1: *********************Referee's Report*************************************
2:
3: The paper 'Evidence for Low Redshift Origin of the Source Subtracted Near
4: Infrared Background' by Thomson et al. shows that the fluctuation power in
5: the HUDF at 1.1 and 1.6 \mum after subtraction of all detectable sources is
6: inconsistent with a large contribution from galaxies at z>8. This conclusion is
7: obtained by comparing the color of residual fluctuation in the 1.1 and 1.6
8: \mum HUDF images with numerically redshifted galaxy SED. Moreover this
9: conclusion is supported by the comparison with recent background fluctuation
10: detection at 3.6 and 4.5 \mum obtained with Spitzer.
11:
12: The paper is timely and original. It warrants publication in ApJ but I do
13: not see the urgency for a publication as a Letter (see also my GENERAL
14: COMMENT a).
15:
16:
17:
18: 1. GENERAL COMMENT
19: a) I find the paper a quite difficult read: the main difficulty arises
20: from the fact that many important pieces of information are not explicitely
21: given in the paper. The reader is often referred to previous papers of the
22: same authors in order to relevant details about data analysis, source
23: subtraction, source redshift distribution, background intensity, SED
24: calculation,... Moreover, the introduction section is too short and generic
25: (see SPECIFIC MAJOR POINT a) .
26: I guess that this problem is partly due to the tight constraints set by the
27: ApJ Letter format. I suggest that the authors consider the idea of expanding
28: the paper and re-submitting it for publication in the ApJ Part 1.
29:
30: REPLY: The referee is correct that we condensed several areas to fit within
31: the Letter constraints. In responding to the referee's suggestions we have
32: relaxed that constraint and the current length is probably beyond the letter
33: allocation. We presume that the editor will send it to the main journal if
34: that is the case.
35:
36: b) I am confused by authors use of the expression 'low redshift galaxies'.
37: The final conclusion of the paper is that fluctuations from unresolved
38: sources can be due to faint galaxies in the redshift range z=1-8. For a large
39: part of the astrophysical community, galaxies at z>3 are high-redshift objects!
40: I suggest to change the expression 'faint, low redshift galaxies' with
41: 'faint galaxies in the redshift range 1<z<8'. Also, I suggest to change the
42: title accordingly.
43:
44: REPLY: We have changed the references to "low redshift galaxies" to "z<8".
45: Our original term was in response to suggestions that the galaxies responsible
46: for the source subtracted fluctuations were at redshifts above 10.
47:
48: 2. SPECIFIC MAJOR COMMENTS
49: a) The introduction should include sufficient introductory material and
50: background explanation for readers who are not specialists in the field.
51: Some more details should be given about the existence (or not) of a NIRB
52: intensity excess and about the detections of residual fluctuations at
53: different wavelengths.
54:
55: REPLY: We have added details of the analysis in our first paper in the
56: introduction.
57:
58: Moreover, a brief summary of the discussion on the interpretation of the
59: residual fluctuation origin appears necessary. Stating that "the
60: interpretation (of Kashlinsky et al. 2007a) has been challenged by Cooray et
61: al, with a reply to the challenge by Kashlinsky et al. (2007)" does not
62: help the reader to understand what are the main points of the discussion.
63:
64: REPLY: We have added a discussion of the interpretation of the fluctuations
65: to the introduction.
66:
67:
68: Finally, reference to other works that have provided different interpretations
69: should be given. For example, Salvaterra+ (2006, MNRAS, 368, L6) have shown that
70: the unresolved fluctuations observed by Spitzer could be accounted for by
71: normal (i.e. hosting PopII stars) galaxies at z>=5. B.t.w., this conclusion
72: is not at odds with (in fact, it supports) the main conclusions of the present
73: paper.
74:
75: REPLY: We have added a reference to the Salvaterra 2006 work in the introduction.
76: Thanks to the referee for calling it to our intention. The paper came out during
77: the extensive discussions with the referee on the first paper. We had intended
78: to include it in this paper but it got missed.
79:
80: b) Expand Sect. 2.1. Since a crucial confirmation of the conclusions
81: comes from the comparison of HUDF and Spitzer residual fluctuation,
82: some more details on the compatibility of different data sets are needed.
83: In particular, the use of ACS images to remove sources from NICMOS images
84: should be illustrated more extensively. Are the subtraction levels really
85: comparable? How does the result depend on the adopted SED? Could the excess of
86: fluctuations in the 3.6 \mum predicted for template 6 be caused by
87: a different subtraction limit?
88:
89: REPLY: We have greatly expanded the discussion on the subtraction levels
90: in section 3.1 which has been moved from its original position to after
91: the discussion of the fluctuation colors. We have added a figure (Figure 4)
92: to show the subtraction levels as a function of redshift for the F775W band,
93: the band we had shown the detection levels for in the Thompson et al. 2006
94: NUDF paper. The answer is that the subtraction levels are equal except for
95: the very highest redshift galaxies that have no flux in the ACS bands.
96:
97: c) Give more details on the adopted SED templates: e.g. specify the adopted
98: IMF and metallicity values. How do authors model IGM absorption?
99:
100: REPLY: We have added more details on the adopted SED templates in section 3.
101: We also added figure 2 to show the template distribution and Table 1 that gives
102: the SED properties. We have also added a reference to Madau et al. 1996
103: which is the source of our calculation of the IGM absorption.
104:
105: d) The color results are quite insensitive to the actual redshift of the main
106: sources of fluctuations, as long as their redshift is <8. Can the authors
107: speculate on what theoretical and/or observational features may be useful
108: to disentagle low-redshift (i.e. z<5) and high redshift (z>5) contributions?
109:
110: REPLY: It may be possible to do a fluctuation analysis on the ACS images
111: as well to get color information to do a fluctuation photometric redshift.
112: Since we used the HUDF team prepared ACS images we did not have
113: the very detailed knowledge of how they were produced and felt that analysis
114: was beyond the scope of this paper.
115:
116: d) In Fig. 3 an upper limit of the fluctuation power from 1.1 mum dropout is
117: given. As stated in Sect. 3.1, this limit is still above the Gaussian noise
118: spectrum. Quantify the fluctuation excess that is possibly due to very high
119: redshift galaxies on the basis of this finding.
120:
121: REPLY: We have added that discussion as the fourth paragraph in section 3.2.
122:
123:
124: 3. MINOR COMMENTS
125: a) The quality of Fig. 2 is very low. Please cut the y-axis to a value of ~10.
126: in order to improve the presentation of the results.
127:
128: REPLY: We have cut the axis to a value of 10.
129:
130: b) Sect. 3: '.. the predominant SED for faint galaxies at Z>0.5'. Is Z
131: metallicity or redshift?
132:
133: REPLY: It is redshift and we now indicate so in the text.
134:
135: c) Please, check carefully the text. I find many typos. Here I list just a few
136: examples:
137: - Sect. 3: \mu instead of \mu m
138: - Sect. 3.1 :'...we cannot make a the same argument...'
139: - References: angstrom instead of A&A
140:
141: REPLY: We have corrected these and other typos. Thanks to the referee for the
142: very helpful comments.
143: