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This page belongs to EPFL's web archive and is no longer updated.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

Time to go deeper!

Back in July 2013, the expedition Leman-Baikal has observed a curious phenomenon on the surface of lake Baikal. A line of regular dark circles was photographed from an ultralight plane near the outflow of the main channel of river Selenga. The circles are 20 meters in diameter and 60 meters in-between. At least 10 circles could be seen in a straight line. The circles could be only observed at a certain angle, on the edge of the sun glint region, which indicates that they are caused by the variations in the ripple pattern on the surface of the lake.

A number of hypotheses have been raised to explain the phenomenon in the year since it was first observed. It was agreed that a more detailed analysis will be required. The investigation included a dedicated wintertime field campaign by the scientific team from the Baikal Institute of Nature Management that involved drilling of the ice at the location of the circles and the analysis of the ice-trapped gas samples. Today, we will try to make one more step in our quest to find a plausible explanation.

We left Istomino in the morning and headed towards the site of the circles. The team includes Natacha, Roman, Vasily, Sasha and Yosef. Once we reached our destination we were joined by the ground team - Galia, Misha, Vincent and Baptiste, who were enroute to collect water samples for subsequent chemical analysis. Finally, the main player for the day - the diving vessel Merlin has arrived.

Merlin is a magnificent vessel purposefully build for long-endurance fishing and diving trips around Baikal. It can host up to 10 passengers and crew. Suffice it to say that it has an onboard sauna.

After a short and enthusiastic introduction, we've set out to our main task. First, we used an underwater camera to try and assess the underwater conditions at the location.

The visibility condition are quite poor - less than two meters. The lake bottom is flat and covered by silt deposited by the outflow of Selenga. There are many pockmarks, which indicate the release of methane. We decide to move a little to the north-west, away from the delta and further explore the bottom of the lake.
Our host Igor and myself are getting ready for the dive, while the others enjoy a swim. The temperature on the surface is nice and comfortable, around 17 degrees, but it's much colder further down. The normal practice is to use dry suits here, but I decide to try a 8mm wet suit - it's much simpler and I don't have much experience with dry suits.
The water is nice and fresh and looks quite clear on the surface.

But once we are at the bottom, the visibility is poor and I feel that dry suites were invented for a good reason.


The bottom is the same sandy silt with many pockmarks and borrows of crustaceans. An empty shell of a shrimp is the only sign of life that we encounter, but this is to be expected in this part of the lake.


Once we are back to the surface, the first thing to do is warming up. "Russian tradition" as they say.

We didn't find anything particularly interesting under the water, but shortly after drying up we are treated to the view of the same phenomenon clearly visible on the surface of the lake. It doesn't necessarily help me understand the cause of this effect, but at least it means that it is real and a good explanation should exist.

After a wonderful lunch and large quantity of alcohol, we thank our generous hosts and leave Merlin. I will be looking for an opportunity to board Merlin again.

On the way home I am trying to reflect on all the various theories of what the circles might actually be. I think the most plausible possibility is that it is an hydrological interference pattern between the outflows of two or more channels of the Selenga. The position and the orientation of the circles seen below clearly supports this hypothesis.

Here is an illustration of a 2D interference pattern from two slits (left) and multiple slits (right).

Of course, this result is far from conclusive and the search for further evidence and better theories will continue. 


Posted by Yosef Akhtman at 17:01
Good things come to those who don't wait.

Although the weather was very unpredictable today, we tried to be ready to fly at any moment and successfully finished a very long flight in the morning:

The hyperspectral system ran smoothly, however, the IMU disconnected and needed a reboot in the midle of the flight, making us miss a couple of flight lines that we'll cover tomorrow:

While this type of error used to ground the ULM and cancel the flight, recent updates to the system allowed us to recover mid-flight and continue data acquisiton. This flight also validated our most recent attempts at improving signal to noise ratio, as the following plots show clearly:

Yesterday, Vincent, Misha and Dragos took radiometric samples along the main road as well as different areas in order to calibrate the camera to actual physical reflectances. The data from this experiment was also very good, showing little variability along the road. Considering the recent wave of good news, the team is motivated and very optimistic about the quality of our data. Very soon we will be able to produce georeferenced hyperspectral maps of actual reflectance values, marking a huge milestone in this project.

Posted by Dragos Constantin at 11:28
Rainy days are not sad days.

Today was the second day of almost continuous rain. While no flights or boat trips were done, we took the time to process data and get some much needed R&R.

During the last days we managed to get our first hyperspectral projection from the Headwall system, showing that the system worked indeed perfectly but the timing of our flights is not ideal:

Due to turbulence and strong wind, we cannot safely fly between 12:00 and 16:00. However, the sun is not strong nor stable enough during early hours or late evenings, giving us underexposed images. We tweaked the system to 30 FPS from 50 FPS, in order to increase the signal to noise ratio, all the while motivating the pilots to be a bit more courageous and fly when there's more sun. Tomorrow we'll test the new setup.

Sergei and Dragos spent most of their working hours implementing a fast demosaicing method for the VISNX cameras. The results are very promising: images processed with the new method have a much higher registration rate and produce much denser point clouds:

Vincent was busy all day long processing spectrometer and ground truthing data from 34 sample points together with Misha, reaching the happy conslusion that the data shows enough variability for extrapolation to hyperspectral images.

While a lot of work was done during this bad weather, we also had nice surprises. We were recently visited by a mother and her kitten, so adorable that we've build a small cat house in one of our houses, especially for them.

Some of us, who've been working almost continuously since our arrival, have taken some time and visited the thermal baths in the east of Selenga Delta. We bathed in warm water filled with sulphur then swam in the chilly Baikal lake. It was some well deserved and very appreciated relaxation time.

Posted by Dragos Constantin at 16:27
Perfect flight

Best flight EVER piloted by Nikolai and operated by Dragos. Over two hours and 400,000 hyperspectral scan-lines without a single error! We are making a huge progress in getting the system stable.

Here is our plan for the entire campaign 2014
Posted by Yosef Akhtman at 15:51
August 5th, 2014

False start in the morning. After taxing to the take-off strip, Vladimir takes off to probe the wind conditions and comes back with bad news. Too turbulant to fly. The morning flight is canceled.

We try again in the afternoon. This time piloted by Vladimir and operated by Yosef. Lots of issues in the beggining of the flight. The system seems to have it's own (very capricious) personality. After multiple reboots, all of a sudden, the system stabilises and works without a glitch for over an hour. 200,000 hyperspectral scan lines in total. I am coming home satisfied with the data collected.

Posted by Yosef Akhtman at 15:49
Pushing forward

The most successful flight to date piloted by Nikolai and operated by Sergei. All systems work.

Rough day on the lake but a lot of progress has been done by two ground teams.

Posted by Yosef Akhtman at 17:15
Our system is now complete

The main hyperspectral platform has arrived yesterday, 2nd of August. The equipment was mounted on the ULM today at 9h30. The complete remote sensing platform now comprises the total of six cameras.

The morning flight took off at 10h00 and lasted for 2.5 hours. The eqiupment was operated by Sergei.

Five out of the six cameras were operating seamlessly, but we still experience problems with the main Headwall sensor. Upon arrival, we have carried out an extensive testing of the various mechanical and electronic aspects of the platform with the aim of pinning down the possible problem. In the end tree possible sources of disturbance were identified and eliminated.

The day proceeded with the visit of the project's sponsor Prof Michail Slipenchuk. A joint presentation was prepared by Natacha, Galina, Michail and Yosef, describing the different work packages comprising the project and the respective collaborative links between the Swiss and the Russian teams.

The afternoon flight was dedicated to the extensive testing of the newly applied modifications of the remote sensing platform. All systems performed seamlessly!

Posted by Yosef Akhtman at 14:49
Making progress

Good progress today for both ground and airborne teams. First airborne data is finally in. The brand new triple-head hyperspectral remote sensing platform looks great and is a joy to work with.

Also a great evening flight for our guest Jeffrey Nittrouer from Rice University.

Posted by Yosef Akhtman at 16:32
Another hard day

Hard day for all teams. Rain in the morning and Baikal is dark and stormy. The weather changes rapidly from blue sky to dark rain clouds several times during the day. The boat team strugles to work in meter-high waves.

All preparations for the airborne team are finished. We are waiting for the next weather opportunity, but no luck today.

Activity shifts into the laboratory, where the filtering of the water samples is ongoing around the clock.

Posted by Yosef Akhtman at 17:19
Fruitful day in the delta

Remote sensing team spent the day in last preparations of the airborn equipment. The weather conditions are difficult and no flights were possible today.

Fruitful day for the joint EPFL-MSU ground team in the delta. The Silver cutter leaves at 10h30 and comes back at 18h00 with a hefty load of sample materials: water, sediments and vegetation.

Galia and Lera are measuring pH, ox-redox potential and total dissolved salinity of the water sample.

Lera is measuring disolved oxygen using the Winkler method.

Misha is measuring the reflective spectrum of the delta vegetation and water surface.

The dragonflies are measuring ...

Baptist is supervising the scientific work by Misha and Rita.

Posted by Yosef Akhtman at 15:14
Page : « Previous 1 2 3 Next »