Microlensing of the broad-line region in the quadruply imaged quasar HE0435-1223
L. Braibant, D. Hutsemékers, D. Sluse, T. Anguita, C. J. García-Vergara
aa r X i v : . [ a s t r o - ph . GA ] M a y Astronomy & Astrophysicsmanuscript no. HE0435_microlensing_BLR c (cid:13)
ESO 2018August 13, 2018 L etter to the E ditor Microlensing of the broad-line region in the quadruply imagedquasar HE0435-1223 ⋆ L. Braibant , D. Hutsemékers , D. Sluse , T. Anguita , C. J. García-Vergara F.R.S.-FNRS, Institut d’Astrophysique, Université de Liège, Allée du 6 Août 17, B5c, 4000 Liège, Belgium Argelander-Institut für Astronomie, Auf dem Hügel 71, 53121 Bonn, Germany Departamento de Ciencias Fisicas, Universidad Andres Bello, Av. Republica 252, Santiago, Chile Instituto de Astrofísica Pontificia Universidad Católica de Chile, Avenida Vicuña Mackenna 4860, Santiago, ChileAugust 13, 2018
ABSTRACT
Using infrared spectra of the z = ff ect in images A and D . While microlensing a ff ects the blueand red wings of the H α line profile in image D very di ff erently, it de-magnifies the line core in image A . The combination of thesedi ff erent e ff ects sets constraints on the line-emitting region; these constraints suggest that a rotating ring is at the origin of the H α line. Visible spectra obtained in 2004 and 2012 indicate that the MgII line profile is microlensed in the same way as the H α line.Our results therefore favour flattened geometries for the low-ionization line-emitting region, for example, a Keplerian disk. Biconicalmodels cannot be ruled out but require more fine-tuning. Flux ratios between the di ff erent images are also derived and confirm fluxanomalies with respect to estimates from lens models with smooth mass distributions.
1. Introduction
Gravitational microlensing by compact objects in lensing galax-ies is a powerful tool for probing the inner regions of distantquasars. Microlenses typically magnify regions of the source onscales of a few tens of light days (e.g., Schmidt & Wambsganss2010). The region that generates the optical and UV continuum(i.e., the accretion disk) is therefore very likely (de-)magnifiedby microlensing, which can be observed as a (de-)magnificationof the continuum spectrum since micro-images are not resolved.The broad-line region (BLR) can also be a ff ected by microlens-ing. Sluse et al. (2012) have shown that microlensing of the BLRis commonly detected in systems exhibiting microlensing of thecontinuum. Interestingly, microlensing of the BLR can signifi-cantly alter the broad emission-line profiles (e.g., Abajas et al.2002, Lewis & Ibata 2004) so that it may be possible to exploitthese spectral di ff erences to probe the BLR structure.We study the broad emission lines of the gravitational lenssystem HE0435-1223 discovered by Wisotzki et al. (2001). Itis composed of four quasar images of similar brightness, sep-arated by about 1 . ′′ and arranged with quasi-perfect symme-try around the elliptical lensing galaxy. The quasar is at red-shift z s = .
693 (Sluse et al. 2012). The redshift of the lens is z l = . ff ect on the line pro-files (e.g., Wisotzki et al. 2003, and references therein).In the following, we mainly focus on the distortions of theH α line profile in the spectra of the quasar-lensed images ob-served between October and December 2009. We also examineMgII line profiles observed in November and December 2004,and in September 2012. We finally draw some conclusions aboutthe geometry and kinematics of the BLR. ⋆ Based on observations made with the ESO-VLT, Paranal, Chile;Proposal 084.B-0013 (PI: Rix).
2. Data collection and processing
We used archive Integral Field Spectroscopy data of HE0435-1223 secured between Oct. 19 and Dec. 15, 2009 at the VeryLarge Telescope (Table 2). The system has been observed withSINFONI, using the 3 ′′ × ′′ field-of-view (FOV) with 0 . ′′ spatial resolution. The H-band grism leads to a spectral cover-age of [1 . , . µ m, which contains the broad H α emissionline. Due to o ff sets between exposures, some lensed images wereoccasionally outside the FOV. Only the spectra of the quasar im-ages entirely enclosed in the FOV were extracted.The data were calibrated and 3D data cubes, in which eachplane is a monochromatic FOV, were built with the ESORexSINFONI pipeline (version 2.2.9). Cosmic rays were removedfrom every monochromatic FOV using the la_cosmic proce-dure (van Dokkum 2001). In addition, the SINFONI FOV isnot uniformly illuminated. Since this is not corrected for by thepipeline, we applied an empirical illumination correction: we di-vided each monochromatic FOV by a normalized median FOVof the infrared empty sky. Nevertheless, as a precautionary mea-sure, we discarded the spectra of the images whose peak is lo-cated on slitlets MPFIT pack-age (Markwardt 2009). In this model, the quasar images weremodelled with identical 2D Gaussian point spread functions(PSFs), whose relative positions were fixed to the astrometry ofCourbin et al. (2011). The lensing galaxy was modelled with ade Vaucouleurs profile with r e f f = . ′′ (de Vaucouleurs 1948),convolved with the PSF. The fitting procedure was iterative: the Article number, page 1 of 6 & Aproofs: manuscript no. HE0435_microlensing_BLR
Table 1.
Flux ratios measured in the continuum.Date Spectra A / B a C / B b D / B c C / A d D / A e /
10 A, B, C, D 1 .
424 1 .
047 0 .
762 0 .
736 0 . /
10 A, C - - - 0 .
658 -19 /
10 A - - - - -19 /
10 A, B, D 1 .
356 - 0 .
937 - 0 . /
12 A, C, D - - - 0 .
823 0 . /
12 A, C - - - 0 .
745 -09 /
12 A - - - - -09 /
12 A, B, D 1 .
354 - 0 .
837 - 0 . /
12 A, C, D - - - 0 .
820 0 . /
12 A, C - - - 0 .
789 -09 /
12 A - - - - -09 /
12 A, B, D 1 .
311 - 0 .
720 - 0 . /
12 B - - - - -10 /
12 B, C - 1 .
045 - - -15 /
12 A, B, C, D 1 .
353 0 .
999 0 .
720 0 .
738 0 . /
12 A, C - - - 0 .
763 -15 /
12 A - - - - -15 /
12 A, B, D 1 .
376 - 0 .
778 - 0 . Notes.
We list for each science exposure acquired with SINFONI in2009 the observing date (col. 1), the images with extracted spectra (col.2), and the flux ratios measured in the continuum (col. 3-7).Error bars : ( a ) ± . ( b ) ± . ( c ) ± . ( d ) ± . ( e ) ± . Fig. 1. (a)
Median spectra of the QSO-lensed images. (b)
The spectraof images A and B normalized so that their continua have a unit averageintensity and are superimposed, and (c) the spectral di ff erences betweenthe H α line profiles in the normalized spectra of images A and B , plottedwith error bars for a wavelength bin of 0 . µ m. (d) and (e) Same forthe pair of images B and C. (f) and (g)
Same for the pair of images Band D. Spectra have been filtered with a five-pixel-wide median filterfor clarity. values of the model parameters determined at a given iterationwere used as starting values for the following fit after they weresmoothed over the wavelength with a median filter and fittedwith a spline function. The PSF widths and the position of thewhole system were forced to vary between physical boundariesthat were shrunk at every iteration.Table 1 provides for each science exposure the list of usablespectra and the measured image flux ratios. These ratios wereobtained by fitting the previous model to a “pure continuum”FOV, derived by co-adding all the monochromatic FOVs whose wavelengths are located in the range [1 . , . µ m. The er-rors on flux ratios were estimated by multiplying the flux ratioswith the dispersion of the F A / F A ratio ( σ A / A = . A secured on Dec. 9, 2009 .The spectra were normalized so that their continua havea unit average intensity. Since the normalized spectra of eachquasar image are compatible within 3 σ -error regardless of theobserving date, we increased the signal-to-noise ratio by com-puting a median normalized spectrum over all the epochs foreach lensed image. The median spectra of images A , C , and D were then multiplied by the mean flux ratio in the continuum, A / B = . ± . C / B = . ± .
11, and D / B = . ± . a . We also examined visible spectra of B and D obtained in 2004with the ESO-VLT FORS1 instrument (Fig.3 a ). The observa-tions are summarized in Table 2. More details about the data re-duction can be found in Eigenbrod et al. (2006) and Sluse et al.(2012).Additional spectroscopic observations of HE0435-1223were acquired in 2012 using the IMACS spectrograph (long-camera mode) at Magellan-Baade (see Table 2). The slit was ori-ented through components A - D on one hand and through com-ponents B - C on the other hand.The Magellan data were reduced using the COSMOS pack-age. The 2D slit spectra were extracted in two steps. First, foreach wavelength bin, we fitted a sum of two identical Mo ff atprofiles on the two quasar images, with no prior on the imagepositions. The image centers and Mo ff at widths, derived afterthat first step, were smoothed with a median filter and fitted witha spline function to set the initial conditions of the second fit. Inthis second fit, the parameters were constrained to agree with thesmoothed values within 5%. The spectra of images D and A arepresented in Fig. 3 c .
3. Microlensing in HE0435-1223
Fig. 1 b − g highlight the di ff erences between the H α line profilesobserved in the median spectra of the di ff erent quasar images,normalized in the continuum. Following Courbin et al. (2011),who identified image B as the “least a ff ected by stellar mi-crolensing”, we assumed that the spectrum of image B is notaltered by microlensing and used it as the reference to which wecompare the spectra of the other quasar images.The spectral-line profiles of A and D di ff er significantly fromthe profiles of the two other images, A having a brighter red wing(see Fig. 1 b , c ) and D a fainter blue wing (see Fig. 1 f , g ). Wecomputed median spectra of the quasar images for each individ-ual observing night. At every epoch, image A displays a promi-nent red H α wing and image D a dimmer blue H α wing thanimage B ; these di ff erences are therefore probably not caused bythe intrinsic variability of the quasar. This interpretation is sup-ported by R-band light curves published in Courbin et al. (2011),which show that the fluxes of the images have varied by less than10% during the period of the SINFONI observations.The line profiles of images B and C only show small di ff er-ences in their blue wing (see Fig. 1 d , e ). The amplitude of the The night with the most observed spectra. http: // code.obs.carnegiescience.edu / cosmosArticle number, page 2 of 6raibant et al.: Microlensing of the broad-line region in the quadruply imaged quasar HE0435-1223 di ff erences between B and C varies from one observing date toanother, which indicates that they may be (partly) caused by in-trinsic variability. Comparison of spectra obtained on Dec. 9 andDec. 15, 2009 allows us to roughly correct for the time delayof 7.8 days (Courbin et al. 2011) between these two images. Wefound that the di ff erences decrease but do not vanish. Becausespectra separated by the exact time delay are not available, wecannot conclude on the nature of the small spectral di ff erencesobserved between B and C (Fig. 1 d , e ).Apart from the spectral di ff erences between the line profilesof some lensed images, the flux ratios do not vary with wave-length over the observed wavelength range, in agreement withthe absence of significant di ff erential extinction (at least over thissmall wavelength range). We used the macro-micro decomposition (MmD) method(Sluse et al. 2007, Hutsemékers et al. 2010, Sluse et al. 2012)to separate the part of the quasar spectrum that is a ff ected bymicrolensing from the part which is not. The MmD interpretsthe di ff erences between the spectra of a pair of gravitation-ally lensed images, and in particular the di ff erences betweenthe line profiles, under the hypothesis that microlensing a ff ectsthe continuum emission, but does not a ff ect the emission line,or at least any wavelength interval of it (cf. appendix A inHutsemékers et al. 2010). This method allows one to determinethe macro-magnification ratio M between the macro-images ofthe quasar and a microlensing factor µ , which quantifies the ad-ditional (de-)magnification caused by microlensing. • Image D : undergoing a large magnification In Fig. 2 (upper panel), we applied the MmD to the median spec-tra of images B and D , computed using all the available spectra.We derived a macro-magnification ratio M = . ± .
03 and amicrolensing factor µ = . ± . D as wellas the red wing of its H α line, which causes the displacement ofits peak towards longer wavelengths (see Fig. 1 f ). Microlensinga ff ects about 50% of the flux of the Balmer emission line.Since the time delay between B and D is 6.5 days(Courbin et al. 2011), we used the spectra obtained on Dec. 9and Dec. 15 to remove intrinsic variability. The results obtainedfollowing this procedure are noisier but consistent with the aboveanalysis. This is expected because the time delay is shorter thanthe time scale of large-amplitude intrinsic variations. This sup-ports the microlensing interpretation of the spectral di ff erencesobserved in the H α emission line.Microlensing has been detected in image D by previous stud-ies. Wisotzki et al. (2003) found signatures of microlensing in D ,using the first spatially resolved spectroscopic observations ofHE0435-1223, which were secured in 2002. Applying the MmDto visible spectra acquired in 2004, Sluse et al. (2012) detectedmicrolensing of the CIII] and MgII lines in the spectrum of im-age D . • Image A : weakly de-magnified The MmD applied to the pair of spectra A - B is displayed in Fig. 2(lower panel). The continuum emission and the core of the H α line are de-magnified by microlensing. The microlensing factoris µ = . ± .
05. The macro-magnification ratio between A and B is M = . ± . A and B areseparated by an eight-day time delay (Courbin et al. 2011), this Fig. 2.
Upper panel: decomposition of the quasar spectrum into amacrolensed-only component ( F M , black line) and a component bothmacro- and microlensed ( F M µ , blue line) by applying MmD to the spec-tra of images B and D . F M µ is shifted upward by three units. The spec-trum of image B ( F B , grey line), assumed to be unaltered by microlens-ing, is shifted downward till its continuum intensity is null and is super-imposed on F M . The spectra are filtered with a five-pixel-wide medianfilter for clarity. The vertical dotted line indicates the redshifted restwavelength of the H α line. Lower panel: same for the pair of spectra ofimages B and A . is only an approximate correction. The MmD applied to thesespectra leads to noisier but consistent results, supporting our mi-crolensing interpretation.Interestingly, microlensing a ff ects the line profile symmet-rically in A , in contrast to the e ff ect that alters the spectrum of D . About 60% of the flux of the H α line is microlensed. This isnot unexpected since de-magnification can act on more extendedregions than magnification (e.g. Lewis & Ibata 2004).That there is microlensing in image A has been sug-gested based on photometric monitoring (Ricci et al. 2011;Courbin et al. 2011). Using narrow-band photometric data ob-tained in 2007, Mosquera et al. (2011) have detected chromaticmicrolensing in the continuum of image A .The comparison of the flux ratio A / B = . ± .
03 measuredin the H band (Table 3) with the contemporary R-band measure-ment A / B = . ± .
05 of Courbin et al. (2011) suggests a weakchromatic trend, but we find no indication of chromaticity whenthe H-band A / B flux ratio is compared with the V-, R- and I-bandmeasurements made by Ricci et al. (2011) at a slightly di ff erentepoch (Aug.-Sep. 2009).
4. Discussion and conclusions
The A / B macro-magnification ratio we determined is consis-tent with the L ′ flux ratio A / B = . ± .
23 measured byFadely & Keeton (2011). Our measurement confirms the A / B flux ratio anomaly (e.g., Fadely & Keeton 2012).The macro-magnification ratio M D / B = . ± .
03 we de-rived is significantly lower than M D / B = . ± .
015 and0 . ± .
02 obtained by Sluse et al. (2012), in the CIII] andMgII line profiles respectively, and than the L ′ flux ratio D / B = . ± .
16 measured by Fadely & Keeton (2011).
Article number, page 3 of 6 & Aproofs: manuscript no. HE0435_microlensing_BLR
Fig. 3. ( a ) Comparison of the MgII line profiles in images B and D ,observed in 2004. ( b ) Comparison of the H α line profiles of images Band D, observed in 2009. ( c ) Comparison of the MgII line profiles ofimages A and D , observed in 2012. In every panel, the continua havebeen scaled so that they overlap. Such a low D / B macro-magnification ratio implies amicro-magnification of the continuum stronger than previouslythought, both in the UV-visible and L ′ band . It also implies thatmicrolensing a ff ects a larger part of the CIII] and MgII linesthan found by Sluse et al. (2012) using MmD. This method min-imizes the part of the BLR emission that is microlensed and thusprovides a lower limit of the microlensing factor. Such a low M D / B may not be easily reproduced by smooth mass models.On the other hand, we also investigated microlensing sce-narios corresponding to M D / B ∼ . M D / B ∼ .
8, obtainedby Sluse et al. (2012) and Fadely & Keeton (2011). In brief, for M D / B ∼ .
8, the micro-de-magnification of only the blue wing ofthe H α line would reproduce the line profile distortions observedin image D , while for M D / B ∼ .
7, microlensing would a ff ect thewhole line profile, magnifying the H α red wing and simultane-ously de-magnifying its blue wing (details in the appendix).None of these scenarios appears completely satisfactory,given the available data but, regardless of the scenario consid-ered, this does not change the strong microlensing dichotomyobserved between the blue and red wings of H α in image D .The microlensing e ff ect causing the displacement of the peakin D has already been observed by Sluse et al. (2012) in the MgIIline profile, in 2004 (Fig. 3 a ). It is still detected in the MgIIline profile obtained in 2012 with Magellan IMACS (Fig. 3 c ) .This large-amplitude microlensing e ff ect hence appears to bestable over time, in agreement with the microlensing time-scale estimated for the HE0435-1223 system: ∼ p M / M ⊙ years (considering a relative transverse velocity of 600 km / s,Schmidt & Wambsganss 2010). A rotating disk is a popular model for Balmer-line-emittingregion (e.g., Smith et al. 2005, Bon et al. 2009), but less flat-tened (e.g. Gaskell 2009, Goad et al. 2012) and biconical(Fischer et al. 2011) geometries have also been proposed. Theway microlensing alters the H α line in images A and D sets con-straints on the BLR geometry and its velocity field.Simulations (Schneider & Wambsganss 1990) showed thatonly a non-spherically symmetric BLR can be at the origin ofa blue / red dichotomous microlensing e ff ect that causes the dis-placement of the peak, as observed in image D . In the L ′ band, 20% of the continuum may come from the accretiondisk; therefore microlensing may not be negligible (Sluse et al. 2013). Spectra of A and D were acquired simultaneously, separately fromspectra of B and C , with a di ff erent integration time (Table 2). Thus,spectra of B and D cannot be safely compared with each other. In image D , the most blueshifted and redshifted parts of theH α line, which correspond to the high-velocity emitting regions,are a ff ected very di ff erently by microlensing. Hence, the regionsof the BLR that produce these parts of the line profile must bespatially well separated in projection. In addition, these highlyDoppler-shifted parts of the line are not microlensed in image A , which implies that the corresponding emitting regions are notonly distant from each other in projection, but also located awayfrom the central continuum source. On the other hand, the coreof the H α line, which corresponds to the low-velocity part ofthe BLR, is a ff ected by microlensing in A , so that it most likelycomes from a compact region close to the continuum source inprojection.A rotating-ring geometry for the H α emitting region fitsthese constraints nicely, favouring Keplerian-disk models for thelow-ionization line-emitting region. Biconical winds cannot becompletely ruled out, but we expect that only specific combi-nations of inclinations and velocity fields of the bicone will con-form with both the type 1 nature of the system and the constraintsderived from microlensing.In future work, we will more quantitatively investigate themicrolensing e ff ects on flattened and biconical BLRs using sim-ulated spectra. New spectra that simultaneously cover a widerwavelength range and contain several broad emission lineswould also allow us to gather information on the ionization struc-ture of the broad line region. References
Abajas, C., Mediavilla, E., Muñoz, J. A., Popovi´c, L. ˇC., & Oscoz, A. 2002, ApJ,576, 640Blackburne, J. A., Pooley, D., Rappaport, S., & Schechter, P. L. 2011, ApJ, 729,34Bon, E., Popovi´c, L. ˇC., Gavrilovi´c, N., Mura, G. L., & Mediavilla, E. 2009,MNRAS, 400, 924Courbin, F., Chantry, V., Revaz, Y., et al. 2011, A&A, 536, A53de Vaucouleurs, G. 1948, Annales d’Astrophysique, 11, 247Eigenbrod, A., Courbin, F., Meylan, G., Vuissoz, C., & Magain, P. 2006, A&A,451, 759Fadely, R. & Keeton, C. R. 2011, AJ, 141, 101Fadely, R. & Keeton, C. R. 2012, MNRAS, 419, 936Fischer, T. C., Crenshaw, D. M., Kraemer, S. B., et al. 2011, ApJ, 727, 71Gaskell, C. M. 2009, New A Rev., 53, 140Goad, M. R., Korista, K. T., & Ru ff , A. J. 2012, MNRAS, 426, 3086Hutsemékers, D., Borguet, B., Sluse, D., Riaud, P., & Anguita, T. 2010, A&A,519, A103Kochanek, C. S., Morgan, N. D., Falco, E. E., et al. 2006, ApJ, 640, 47Lewis, G. F. & Ibata, R. A. 2004, MNRAS, 348, 24Markwardt, C. B. 2009, in Astronomical Society of the Pacific Conference Se-ries, Vol. 411, Astronomical Data Analysis Software and Systems XVIII, ed.D. A. Bohlender, D. Durand, & P. Dowler, 251Morgan, N. D., Kochanek, C. S., Pevunova, O., & Schechter, P. L. 2005, AJ, 129,2531Mosquera, A. M., Muñoz, J. A., Mediavilla, E., & Kochanek, C. S. 2011, ApJ,728, 145Ricci, D., Poels, J., Elyiv, A., et al. 2011, A&A, 528, A42Richards, G. T., Keeton, C. R., Pindor, B., et al. 2004, ApJ, 610, 679Schmidt, R. W. & Wambsganss, J. 2010, General Relativity and Gravitation, 42,2127Schneider, P. & Wambsganss, J. 1990, A&A, 237, 42Sluse, D., Claeskens, J.-F., Hutsemékers, D., & Surdej, J. 2007, A&A, 468, 885Sluse, D., Hutsemékers, D., Courbin, F., Meylan, G., & Wambsganss, J. 2012,A&A, 544, A62Sluse, D., Kishimoto, M., Anguita, T., Wucknitz, O., & Wambsganss, J. 2013,A&A, 553, A53Smith, J. E., Robinson, A., Young, S., Axon, D. J., & Corbett, E. A. 2005, MN-RAS, 359, 846van Dokkum, P. G. 2001, PASP, 113, 1420Wisotzki, L., Becker, T., Christensen, L., et al. 2003, A&A, 408, 455Wisotzki, L., Christlieb, N., Liu, M. C., et al. 2001, in Astronomical Societyof the Pacific Conference Series, Vol. 237, Gravitational Lensing: RecentProgress and Future Go, ed. T. G. Brainerd & C. S. Kochanek, 63 Acknowledgements.
TA acknowledges support from FONDECYT grant number11130630. DS is supported by the German Deutsche Forschungsgemein-schaft,DFG project number SL172 / Article number, page 4 of 6 &A–HE0435_microlensing_BLR,
Online Material p 5
Table 2.
Scientific observations of HE0435-1223.
Date Images Grating / Grism / Filter Instument Exp. time (s) Seeing (”) Airmass11 / / + GG435 FORS1 (ESO-VLT) 4 × / / + GG435 FORS1 (ESO-VLT) 2 × / / ×
600 0.74 1.0619 / / ×
600 0.91 1.0309 / / a ×
600 0.48 1.1109 / / ×
600 0.49 1.2010 / / ×
600 0.58 1.0415 / / ×
600 0.73 1.1515 / / ×
600 0.76 1.2420 / / / mm IMACS (Magellan-Baade) 2 × / / / mm IMACS (Magellan-Baade) 1 × / / / mm IMACS (Magellan-Baade) 1 × ( a ) One exposure has been discarded because of target misalignment.
Table 3.
Flux ratios of the HE0435-1223 components measured at several epochs (compiled from the literature).
Date HJD a range Band A / B C / B D / B References2003 Aug. 2870 V . ± .
22 0 . ± .
08 0 . ± .
12 12003 Aug. 2870 I . ± .
11 0 . ± .
04 0 . ± .
06 12004 Jan. 3013-3036 R . ± .
05 1 . ± .
02 0 . ± .
02 22004 Jan. 3015 H . ± .
10 1 . ± .
03 0 . ± .
05 12007 Sep. 4366 J . ± .
09 1 . ± .
08 0 . ± .
06 32007 Sep. 4366 H . ± .
08 1 . ± .
08 0 . ± .
06 32007 Sep. 4366 K . ± .
03 1 . ± .
04 0 . ± .
03 32007 Oct. 4316-4340 R . ± .
04 1 . ± .
02 0 . ± .
02 22008 Jul.-Oct. 4675-4744 V . ± .
08 1 . ± .
06 0 . ± .
05 42008 Jul.-Oct. 4675-4744 R . ± .
05 1 . ± .
05 0 . ± .
04 42008 Jul.-Oct. 4674-4766 R . ± .
04 1 . ± .
03 0 . ± .
02 22008 Jul.-Oct. 4675-4744 I . ± .
08 1 . ± .
06 0 . ± .
04 42008 Aug.-Dec. 4682-4829 R . ± .
05 1 . ± .
03 0 . ± .
02 22008 Aug. 4709 K . ± .
18 0 . ± .
04 0 . ± .
07 52008 Dec. 4822 L ′ . ± .
23 1 . ± .
08 0 . ± .
16 52009 Aug.-Sep. 5064-5094 V . ± .
04 1 . ± .
02 0 . ± .
02 42009 Aug.-Sep. 5064-5094 R . ± .
05 1 . ± .
03 0 . ± .
03 42009 Aug.-Sep. 5047-5099 R . ± .
03 1 . ± .
03 0 . ± .
02 22009 Aug.-Sep. 5064-5094 I . ± .
05 1 . ± .
03 0 . ± .
03 42009 Oct.-Dec. 5111-5176 R . ± .
05 1 . ± .
04 0 . ± .
03 22009 Oct.-Dec. 5123,5174,5175,5180 H . ± .
03 1 . ± .
03 0 . ± .
02 6
Notes. ( a ) HJD = Julian Day - 2450000
References. (1) Kochanek et al. (2006); (2) Courbin et al. (2011); (3) Blackburne et al. (2011); (4) Ricci et al. (2011); (5) Fadely & Keeton (2011);(6) This work: the photometric flux ratios were computed by multiplying the spectrum of each quasar image by the transmission of the H -bandfilter and then integrating the flux. &A–HE0435_microlensing_BLR, Online Material p 6
Appendix : Microlensing scenarios fordi ff erent D / B macro-magnificationratios We investigated microlensing scenarios corresponding to themacro-magnification ratios derived by Sluse et al. (2012), M D / B ∼ .
7, and Fadely & Keeton (2011), M D / B ∼ . M D / B ∼ . µ ∼ . ff ected by microlensing so that a rigorousMmD cannot be performed. Simultaneous micro-magnificationof the H α red wing (although to a lesser extent than for M D / B ∼ .
5) and micro-de-magnification of the blue wing would repro-duce the distortions observed in the H α line profile of image D (red curve in Fig. A.1). Caustics can cause such microlensing ef-fect since they can delineate magnification and de-magnificationareas.For M D / B ∼ .
8, only the blue wing of H α is micro-de-magnified (magenta curve in Fig. A.1) and the continuumis not microlensed ( µ ∼ ff ect that magnifies only the broad emission line re-gion, has presumably been observed for the system J1004-4142(Richards et al. 2004). In that interpretation, a de-magnificationarea stands close enough to a ff ect the BLR but far enough notto a ff ect the continuum source. This agrees with the stability ofthe D / B flux ratio over time and wavelength (see Table 3). How-ever, considering the unification model for active galactic nuclei,in which a dusty torus surrounds the BLR (regardless of its ge-ometry), we would expect a broad de-magnification region to de-magnify the dusty torus as well. Thus, the L ′ -band flux shouldbe a ff ected, which is not observed. Wavelength (µm) M D/B =0.5M
D/B =0.7M
D/B =0.8 C o n t i nu a h a v e b ee n s u b t r a c t e d f r o m F B a n d F D F D / ( M D / B ∗ F B ) Fig. A.1.