Earth's-field magnetometry

GELLER FDM-PPM in my backyard

Listening to the music of the spheres

All of us have, at one time or another, used our planet's magnetism when reading a compass, but how many of us are aware that this seemingly reliable force is actually in constant flux? Generated in Earth's molten core, the magnetic field waxes and wanes and moves around, even flipping its direction every several-hundred-thousand years. If you could visualize it from a great distance in space, the field would look like a bubble swept into a long tear-drop shape by the solar wind--charged particles streaming from the sun. Gusts in the solar wind, un-inspiringly named “coronal mass ejections,” buffet the field, compressing it but also interacting with it in complex ways as the magnetic fields carried inside the traveling plasma react with our own field. The resulting fireworks can be watched from your own back yard, given a measurment device with enough precision.

These “magnetic storms” are of more than scientific interest; they affect our wired world in all sorts of ways, some quite damaging. A rapidly changing magnetic field will induce enough electrical current in really long wires, like those of the power grid, to destroy transformers. Shortwave radio communications can be lost and GPS accuracy is degraded. The high-energy solar radiation damages electronics in space and even puffs out our atmosphere, increasing drag and causing satellites to decay sooner. Auroras (northern and southern lights), the most spectacular and beautiful sight in all the night sky, grow brighter, more active, and closer to the equator.

There is something undeniably magical watching these processes with an instrument you built yourself—normally invisible events of cosmic scale, suddenly registered on your computer screen. I suppose that is why so many of my hobby projects involve astronomy or geophysics in one way or another.

The story

The history of the magnetometer pictured above actually goes back to at least 1989. In March of that year, an especially potent magnetic storm resulted in the temporary collapse of Hydro-Quebec's power grid and an aurora visible in the southern United States. Growing up in California I had never seen the aurora, and by the time I heard about this one, it was over. I thought a magnetometer might help me next time, so I did some library research and posted a query on USENET. That message resulted in a years-long correspondence with Joe Geller, an engineer at Brookhaven National Lab. Looking back now I find it extraordinary that Joe was willing to invest so much time in a kid who was, after all, still in high school! We went through a few iterations of a design for a fluxgate magnetometer but were unable to get it thermally stable enough to record magnetic storms. Eventually I went off to college and grad school, and we lost touch.

Twenty years later, I was thinking about magnetometers again when I found Joe's web site for Geller Labs, with a complete set of plans for a nice proton-precession magnetometer! I emailed and he generously offered to send me a kit for beta testing. In the photo above, the sensor coils are in the frame covered in black plastic next to the fence. This is connected by a cable to boxes housing the switch control and narrow-band LNA, which feeds a simple National Instruments ADC. The software running on the netbook is a custom Labview app for running the acquisition and saving data. Cleverly, the software is able to optimize the coil polarization time, which depends on ambient temperature, in a feedback loop, and to track a number of other quality metrics.

Joe has an excellent article in EDN explaining the principles of PPM operation. His design is noteworthy for using a very sophisticated spectral estimator called the Filter Diagonalization Method (FDM), and for giving careful attention to the proper (smooth) ramping down of the coil polarizing field at the beginning of each measurement interval. More details are available at the FDM-PPM project page. Sadly, kits are no longer being offered but the design is still available for personal use.

Here are some more pictures of the sensor installation and siting. It does not have to be too far away from the house to get good recordings.

Counter-wound sensor coils Magnetometer site area

The data

National Resources Canada's Victoria Magnetic Observatory is less than 100 miles from my location north of Seattle. The NRC data are available online and helpful for quantifying the performance of my backyard observatory. I was pleased to find that my recordings match theirs almost perfectly, as during this series of small magnetic storms:

(These are SVG plots; to zoom in, right click on one, choose “view image,” and maximize the browser window.)

Zooming in slightly and adding 558-nm photometry from the aurora monitor in Walla Walla, one can clearly see a faint aurora, masked by bright moonlight. And the brightness peaks in the aurora correspond nicely to features in the magnetogram!

This is the closest I've come to that initial goal of so long ago, using a magnetometer to help me see the northern lights. (Direct optical detection proved much easier, and I saw many beautiful displays in Illinois and Washington.) Now the sun has gone quiet again and we are in a prolonged drought of good auroras.