I travelled to the high north again. On Senja, an island in Norway situated 300 north of the Arctic Circle, me and my band of brothers desired to experience the wonderful winter light of these latitudes, a real winter with lots of snow – and of course the alluring aurora borealis. At the same time, I was equipped with a drone for the first time and wanted to test my flying skills and the photographic potential of the birds eye view first hand. Several incidents, sometimes exhilarating and sometimes nerve-wrecking, made me think about geomagnetic storms during this trip. Back home I decided to dig deeper than Wikipedia and consult some experts. Here is exactly how the earth’s magnetosphere allows solar winds to paint the northern lights and why you should not fly a drone at Kp levels higher than 4.
A primary motivation for me to go out and explore is to push the boundaries of what I have seen, doing and experiences. On a quest for new horizons I decided to book me into another adventure in the sub-arctic north of Norway. I’ve visited the Lofoten islands three times before and had opportunities to photograph many of the natural sights and phenomena that draw photographers from all over the world to this region. But this time I brought my family. The few minutes of aurora experience I had in one night four years ago was essentially all our experiences combined, meaning that five hungry aurora novices (and me who could stand some more sightings for sure) headed to the remote Northwest of Senja where the jagged peaks above the fjords are most stunning. We arrived to glorious winter weather and a perfect spot to take it all in: a holiday home right at the shores of the tiny island of Husøy in Øyfjorden. Overlooking the fjord with a clear view of Segla, Senja’s most iconic mountain, and a terrace facing northwest we were prepared for a show. The terrace was useful in a different way too, as a runway for the drone that I brought with me.
The first night delivered. We played card games well into the evening repeatedly stepping outside scanning the sky for hints of green. At half past eleven my wife got all excited as she noticed her first northern light. Poorly dressed we stumbled outside, turned of all lights and gazed into the sky. The aurora I had seen before – at Lake Solbjørnvatnet and in the village of Reine in Lofoten – was pronounced but well dosed. Singular bands of bluish green slowly moving in the sky for some minutes. This time was different: After a slow start the green lady decided to let go of all inhibitions and put on a real show. For an hour we witnessed all forms of aurora dances. Broad sweeps, tiny swirls, fast ribbons, and very bright bands – so bright in fact I had to stop down my exposure several times. First night, awesome show, we were safe.
Prior to the trip I took a look at Kp level predictions and found a consistent forecast of level 3 geomagnetic storms for the first few nights and then a slight drop to level 2 for the remainder of the trip. I knew roughly what these numbers meant, but as I am naturally curious I decided to learn about it a little better after returning home. I contacted an old friend of mine who now works as a communication officer at the German Research Centre for Geosciences (GFZ Potsdam). The folks at GFZ are experts on geomagnetism sporting a designated research group on this topic and calculating a three hour range Kp index. By doing so they contribute to the International Service of Geomagnetic Indices (ISGI) of the International Association of Geomagnetism and Aeronomy. Here is their explanation of the aurora phenomenon in a nutshell.
The northern lights occur because of a combination of two events or constellations. The first is the earth’s magnetic field which is mainly created by geodynamic processes in the earth’s core. The outer fluid of the core is rotating and thereby producing electric magnetism. There are other sources of geomagnetism which overlay the core field (like crustal field or external contributions) but they are responsible for less than 10 percent of the earth’s geomagnetic field. The shape of the field can be described as a stacked ring of field lines around the earth with connecting lines at the magnetic poles. So in fact it looks like a giant doughnut: Over the equator the magnetic field lines are horizontal, meaning they are parallel to the surface of the earth, while near the magnetic poles (which are not entirely congruent with the geographical poles) they intertwine and ultimately point to the earth’s core.
Now comes the second part: Solar winds significantly compress the magnetic field on the day-side of the earth. The solar wind is a stream of charged particles from the corona of the sun reaching speeds of 250 to 750 kilometres. This giant blower presses together the field lines on the day-side while on the night-side the magnetic field extends like a giant tail. This “magnetotail” is the primary source of the aurora near the magnetic poles. So yes, it’s a phenomenon that is dependent on the solar winds AND the magnetic field of our home planet. And yes, it mainly occurs at night (it’s not the case that it is there during the day and just not visible).
Earth’s magnetic field is relatively stable, although there are changes on various time-scales. Right now, the magnetic north pole is on the move distancing itself from the geographic pole. But the major dynamic component of the aurora phenomenon is the solar wind. Owing to the 27-day rotation cycle of the sun there is a 27-day cycle for every mass ejection to hit the earth. How many of the particles reach the earth determines how much interaction there is in the earth’s atmosphere – and how strong visible auroras might be. This is measured by the Kp index. Here are a few links that go into great detail on these issues:
The next day I started another venture that first seemed unrelated to our extraordinary experience of the first night. I assembled the drone, placed it onto the terrace and started to fly. I was late morning which means 11-ish in these latitudes. The aircraft quickly gained height and I was thrilled by the change of perspective. Hovering 80 meters above the fjord with the light coming in from the far left it was an attractive sight. I did not perform any stunts at my first flight and rather concentrated on getting used to the controls. After a few awkward moves to “compose” a shot I fired a series of photos. They are by no means very mindfully composed or making use of the perspective a drone offers a lot, but hey they are my first-ever photos from a remote controlled aircraft and the scenery could not be more stunning.
Two days later I decided to venture further out and fly over the fjord to the other shore. For the first time I was flying the beast – a DJI Phantom 4 pro – without eyesight and solely relying on the screen. Although there was clear views and only a few hundred meters to the opposite shore I noticed quite a few glitches in the communication between the aircraft and the controller in my hands. Video signal was weak, and from time to time I was told that the aircraft disconnected. A few minutes in, before shooting some of the scenes that I envisioned on the mighty ridge between the fjords I pushed the return home button and was happy that I saw the drone coming back. I landed with 7 percent power left. Wow, that was far too much stress on my sight and I could not explain the weird behaviour. Maybe it was too cold? Maybe communication was disturbed by something? I decided to call it quits for the trip and leave the drone in the case until I practised a little more at home. I should have sticked to it!
But on the final evening, meaning 3 in the afternoon, I spontaneously decided to try two more things with the drone. I unpacked it, mounted the wings, started the devices, let them find each other and after a few minutes I launched the copter. My first plan was to emulate a photo that was a land-based perspective but from a point where I could not walk due to ice and snow. 30 meters in front of our house was a rocky stretch of coast featuring a small narrow where some nice wave action happened. I hovered above the ice and snow and very carefully put the drone less than a meter above ground to come as close to the waves as possible. Everything was alright although I did not really tracked down what I consider possible within the few minutes of flight.
Giving in to an impulse, mainly motivated by the nice soft light, I decided it’s time for a goodbye photo or video from our tiny island. I quickly gained height to around 100 meters and directed the drone towards the northern tip of Husøy. I had hiked there on the first day, I was a rocky stretch of land covered in deep snow. There is a small lighthouse which I aimed at, roughly 350 meters away from our holiday home. I noticed the same glitches in communication again and opted for a quick abortion of the adventure. I rotated the camera back towards the island and started a video recording. Then, interrupted by a few disconnections I managed to call the drone to return home. Everything would be good, I thought. But it was not. Ten seconds later I lost contact with the aircraft again and would not re-establish a connection for the next half hour. First I thought the drone must be on the way back to me so I just have to wait until it’s in reach again. After ten minutes I became very nervous.
Five minutes later and me getting even nervous bordering panic, my wife convinced me to start walking around the island. We moved towards the lighthouse where I last spotted the drone on my screen. Endless minutes of wandering around brought one tiny spot of light at the end of the tunnel. At one position, only with a range of two meters on the narrow road, I received a signal from the aircraft. No image from the camera, battery at 49%, error messages stating something was wrong with the propellers and the magnetic compass. 49% after half an hour, the thing could not be airborne anymore. But how the hell should I find it? All positioning systems were not working and dusk was closing in. Panic broke loose again. Finally my wife saved the day: She asked me about whether there are any lights on the drone and I said yes, two green and two red ones. She said, look over there on the cliff near the lighthouse, there is a tiny red dot, could this be it? The dot was approximately 200 meters away on the eastern slope of the rocky outcrop that is the northern end of Husøy. I did not have a better idea than giving it a try climbing through wrist-deep snow over 20 meter high cliffs near the point where the red dot originated. A few minutes later I had certainty: This is the drone. I had to climb a few more meters – with absolutely inadequate gear by the way – to reach it. Nestled below a small rock in the deep snow I found it. The majority of the island was between the aircraft and our house. How the hell did it crash here?
Luckily the video recording was still active shooting exciting black movies inside the snow. But that means I got the whole thing on video. I hoped this would clarify the events in the last 30 minutes. And it did. Watch the sequence here before I continue.
Several things become obvious and some of them are addressing the painful subject that I am an absolute beginner and made corresponding mistakes. First was the return home altitude. This specification can be made in the menu and should be adapted according the environment. It was set for 50 meters which normally is a good choice, even with some trees. Here it might have hampered a safe return. Altitude is relative to ground level, so 50 meters means 50 meters either above launch site altitude or ground level when pushing the return home button (I’m not sure actually). Both alternatives where bad choices because the cliffs seemed to exceed this height. This means the aircraft came down and suddenly had to climb when coming close to the rock walls. This leads to the second beginner mistake. The camera sensors that prevent collision were starting to malfunction due to low light. I was flying at dusk and planned to return much earlier. Some of the extra loops the drone was flying may have been induced by collision sensors not working perfectly in the approaching dusk.
But the fundamental question remains: Why were coordination and communication so error-prone? The message that something is wrong with the magnetic compass incubated a line of thought in my brain: Could it be that the very reason for our northern lights show is at the same time responsible for the malfunction of the drone? It certainly makes sense from my layman’s position. To clarify this I approached the experts at the German Research Centre for Geosciences (GFZ). They largely confirmed what I was thinking:
“Magnetic compasses rely on stable geomagnetic conditions in order to correctly determine where north is”, says Claudia Stolle, section head for geomagnetism at GFZ. “However, most drones are equipped with a GPS system which are satellite based and do not rely on magnetic orientation. Yet even GPS system can be affected by solar winds: Variations in the upper atmosphere and the ionosphere disrupt signal transmission via GPS.” The Kp index measures irregularities in the magnetic field of the earth, but since they are largely caused by the solar winds the index is also a reliable indicator for those winds and hence the possible malfunction of GPS systems. A Kp value of 1 to 3 is considered trouble-free, but already Kp 4 is critical and experts advocate a strict no-fly policy for Kp 5 or higher. “The risk for the radio connection between copter and control device being disrupted and for unintended manoeuvres is very high”, concludes Stolle.
Well, this puts my adventure into a whole new perspective. Knowing that the Kp index is a three hour averaged estimate the possibility of some short peaks within a Kp 3 window is clearly there. I payed with 45 minutes of panic and anger, luckily I did not pay with a complete loss in the end. There is a lot to learn out there and I hope you found some bits of info in here valuable. Feel cordially invited to add your own experiences here and to provide some more info that might help beginners and pros alike.