My group from Williams College, including my colleague Bryce Babcock and 8 students, successfully observed the annular eclipse from the National Radio Astronomy Observatory's Jansky Very Large Array, sixty miles west of Socorro New Mexico. I was working with Dale Gary of the New Jersey Institute of Technology, Tim Bastian of NRAO, and Stephen White of Air Force Research Laboratory to make high-resolution radio observations of the annular eclipse with the JVLA and with the CARMA array in the Inyo Mountains of California. I was also working with Dr. Thomas Kuiper of JPL on using the 34-m Goldstone antenna, also to get high-resolution solar observations as the advancing edge of the moon defined the time and position on the Sun of solar active regions as seen in the radio, improving on capabilities outside of eclipse. Our observations were successful, and it will take months to analyze the data.
The following summary is based on evaluations by Dale Gary of our group at the JVLA and Alphonse Sterling of NASA, a Hinode scientist:
What we did was looking at the JVLA is to take radio images of an active region (AR), using the normally relatively-poor radio spatial resolution. We are then using the Moon's limb as a knife edge; as the Moon moves, we will look for sudden changes in the radio intensity, such as a sudden drop-off over some frequency range. Via the timing we will know the location on the AR of that feature to better than the normal radio spatial resolution. By using Hinode and other opticla imaging spacecraft and ground-based telescopes, we can see precisely what that feature was.
Another advantage in addition to spatial resolution is that we also get better image quality because by differencing the signal we get only signals from the narrow window covered by the Moon. (At the frequencies we will be using, the dominant signal is from the AR, rather than from the quiet-sun corona north and south of the AR along the sliver that represents the differentiation.) The differencing removes the confusion from the entire remainder of the disk.
At the frequencies we are using (corresponding to wavelengths from 20 cm down to 6 mm), the resolution is normally 6-12 arcsec. The limb of the Moon will move about 0.5" per s, so if we have a stable enough signal we could get better than a factor of 10 improvement in resolution.