-
Notifications
You must be signed in to change notification settings - Fork 1
/
referee.reply
154 lines (116 loc) · 8.21 KB
/
referee.reply
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
# Referee report replies for AA/2014/23668
# Editor comments:
# Also make sure that the figure and table captions are formatted in accordance
with the A&A instructions (see http://www.aanda.org/index.php?option=com_content&view=article&id=136&Itemid=200&lang=en_GB.utf8%2C+en_GB.UT#paper_organisation and
the current A&A macro for examples).
# - In your cover letter, please indicate precisely all the changes made in the revised version,
# - Mark all the changes clearly (using boldface) in your manuscript.
# Referee comments:
This paper describes the calibration process for the AARTFAAC ASM which is a
passive backend to the LOFAR telescope front-end hardware, designed for the
detection of (bright) transients in real-time. This paper is suitably complete in
its overview of the system, which is described nicely in plain style appropriate
for a general readership. I have no major issues with the draft as it stands but
only a few more minor comments. Some general observations:
(1) The order of the paper is not optimal. Figures are frequently referenced
(a long way) out of sequence. e.g. Sec 3.2, Fig. 10 is referenced before Fig. 2.
Reply:
- The forward references to Fig. 10 (in the last line of Sec. 3) and Fig. 11
(Sec. 4.1.3, second last line under Model source flux estimation, and again
in Sec. 4.1.3 above 'Bias of the WSF estimator') have been removed.
- All figures are now in the same or the next page of being referenced. Fig. 11
is too big to afford much movement, and has been placed 2 pages away by
latex.
(2) It is not made clear until quite far into the paper that the calibration
pipeline is not being tested on the actual AARTFAAC correlated data, since the
Uniboard system is not in place yet, but is instead being tested on a subset of
LOFAR data. I think this should be made clear earlier in the paper.
Reply: Added a sentence above Section 2 (The AARTFAAC All Sky Monitor System):
"The results presented in this paper have been obtained on test
observations carried out using existing LOFAR system infrastructure
while the AARTFAAC Uniboard \footnote{http://www.radionet-eu.org/uniboard}
based piggyback recording system was being built."
(3) The figures are not suitable for B&W viewing. The best example of this is
Figure 12, where it is not possible by eye to distinguish the raw and
difference image data when the figure is in B&W. It would be better to use
different shapes for the data markers.
Reply: The following figures have been changed to incorporate different symbols
and colors for the different parameters:
- Fig. 9: Temporal stability of the complex gain of a randomly chosen set
of 6 antennas from the 288 AARTFAAC antennas over a period of ~3 hrs.
- Fig. 11: Short term ionospheric ... separated by ~30MHz.
- Fig. 12: Reduction in ... increasing cadences.
Minor Comments:
Sec 1: para 1: LOFAR reference should be van Haarlem et~al. 2013
Reply: Corrected.
Sec 1: Table 1: There should be a reference in the table caption to Sec 2 where
the form of the FOM is explained.
Reply: Table footnote d added with a reference to Sec. 2, which includes the
definition.
Sec 2.1: para 5: the wording here is rather convoluted and could be clarified.
Reply : This paragraph attempts to justify spectral integration by AARTFAAC
while not de-dispersing.
The paragraph has been changed to:
"The AARTFAAC imager does not carry out DM measure searches via de-dispersion on trial DMs. Thus, attempts to increase the transient detection sensitivity of AARTFAAC by spectral integration over its full 13 MHz band would leads to a lowered detection SNR for any pulses with DMs greater than a few. This is because dispersion broadening would spread the pulse beyond the 1 second imaging cadence, restricting the transient search spatial radius. However, as stated above, scattering along Galactic lines of sight broadens pulses with DM ~100 to beyond 1 second. Thus, some level of spectral integration can be afforded by this limit before scattering and dispersion broaden the pulse beyond the AARTFAAC cadence. Finally, as shown by Hassall et al. (2013), de-dispersion is not beneficial for highly scattered sources. The scatter broadening of the pulse increases more rapidly with DM than the dispersive delay. At high DMs, this makes the inherent pulse width (due to scatter broadening)larger than the dispersion broadening. This would allow the AARTFAAC to probe the high DM domain with full spectral sensitivity."
Sec 2.1: para 8: confusion limit calculations are model dependent. The assumed
source counts model should be cited, as should the specific definition
of confusion used here since this varies in the literature.
Reply : The following was added:
"Assuming the confusion noise limit to be reached with the presence
of one source per ten PSFs and the source counts from
\citet{bregman2012system}, results in a classical confusion noise of
$\sim$$10$ Jy."
Sec 2.1: para 9: w-projection should be referenced.
Reply : Reference added.
Sec 3.2: para 1: "temporal variation in the PSF" - I am not sure exactly what
this refers to. Do the authors mean temporal variation in the PB, i.e.
A-projection?
Reply : Yes, the PSF should be replaced by PB. The implication is that
calibration for snapshot imaging from a coplanar array can be carried out
more efficiently by implementing a stripped down calibration and imaging
algorithm than reusing available tools, which are overkill for this
application.
The relevant sentence has been modified to:
"Their algorithms are further complicated by the need to compensate for higher order effects like the temporal variation of the primary beam
during long synthesis due to changing array geometry, or the beam
rotation of altaz mounts etc. which are moot in our application."}
Sec 4: bullet 5: this reference to bandpass calibration is made before bandpass
calibration is described. In neither section is the stability of the
bandpass discussed.
Reply: The forward reference to bandpass calibration is part of the algorithmic
summary. In section 4.1.5, the bandpass temporal variation is sampled
per channel and per second by calibration, and hence can be unilaterally
corrected. Thus, bandpass stability is not discussed.
Sec 4.1.1.:para 1: "brightest sources in the sky" -> "brightest sources in the
Northern sky"
Reply : Corrected.
Sec 5.1: para 4: when referring to Fig. 7, "the absence of the Galactic centre
due to spatial filtering". Where is this filtering described? Is it a
consequence of the calibration method?
Reply : Yes, the filtering of the low spatial frequencies is carried out during
calibration in order to use a point source model sky for
selfcalibration.
This is described in Sec. 4.1.1 (Model generation), paragraph 2:
"Galactic emission, having a steep spectrum, is extremely bright
at LOFAR frequencies and also difficult to model due to the detailed
structure resolvable at the AARTFAAC's resolution and sensitivity.
Hence, we filter out the low spatial frequencies to suppress this
emission during calibration. The model can then account for a large
fraction of the received flux on the filtered visibilities, and
is constrained by the fluxes and positions of the model sources,
as well as the noise model."
--------------- Other corrections ---------------
- Added a reference to a proceedings:
@inproceedings{salvini2014stefursi,
title={StEFCal - An Alternating Direction Implicit Method for Fast Full Polarization Array Calibration},
author={Salvini, Stef and Wijnholds, S. J.},
Booktitle={Proc. URSI General Assembly, Beijing},
year={2014}
}
- Acknowledgements: "The Netherlands Foundation for Radio Astronomy" -> "the Netherlands Institute for Radio Astronomy"
- Clarified (in Section 4.1.3) that WSF source position estimation is carried out
just once, and not in every major cycle:
"The cycle begins with a one-time estimation of the model source positions via
the WSF algorithm."
- Reordered the WSF description to be above the model source flux estimations.
- Modified Fig. 2 to show WSF being done only once, outside the major cycle.