Cette page appartient aux archives web de l'EPFL et n'est plus tenue à jour.
This page belongs to EPFL's web archive and is no longer updated.

Leman-Baikal project blog

Mission complete

On Thursday we continued working in the lagoon. The whole area was scanned several times in order to detect changes caused by strong water streams. In addition, several solitary points were scanned at different altitudes to get resolution dependent spectral information. 

While we were flying, the French team was acquiring water samples and spectral signatures of the coral reefs.


Upon our arrival to the hotel, Dragos processed the previously acquired images. Thanks to the integration of inertial measurement unit, we were immediately able to project all images. The results exceeded our expectations:


Martin spent Thursday afternoon with the French team collecting water samples from a boat close to Capmohard village.


Pascal was regurarly checking the spectrometer and other sensors.


Several dolphins and whales were spotted. As it's often the case, the dolphins were sometimes swimming along with the boat.


 Friday was the most intensive day in the week. Right after the sunrise at 6am, both teams were ready for the final data acquisition. We flew both in the morning and in the afternoon, totaling 13 flights,  both over stationary points as well as scans of areas. The drone survived unharmed and was happy to finish the work with warm but working engines.


... the news about our activities spread quickly and lots of people came to meet us ....


In the afternoon, we discovered a wonderful underwater world of the coral reef. While snorkeling, we met local inhabitants protecting their home :)


and a beautiful and majestic sea turtle which nonchalantly swam away faster than Dragos could follow.


For us, this is where the work ends and the holidays begin. We will be in the mountains for the next week, enjoying the magnificent views either by foot or mountain bike.


Posted by Martin Rehak at 19:55
From Siberia to the tropics: Hyperspectral in Reunion

Following the Baikal campaign and continuing collaboration with French colleagues from the University of Brest, we came to Reunion Island, more than 9000km from home, to help with hyperspectral data acquisition.

And so it begins:

After a long flight from Paris, we finally arrive to our destination, Saint-Denis. The weather is beautiful, sunny without wind, however quite humid. Tristan gives us a ride to a hotel where the whole team stays. The main area of interest – Le lagon de Saint-Gilles is only 200m far away.


After preparing the drone we rest and enjoy a local Creol band during a nice and warm evening while stil being tired from our long journey.


Next morning we realise we're surrounded by breath taking landscapes. The lagoon we needed to scan is aprox. 300m wide, 0.5-1.5m deep and an unique coral ecosystem.   


On the second day we test the drone and do first data acquisition flights once safety is assured. Although the area of interest is relatively small, the flight plans had to be divided into four smaller missions. Although each flight takes about 6 minutes to complete, we needed about an hour to complete our task as setting up the drone and acquisition after each flight was time consuming.

The drone with the VISNX cameras worked great and we got our first hyperspectral images !

Dragos was resting after a successful morning of data acquisition and too little sleep the previous night. Swimming in the lagon was the best reward.

Martin, Nicolas and Touria were hiding from the strong sun in our provisional base camp.


On Tuesday morning, a local TV came to take interviews from all the teams. Dragos was especially enthusiastic while explaining our tech & methodology.

On Wednesday we had another day of successful data acquisiton after which we went to do some special shopping in the south of the island. While driving to St. Pierre from our hotel, we were rewarded with a wonderfully warm sunset against the black volcanic rocks of the island: Our pictures cannot do it justice.

Posted by Dragos Constantin at 20:46
Programme on lake Gusinoe almost finished

Completed tracks on the lake Gusinoe

Posted by Yosef Akhtman at 15:53
Great results by the Brest bathymetry team

Good news from the team of Niclolas, Marco and Mideg, who have successfully completed a high-res bathymetry survey around the Selenga delta aboard the vessel Musson. They have located and mapped a previously unchartered underwater canyons, which pose a great interest for the study of the delta.

Nicolas and Marco are now on lake Gusinoe to push forward with there bathymetry surveying plans.

Posted by Yosef Akhtman at 4:22
Water shortage on lake Baikal?!

The level of lake Kotokel situated less then 2km from the shore of lake Baikal has fallen two meters over the past year. The reson is most likely the logging of the surrounding forests and the subsequent disappearance of the springs that feed the lake. It is increadible to find water shortage problems in such a close proximity to the largest body of fresh water in the world.

Posted by Yosef Akhtman at 4:03
First glimpse of the hyperspectral data

A composite image - RGB (left) and Principle Component Analysis (PCA)-based false colour (right) of the hyperspectral imaging data collected during our field campaign. I thank everyone involved for helping collecting this unique and beautiful data.


Posted by Yosef Akhtman at 18:16
Arriving at Gusinoe

The remaining team of the Leman-Baikal 2014 expedition has arrived in the end of the world. Wind speed 18 m/s and growing.

Posted by Yosef Akhtman at 15:21
Mission completed

World's first high-resolution hyperspectral survey of the Selenga delta successfully completed.

  • 1000 sq. km
  • 4,000 km of flight tracks
  • 40 hours
  • 500,000 images

Posted by Yosef Akhtman at 16:27
Programme almost done

We've had a few challenges and a couple of days of difficult weather, but our programme for the first ever high-resolution hyperspectral mapping of the Selenga delta is almost finished.

Posted by Yosef Akhtman at 4:07
ULM tracks so far
Posted by Yosef Akhtman at 16:13
Page : 1 2 3 Next »