18-Months Operation of Lunar-based Ultraviolet Telescope: A Highly Stable Photometric Performance
J. Wang, X. M. Meng, X. H. Han, H. B. Cai, L. Cao, J. S. Deng, Y. L. Qiu, S. Wang, J. Y. Wei, J. Y. Hu
aa r X i v : . [ a s t r o - ph . I M ] O c t J. Wang , X. M. Meng , X. H. Han , H. B.Cai , L. Cao , J. S. Deng , Y. L. Qiu , S.Wang , J. Y. Wei , J. Y. Hu Abstract
We here report the photometric performance ofLunar-based Ultraviolet telescope (LUT), the firstrobotic telescope working on the Moon, for its 18-months operation. In total, 17 IUE standards havebeen observed in 51 runs until June 2015, which re-turns a highly stable photometric performance duringthe past 18 months (i.e., no evolution of photometricperformance with time). The magnitude zero point isdetermined to be 17 . ± .
05 mag, which is not onlyhighly consistent with the results based on its first 6-months operation, but also independent on the spectraltype of the standard from which the magnitude zeropoint is determined. The implications of this stableperformance is discussed, and is useful for next gener-ation lunar-based astronomical observations.
Keywords space vehicles: instruments — telescopes— techniques: photometric — ultraviolet: general
As the first robotic astronomical telescope working onthe lunar surface in the history of mankind, Lunar-based Ultraviolet Telescope (LUT) on board the Chi-nese first lunar lander (Chang’E-3, Ip et al. 2014) hassmoothly worked for 18 months on the Moon up to thetime when this paper is prepared. About 10,000 imageshave been acquired per month by LUT in the past 18months.
J. Wang, X. M. Meng, X. H. Han, H. B. Cai, L. Cao, J. S. Deng,Y. L. Qiu, S. Wang, J. Y. Wei, J. Y. Hu National Astronomical Observatories, Chinese Academy of Sci-ences Key Laboratory of Space Astronomy and Technology, NationalAstronomical Observatories, Chinese Academy of Sciences
A detailed description on the scientific goals andmission conception of LUT can be found in Cao etal. (2011). Briefly speaking, the telescope is dedi-cated to 1) continuously monitor bright variable stars inthe near-ultraviolet (NUV) band for as long as a dozendays; 2) perform a sky survey at low Galactic latitudein the NUV band. LUT consists of a 2-dimensionalgimbal with a flat mirror used for pointing a given tar-get and a telescope with an aperture of 150mm. Bothcomponents are horizontally mounted in a cabin of thelander (see Figure 1 in Wang et al. 2015a for a cross sec-tional view). The telescope adopts a Ritchey-Chretiensystem with a focal ratio of 3.75, and is equipped withan UV-enhanced 1024 × µ m, which corresponds apixel scale of 4.76˝pixel − and results in a filed-of-viewof 1.36 × . Two LED lamps with a centerwavelength of 286 nm are equipped to provide an in-ternal flat field. The efficiency curve of LUT that ismeasured in laboratory peaks at around 2500˚A with apeak value of ≈ Fig. 1
The normalized throughput of LUT plotted as afunction of wavelength. months by including the new observations taken in oneyear. The importance of monitoring the photometricperformance of LUT is described as follows. At first,the magnitude zero point that is used to obtain theactual brightness of an observed target might changewith the time. This change is essential not only for asurvey program with a ling time duration, but also fora re-visitation of the same object separated by severalmonths. Secondly, the study of evolution provides anunique opportunity for the first time to examine theeffect on lunar-based astronomical observations causedby the extreme environment on the Moon, which is use-ful for next generation of lunar-based astronomical tele-scope working in optical/NUV band.The paper is organized as follows. The adopted cal-ibration strategy is briefly described in Section 2. Theobservations and data reductions are given in Section3. Section 4 presents the results and discussions.
Transforming the observed instrumental magnitudeof an object to its actual brightness requires a pre-determined magnitude zero point. The zero point(zp) can be determined through the relationship zp = m LUT − m inst for a series of standard stars, where m inst is the observed instrumental magnitude of a standardstar and m LUT its actual brightness. By adopting theAB magnitude system (Oke & Gunn 1983; Fukugita etal. 1996), m LUT is calculated for a given standard as m LUT = − . R f ν S ν d ln ν R S ν d ln ν − . f ν is the specific flux density of the standardstar in unit of erg s − cm − Hz − , and S ν the LUT’stotal throughput at frequency ν which is needed to bedetermined in laboratory at pre-launch (see Figure 1). The IUE standard stars are adopted in our cali-bration both because of their uniform distribution onsky and because of the limit available sky region ofLUT. Basing upon a combination of the location (i.e.,13.31˚S and 14.12˚E on the Moon) where it waslanded and the available range of the gimbal rotation,LUT can only cover a cap around the north pole of theMoon (see Figure 3 in Wang et al. 2015a). The totalavailable sky area is only about 3600 degrees , and theavailable sky area at a given time is ∼ . Intotal, 44 IUE standards (Wu et al., 1998) located in theavailable sky region have been selected by Wang et al.(2015a). For each of the standards, its absolute specificflux in NUV band obtained by IUE has to be extendedto optical band (i.e., has a final wavelength coveragefrom 2000˚A to 8000˚A), both because of the relativelybroad wavelength range of LUT (see Figure 1) and be-cause of the red cutoff at 3200˚A of the IUE data. Werefer the readers to Section 4.2 in Wang et al. (2015a)for the details of the construction of the spectral datasets. Briefly speaking, given the effective temperature,surface gravity and abundance, the extension is realizedby a 3-dimensional linear interpolation basing upon theATLAS9 model atmospheres (Castelli & Kurucz 2003).The absolute specific flux level of each extracted modelspectrum is determined from its V-band magnitude, af-ter a reddening by its color excess E ( B − V ). We have observed 17 standards in 51 observational runsuntil June 2015. Table 1 lists the details of the 17 stan-dards. Columns (7), (8) and (9) tabulates the effectivetemperature, surface gravity and abundance collectedfrom the literature, except for two standards whose val-ues are adopted from the suggested models with a solarmetallicity for specific stellar types. Columns (10) liststhe magnitudes calculated from Equation (1). The ob-servations strictly follow the strategy described in Wanget al. (2015a) and Meng et al. (2015), which is designedto properly remove the strong stray light caused by thesunshine. Each observational run with a duration ofabout 30 minutes consists of a series of short expo-sures with duration of 2 to 10 seconds depending onthe brightness of the standards. The telescope pointingwas fixed with respect to the Moon in the run. Thismeans the shift of a star on the focal plane within eachexposure is negligible compared with the size of thepoint-spread-function due to the slow rotation of themoon, and the total shift within the run is ∼
100 pix-els.A general pipeline is developed by us to reduce theraw data. We refer the readers to Meng et al. (2015) for the details of the pipeline. The procedures of thepipeline includes overscan correction, removal of straylight (including bias and dark current), normalizationby a composed flat filed, source extraction and cosm-icrays rejection and final aperture photometry. Theused composed flat field is derived from a combinationof the internal flat field provided by the internal LEDsand superflat from a dithering observation of a singlestandard. The instrumental magnitude measured witha fixed radius of 7FWHM is adopted in subsequent anal-ysis, since this aperture is believed to enclose almost allthe signal from a bright star (see Figure 6 in Meng etal. 2015 for a growth curve of aperture photometry). σ significance level. Onecan learn from the figure that the photometric perfor-mance of LUT is highly stable during its 18-monthsoperation. Based on all the observations of standarduntil June 2015, a statistics returns an average value ofzp = 17 . ± .
05, which is highly consistent with theresults based on the first 6-months operation (Wang etal. 2015a). The highly stable photometric performanceis further illustrated in Figure 3, which shows the stablelight curves (transformed to zp) of four standards thathave been repeatedly visited for at least four times inthe past 18 months. In addition, one can learn from thefigure that the zp values based on different standardsare consistent with each other within their uncertain-ties, although a slightly lower value is yielded from thelate type-K star HD 164058 (see discussion given be-low).The revealed highly stable photometric performancetherefore indicates that the performance degradationdue to the reaction with the environment is negligiblefor LUT. Possible effects include the bombardment ofthe high energy protons from solar wind, the oxidationof the coating, and the deposition of the charged lunardust on the the mirrors.At first, although the damage to the coating on themirrors/lens by the protons from solar wind has beenalleviated significantly by a designed shielding. the firstreflection mirror of LUT is nevertheless fully exposedto the environment when an observation is carried out.
Fig. 2
The determined magnitude zero point in eachmonth plotted against time in unit of month since the be-ginning of 2014. The average value of magnitude zero pointis marked by the solid horizontal line. The two dashed hor-izontal lines mark the corresponding error at 1 σ confidencelevel. Fig. 3
The light curves (transformed to magnitude zeropoint) of four standards that have been repeatedly observedfor at least four times in the past 18 months.
Assuming a typical inclination of θ =45˚ with respectto the local zenith, the total number of protons bom-bard on the mirror is N ∼ cos θjA ∆ t ∼ p + in thepast 18 months, where j ∼ p + cm − s − is the pro-ton flux of the solar wind, A = 20 × . thearea of the mirror, and ∆ t the total observational timeelapsed in the past 18 months. Secondly, not as the tele-scopes on the ground or on a low orbit around earth,the degradation of reflection efficiency due to the oxida-tion is negligible on the Moon thanks to the extremelytenuous lunar atmosphere (e.g., Stern, 1999; Wang etal. 2011, 2015b). Finally, the stable photometric per-formance in the past 18 months suggests that LUT didnot suffer the influence of the charged lunar dust thatcan continuously decrease the reflection/transparencyefficiency by a deposition on the mirrors, although this Table 1
IUE standards observed by LUT until June 2015Star s.p. type α (J2000) δ (J2000) m v E ( B − V ) T eff log g Fe/H m ck04 Ref.mag mag K mag(1) (2) (3) (4) (5) (6) (7) (8) (9) (10) (11)HD123299 A0III 14 04 23.6 +64 22 31 3.66 0.00 10371 3.95 -0.19 4.61 3HD127700 K4IIIBa0.3 14 27 31.2 +75 41 43 4.27 0.00 4395 1.86 -0.08 8.30 6HD131873 K4III 14 50 42.3 +74 09 20 2.07 0.04 4077 1.7 -0.10 6.52 1HD132813 M5III 14 57 35.6 +65 55 54 4.59 0.00 3500 1.34 0.00 9.11 . . . .HD137759 K2III 15 24 55.7 +58 57 57 3.29 0.01 4520 2.61 0.12 7.19 3HD139669 K5III 15 31 25.4 +77 20 56 4.96 0.07 3962 1.44 0.18 9.73 3HD147394 B5IV 16 19 44.4 +46 18 45 3.90 0.01 14906 4.06 0.14 3.98 1HD153751 G5III 16 45 58.2 +82 02 14 4.23 0.00 5150 2.54 0.00 7.35 2HD159181 G2Ib-IIa 17 30 26.0 +52 18 05 2.80 0.09 5325 1.51 -0.02 6.09 1HD164058 K5III 17 56 36.4 +51 29 20 2.23 0.01 3990 1.64 0.11 6.87 3HD166205 A1Vn 17 32 12.1 +86 35 08 4.36 0.00 9230 4.10 0.00 5.63 . . . .HD185395 F4V 19 36 26.5 +50 13 16 4.48 0.00 6700 4.30 0.01 6.34 4HD188209 O9.5Iab 19 51 58.9 +47 01 39 5.63 0.20 31910 3.36 0.00 5.19 5HD188665 B5V 19 53 17.4 +57 31 25 5.14 0.02 14893 3.86 -0.17 5.26 5HD198149 K0IV 20 45 16.6 +61 49 38 3.42 0.01 4888 3.19 -0.21 6.64 3HD214470 F3III-IV 22 35 46.1 +73 38 35 5.08 0.00 6637 3.59 0.09 7.18 2HD203280 A7IV-V 21 18 33.6 +62 35 05 2.46 0.00 7773 3.45 0.09 4.23 7
Note : Column (1): Star name; Column (2): Spectral type; Column (3) & (4): Right ascension and declination in J2000 mean equatorcoordinate; Column (5): apparent magnitude in V-band; Column (6): Color excess; Column (7): Surface effective temperature; Column(8): Surface gravity; Column (9): metal abundance; Column (10): LUT magnitude calculated through Equation (1). Reference: [1]Koleva & Vazdekis (2012); [2] Soubiran et al. (2010); [3] Prugniel et al. (2011); [4] Cunha et al. (2000); [5] Fitzpatrick & Massa(2005); [6] Luck & Heiter (2007); [7] Gray et al. (2003). effect has been indeed observed by the Apollo 12 astro-naut (e.g., Colwell et al. 2007). The avoidance of theinfluence of the dust can be understood by the followingtwo reasons. At the beginning, LUT is mounted withina cabin of the CE-3 lander. Secondly, for each lunarday, the cabin was closed at both sunrise and sunsetwhen the dust particles are believed to be launched atterminator (e.g., Berg et al. 1976).4.2 UncertaintiesThe calibration based on the IUE standards results in amagnitude zero point with an accuracy of 5%. The re-sulted final uncertainty contains the contributions fromthe shot noise from incident light, the residual nonuni-formity in CCD quantum efficiency after the flatfield-ing, the error in the determined throughput curve, andthe modeled standard spectra (i.e., the uncertainties ofthe parameters defining individual stellar atmosphere,see discussion given in Section 4.3). Because the usedstandards are so bright, the shot noise is generally esti-mated to be as low as 0.001-0.003 magnitudes for theseobserved standards. The magnitude uncertainty causedby the error of the throughput curve is estimated froma Monte-Carlo simulation with 100 iterations, in whicha random curve is produced by a random sampling at each wavelength according to the measured error of S λ .The simulation shows that the resulted magnitude er-ror is ∼ .
001 mag. As indicated by Meng et al. (2015),the residual medium-scale nonuniformity in CCD con-tributes a typical error of ∼ . of pixels discussed above, the error of the determinedzp is additionally contributed by the uncertainty of thecalculated magnitude of a standard. In addition to de-pend on the IUE spectral/ATLAS9 model calibration,the calculated magnitude depends on the V-band mag-nitude, the level of extinction and the parameters defin-ing the atmosphere model. The level of this dependenceis expected to be more significant for a late type starthan for an early type star, because the peak of theSED of a star shifts toward a longer wavelength whenthe stellar surface temperature decreases. The photometric performance of LUT is reported herefor its 18-months operation on the Moon. The observa-tions of 17 IUE standards in 51 runs allow us to claima highly stable photometric performance of LUT dur-ing the past 18 months. The magnitude zero point isdetermined to be 17 . ± .
05 mag, and is found to beindependent on the spectral type of the used standard.
Acknowledgements
We would like to thank theanonymous referee for his/her suggestions that improvethe manuscript. The authors thank the outstandingwork of the LUT team and support by the team fromthe ground system of the Change-3 mission. This studyis supported by the Key Research Program of ChineseAcademy of Sciences (KGED-EW-603). JW is sup-ported by the National Natural Science Foundation ofChina under Grant 11473036. MXM is supported bythe National Natural Science Foundation of China un-der Grant 11203033.
Table 2
The resulted magnitude zero points from the stan-dards with different spectral types.s.p. type zp Num. of runsmag(1) (2) (3)O, B, & A 17 . ± .
04 11F & G 17 . ± .
07 17K & M 17 . ± .
13 23
Note : Column (1): Spectral type; Column (2): Determinedmagnitude zero point; Column (3): Total number of the usedobservational runs.
References
Berg, O. E., Wolf, H., & Rhee, J.: Lunar and PlanetaryExploration 48, 233 (1976)Cao, L., Ruan, P., Cai, H.B., et al.: ScChG 54, 558 (2011)Castelli, F. & Kurucz, R.: IAU Symposium 210, Mod-elling of Stellar Atmospheres, Uppsala, Sweden, eds. N.E.Piskunov, W.W. Weiss. and D.F. Gray, ASP-S210 (2003)Colwell, J. E., Batiste, S., Horanyi, M., Robertson, S., &Sture, S.: Reviews of Geophysics 45, 2006 (2007)Cunha, K., Smith, V. V., Boesgaard, A. M., & Lambert, D.L.: Astrophys. J. 530, 939 (2000)Fitzpatrick, E. L., & Massa, D.: Astrophys. J. 129, 1642(2005)Fukugita, M., Ichikawa, T., Gunn, J.E., et al.: Astron. J.111, 1748 (1996)Gray, R. O., Corbally, C. J., Garrison, R. F., McFadden, M.T., & Robinson, P. E.: Astron. J. 126, 2048 (2003)Ip, W.-H., Yan, J., Li, C.-L., Ouyang, Z.-Y.: Research inAstronomy and Astrophysics 14, 1511 (2014)Koleva, M., & Vazdekis, A.: Astron. Astrophys. 538, 143(2012)Luck, R. E., & Heiter, U.: Astron. J. 133, 2464 (2007)Meng, X. M., et al.: astro-ph/arXiv:1505.07951 (2015)Oke, J. B., & Guun, J. E.: Astrophys. J. 266, 713 (1983)Prugniel, P., Vauglin, I., & Koleva M.: Astron. Astrophys.531, 165 (2011)Schneider, D. P., Gunn, J. E., & , Hoessel, J. G.: Astro-phys. J. 264, 337 (1983)Soubiran, C., et al.: Astron. Astrophys. 515, 111 (2010)Wang, J. et al.: AdSpR 48, 1927 (2011)Wang, J. et al.: Research in Astronomy and Astrophysics7, 1068 (2015a)Wang, J. et al.: P&SS 109, 123 (2015b)Wu, C. C., et al.: ESASP 413, 751 (1998)