Correcting Ionospheric Faraday Rotation for the VLA and MeerKAT

Abstract

We report here on studies to determine the accuracy of estimated corrections of Ionospheric Faraday Rotation Measure (IFRM) using observations of the Moon with the Very Large Array (VLA) and MeerKAT telescopes. To estimate the IFRM requires an estimate of the total electron content along the line-of-sight to the observed sources (the so-called Slant Total Electron Content, or STEC). Estimating the STEC requires an estimate of the global 2-D map of Vertical Total Electron Content (VTEC) along with the ray path from the telescope to the source. Traditionally, these global VTEC maps have been utilized along with an assumption that the electrons are in a thin shell at a given altitude to provide an estimate of the IFRM as a function of time. We find that this traditional technique significantly overestimates the IFRM - typically by 0.5 to 1.1 rad/m^2 for the VLA, and -0.3 rad/m^2 for MeerKAT. Alternatively, the software package ALBUS utilizes raw data from nearby Global Navigation Satellite System (GNSS) stations, to generate a local estimate of the IFRM as a function of time. ALBUS provides considerably better estimates of the IFRM - accurate to 0.1 rad/m^2 for both the VLA and MeerKAT, provided the stations utilized have known receiver bias values. A byproduct of our study is the establishment of the intrinsic electric vector position angle (EVPA) of the standard polarized calibrators 3C286 and 3C138 from 500 MHz to 50 GHz, using additional VLA observations of the Moon, Venus, and Mars.

Figures (16)

• Figure 1: Locations of the GNSS stations for the VLA (left) and MeerKAT (right) potentially utilized by ALBUS are shown by the small dots. The larger blue dots show the locations of the two arrays.
• Figure 2: Color-coded MeerKAT images of the Moon at 1002 MHz. The Stokes I brightness (left) is nearly featureless while the fractional polarized brightness (center) is strongly peaked toward the limb, with fractional polarization reaching 30% near the lunar limb. The EVPA vectors (right) are nearly radial. The slight ripple seen in the Stokes I image is from a deconvolution instability.
• Figure 3: Showing the image products from the C3 observation. The bright spot in the three leftmost panels is due to reflection of solar emission.
• Figure 4: Examples of the estimated lunar IFRMs by TECOR, using the 'jplg' gobal VTEC maps. Time is in UTC hours -- subtract seven for local MST. Sources are color-coded: red for the Moon, blue for DA240, purple for 3C303, and green for 3C345. The vertical line indicates time of local sunrise or (for the DC observation) sunset.
• Figure 5: The lunar IFRM estimates from three global and two local models, for the 'D2' observation (left) and the 'DC' observation (right). The three displayed global estimates are representative of all seven global maps we have investigated. The vertical dashed line marks sunrise or sunset. The plotted data points (in maroon, with error bars) are the actual IFRM values determined from the lunar observations.