The Solar Dynamics Observatory launched last week for the thrilled scientists and engineers who have worked for years on this mission, some happy #SDOisGO TweetUp participants, and countless other space fans around the world. (The who? The wha? Oh, pretty!)
SDO’s EVE instrument (Extreme-ultraviolet Variability Experiment) is particularly interesting to me since these EUV photons from the sun are what drives the Earth’s ionosphere. The ionosphere is the outermost layer of the Earth’s atmosphere and consists of ionized, or charged particles. As new low frequency radio telescope capabilities have been coming online at the VLA, GMRT, LOFAR, LWA, MWA, and, my home, PAPER*, the ionosphere is gaining more attention, and not the good kind. Just as the lower levels of the atmosphere cause all kinds of scintillation and “twinkling” to annoy visible light observers, the ionosphere refracts and distorts light coming in at low radio frequencies. (Well, low for astronomers, that is… less than a few 100 MHz!) This is particularly troubling since all of these telescopes are interferometers, or systems of multiple radio antennas linked together to make one telescope. Images are made by measuring the difference between the arrival of light at different antennas, and this difference can be skewed by a turbulent ionosphere. And the ionosphere changes in density and turbulence, based on the solar output! See, it all ties together.
I have to admit, of course, there are more pressing concerns than radio astronomy. The ionosphere will also have an effect on GPS signals. As we start the upswing in the solar cycle, more turbulence in the ionosphere will mean larger position errors and even times when the signals cannot propagate at all. Monitoring systems can do their best to account for changes in the ionosphere, but an early warning system will help those who have to make very precise GPS measurements (not just you and your TomTom) plan their activities. EVE will measure the extreme-ultraviolet output of the sun on 10-second time scales, 30 times better than previous instruments could. Therefore, solar physicists will be able to better understand the signs and signals of a sun that is about to make our ionosphere dance around.
That’s it, for now, for my sciencey SDO posts. I can’t wait to see what new discoveries start rolling in when science operations begin. I’ll probably have one more SDO post soon, about the TweetUp itself!
How hard is it to calibrate out the ionosphere with these low-frequency instruments? At L-band, the VLBA does fine if there’s a calibration source in the primary beam. The low-frequency telescopes mostly have small dishes and so enormous primary beams, so that shouldn’t be a problem, but at what point do you run into problems where the piece of ionosphere you’re looking at your calibrator through is different from the piece of ionosphere you’re looking at your target through?
On the other hand, for amateur radio operators a thickening ionosphere is good news: higher-frequency signals (in this case more than about 10 MHz) will skip off the ionosphere and can be used to talk over the horizon.
Yeah, hams like the kind of ionosphere we despise 🙂
It’s true that you have lots of calibrators in your primary beam, but the isoplanatic patch is now much smaller than your primary beam. One way this has been dealt with at the VLA at 74 MHz is with a field-based phase calibration scheme, using multiple phase centers within the FOV to calibrate each “patch.” Many of the new instruments are not just wide-field, however, but all-sky, so that method needs to be pushed and tested out there.
I don’t think it’s clear yet how badly the ionosphere will affect the ability to get to EoR sensitivities, and that is what’s really interesting to me.
I’ve never done L-band, and that’s, I think, just before the ionosphere becomes a big factor? (I was from C up to… ugh, W… does anyone even call it that?)
Oh, one of the other things is that there are times when scintillation is so bad (again, from VLA74) that making any image was just impossible. That’s expected to happen more frequently as the sun starts to wake up in its cycle, but how much time that kills in a multi-month integration, I don’t know if it’ll make much of a difference.
How much of a factor the ionosphere is at L-band depends what you’re doing. If you’re just pointing a single dish at a pulsar, of course imaging distortions don’t bother you (and the ionosphere turns out not to affect dispersion measure and hence timing much, even at the microsecond level). If you want to use the VLBA to get a parallax good to 40 microarcseconds, well, then you’d better worry about the ionosphere. But for most pulsar applications, at L-band or even down at 350 MHz we just don’t worry about the ionosphere. Interstellar dispersion and scattering are much more serious problems, but obviously SDO won’t help us much with them…