Test observations below 1200 MHz have recently been carried out
with the EVLA to assess the feasibility of pursuing spectral-line
observations for red-shifted HI 21cm or other spectral-line
transitions at frequencies where the VLA has severe limitations and/or
simply cannot observe.
We have observed two known HI 21cm absorbers, one at z=0.2467
(~1139 MHz) and the other at z=0.3127 (~1082 MHz). Each target was
observed for about 45 minutes in B configuration using bandwidths of
3.125 MHz with 127 spectral channels. Both VLA and EVLA antennas (a
total of 26) were used for the source at 1139 MHz. The inclusion of
the VLA antennas in these observations was possible because the
frequency falls within one of the "good VLA frequency windows" (see
Figure 4 in EVLA
Memo #119 for the sensitivity of the VLA between 1100 and 1225
MHz. The source at 1082 MHz, however, was observed using only the EVLA
antennas (total of 13), because the VLA antennas cannot be tuned to
frequencies below 1100 MHz.
Figures 1 and 2 show the HI absorption spectra of the sources at
1139 and 1082 MHz, with rms noise values of 4.7 and 10 mJy,
respectively. We note that sensitivities below 1200 MHz will improve
dramatically when the final L-band receivers are installed through
2009.
Figure 1: The HI 21cm absorption line toward the source PKS
1413+135 at z=0.25 with the EVLA+VLA
Figure 2: The HI 21cm absorption line toward the source PKS
1127-145 at z=0.31 with the EVLA
Figure 3 shows a representation of the point source rms values
obtained by observing a blank field every 5 MHz between 1000 and 1200
MHz. Values obtained at 1350 and 1440 MHz are also included for
comparison purposes. The observations were carried out at night time
with 10 EVLA antennas in CnB configuration. The imaged bandwidth at
each frequency setting is 24 kHz and the integration time is 100
sec. It should be noted that for observations below 1200 MHz, only IF
pair B and D can be used at this time. Furthermore, when loading the
data into AIPS, make sure online flags are ignored by specifying
CPARM(3)=16 in FILLM. Finally, we note that the polarization
performance below 1200 MHz is severely compromised, with the
polarization impurity reaching 40% at 1 GHz.
Figure 3: The point source rms values for L-band frequencies
between 1000 and 1440 MHz.
These tests show that successful observations are possible with the
EVLA at frequencies where the signals are believed to be dominated by
RFI due to aeronautical navigation transponders. Interference from
these transmissions is more evident on spacings of ~1 km or less.
Thus the shorter spacings in extended configurations, and all spacings
for the D-configuration, are affected. Recent tests carried out
during day-time hours with the D-configuration at 1139 and 1082 MHz,
using a bandwidth of 3.125 MHz, show that obtaining useful
astronomical results is still possible at these frequencies, but that
it is much harder compared to other array configurations. This
assumes, however, that the targeted spectral-line feature is
sufficiently narrow (a few 10 km/s), and does not fall directly on any
of the frequencies used by the transponders. Because of the above
mentioned RFI, D-configuration day-time observations of sources with
wide spectral features (a few 100 km/s) are not recommended in the
frequency range 1028 to 1150 MHz. However, except for 1090 ± 3 MHz,
such observations can be carried out at night.
NRAO will consider proposals to use this new 1.0 - 1.2 GHz
capability on the EVLA. It will be assumed that all proposals
submitted are suitable for dynamic scheduling unless explicitly
justified in the proposal and in the "comments" field of the sessions
section of the cover sheet. At present this excludes the use of
subarrays, although this may change in the future. Potential users
are advised that these are "shared risk" observations. We cannot
guarantee success, especially for daytime observations that might be
severely affected by RFI, but we will provide reasonable assistance.
Users are also reminded of the aliasing of continuum emission in the
bottom 0.7 MHz of baseband that affects narrow-bandwidth observing
with the EVLA antennas, and should consult the EVLA returns web page
for recommendations on its mitigation.
Traditional observe file preparation software such as observe and
JObserve do not support frequencies in this new range. We recommend
those preparing an observe file for observations in this new frequency
range to prepare it for frequencies in the traditional L-band range
first, and then hand-edit the resulting observe file with the LO
settings given by our online
LOSER utility. We strongly recommend that any observe file thus
created be sent to vlahelp@nrao.edu prior to
submitting it.
Technical questions about observing below 1200 MHz should be
directed to E. Momjian.
|