Daily Image
07-03-2014
A scintillating radio movie
| Submitter: | Ger de Bruyn and the LOFAR EoR team |
| Description: | The two movies above each show a series of 720 images, 6 hours at 30s time resolution, of LOFAR high-band observations of the same EoR-3C196 field, taken in December 2013. The movies are 600 times faster than real time. Both show the same field of view (2.5x 2.5 deg) with a 3' PSF. If you stare at them for a while you will notice that the right images are rather 'unstable', as compared to those on the left. So what is causing the difference? We are all familiar with optical seeing: the intensity (and position) fluctuations of stars due to light propagation through the turbulent neutral troposphere. It is also called scintillation. This phenomenon can also occur at radio wavelengths where it is caused by turbulence in the ionized plasma between source and observer: the Galactic interstellar medium, the interplanetary medium and the ionosphere. Ionospheric scintillation is a relatively rare phenomenon. If it occurs, it is mostly in the late evening to midnight hours, but it can persist for many, many hours. During LOFAR Cycle-0 observations, ionospheric scintillation was relatively rare, so we got spoiled. However, since November 2013 we see a marked increase in the frequency of ionospheric scintillations, probably related to enhanced Solar activity during the secondary maximum in this Solar cycle. Most discrete radio sources will exhibit the phenomenon and can easily exhibit intensity variations of 20% (rms) at 150 MHz, accompanied by large image motion. Both effects combined make it impossible to make high dynamic range images and these data will therefore not be useful for the primary goals of the LOFAR EoR project. The movies give us fascinating insights into small-scale ionospheric phenomena. During periods of scintillation the lateral scale of significant (say 1 radian) phase fluctuations in the ionosphere approaches the projected Fresnel scale at the height of the ionosphere which is a few km at 2m wavelength. The wavefront then gets wrinkled so much that different parts of the wavefront interfere (either constructively or destructively) on the ground. Both the intensity and the apparent position of the source then vary. Some technical details on the making of the movies: To avoid that the images are dominated by the side-lobes of 3C196 - an 80 Jy source - we have peeled (= self calibrated and subtracted) it from the data. The residual visibilities were imaged using Sarod Yatawatta's impressive ExCon imager. Snapshot images were combined into a movie using scripts created by V. Pandey and Vibor Jelic. By using only the core stations, yielding a resolution of about 3', the snapshot images could be made without selfcalibration (which would be a real challenge anyway !). To reduce the array side-lobe pattern, a total of 10 sub bands (of 0.18 MHz each), spanning the frequency range from 115-160 MHz, were combined. During the periods of very rapid image motion (right movie) the images become blurred. This is partly because ionospheric refraction scales inversely with the square of the frequency. The careful observer can also see the side-lobes of the brightest sources rotate around them during the 6 hour synthesis time. During some of those periods, the signals of 3C196 become decorrellated, and very faint (1-2%) residuals of 3C196 are then visible in the lower left corner (at 08h13+48d13'). At least a dozen sources can be detected in the image, with flux densities ranging from 1 - 4 Jy. The sources show mostly uncorrelated movements and intensity fluctuations. There is obviously a wealth of information in these movies on the motions and turbulence in the ionosphere on angular scales down to 30', which corresponds to linear scales of 3 - 4 km in the ionosphere at an altitude of 300 km. |
| Copyright: | Ger de Bruyn |
| Tweet |  |