Hagolle on 05/11/2018 at 08:22 said:

      Thanks Clement !
      I fully agree with the interest of TIR measurements. But there are already several projects of TIR sentinels (such as Trishna,CNES+ISRO) and LSTM (ESA). They aim a 50-60m resolution and a 1-3 days revisit. On Sentinel-HR, we would have at best a revisit of 20 days, which is probably not enough given the fast changes of earth temperatures. If we ask too much from Sentinel-HR, I am afraid it might be too expensive...
      Best regards,
      Olivier

Leave a Reply

Your email address will not be published. Required fields are marked *

*

You may use these HTML tags and attributes: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <strike> <strong>