Solar interference

Short baselines projected towards the Sun, particularly at sunrise and sunset, will detect solar fringes. This is seen in both UHF and L-band observations, but is generally more prominent at UHF. The worst affected baselines are

M000 - M001; M002 - M003; M003 - M000; M000 - M004; M000 - M017; M002 - M006; M003 - M029; M004 - M005; M011 - M020; M012 - M020; M018 - M020; M038 - M039

Continuum observations are generally minimally impacted by this effect. However, spectral line observers may be adversely affected, particularly if the science interest is in low surface brightness emission. In the case of relatively bright, unresolved sources, it is advised to simply flag these few baselines, the loss of which will have relatively low impact on overall sensitivity.

Figure 1 shows an example of the ripples in phase, which cease abruptly at sunset. Figure 2 shows the RMS amplitude across channels. More details can be found in the report on  Solar interference on short baselines in the UHF band.

 

Figure 1: Visibility phases on the M000-M002 (29 m) baseline during a stability observation on J0408-6545, started two hours before sunset. The x-axis is the time since the start of the observation. This data are at 8 second integration, with 32k channels across the band.

 

Figure 2: RMS of amplitudes computed over 50 channels at the lower end of the band plotted against time. Only the 25 baselines with the largest RMS are shown. The total fraction of baselines affected is less than 5%.

The affected baselines are those with the shortest projected lengths towards the Sun (Figure 3). These constitute less than 5% of the total number of baselines. These baselines can be flagged for the duration of the interference or, if they happen to be vital for a particular science case, observations should be scheduled after sunset.

 

Figure 3: The inner uv coverage (corresponding to the mid point of the observation) as computed towards the Sun. The colours represent the RMS of the visibility phases for each baseline calculated over the duration when the Sun is above the horizon (redder implies a higher RMS). The circle represents the first null of the jinc function corresponding to 1.3 times the optical solar disc.

 

 

 

Figures 4 and 5 below shows an example at L-band. These plots come from an observation on the DEEP2 field on 14 September 2018. Note the ripple which appears on short baselines after SAST 06:00 (sunrise). This is due to solar fringes and is clearly visible on shorter baselines.

Interference from DME at ~1100 MHz is clearly visible from early morning through to the late evening, but absent once local flights are no longer in the air.

Figure 4: Dynamic spectrum of amplitudes on a short baseline; M002-M003 (40 m).
Figure 5: Dynamic spectrum of amplitudes on a long baseline; M058-M063 (7700 m).