So where are the #meteorites from the most recent fireball in Norway?

On November 18, 2013 by Mike

BurstWithMets2After writing the previous post on determining the location of a fireball burst with seismic data, I had a discussion with Steinar Midtskogen with the Norwegian Meteor Network. We discussed what we knew from the seismic and what additional information could be gleaned from the visual reports of the fireball. In addition, I also spoke with an old (not in age) office mate from the University of Calgary – Dr. Wayne Edwards. As it turns out, his PhD work dealt with the issue of locating bolides with the help of seismic and infrasound data. From these discussions, it became clear that the initial seismic work described in a previous post was leading us in the right direction. So, what follows is an update of what I believe to be the most-likely burst location, an area to search, and a possible orbital solution.

Before you go on, please considering leaving a comment or ask questions when you get to the end. Feedback is good…

20131107_006.nc302zYou may recall from the previous post on this event, that there was a clear seismic signal observed by the NOA earthquake seismometers. The signals are characteristic of bolide bursts with a possible indication of multiple bursts (4?) along the fireball trajectory. Cool. By assuming a time for the event of 19:19:00 (now known to be 19:18:17 UT) and a range of velocities (300m/s – 350m/s) for sound travelling from the burst(s) to the seismometers we’re able to compute the rough distance of the burst(s) from each of the seismometers. We then project this distance along the ground by drawing circles corresponding to 300m/s and 350m/s distances at each of the recording stations. What we saw from this initial work was that the circles converged rather neatly. We then assumed that the areas where they converged were possible burst locations.

But can we do more with the seismic data? Well, as it turns out, yes. The NAO group of seismometers are configured such that we can also determine the direction from which the airburst energy arrived. And this is where Dr. Wayne Edwards’ work comes in. Using his code, he determined that,

  • The energy arriving at NAO had come from an azimuth of 277.2 degrees at a velocity of 349m/s
  • The energy arriving at NB2 had come from an azimuth of 263.3 degrees at a velocity of 325m/s
  • The energy arriving at NBO had come from an azimuth of 258.5 degrees at a velocity of 368m/s
  • The energy arriving at NC2 had come from an azimuth of 342.9 degrees at a velocity of 324m/s

Three of the above are pretty reasonable. The fourth, however, is clearly off so we discount it. The remaining solutions are rather interesting as they allow us to confirm that the original range of velocities was reasonable and that our areas of convergence are in the right spot. Hats off to Dr. Edwards.

In addition to the seismic data, we also have infrasound data. This data is much lower frequency than seismic and, as a result, can be ‘seen’ over much larger distances. In the case of this fireball, both the Kiruna and Lycksele infrasound stations appear to have detected the event.


The infrasound data from the Lycksele, Sweden infrasound array shows the Røn fireball event as the blue cluster of points near the top left corner of the plot. From this, we have an azimuth and arrival time so we can roughly locate the burst.

Now the main reason that we’re aware of this event is not because some seismic ‘geek’ is looking through the data for interesting events like this. It’s because it was visible to the naked eye. So visible that people thought that it was special enough to report to the Norwegian Meteorite Network. Brighter-than-the-full-moon bright. For meteorite searchers, this is great news. Visual reports give us first-hand accounts of where the bolide was first seen and where it ended. And it is the coordinates from these accounts that allow us to calculate a number of interesting things about the flight of the fireball.

One of the main problems with eyewitness reports, though, is that they are notoriously inaccurate. What we think that we saw and and what we actually saw are often two different things. Our interpretation of angles (especially altitude) without objects of reference is typically poor. And, when it comes to distance, we have serious problems. It is quite common for someone to see a fireball approach the horizon and believe that it ‘landed just behind that hill in the distance.’ Often, it is very difficult to convince the observer otherwise. In their mind, the event was so bright that it must been very close. However, in actual fact, if you see a fireball approach the horizon, it is many hundreds of km away from you and 10’s of km above the ground. As odd as that sounds, you’ll just have to trust me on this one.


Adding eyewitness reports into the mix gives a number of solutions for the start (within blue polygon) and end points (within orange polygon) of the fireball. By combining these measurements with the seismic azimuths (magenta lines), most-likely start and end positions are derived.

So, in this case, what do the eyewitness reports tell us? Well, I don’t want to spoil Steinar’s story of how he interprets the data so I’ll just say that my results are shown on the map to the right. By looking at all of the possible start and end locations from the visual and seismic observations, I’m able to (statistically) derive ‘most-likely’ start and end points for the fireball.

With the start and end locations in hand, I make a number of assumptions about the fireball. The first is that the fireball first became luminous at about 70km in altitude. Likely it was somewhat lower than that before people realized what it was, but it’s a starting point. The second assumption is that it made it down to at least 30km above the ground before bursting. Normally, we might assume a burst at a greater altitude for something of this (assumed to be) small size, however, we have eyewitness reports… Two of which point towards a burst at 31km (or less) altitude. The third assumption is that it wasn’t moving terribly quickly. Since it was visible for a couple of seconds and its distance of visible travel was just over 40km, 20km/s seems to be a pretty reasonable as a starting point. There is, however, one hitch. And that is that the only video shows a luminous fireball for all of about 1.5s. This translates into a velocity of about 28km/s which is somewhat above the 23km/s or less that we’d like to see. I’m an optimist, though, so I’m going to go with 20km/s so that I’ve got a good excuse to go out meteorite hunting in the spring. Anyway, with these assumptions and the relatively hard ground truth of seismic, infrasound, and eyewitness observations, we can make an attempt at calculating the location of meteorites on the ground and, finally, an orbit for the object that created the fireball.


A very approximate solution for dark-flight shows that any potential meteorites will continue downrange after the terminal burst. Notice how rapidly the rocks slow down and how their fall becomes near vertical as they approach the ground.

When a small object like this explodes in the atmosphere, the pieces normally don’t just land directly below the burst. If they did, it would be too easy. Instead, any pieces surviving the burst and subsequent ablation continue to travel along a path that is more arcuate than it is a straight line. As the velocity decreases rapidly from 10+ km/s to something on the order of a couple of hundred metres per second, the paths of the soon-to-be-meteorites become affected more and more by atmospheric winds (and, quite possibly, fairies). It is this part of the final part of the trajectory, termed dark-flight, that is critical to our understanding of where meteorites could be found. Unfortunately, we don’t really know much about the winds between the ground and 30km altitude on the evening of the fireball. So, what follows is really just a ‘hand-wavy’ approximation of how potential meteorites may have travelled during dark-flight. Roughly speaking, we might want to start looking downrange from the most likely burst point by the following amounts…

  •  1g – 750m
  • 10g – 1050m
  • 100g – 1550m
  • 1kg – 2200m
  • 10kg – 2750m

Or in map view…


‘Most-likely’ fall location based on seismic, infrasound, and visual observations and very crude dark-flight modelling. Take it with a grain of salt and plan on searching a wide area to the SW and NE if you’re going out! And, please, ask for permission before going onto somebody else’s property.

Keep in mind though that there could be a huge amount of error here. In reality, anywhere between my P90 and P10 burst points is probably fair game.

The orbit is calculated by populating a very clever spreadsheet (written by Marco Langbroek of the Dutch Meteor Society) with the following information.

  • Start Location: 61.053324d N, 8.799149d E
  • End Location: 60.938356d N, 8.861259d E
  • Start Altitude: 70km
  • End Altitude: 30km
  • Ground Distance: 12.7km
  • Entry Angle: 72 degrees
  • Entry Azimuth: 244 degrees
  • Velocity: 20km/s

Which gives orbital elements that can be used to plot an orbit with a program like Celestia…

  • q: 0.724 (closest approach to sun)
  • a: 1.01 (semi-major axis)
  • e: 0.283 (eccentricity of orbit)
  • i: 21.69 (inclination of orbit)
  • w: 282.73
  • W: 221.892
  • P: 1.02 years (orbital period)

Orbital elements derived for a fireball can be plotted using a program like Celestia. The Røn orbit is the white ellipse.

Now, to me, that’s pretty cool. From a few photons of light and some acoustic energy we are able to not only figure out where to look for meteorites from this fireball, but also, (roughly) where it came from in the solar system. Neat, eh?

Comments are closed.

%d bloggers like this: