Mem. S.A.It. Vol. 81, 63
SAIt 2010 c
MemoriedellaDi use UV background: GALEX results
Richard Conn Henry
Henry A. Rowland Department of Physics & Astronomy, The Johns Hopkins University, 3400 North Charles Street, Baltimore, Maryland, 21218-2686 USA
e-mail: henry@jhu.edu
Abstract.
A bright UV GALEX image in the direction of a dense high galactic latitude interstellar dust cloud is examined to test (and to reject) the idea that a bright extragalactic UV background radiation field exists. A GALEX “Deep Imaging Survey” image of a second high latitude region (a region almost totally free of dust) shows a similar bright background, which, clearly, cannot be due to starlight scattered from interstellar dust. I speculate that the background is due to dark matter particles interacting with interstellar gas/dust nucleons.
Key words.
techniques: ultraviolet – cosmology: diffuse background
1. Introduction
Sujatha, Murthy, Karnataki, Henry, & Bianchi (2009) have used the GALEX ultraviolet im- agers to study the diffuse UV background at the high-galactic-latitude location “SANDAGE,”
discovered by Allan Sandage (1976) to be ex- tremely dusty (abscissa, Fig. 1). They showed that the large observed signal (∼750 photons cm
−2s
−1sr
−1Å
−1) might be due, in the main, to starlight scattered from dust. In the present paper I use these same data to test the sug- gestion by Henry (1991, 1999) that the ubiq- uitous UV background observed longward of 1216 Å is of extragalactic origin. (This target, at b = 38 .
◦6, was proposed to NASA expressly for the purpose of testing Henry’s idea.)
Fig. 3 gives the values of E(B-V) across SANDAGE. If I assume that the only signal from this direction is a uniform extragalac- tic (i.e., from beyond the ∼ 100 pc distance of the dust) background (and that the dust is forward-scattering) I predict that GALEX should see the FUV image shown in Fig. 4.
In fact, the actual GALEX FUV image ap- pears in Fig. 5. It bears no resemblance to our prediction. Because the dust is optically thick, starlight from foreground stars back-scattering from the dust could not “fill the hole” that is predicted under the extragalactic hypothesis, which we can thus now rule out.
2. A high latitude dust-free target The GALEX FUV image of the diffuse back- ground at SANDAGE is remarkable in appear- ance, but is nevertheless similar in appearance to every other such image that I have seen.
What can be the source of this strange back-
ground radiation? Could it be starlight scat-
tered from dust? To find out, consider Deep
Imaging Survey target ELAISS1 00 which
(Fig. 6) is almost free of interstellar dust. The
image (Fig. 7), despite the lack of dust, closely
resembles that of SANDAGE. I have written a
simple fortran program to model the expected
dust-scattered starlight. For a dust albedo of
0.28 and a Henyey-Greenstein scattering pa-
0.0 0.1 0.2 0.3 0.4 0.5 600
650 700 750 800 850 900
GALEX FUV Brightness
E(B-V)
Fig. 1. SANDAGE: lack of correlation of FUV (∼
1500 Å) GALEX brightness with E(B-V).
rameter g = 0.61 (Sujatha et. al. 2007), I calcu- late an expected background at 1500 Å for this target of only 17 photons cm
−2s
−1sr
−1Å
−1.
There are GALEX images in which dust- scattered starlight is clearly identifiable (Henry 2006), but the ELAISS1 00 data prove that the ubiquitous strong cosmic FUV background is not due to dust-scattered starlight. And, as we have just seen, that background cannot be ex- tragalactic in its origin).
Fig. 8 shows the correlation between the NUV and FUV intensities for ELAISS1 00.
The correlation coefficient is 0.63. I have made no correction for the large zodiacal light con- tribution to the NUV intensity (the FUV image has no zodiacal light contamination).
3. Source of the UV background I began my investigation of the UV back- ground in Henry (1973). By 1995 the situation was as shown in Fig. 9—which allows one to understand, I hope, why I believed the radiation longward of 1216 Å to be redshifted recombination radiation from the intergalactic medium. But, the SANDAGE data show that idea to be wrong. What, then, is the source of the radiation? It cannot be dark matter annihilation radiation, for the photon energy would be much too high—also, most such radiation would originate beyond the SANDAGE cloud. I suggest that the radiation results from collisions of dark matter particles with interstellar medium nucleons. The weak interaction conveys energy to the electrically-
140°
145°
150°
45°
45°
40°
40°
35°
35°
Fig. 2. SANDAGE: circle: GALEX fov; rectangle:
Voyager spectrometer slit; white dots: TD1 stars; red color: IRAS E(B-V). δ = 70 .
◦4
charged quarks, which radiate. See the lattice gauge models of Derek B. Leinweber:
http://www.physics.adelaide.edu.au/theory/→
staff/leinweber/VisualQCD/Nobel/ Because the interaction cross section is expected to go as the square of the atomic number, hydrogen, helium, and metals would contribute about equally to the background radiation.
4. Contamination in GALEX images Signal in both the FUV and NUV GALEX imagers (while pointed at a fixed target) al- ways decline until local midnight, and then rise as dawn approaches (see Fig. 2 of Sujatha et al. 2009). Murthy (private communication) has discovered that the magnitude of the decline correlates with solar activity. The GALEX im- ages in Fig. 5 and Fig. 7 look like they could be 100% contamination. (The NUV images are seriously contaminated with zodiacal light, so in the present paper I have largely confined my attention to the FUV images.)
The FUV imager may transmit some OI
1356 Å airglow, and there could also be a back-
ground due to particle fluorescence in the win-
dow. Each of these would be expected to vary
with both time and solar activity. In the next
section, however, I will demonstrate that only
a small fraction of what appears in Fig. 5 and
9 20.4
h m
9 22.4
h m
9 24.4
h m
9 26.4
h m
9 28.4
h m
9 30.4
h m
9 32.4
h m
9 34.4
h m
69 47
o69 59
o70 11
o70 23
o70 35
o70 47
o70 59
oDeclination
Right Ascension
SANDAGE E(B-V)
.307.301.284.267.251.240.248 .296
.257.275
.242.255.265.288.300.298.275.247.221.217.228.244.244 .241.246
.241.251.262.262.273.283.293.303.299.269.229.210.201.214.234.231.223 .249.258.269.275.280.283.287.295.300.287.263.229.199.197.210.220.216.202.186 .248.261.270.278.281.279.280.279.283.288.273.244.217.194.195.202.205.195.176.173.173 .243.254
.160.165 .160.156 .182.168 .187.185 .202.191 .261.235 .264.269 .264.260 .273.265 .272.275 .263.268 .257.261
.145.144 .145.144 .159.155 .169 .181.176 .197 .218 .257.250 .250 .244 .253.242 .260 .266 .261.267 .263.260 .268.267
.131.131 .137.134 .138 .151.145 .165.157 .173 .204.182 .238 .252 .242 .230.232 .234 .255.243 .252.256 .252.248 .270.260
.127.129 .132.128 .137 .149.142 .153 .167.160 .197.174 .225 .243.250 .223.232 .223.220 .240.236 .243.245 .248.243 .264.261
.128.132 .133.129 .143.136 .156.149 .168.162 .194.178 .248.228 .237.249 .215.228 .212 .217 .237.230 .243 .246 .255.249 .250.254
.134.141 .135.133 .140.137 .158.149 .173.167 .194.178 .251.220 .253.259 .238 .213.227 .220.214 .244.235 .253 .260.261 .241.252
.158 .142.152 .138.141 .137 .146.138 .170.156 .181.175 .223.198 .264.250 .263.272 .238.249 .224 .234 .251.240 .266 .268 .255.267 .222.238
.162.166 .147.156 .137.143 .140.135 .167.155 .185.178 .199 .251.227 .286.272 .282.291 .254.276 .266.264 .280.273 .278 .254.271 .223.235
.171 .162.170 .144.156 .130.136 .147.136 .174.166 .206.186 .230 .281.262 .313.301 .293.316 .295 .296.297 .288.298 .261.277 .229.243 .221.222
.163.166 .142.155 .131 .125 .140.127 .169.156 .197.180 .225 .255 .307.281 .328 .327.334 .325.319 .308.321 .268.291 .232.250 .223 .222 .226
.150 .144.151 .126.133 .121.120 .145.133 .169.160 .210.184 .279.234 .307 .344 .356 .346.356 .327.339 .313 .258.291 .226.239 .222.221 .226
.132.134 .121.127 .117.117 .129.120 .149.142 .162 .187.166 .229 .264 .312 .378.346 .376.384 .326.347 .273.297 .232.250 .216.220 .216 .218
.117.115 .112.114 .114.112 .131.122 .147.138 .158.150 .206.175 .306.269 .353 .388 .405 .368.391 .314 .257.278 .221.232 .204.211 .204.202
.104.101 .103.106 .111.105 .123 .140.131 .147.146 .152 .200.163 .292.249 .330 .404.369 .381.412 .268.325 .215.230 .202.208 .191.197 .185
.091 .095.094 .113.100 .140.125 .154.149 .157.154 .194.169 .278.235 .345.302 .398.368 .336.383 .266 .221 .193.195 .194 .191 .184 .167.172
.086.083 .093 .121.105 .149.137 .160.157 .168.164 .180 .244.207 .268 .291 .339.305 .366 .356 .316 .203.261 .178.179 .186.185 .168.179
.096.086 .126.109 .151.142 .159.154 .174.166 .208.186 .268.238 .279 .284.282 .313.313 .235.296 .164.193 .170.162 .176 .162.170
.106.096 .131.118 .145.141 .150.146 .170.161 .204.186 .267.239 .262.277 .251.253 .255.265 .217 .155.183 .150.150 .156.152 .153
.129.120 .136.134 .134.134 .150.137 .174.165 .220.200 .267.255 .252 .210.228 .215.217 .171.195 .151 .138 .131.133 .133
.133.117.116.118.123.132.144.167.184.189.180.185.206.234.247.235.210.179.168.149.140.132.125.126.131.134.133 .135.134 .133 .122.125 .129.124 .143.134 .157 .178 .220.207 .192.217 .171 .169 .184.178 .151.168 .123.134 .110.115
.135.138 .122.128 .119.119 .119.118 .131 .188.146 .200.201 .163.179 .171.162 .177.184 .163 .132.139
.110.121.132.139.153.152.161.163.166.171.164.152.180.174.162.149.180.171.156.143.135.169.157.149.143.156.136.151.147.145.142.156.153.154.153.156.167.162.164.163.184.161.177.169.159.156.190.171.157.145.164.130.149.131.112.111.139.114.101.099.105.090.096.093.093.088.104.097.097.100.107.104.107.108.112.110.112.115.113.112.119.116.125
average = 0.214
maximum = 0.412
minimum = 0.100
Fig. 3. Values of E(B-V) for GALEX SANDAGE (Schlegel, Finkbeiner, & Davis 1998). The heavy line (one degree diameter) shows the portion of the data that can be interpreted.
Fig. 4. Predicted GALEX FUV image if the back- ground radiation comes from beyond the dust. (Any additional spatially uniform contributor would re-
duce the predicted contrast.) Fig. 5. The actual GALEX FUV image bears no re-
semblance to the prediction of Fig. 4.
Photons cmssr-2-1-1-1Aº
SANDAGE
500 550 600 650 700 750 800 850 900 950 1000 1050 1100 1150 1200
1000.
900.
800.
700.
600.
500.
400.
300.
200.
100.
0.
-100.
-200.
-300.
-400.
-500.
Wavelength in Aº
Fig. 11. Voyager UV spectrum of the diffuse cosmic background in the direction of SANDAGE (Murthy et al. 1999). Murthy et al. assumed that there is no cosmic signal short of the interstellar hydrogen ionization edge. We see that for SANDAGE there is also no cosmic signal longward of that edge, but beyond ∼1000 Å a signal appears which rises to meet the observed GALEX SANDAGE image intensity (yellow band) of Fig. 1. This strongly suggests that there is minimal contamination in the GALEX image.
0.000 0.002 0.024 0.006 0.008 0.010
375 400 425 450 475
GALEX FUV Brightness
E(B-V)
Fig. 6. A strong FUV background is seen at this
b = −79 .◦9 ELAISS1 00 location (which is almost free of dust). Error bar, as in Fig. 1, shows statistical uncertainty in a typical data point.
Fig. 7 can, in fact, be due to contamination: we are mostly seeing cosmic signal.
Fig. 7. The pedestal shows where data have been omitted from the ELAISS1 00 image.
5. Discussion
In Fig. 11, I plot the Voyager spectrum for
SANDAGE (Murthy et al. 1999). Fig. 2 lo-
cates the Voyager spectrometer slit (28 Å reso-
380 400 420 440 980
1000 1020 1040 1060
NUV Brightness
FUV Brightness 0.63
Fig. 8. There is significant correlation between the FUV (1350-1750 Å) and NUV (1750-2750 Å) in- tensities for the ELAISS1 00 images.
1000 1200 1400 1600 1800 2000 2200 2400 2600 2800 3000
0 100 200 300 400 500 600
Photons cmssr-2-1-1-1Aº
Wavelength in Aº F A
J
K J
D S
W H
W A
T
M
Fig. 9. Henry & Murthy (1995) exhibited this strange spectrum of the diffuse UV background ra- diation. The red triangle is the Voyager upper limit of 30 photons cm
−2s
−1sr
−1Å
−1that has been de- fended by Murthy et al. (2001). The Aries rocket spectrum (A) by one of my students (Anderson et al.
1979) shows continuum radiation, as does (at longer wavelengths) the Aries rocket spectrum (T) by an- other of my students (Tennyson et al. 1988); we cor- rected the Tennyson et al. observation meticulously for airglow, and for zodiacal light.
lution) on SANDAGE. The Voyager spacecraft was powered by an RTG nuclear source that contributed a strong wavelength-independent background to the UV spectrometers. Murthy et al. subtracted a wavelength-independent background— just enough to give zero signal shortward of 912 Å. But now notice that from
+90º
+60º
+30º
0º
-30º
-60º -90º
-60º -30º 0º
+30º +60º
Fig. 10. This Galactic-coordinate map shows the Voyager (Murthy et al. 1999) diffuse UV back- ground at ∼1000 Å. Filled circles show dust- scattered starlight; open circles have only an upper limit (as in Fig. 9). So, at ∼1000 Å the UV back- ground shows high contrast, while the GALEX im- ages at ∼1500 Å show, thus far, no locations with low background.
912 Å, to beyond 1000 Å, there is also no sig- nal: a clear indication that a) the subtraction of background has been done right, and b) there truly is no celestial background radiation just longward of the interstellar hydrogen ioniza- tion limit. Finally, notice that the Voyager spec- trum then rises toward the level (yellow band) that we see in our GALEX image of this target.
Our GALEX SANDAGE image covers the range 1350 Å to 1750 Å with average signal about 750 photon units (top of yellow band) which should be reduced to ∼600 photon units (bottom of yellow band) to allow for contam- ination, following Murthy’s correlation. This Voyager spectrum surely rules out any larger contribution from geophysical contamination to the GALEX image. The ELAISS1 00 data shown in Fig. 6 were acquired in 2003, when solar activity was high. A similar plot for ob- servation of the same target in late 2006, when solar activity was much lower, shows a decline to about 400 photon units—hardly a substan- tial decrease.
Finally, the data shown in Fig. 9 include no GALEX data, and are from a wide variety of instruments. Like GALEX, they show a back- ground of order 400 photon units in every case.
So, there is a ubiquitous cosmic background longward of about 1200 Å, and it has the ap- pearance that is exhibited in Fig. 5 and Fig.
7—which look unlike any other astronomical
photographs.
6. Conclusions
I think in Figs. 5 and 7 we are, for the first time, seeing the bulk of the interstellar medium by means of radiation (“radiative corrections”) from dark matter particles impacting interstel- lar medium nucleons. Assuming that the nu- cleons are destroyed in the interaction (which may not be the case), some thousands of UV photons must be emitted per interaction, if the interstellar medium is not to be depleted over the age of the universe.
Our FUV background, we know, continues into the NUV (Tennyson et al. 1988) and, per- haps, even into the visible (Henry 1999)—most of the photons in the Hubble Deep Field image come, not from the galaxies, but from a dif- fuse optical background, which, I suggest, is the continuation to the optical of the spectrum that appears in Figs. 9 and 11.
Acknowledgements. I am grateful to my colleagues,
