A new year of time series


Dear readers of this blog, all my best wishes for 2014 !


May this new year bring you nice image time series, but spare you some time to process those recorded last year (among which SPOT4 (Take5) ;) ). Even if 2014 will be another year without Sentinel-2 and VENµS, let's guess that with LANDSAT-8 et SPOT4 (Take5),  the data sets produced by THEIA et maybe the new experiment SPOT5 (Take5), we won't have time to get bored.

This new year is also the first birthday of this blog, and though I only imagined to keep it during SPOT4 (Take5) experiment,  the large number of readers incite me to go on.  And a blog about time series should at least last for some time...


At least one visit from all the countries in blue has been recorded

Country Pages Visits
19 228 4 272
1 306 177
1 250 515
781 218
698 290
684 243
682 162
549 220
493 277
458 84


During its first year, this blog received 14000 visits, which is 60 visits per day of work (some of you even read it during week-ends, but I hope this blog is not the reason why you are not going for  a walk, to the movies or don't read good books !). 31000 pages were visited (with a click), but as it is a blog you may read several posts with only one click. 60% of the visits come from France and 40% from abroad, and the blog received at least a visitor from nearly every country.

Apart from the main pages, the most successful posts are :





Venµs launch contract just signed


We could have written a real cliffhanger serial on this blog, describing the list of rockets that were someday supposed to launch the Venµs satellite : the first one was Dnepr (an Ukrainian rocket derived from intercontinental missiles) at the time the project was decided, a long time ago, then the PSLV Indian launcher during a few years (with a period when the Venµs 2 days orbital cycle had to be changed to a 3 days cycle, degrading the repetitivity of Venµs observations), then it was Space-X (an American private company) Falcon 1E (which was finally dropped by Space X), then Soyuz from Kourou with Pleiades 1B (whose launch was finally advanced), then Space-X Falcon 9, and the VEGA (but again with a modified orbit, with an ascending part during daylight instead of the descending one). All these possibilities finally failed, for reasons that may have been funny if they had not contributed to the large delay of our mission (sorry, I am not sure I am allowed to tell these stories...).


The life of a small space project is not easy : to lower the launch cost, a small satellite must find a principal co-passenger for the launch. And this co-passenger must have an orbit close to the one of Venµs, with an expected launch date close enough to the small satellite. The absence of a launch contract made the launch date hypothetic, and it was not possible to use this date to put pressure on the industry that considered Venµs with a low priority : this fact also contributed to the satellite delays.


Finally, thanks to the patience of Venµs project managers at CNES, ISA and IAI, we now have a launch contract with VEGA, on Venµs nominal orbit. The Venµs satellite should join its 720 km orbit within a launch window that extends from October 2015... to December 2016. This broad window should be narrowed to three months by April 2015.


We will thus shortly issue a new call for site proposals, as a large part of the sites proposed in 2006 may not be in activity anymore. We will of course tell you more about that on this blog.


Oliver Hagolle, Gérard Dedieu


New satellites added to SMAC atmospheric correction


New coefficients have been added to the CESBIO repository for SMAC coefficients. The new sensors taken into account are :

  • Landsat 8, RapidEye
  • Quickbird, Worldview2, Ikonos
  • Pléiades1A (PHR1A)


The Simplified Model for Atmospheric Correction (SMAC) is the perfect model to perform easy, quick and not too dirty atmospheric corrections. It is based on very simple analytic formulas, based on the 5S model. The 49 coefficients of this model are fitted using a large number of radiative transfer simulations with the 6S model (the old historic version, not the recent vector version). This software is not very accurate (much less than MACCS), and it requires in-situ measurements for the aerosol optical thickness, and weather analyses for ozone and water vapour. If these data are available,  in most cases, its accuracy is within 2 and 3 percent, if we do not account for adjacency effects and slope effects, and it may be worse for large viewing and solar angles (above 70°) or within strong absorption bands.


SMAC is very easy to use:

#read the 49 coefficients in smac_soefs table
nom_smac ='COEFS/coef_FORMOSAT2_B1_CONT.dat'
#read the TOA reflectance image in r_toa variable
#depends on the file format
#read the angle values in the image metadata
# compute pressure at pixel altitude
#find the values of AOT, UO3, UH2O
#compute the atmospheric correction

where :

  • theta_s, phi_s are the solar zenith and azimuth angles
  • theta_v, phi_v are the viewing zenith and azimuth angles
  • AOT is the aerosol optical thickness at 550 nm which may be obtained from an Aeronet stations, or guessed, or equal to 0.1 for a really dirty atmospheric correction.
  • UO3 is the ozone content in cm.atm (0.3 is OK)
  • UH2O is the water vapour integrated content in kg/m². When I do quick and dirty atmospheric correction, I often use a value equal to 3, but I do not process spectral bands with strong water vapour absorption bands.

[1] Rahman, H., & Dedieu, G. (1994). SMAC: a simplified method for the atmospheric correction of satellite measurements in the solar spectrum. REMOTE SENSING, 15(1), 123-143.
"[2]"Tanré, D., Deroo, C., Duhaut, P., Herman, M., Morcrette, J. J., Perbos, J., & Deschamps, P. Y. (1990). Technical note Description of a computer code to simulate the satellite signal in the solar spectrum: the 5S code. International Journal of Remote Sensing, 11(4), 659-668.

"[3]"Vermote, E. F., Tanré, D., Deuze, J. L., Herman, M., & Morcette, J. J. (1997). Second simulation of the satellite signal in the solar spectrum, 6S: An overview. Geoscience and Remote Sensing, IEEE Transactions on, 35(3), 675-686.>
"[4]"Kotchenova, S. Y., Vermote, E. F., Matarrese, R., & Klemm Jr, F. J. (2006). Validation of a vector version of the 6S radiative transfer code for atmospheric correction of satellite data. Part I: Path radiance. Applied Optics, 45(26), 6762-6774.
"[5]"Kotchenova, S. Y., & Vermote, E. F. (2007). Validation of a vector version of the 6S radiative transfer code for atmospheric correction of satellite data. Part II. Homogeneous Lambertian and anisotropic surfaces. Applied Optics, 46(20), 4455-4464.


What about playing Take Five again with SPOT-5 ?


This post is regularly updated with news (the official date of Sentinel-2 launch) or to add new arguments.

SPOT5 will soon end its career. After 12 years of image acquisitions, the satellite will retire in 2015. CNES might launch a call for scientific experiments with SPOT5 before the satellite de-orbitation. I am quite sure it would be useful to repeat the SPOT4 (Take5) experiment, for the following reasons :


  • The official Launch date of Sentinel-2A in now March the 30th, 2015. Even if the satellite is launched in time, the routine acquisition on the 10 day cycle orbit will start one or two months later, and it will take several months to distribute data operationally. Data from the second Sentinel-2 satellite will probably only be available the year after, we will thus have to wait until 2016 to get data with a 5 days repetitivity.
  • A new Take5 experiment based on SPOT5 would allow to go on preparing uses, methods and applications based on time series, to get ready to make an operational use of Sentinel-2 just after the data are released.
  • SPOT5 provides multispectral images with a 10 m resolution, just like Sentinel-2. A new Take5 experiment would allow better simulations of Sentinel-2 data.
  • SPOT4 (Take5) experiment was held in Spring and ended in June, we might this time try to extend the time period towards the Summer to monitor the summer crops.

For the first presentation of SPOT4(Take5) proposal, I had been told that it had no chance to succeed, and I had even used this drawing as my last slide. SPOT5 (Take5) chances of success are the same, but isn't it worth trying ?

  • The SPOT4(Take5) experiment was set up very quickly and we lacked time to convince international partners to take part to the experiment. This time, thanks to SPOT4(take5). Now, thanks to SPOT4(Take5) little celebrity, it should be easier to involve new international partners, and to reach new users. The cost for international partners to get access to data for one site every fifth day during five months was about 4000€. Please contact me if you think your organism might participate.
  • SPOT4 (Take5) sites were chosen very quickly, and many users complained they had no time to set-up ambitious measurement campaigns, to hire people for ground truth measurements and define protocols. This time, we might obtain a longer prior notice period to set things up properly.
  • The dreadful weather we had in Europe during SPOT4(take5), nearly spoiled a few experiments, but we may hope that it would not happen again if we tried it one more time.
  • ESA and JRC excepted, we did not have time to involve European partners in SPOT4 (Take5), and 95% of the sites chosen by ESA and JRC were outside Europe. This time, we could focus part of the experiment on Europe.
  • We might try to have the experiment running for a longer period that in the first time, but it will mean a  larger cost for CNES, and we willl need to have a convincing set of experiments to convince CNES.

Even if SPOT4(Take5) was a success, we will need to build an excellent proposal in order to convince CNES, in a constrained funding context. After the experiment was already funded once, it is not a premiere anymore, and its impact will be less straightforward.


We thus need to compensate with original ideas and a large support.  If you are interested to participate to a possible SPOT5 (Take5), please leave messages on this blog or on my email, Please do not forget to provide us the results you obtained with SPOT4(Take5).


Using High Spatial Resolution Time Series to monitor forage production

An index-based insurance solution is developed to estimate and monitor the near real-time forage production in France. In this system, payouts are indexed on an indicator, called Forage Production Index (FPI), calculated using a biophysical characterization of the grassland from medium spatial resolution remote sensing time series.

Figure 1: fCover mean profile on parcel 4 derived from remote sensing images of multiple sensors.

Figure 1: fCover mean profile on parcel 4 derived from remote sensing images of multiple sensors.

We used the fCover integral as a surrogate of the forage production. fCover is a biophysical parameter that estimates the fraction of ground covered by green vegetation, looking in a vertical direction, independently of the actual image sensor viewing or illuminations conditions.


The first step of the validation process is to compare local ground measurements of biomass production and FPI values obtained from high spatial resolution space-based images. A field protocol (from PV PROTIN, 2010. ARVALIS –Arvalis, Institut du Végétal) was applied to 6 plots of grassland located in the area of Toulouse. These plots were selected to represent variations on pasture management practices and to consider different types of grassland species (Table 1). From March, 7th to June, 17th 2013, biomass was measured every 15 days using a sickle bar mower with a 110 cm cutter bar. Finally, the dataset contains 320 plots. It corresponds to one production data per hectare per plot every two weeks in average.

Name Cover Surface (ha) Pasture management practice Images used
Parcel 1 Alfalfa 7,3 Hay Take 5 : 6 / Spot 6 : 2 / Landsat 8 : 1
Parcel 2 Natural grassland 9,4 Hay Take 5 : 4 / Spot 6 : 1 / Landsat 8 : 1 / Formosat2 : 4
Parcel 3 Ray-Grass 8,6 Silage in May /  Hay in June Take 5 : 5 / Spot 6 : 3 / Landsat 8 : 1
Parcel 4 Natural grassland 6,0 Hay Take 5 : 7/ Spot 6 : 2 / Landsat 8 : 1 / Formosat2 : 1
Parcel 5 Fescue/ Orchard grass / White clover 11,5 Hay then pasture Take 5 : 5 / Spot 6 : 2 / Landsat 8 : 2 / Formosat2 : 1
Parcel 6 Fescue/ Orchard grass 6,8 Hay Take 5 : 3 / Spot 6 : 2 / Landsat 8 : 1 / Formosat2 : 1

Table1 : Characteristics of the 6 parcels


The 6 plots are in the SudMipy area defined in the framework of SPOT4 -Take5 project. So we used SPOT4 images acquired in the context of this program to build a high spatial resolution time series. Due to the climatic conditions during spring 2013, we had to complete the dataset with images from other sensors Landsat-8, SPOT-6 and Formosat-2. Finally, from February, 16th to June, 26th 2013, we have one image every 15 days over the 6 parcels (Table 1).

Figure 1 presents the fCover mean profile for parcel 4 and derived from remote sensing images of multiple sensors. Figure 2 presents the relation between the biomass measured on the ground and estimated from remote sensing images time series. For the whole dataset, the scatter plot between FPI and ground biomass shows an acceptable correlation (R²=0,724; α < 0,0001). However, it remains a substantial dispersion with a RMSE going up to 0.128.

Figure 2: Regression between FPI and local ground biomass measurements

Figure 2: Regression between FPI and local ground biomass measurements

If we take into account only data recorded during the growing period, the results are improved (R2= 0,811; α < 0,0001 and RMSE 0,101). This can be explained by the way FPI is designed: by definition, when the fCover integral is calculated, the brown fraction of the cover is not considered. In the framework of the research activities developed to create the index-based insurance product, these results enable to conclude that High Spatial Resolution images can be used to perform an indirect validation of the FPI produced from medium spatial resolution remote sensing time series.


Anne Jacquin

Antoine Roumiguié

Université de Toulouse, Institut National Polytechnique de Toulouse, Ecole d’Ingénieurs de PurpanUMR 1201 DYNAFOR, France.

Schedule for SPOT4 (Take5) reprocessing

We should be able to start reprocessing  SPOT4(Take5) data in January, and to release them in February (it is always risky to announce a date, we have already announced November, then December). This new version will also be produced on MUSCATE prototype processing center, implemented at CNES, for the THEIA Land data center.


Why does it take so long ?


  • The longest part was the negociation between CNES and CAP GEMINI about the modifications of MUSCATE.
  • The modification itself took a while, as it includes :
    • A major version change of SIGMA, which is CNES software for ortho-rectification. This new version will allow us to improve geometric accuracy using new reference images to take ground control points (GCPs), while the previous version only allowed us to work with LANDSAT 5 or 7, while the other solutions did not work because of a strange bug. Instead of LANDSAT 5 and 7, we will use the GEOSUD mosaic over France, which is a high quality mosaic produced by IGN, that covers the whole France. Elsewhere, we will use LANDSAT 8, which also provides a large gain in accuracy. And for the 4 sites, whose ortho-rectification was most difficult, we will the most cloudfree SPOT4(Take5) image as référence, which should provide us better quality GCPs.
    • We will be able to tell normal Level 1C failures due to cloud cover from possibly abnormal ones.
    • We will at last be able to process NASA's maricopa site, y separating it in two sites, one taken from the East, and the other from the West.
    • The level 2A chain was updated, and we will use a new aerosol model, with slightly larger particle sizes, which provide better results.
    • The Level 2 will provide 2 new flags that indicate the pixels for which the slope correction cannot be performed, for instance because these pixels lie in the shadow. We had forgotten these flags in our first version.
    • I do not know whether you are using the saturation masks or not. But these ones had a few defects that will be corrected in this new version.
    • The level 2A products will also contain nice quicklooks, on which clouds and shadow will be circled in colors. We will use them to perform a quicker validation of the processing, and we hope they will be useful for you too.
  • Until the end of the year, we will install the new version of MUSCATE prototype, then the processors, the parameters, and the new image references. The preparation of 45 reference images is quite some work.
  • Then, banzaï !,  we will validate all this stuff and start the reprocessing of version 2.0 of  SPOT4 (Take5) data.



La version v3.2 de l'OTB est sortie/ OTB v3.2 is out

Les collègues du CNES et de CS-SI viennent de sortir une nouvelle version de la bibliothèque open source Orfeo Tool Box. Parmi les améliorations, je suis sûr que vous serez nombreux à apprécier la segmentation "Large Scale Mean Shift", qui permet un traitement par tuiles qui donne le même résultat que celui qui serait obtenu sur l'image traitée en un seul bloc. Cette caractéristique sera très utile pour segmenter les volumineuses séries temporelles issues de Sentinel-2.




CNES and CS-SI colleagues juste released a new version of the open source library Orfeo Tool Box. Among the enhancements, I am quite sure that many of you will like the new "Large Scale Mean Shift" segmentation application,  "which allows to perform tile-wise segmentation of very large images with theoretical guarantees of getting identical results to those without tiling". This feature will be especially useful to process Sentinel-2 huge time series.

SPOT4(Take5) data downloads statistics

Three and a half months after the first release of SPOT4 (Take5) products, we can draw a first analysis of the SPOT4(Take5) data downloads (from THEIA distribution website : http://spirit.cnes.fr/take5/)


  • 160 different users have downloaded data, from all over the world.
  • 75% of downloads are Level 2A products, and most of the users who downloaded Level 1C products also downloaded Level 2A.
  • 40% of downloads concern whole time series, 60% concern mono-date images, which means that, with an average number of 14 dates per site, about 90% of product downloads are downloaded as whole time series.
  • Each time series has been downloaded 12 times on average.
  • Two users (I would bet they are French) are named Titi and Toto

Number of downloads per site (only whole site archives)

We have also obtained a ranking of the 45 sites versus the number of downloads, and notice that all 45 sites where at least downloaded once. Well, although CESBIO sites come first (!), this is not a competition, and it is probably biased, as some well organised user groups centralise their downloads, while other labs (among which CESBIO...) have downloaded their sites several times. The least demanded site (Rennes), was added very recently to the list.


The number of communications or papers per site will be much more interesting, but as far as I know, the counter is still set to 3 (3 communications at the Living Planet Symposium). Please remember that users are requested to send us all their communications based on SPOT4(Take5) data.


PS : this information is gathered thanks to the information provided by users at each download, which is collected on the download site data base (designed by J.Gasperi (CNES)), and then provided to me each month by B.Specht (CNES).

Next Sentinel-2 Conference

Our colleagues from ESA/ESRIN are organizing a new workshop to discuss the status of the preparation of Sentinel-2 scientific applications and methods. The three days workshop will be held at ESA/ESRIN premises, next May in Frascati, near Rome. It would be good to have some talks and posters about the results obtained from SPOT4(Take5) experiment. The due date for submitting abstracts is January 31st.


Nos collègues de l'ESA/ESRIN organisent un nouveau workshop pour faire le point sur la préparation des applications et méthodes scientifiques pour l'utilisation de Sentinel-2. L'atelier, qui durera trois jours, se tiendra dans les locaux de l'ESA à Frascati, près de Rome. Ce serait bien d'y présenter des résultats obtenus à partir de l'expérience SPOT4(Take5). La date limite pour soumettre vos résumés est le 31 janvier.



SPOT4(Take5) aerosol optical thickness validation results

We are currently preparing a data reprocessing of all SPOT4 (Take5) data, to be released before the end of 2013. For this, I tested several aerosol models and compiled all the validation results for our multi-temporal Aerosol Optical Thickness (AOT) estimation method named MACCS. Our estimates are compared to AERONET in-situ AOT measurements.

The MACCS method applied to SPOT4(Take5) data, which lacks a blue band, uses two procedures to estimate AOT :

  • either the AOT is estimated by a multi-temporal method
  • or it is gap-filled. The presence of gaps may be due to clouds, water or snow, or because the pixel reflectance is too-high for an accurate estmate, or because of a too large variation of reflectance with time is detected.


Comparison of MACCS AOT estimates with the in-situ measurements from AERONET. The blue dots correspond to cases for which the atmosphere is stable and for which there are no clouds in the neighborhood of the AERONET site. The red dots correspond to situations when the AERONET optical thickness varies around the satellite overpass time, or when clouds are detected in the image neighbourhood (20*20 km).
On the left plot, only the dates and sites for which less than 60% of the pixels were gap-filled; wheras the right plot only tolerates 20% of gap-filled pixels. The gap-filling method does not seem to introduce large amount of errors in wases when the atmosphere is stable, but it is less accurate in unstable cases..


The aerosol estimates have been obtained with MACCS prototype which is developed and maintained by Mireille Huc at CESBIO. The aerosol model is not the same as the one used for SPOT4 (Take5) first processing. This model is based on greater particles (with a modal radius of 0.2µm, compared to 0.1µm in the initial processing), as it provides a better overall agreement with AERONET measurements. We will use this model for most sites for SPOT4(Take5) reprocessing.


The RMS error of AOT estimates is 0.06, which is a state of the art performance, obtained in a very difficult condition with no blue band available. Moreover, in order to show more validation points, a few validation sites (Bruxelles, Gwangju, Ouarzazate, Wallops, NASA_LaRC) are in fact distant by more that 60 kilometers from the image footprint, which tends to degrade the performances.


The AERONET sites used in this study are :


SPOT4 Take5 Site
Aeronet Site
Belgium Brussels
South Great Plains Cart_Site
Korea Gwangju_GIST
Chesapeake NASA_LaRC
Chesapeake Wallops
Versailles Paris
Versailles Palaiseau
Tunisia Ben Salem
Maroc Saada
Maroc Ouarzazate
Sudmipy-Est Seysses + Le Fauga
Sudmipy-Ouest Seysses
Provence Carpentras
Provence Frioul


The worst results are obtained for the following sites :
  • Gwangju (Korea): The SPOT footprint in on the coast, while the AERONET site is 70 km inland, near a large town.
  • Ben Salem (Tunisia): this site was very cloudy in Spring, and large reflectance variations are observed between the remaining clear dates.
  • Palaiseau and Paris : In that case, the aerosol model seems to be inappropriate, and absorbing pollution aerosol should be introduced.

On the contrary, several sites provide very accurate results, for instance in Morocco (even the desertic Ouarzazate), Provence (including the Frioul Island where the AOT is extrapolated from the coast), and also Sudmipy, Wallops et Cart_site. Some SPOT4 (Take5) users reported inaccuracies on some tropical sites but we do not have an AERONET validation site near these SPOT4(Take5) sites.