Stars and Stripes of Polar Lights

There is still some light as we directly drive from UNIS to the other side of Adventdalen, across from Longyearbyen. Steam rises from the fjord.

The air near the ground (where we stand) is -20°C cold. It is ok as there is no breeze. The mini Talon is having fun at -12°C in 50m altitude. This has often been the case here in the valley: we are freezing and the plane is enjoying the warmer air further up.

The sky is great, full of stars. The LEDs of the mini Talons mark his path. Every now and then a touch of a polar light draws across the sky. Above the contours of the mountains to the south you can see a light green at the distance. We are too far north on Svalbard for strong polar lights today.

For the first time I am cold. Standing motionless in the cold can be unpleasant even with good equipment. The mini talon flies profiles over valley and fjord.

We warm up in the old polar light station after two flights. Then we head out for another profile to 1000m and stacked lines.

Flying in daylight

The program is different today: The airport is opened in the evening/night and we are allowed to fly in the morning.

The sun is not yet appearing over the mountains. The small amount of light gets reflected by the snow and wonderfully illuminates the valley. The horizon is filled with colors. A new flightplan is created with the students. The mini Talon flies different profiles from the old polar light station far into the valley. Air temperature is only -15°C but perfect to be in as there is almost no wind close to the ground.

The air gets warmer further up and there is a layer with stronger wind. The mini Talon flies at different heights on a line and perpendicularly through the valley.

Simultaneously another group of students follows the track with sensors on snowmobiles.

A great day.

Cold Wind

We head out to the station in the afternoon to test a new plane: the mini Talon. It already did some test flights in Finland and should now show its capabilities in strong winds. It can fly faster (40m/s), allows a larger payload and is more efficient than the SUMO. For a flight at a constant altitude it needs about 50% less power. But it is much harder to fine tune and fly. The main application is to fly at high wind speeds and operate turbulence sensors that give a more accurately signal at higher speeds.

The advantage of the SUMO is that is possible to operate at up to 15m/s wind even with low model aircraft skills. The autopilot keeps the aircraft level in the assisted “Auto1” mode. The mini Talon needs more active flying by the pilot. The approach has to be precisely planned as it glides very well.

In contrast to yesterday there is strong wind with speeds up to 13m/s. The wind speed is lower at higher altitudes. We seek shelter in the lee of the Bandvagn at the beach. The plane can cope with the turbulent gusts in 40m altitude only after we have reduced the control gains of the autopilot. Profiles are flown at different places in the valley entrance.

Cold Clear Air

The clouds disappear during the night. There is an interesting temperature drop at midnight. Icing on the plane should no longer be an issue. We drive to the old polar light station in the afternoon and fly a 500m and 1500m profile as a test. The Bandvagn brings us to the “beach” at Adventfjorden.

We are flying several profiles from the fjord to the valley as in February 2016 to measure the atmosphere in different parts of the valley.

We use Panasonic Toughbooks as ground station, mostly the CF-19. The computer really cares about its components. It does not charge too cold batteries and heats the hard drive before a boot with a built-in heating foil. This is no longer necessary with the SSDs that replaced the magnetic disks – but it can not be switched off.

The lithium polymer batteries are preheated to about 40°C with a camping cooler/heater. The entire SUMO is made of EPP (a soft polystyrene foam) and the battery compartment is sealed from airflow. During the flight the battery heats itself through the electrical current flowing through it.
The air inside the battery case of the bebop2met is sufficiently insulating. In addition, the powerful computer directly under the battery heats it.

Thin wool gloves with fleece fold-down finger tips above are being used to control the manual takeoffs and landings. Scooter suits are worn against the cold wind.

Chocolate plays an important role in heating from the inside. On special occasions and to match the weather it was “Kalter Hund” tonight.

Oh No!

The aircraft are checked and repaired in the morning. We exchange one engine that had behaved funny during the last campaign with the Polarstern. In the afternoon we use the Swedish-made “Bandvagn” vehicle to go out to the old Aurora station.

What a great device. You can go everywhere with it. But it is horribly loud and moves only slowly. The engine is mounted in between the front seats and makes everything vibrate.

The machine needs a lot of love. We carefully check the oil level of the Daimler engine before running it.

The shock at the old Aurora station: The workshop is gone. There is a darkened room in which tourists can watch polar light films instead. The beautiful workshop in which we spent so many nights staring at aircraft icons circling, just gone.

We move to the kitchen and make the SUMOs ready to fly. It is snowing slightly. Propeller and the wings leading edge freeze on 400m, surprisingly. Svalbard is very dry and we had seen icing only once in all the years, even when we flew in clouds. The ice is rougher and not as clear as during the flights over Finland.

The tethered balloon is set up. The profile is quite unspectacular as expected. On the way home there is a glimpse of polar light over the mountains behind Longyearbyen.

Landing, thrilling

Everything as always, back to Svalbard. The wind blew from the southwest. We flew over the cloudless east of the main island. What a great view of the sea and the glaciers. During the final approach through Adventalen it shook vigorously and the visibility was bad as there were clouds and snow.

The weather also caused delays in the set up of the masts. We shall start tomorrow, though.

Had never realized that there are airports with underfloor heating.

More North

The measurement campaign here in Finland is unfortunately over for us. We are heading south to Helsinki, directly west to Oslo and tomorrow morning north again to the next campaign on Svalbard. This travel could in theory have been done in a much shorter time if we flew directly from Oulu to Longyearbyen.

The campaign in Hailuoto will continue until the end of February, check out the blog from GFI.

The ice road is still officially closed.

Mother Earth

The SUMO uses an acceleration sensor, a gyro rate sensor and a GPS receiver for orientation and position. They are largely independent of external influences.
A pressure sensor for altitude measurement, a pressure tube for the velocity relative to the air or a magnetic compass for determining the orientation towards the earth is not used.

Altitude, speed (only relative the earths surface) and direction are taken only from the GPS receiver. This works as long as the SUMO moves forward so that the GPS can measure a correct direction.

This does not work for a quadcopter. It needs directional information even when it is hovering (which it is most of the time). Therefore a magnetic compass is installed. Our bebop2met is based on the Parrot Bebop2 and unfortunately the sensor in it is installed very close to the rest of the electronics, It provides imprecise data. Another problem is apparently the wide scattering of the sensors in the various individual quadcopters. In addition, we are at over 65° north and the x-y component of the earths geomagnetic vector is very small.

In the end we spend much of the day magnetically calibrating the quadcopters newly purchased for the campaign by holding and spinning the quadcopter at various angles to the earth in a dance-like performance.

The forecast has unfortunately come true and the weather was inappropriate for our measurements throughout the week. The SUMO flies profiles to collect comparison values anyway. In a cloud at 1000m we are back to the icing. The plane reverses and lands safely.

The measurements take place in parallel with several systems. For the lower 10m the mast is set up, above that the bebop2met measures from 2m to 200m, from 40m to 1800m the SUMO collects data and parallel to all that is the SODAR. At high wind speed or humidity, however, the acoustic pulse does not reach as high and the measurable altitude decreases significantly.

Somewhat troubling is water that appears close to the beach. It leads to little seas of slush ice. The locals tell us that the south wind pushes the seawater against the land and the water is pumped up.


We are housed in small holiday homes (Hailuodon Majakkapiha) on the western edge of Hailuoto.

Data from OpenStreetMap

Food is served 3x a day in the small restaurant. The logistics and operations are super-efficient. Short paths between sleeping, working and eating.

There is a small house at the harbor from which we can fly the SUMO and work on the equipment.

The bebop2met quadcopter is operated directly from the ice because of the limited flight time.

Danger Area

Danger Area EFD451 has been created for beyond line of sight flights. That allows operations at up to 6500 feet in a 4 x 4km area.

The measured data can be analyzed on a second laptop during flight.

The flight pattern consists of a cork screw on the way up, some circling and another cork screw down. The SODAR works up to about 650m and the readings can be compared.