New observation mode, phase referencing (relative astrometry) mode, is available from the 2019A semester. Details are summarized in the following subsection.
KaVA phase-referencing mode is opened as a risk sharing. We have conﬁrmed that relative astrometric observation with KaVA can potentially achieve positional accuracies of ∼20 microarcseconds (µas) in right ascension and ∼40 µas in declination, where K-band (22 GHz) data was used for the evaluation (see Fig. 12 and Table 15)1.
In an astrometric observation, the observation using dual-beam for VERA and singlebeam for KVN is recommended. Since KVN should be operated by fast-switching mode, antenna scan diﬀerence between KVN and VERA should be cared in data reduction. For the data reduction, a ﬂag ﬁle is generated based on an antenna log and is provided to an observation PI. Also, it is strongly recommended to observe pair sources with small separation angle (e.g., less than 1 degree) at high elevation (see Table 16).
This will reduce the position errors caused by the residuals in atmospheric zenith delay. Note that Table 16 could be used to estimate accuracies of relative astrometry for VERA and KaVA, since the maximum baselines of both arrays are consistent with each other (i.e. positional accuracy of relative astrometry can be given by ∼ θsep c∆τ |B| where θsep is the separation angle between target and phase reference sources, c is the speed of light, |B| is the baseline length and ∆τ is the uncertainty in a delay measurement, see Reid & Honma 2014).
In the relative astrometric observation with KaVA, digital ﬁlter mode of “VERA1” (see Table 7) can be used in the case of pair observation of continuum sources (same as Fig. 12). Digital ﬁlter mode of “VERA7” or “VERA7MM” (see Table 7) could be used in the case of pair observation of line (e.g. H2O maser) and continuum sources, although the setup has been under evaluation for KaVA array. 2
Generally, users are encouraged to carefully carry out data reduction in consultation with a contact person in the KaVA project group. Procedure of astrometric data reduction for VERA data has been summarized in previous papers (e.g. Fig. 11 of Kurayama et al. 2011; Fig. 5 of Imai et al. 2012). Basically, the procedure of data reduction for KaVA data is consistent with that for VERA data, expect for some points. For instance, parallactic angle should be corrected for KVN array (e.g. with the NRAO AIPS task ”CLCOR” combined with the OPCODE = "PANG"), while no need to calibrate the parallactic angle for VERA array (ﬁeld rotator is equipped in the VERA array).
After the huge earthquake on 11 March 2011, the crustal motion at Mizusawa station is still unstable and may not be simply approximated by linear motion, and this could potentially degrade the astrometric accuracy.
1 We have not ﬁnished for evaluation of parallax accuracy with KaVA
2 The digital ﬁlter modes, “VERA7” and “VERA7MM”, have been used regularly for VERA astrometric observations.