Near-infrared Observations of Nova V574 Puppis (2004)
Sachindra Naik, Dipankar P. K. Banerjee, N. M. Ashok, R. K. Das
aa r X i v : . [ a s t r o - ph . S R ] J a n Mon. Not. R. Astron. Soc. , 000–000 (0000) Printed 23 November 2018 (MN L A TEX style file v2.2)
Near-infrared Observations of Nova V574 Puppis (2004)
Sachindra Naik ⋆ , D. P. K. Banerjee † , N. M. Ashok ‡ , & R. K. Das § , Astronomy and Astrophysics Division, Physical Research Laboratory, Navrangapura, Ahmedabad - 380009, Gujarat, India
Accepted for publication in MNRAS
ABSTRACT
We present results obtained from extensive near-infrared spectroscopic and photomet-ric observations of nova V574 Pup during its 2004 outburst. The observations wereobtained over four months, starting from 2004 November 25 (four days after the novaoutburst) to 2005 March 20. The near-IR
JHK light curve is presented - no evidence isseen from it for dust formation to have occurred during our observations. In the earlydecline phase, the
JHK spectra of the nova are dominated by emission lines of hydro-gen Brackett and Paschen series, OI, CI and HeI. We also detect the fairly uncommonFe II line at 1.6872 µ m in the early part of our observations. The strengths of theHeI lines at 1.0830 µ m and 2.0585 µ m are found to become very strong towards theend of the observations indicating a progression towards higher excitation conditionsin the nova ejecta. The width of the emission lines do not show any significant changeduring the course of our observations. The slope of the continuum spectrum was foundto have a λ − . dependence in the early stages which gradually becomes flatter withtime and changes to a free-free spectral dependence towards the later stages. Recom-bination analysis of the HI lines shows deviations from Case B conditions during theinitial stages. However, towards the end of our observations, the line strengths are wellsimulated with case B model values with electron density n e = 10 − cm − and atemperature equal to 10 K. Based on our distance estimate to the nova of 5.5 kpcand the observed free-free continuum emission in the later part of the observations,we estimate the ionized mass of the ejecta to be between 10 − M ⊙ and 10 − M ⊙ . Key words: infrared: stars – novae, cataclysmic variables – stars: individual(V574 Pup)– techniques: spectroscopic
V574 Pup was independently discovered to be in outburstby Tago (Nakano et al. 2004) on 2004 November 20.67 UTat V ∼ ∼ α and H β emissionlines with P Cygni components, along with the strong Fe II(multiplet 42) in absorption indicating that V574 Pup is a“Fe II” class nova near maximum light (Ayani 2004). Sub-sequent near-infrared spectroscopic observations of the novaon 2004 November 26.98 UT showed strong HI emission linesfrom the Paschen and Brackett series, OI lines at 1.1287 &1.3164 µ m and a blend of NI & CI lines in the spectral ⋆ [email protected] † [email protected] ‡ [email protected] § [email protected] region 1.175 to 1.25 µ m (Ashok & Banerjee 2004). Theoptical spectra of the nova on 2004 November 26 and De-cember 12 obtained by Siviero et al. (2005) were found tobe dominated by Balmer hydrogen and FeII emission lines;no nebular lines were present in the spectra of December 12.One year after the outburst, the nova was found to be wellinto the coronal phase with the detection of [Si VI], [Si VII],[Ca VIII], [S VIII], and [S IX] lines in its spectrum (Rudy etal. 2005). Along with these lines, unidentified nova featuresat 0.8926, 1.1110, 1.5545 and 2.0996 µ m were also present inthe spectrum. However, no evidence for emission from dustwas seen in these observations.V574 Pup was observed by the Spitzer Space Obser-vatory one year after the outburst revealing strong coronallines (Rudy et al. 2006). The spectroscopic observations byLynch et al. (2007), three years after the outburst, showedthe persistence of the coronal phase. Rudy et al. (2006) haveremarked that the presence of strong coronal emission linessuggests similarity with a He/N nova. Siviero et al. (2005)have inferred that V574 Pup suffers negligible interstellarextinction though it is close to the galactic plane (b=-2 ◦ ). c (cid:13) Naik et al.
350 400 450 500121086 M a gn it ud e Time (JD − 2453000)
V−band data from AAVSOV−band data from AFOEVDates of our observations
Figure 1.
The V-band light curve of V574 Pup obtained fromthe AAVSO (shown by filled circles) and AFOEV data (shown byopen circles). The arrow marks show the days of our near-infraredobservations.
They estimate that the nova is located at very large distanceof 15 to 20 kpc. In a later subsection we determine the dis-tance to V574 Pup using the optical light curve and MMRDrelation. V574 Pup is one of the novae detected with X-rayemission, among 12 classical novae studied by Ness et al.(2007) using Swift observations. It was observed on severaloccasions between 2005 May and 2005 August by Swift. TheX-ray spectra showed it to be in the super soft X-ray phase.Ness et al. (2007) estimate the color excess E ( B − V ) = 0.5and the distance to be 3.2 kpc. Near-infrared spectroscopic and photometric observations ofV574 Pup were carried out fairly extensively using the 1.2-m telescope of the Mt Abu Infrared Observatory. The V band light curve is shown in Figure 1 with the epochs ofour observations marked by arrows - the log of the obser-vations is given in Table 1. The Mt. Abu spectra were ob-tained at a resolution of ∼ ×
256 HgCdTe (NICMOS3) ar-ray. Photometric observations of the nova were carried outon several nights (Table 1) in photometric sky conditions us-ing the NICMOS3 array in the imaging mode. Several frameswere obtained in four dithered positions, typically offset by ∼
30 arcsec from each other, with exposure times rangingfrom 0.4–100 s depending on the brightness of the nova. Thesky frames were generated by median combining the averageof each set of dithered frames and subsequently subtractedfrom the nova frames. A nearby field-star SAO 174367, ob-served at similar airmass as the nova, was used as the stan-dard star for photometric observations. Aperture photome-try was done using the APPHOT task in IRAF.Spectral calibration was done using the OH sky linesthat register with the stellar spectra. The spectra of thenearby field star SAO 174400 (1 Pup) were taken in
JHK bands at similar airmass as that of V574 Pup on all the ob-servation nights to ensure that the ratioing process (nova M a gn it ud e H−Band350 400 4501086 Time (JD − 2453000)K−Band
Figure 2.
The light curves of V574 Pup in the
JHK bandsobtained from the Mount Abu Observatory. The J , H and K band light curves are shown in top, middle and bottom panelsrespectively. spectrum divided by the standard star spectrum) removesthe telluric lines reliably. 1 Pup, although an emission linestar (spectral type A3 Iab), was chosen as the standard stardue to its proximity to V574 Pup to minimise the effectsof differential airmass between V574 Pup and the standardstar. We have carefully removed the hydrogen lines in thespectra of 1 Pup (Pa β and Br γ were seen to be in emis-sion; other Brackett lines are in absorption) before takingthe ratios. The ratioed spectra were then multiplied by ablackbody curve corresponding to the standard star’s effec-tive temperature to yield the final spectra. Extraction andreduction of the spectra were done using IRAF tasks. We make a distance estimate to V574 Pup by analysingthe V band light curve using archival data from the Ameri-can Association of Variable Stars (AAVSO) and AssociationFrancaise des Observateurs dEtoiles Variables (AFOEV).The V band light curve is presented in Figure 1 with AAVSOdata shown in filled circles and AFOEV data shown inopen circles. The present light curve shows that the dis-covery of V574 Pup took place on its way to the maxi-mum that was reached on 2004 November 22.2611. The pre-maximum brightening lasted for one day and subsequent tothe maximum there was a sharp drop by about 2 magnitudesagain within a day. Following this drop V574 Pup showeda relatively slower rise reaching V=8.06 on 2004 November26.2458 and steadily decreased in brightness thereafter. Forthe purpose of calculation of t and t , the time for a de-cline of 2 and 3 magnitudes respectively, we assume thatthe sharp drop of 2 magnitudes lasting for a day can be c (cid:13) , 000–000 ear-infrared Observations of Nova V574 Pup Table 1.
Log of the Mt. Abu near-infrared observations of V574 Pup. The date of outburst is taken as 2004 November 20.67 UT.Date of Days since Integration time (s) Integration time (s) Nova MagnitudeObservation outburst J-band H-band K-band J-band H-band K-band J-band H-band K-bandSpectroscopic Observations Photometric Observations2004 Nov. 25 5 40 40 60 28 28 24 5.65 ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ignored and take V max = 6.93 on 2004 November 22.2611UT (JD 2453331.7611). We then derive from the V-bandlight curve a value of t =10 ± t =25 ± M V = − B and V magnitudes as 7.79 and 6.93 respectively. The mean intrin-sic color of novae at maximum is estimated to be ( B − V ) = +0.23 ± E ( B − V ) = 0.63 and the extinction as A v =1.95. We thus es-timate the distance to V574 Pup to be d =5.5 kpc and adoptthis value in future calculations. Ness et al. (2007) have de-rived a distance of 3.2 kpc which is closer to the value ob-tained in the present analysis compared to the 15 - 20 kpcestimate by Siviero et al. (2005). The significantly highervalue for the distance estimated by Siviero et al. (2005) isdue to the fainter value of V max considered by them andalso their assumption of a negligible interstellar extinctiontowards the nova. JHK light curves of V574 Pup
The
JHK light curves of V574 Pup, obtained from thepresent photometric observations are presented in Figure 2.A gradual fading is seen in all the three bands similar to the V band behavior. Further, no rise is seen in any of the near-IR bands indicating the absence of dust formation duringthe period of four months following the outburst. We notefrom the figure, and also from the data in Table 1, thatthe ( J − H ) color index is generally found to be negative- specially so during the later stages of the observations. Anegative ( J − H ) index is characteristic of novae as shown byWhitelock et al. (1984). The reason for this is the presenceof strong emission features in the J band spectra such as the Pa β and Pa γ , HeI 1.083 µ m and the OI 1.1287 µ m lines. Asdiscussed in the following subsection, the HeI 1.083 µ m andthe OI 1.1287 µ m lines are specially strong in V574 Pup giv-ing rise to the observed behavior of the ( J − H ) color index(the peak-to-continuum ratio of the HeI 1.083 µ m line goesas high as ∼
150 in March 2005). However, it may be notedthat a negative ( J − H ) index is not expected if dust for-mation takes place (Whitelock et al. 1984; an example isthe dust-forming nova V1280 Sco (Das et al. 2008)). Theobserved K band brightness of the nova is also affected bysignificant contributions from the HeI 2.0585 µ m and Br γ lines. JHK spectra
The
JHK spectra of V574 Pup are presented in Figures 3,4 & 5, respectively. The early spectra cover the epoch of re-brightening seen at optical wavelengths and display typicalemission lines of Paschen and Brackett series lines from hy-drogen and the OI lines at 1.1287 µ m and 1.3164 µ m. Alsoseen prominently are lines of carbon and nitrogen, particu-larly in the J band. The details of the line identification aregiven in Table 2. The large observed ratio of the OI 1.1287/ 1.3164 µ m lines indicates that Ly β fluorescence is thedominant process contributing to the strength of the 1.1287 µ m line. The HeI line at 2.0581 µ m is clearly detected in the K band spectra taken on November 29. This, and the J bandHeI line at 1.0830 µ m, gain in strength as the nova evolves.Their strength becomes comparable to hydrogen lines by 24-25 December and exceed them in strength by early Januaryindicating a progress towards higher excitation conditionsin the ejecta. Towards the end of our observations these HeIlines become very strong while other weaker He I lines at1.7002 µ m and 2.1120-2.1132 µ m are also seen. We donot detect any coronal line features during the four months c (cid:13) , 000–000 Naik et al.
Figure 3.
The J -band spectra of V574 Pup at different epochs with the continuum being normalized to unity at 1.25 µ m. of our observational campaign. As mentioned earlier, manycoronal features were detected in V574 Pup one year afterthe outburst that were seen to last for the next two years(Rudy et al. 2005, Lynch et al. 2007).During the course of our observations of V574 Pup,there were no significant changes seen in the width of theemission lines in the J , H , & K band spectra. This impliesthe absence of any significant change in the expansion veloc-ity of the ejected envelope. In order to quantify the changesin the line widths in V574 Pup, the evolution of Brackettseries lines which are prominent was investigated. The over-all width of the hydrogen Brackett series emission lines didnot change appreciably during the observations. The meanvelocity of the expanding gas, from the Br lines, is estimatedto be 1830 ±
400 km s − which agrees well with the findings by Rudy et al. (2006) who determined an average velocityof 1800 km s − .It is also noted that no lines from low ionization specieslike NaI or MgI are seen in the JHK spectra. These low ion-ization lines, which are indicative of low temperature condi-tions, have been suggested as potential diagnostic features topredict dust formation in the nova ejecta (Das et al. 2008).The absence of these lines in V574 Pup is consistent withthe lack of dust formation in this nova. In this context, thereis a line at 2.1452 µ m which matches a NaI transition atthat wavelength. However, for reasons discussed in Das etal. (2008), we are doubtful whether this line should be at-tributed to NaI. A notable feature in the H band spectra,is the structure of the Br11 line at 1.6806 µ m which is seento be distinctly different from other Brackett series lines in c (cid:13) , 000–000 ear-infrared Observations of Nova V574 Pup Figure 4.
The H -band spectra of V574 Pup at different epochs with the continuum being normalized to unity at 1.65 µ m. terms of both width and shape. This is most likely causeddue to the blending of an additional nearby emission line ofsignificant intensity. In the following subsection, we attemptto identify this additional line and show that it most likelyis due to a FeII line at 1.6872 µ m. µ m emissionline in V574 Pup We examine the distinct difference seen in the structure ofBr11 line vis-a-vis other Br lines (Figure 4). This differencehas persisted till the middle of 2005 January. The veloc-ity of 3000 ±
400 km s − corresponding to the full width athalf maximum (FWHM) of the Br11 line is larger than theaverage value of 1830 ±
400 km s − corresponding to other Brackett series lines. This indicates that there is a definitecontribution to Br11 line from an adjacent emission feature.It is noticed in the early spectra that there is a centralenhancement (spike) in the Br11 line which gradually de-creases in strength and by late 2004 December, Br11 linestarts showing a prominent redward wing. Considering thisbehavior, we have looked for an emission line on the higherwavelength side of Br11. In two recent novae studied by us,namely, V2615 Oph (Das et al. 2009) and RS Oph (Banerjeeet al. 2009) an emission line at 1.6872 µ m has been clearlydetected which is attributed to FeII. We suspect that thisFeII line is present here too. To study the effect of this lineon Br11 , we have generated synthetic line profiles by addingtwo lines viz. a primary line centered at 1.6806 µ m corre-sponding to Br11 and a second line at 1.6872 µ m corre- c (cid:13) , 000–000 Naik et al.
Figure 5.
The K -band spectra of V574 Pup at different epochs with the continuum being normalized to unity at 2.2 µ m. sponding to FeII. Since there is no apriori knowledge on theshape of the FeII line, we assume that its shape can be sim-ulated by the the Br12 line profile (we choose Br12 becauseit is free from blending with other lines). We have similarlyassumed that Br11 is well simulated by the observed lineprofile of Br12 line. Keeping the peak intensity of syntheticBr11 line constant we have varied the intensity of the FeIIline to simulate the observed temporal evolution. Two suchsynthetic spectra are shown in Figure 6. In the figure, weshow the individual line profiles of Br11 and FeII with dot-ted lines and their resultant, co-added profile with a dashedline along with the observed profile (solid gray line). Theleft and right panels show the profiles for 2004 November26 and 2004 December 9 respectively. The resultant and ob-served profiles are shown with some offset from the individ- ual profiles for clarity. It is seen that the synthetic profilesresulting from the combination of Br11 and Fe II match theobserved line profile reasonably well. This indicates that theFe II 1.6872 µ m line is present in V574 Pup. Based on thework of Banerjee et al. (2009), another Fe II line at 1.7413 µ m could also be expected in the spectrum. It is possi-ble that this line is also there, but it is difficult to draw adefinitive conclusion regarding its presence since it could beblended, rather too closely with Br10 and also a cluster ofCI lines in the 1.74-1.77 µ m region. We note that these FeIIlines are not too commonly reported in the spectra of no-vae. Apart from V2615 Oph (Das et al. 2009) and RS Oph(Banerjee et al. 2009), there are two more novae, namelyV2540 Ophiuchi (Rudy et al. 2002) and CI Aquila (Lynchet al. 2004) where these lines appear to be detected. The ex- c (cid:13) , 000–000 ear-infrared Observations of Nova V574 Pup Figure 6.
The profiles of Br11 line at 1.6806 µ m (dotted line),Fe II line at 1.6872 µ m (dotted line), the added line (Br11+Fe II)intensity (dashed line) and the observed data (gray solid line) areshown for 2004 November 26 (left panel) and for 2004 December 9(right panel). The intensity of the Fe II line was adjusted to matchthe co-added profile with the observed profile. The co-added andobserved profiles are shown with some offset in order to make itclear for comparison. citation mechanism for these lines is believed to be Lyman α and Lyman continuum fluorescence coupled with collisionalexcitation (Banerjee et al. 2009 and references therein). We analyse and discuss the evolution of the continuum spec-tra of V574 Pup here. At the time of outburst, a nova’scontinuum is generally well described by a blackbody dis-tribution from an optically thick pseudo-photosphere cor-responding to a stellar spectral type A to F (Gehrz 1988).The spectral energy distribution then gradually evolves intoa free-free continuum as the optical depth of the nova ejectadecreases (Ennis et al. 1977; Gehrz 1988). The evolution ofthe continuum of V574 Pup is shown in Figure 7 whereinwe have shown representative spectra sampling the dura-tion of our observations. The spectra in Figure 7 were fluxcalibrated using the broadband
JHK photometric observa-tions presented in Table 1. During this process of calibratingthe flux in the continuum, we note that the observed broad-band flux is a sum of contributions from both the continuumand also from emission lines. As the emission lines of HI, OIand HeI are strong and contribute significantly to the broad-band fluxes, we have first calculated the contribution of allthe prominent emission lines to the observed spectra andremoved this contribution from the broad band photometricfluxes measured from the
JHK photometry. This gives thetrue continuum flux which was used to calibrate the spectrain Figure 7.
Table 2.
List of prominent lines in the
JHK spectraWavelength Species Remarks( µ m)1.0830 He I1.0938 Pa γ µ m1.1748 - 1.1800 C I strongest lines at1.1748, 1.1753, 1.1755 µ m1.1819 - 1.1896 C I strongest lines at1.1880, 1.1896 µ m1.2187 - 1.2382 C I, N I blend of N I 1.2187,1.2204, 1.2329, 1.2382,& CI 1.2249, 1.2264 µ m1.2461, 1.2469 N I1.2562 - 1.2614 C I blend of C I 1.2562, 1.25691.2601, 1.2614 µ m1.2818 Pa β µ m2.1452 Na I?2.1655 Br γ µ m2.2906 C I We have tried to fit the spectra in Figure 7 with powerlaw fits i.e. F λ ∝ λ − α . In the beginning of our observationi.e. on 2004 November 25 (4 days after outburst) and 2004December 01, the continuum spectrum approximately fitsa spectral index of α ∼ α = 4.0 at longer wavelengths, does notsimulate the data too well. The subsequent spectra graduallybecome flatter with a slope of α ∼ K. c (cid:13) , 000–000 Naik et al.
Figure 7.
The composite
JHK spectra of V574 Pup for 25 November 2004 (A), 1 December 2004 (B), 14 December 2004 (C), and 25February 2005 (D) from near-infrared observations with the Mt Abu telescope. Model fits to the data (using either a power law or afree-free spectral dependence) are shown by the continuous lines; the broadband fluxes (corrected for contribution from line emission -see text) are shown by filled circles. The slope of the continuum spectra of A, B, and C are compared with power law fits ( F λ ∝ λ − α )with slopes α = 2.75, 2.75 and 2.0, respectively. A free-free emission function at temperature 10 K is plotted along with the fourthspectrum (noted as D). It appears that the nova continuum has flattened gradually to a free-free type of emission towards the end ofour observations.
The recombination case B analysis for the HI lines were car-ried out for all the observed spectra and the representativeresults for five epochs covering the first 80 days of our obser-vations are shown in Figure 8. We have plotted in Figure 8the observed relative strength of Brackett series lines withthe line strength of Br12 as unity along with the predictedvalues for three different recombination case B emissivityvalues from Storey & Hummer (1995). These predicted val-ues cover a representative temperature of T = 10 K and the electron densities of n e = 10 cm − , 10 cm − and 10 cm − . High electron densities are considered because theejecta material is expected to be dense in the early stagesafter the outburst. For the early epochs,namely 2004 Novem-ber 25 and 2004 December 1, the Br10 line is not includedas it is at the edge of the observed spectra and not ade-quately covered. The errors in the estimated line strengthsare ∼
10% for the Br γ , Br12 & Br13 lines and ∼
20% for theBr14, Br15, Br16 & Br17 lines. The errors for the Br11 lineare ∼
30% for 2004 November 25, ∼
20% for 2004 December c (cid:13) , 000–000 ear-infrared Observations of Nova V574 Pup ∼
10% for observations on the other days. The vari-able error assigned to the Br11 is due to the presence of aFe II line at 1.6872 µ m that was strong in the initial phaseof our observations, gradually weakened and finally becameundetectable.Figure 8 shows that the observed line intensities clearlydeviate from case B values in the initial phase of our ob-servations. Specifically, Br γ which is expected to be therelatively stronger than the other Br lines, is observed tobe considerably weaker in the early observations. This ismost likely due to optical depth effects in the Brackett lines(Lynch et al 2000). Such deviations from the recombinationcase B conditions during the early stages after outburst canbe expected and have been observed in other novae too,for example, V2491 Cyg and V597 Pup (Naik et al. 2009),RS Oph (Banerjee et al. 2009) etc. However, towards theend of the observations, on 2005 February 13 and 25 (fourthand fifth panels of Figure 8) there is an indication that CaseB conditions have begun to prevail. For these last two dates,it is found that the observed data matches well with the pre-dicted values for the recombination case B values of T = 10 K and an electron density in the range n e = 10 − cm − . We estimate the mass of the ionized gas in the ejecta usingthe fact that on 2005 February 25, the observed SED ofV574 Pup is well fit by a free-free flux distribution at atemperature of T = 10000K (Figure 7).The free-free volume emission coefficient from anionized gas is given by j λ ff = 2.05 × − λ − z g T − / n e n i e ( − c /λ T) W cm − µ m − where λ is the wavelength in µ m, z is the charge, g is theGaunt factor (assumed equal to unity), T is the temperature,n e and n i are the electron and ion densities respectively andc = 1.438 cm K. The total continuum emission from thenova ejecta can then be estimated by multiplying the fluxgiven in the above equation with the shell volume V s andequating it to the observed flux F λff which equals: F λ ff = j λ ff × V s /4 πd where d is the distance to the object. We use d = 5.5kpc, T = 10 K and n e = n i assuming a pure and completelyionized hydrogen ejecta ( z = 1). At K band center ( λ =2.2 µ m), using the observed flux on 2005 February 25 to be1.325 × − W cm − µ m − and n e = 10 cm − derivedfrom the recombination analysis, we obtain the volume ofthe emitting HI gas to be V s = 2.2 × cm . Similar valuesfor V s are obtained if the J and H band observed fluxes andcorresponding central wavelengths of these bands are usedinstead.The mass of the ionized gas can then be calculated usingM gas = V s n e m H , where m H is mass of the hydrogen atom.Taking n e =10 cm − gives a value of M gas = 1.8 × − M ⊙ .A similar calculation for n e =10 cm − , which could also bea valid estimate in V574 Pup (as indicated from case B anal-ysis for 25 Feb 2005), yields M gas = 1.8 × − M ⊙ . Thus,within the uncertainties associated with the parameters in-volved, we would estimate the mass of the ionized gas to bein the range of 1.8 × − to 1.8 × − M ⊙ . This estimate isreasonably in agreement with the observed range of the mass R e l a ti v e I n t e n s it y
25 December 200402468 13 February 200510 1502468 Upper level number25 February 2005
Figure 8.
Recombination analysis for the hydrogen Brackett linesin V574 Pup on selected dates of our near-IR observations (asnoted in the figure). The abscissa is the upper level number ofBrackett series line transition. The line intensities are relativeto that of Br 12 which is normalized to unity. The errors in theestimated line strengths are ∼
10% for the Br γ , Br12 & Br13 linesand ∼
20% for the Br14, Br15, Br16 and Br17 lines. The errorsfor the Br11 line are ∼
30% for 2004 November 25, ∼
20% for 2004December 1 and ∼
10% for observations on the other days. Thecase B model predictions for the line strengths are also shown for atemperature T = 10 K and electron densities of n e = 10 cm − (dot-dash line), 10 cm − (solid line), and 10 cm − (dotted line). of novae ejecta (1 to 30 × − M ⊙ ) and also the theoreti-cally calculated range of 5.3 × − to 6.6 × − M ⊙ in theextended grid of models computed by Yaron et al. (2005).The mass estimate made above can be checked for con-sistency through an alternative approach. As discussed insection 3.6, on 13 and 25 February 2005 there is a rea-sonably good match between the observed line intensitiesof HI Brackett series lines with the theoretical case B val-ues listed in Storey and Hummer (1995) for T=10 K and n e = 10 − cm − . For these values, the emissivity in Br γ is expected to be j ( Brγ ) ∼ × − W cm − (Storey andHummer, 1995). From the observed data of 2005 February13 and 25, the Br γ line is measured to have a mean lineluminosity of ∼ × − W cm − . Using d = 5.5 kpc, thetotal power in the line is thus 1.8 × W. With j ( Brγ )as estimated above, this yields the volume of the emittinggas to be V s = 4.0 × cm (consistent with the value c (cid:13) , 000–000 Naik et al. of 2.2 × cm derived from the free-free analysis) whichthereby leads to a similar mass estimate as made earlier. We have presented an extensive set of spectroscopic andphotometric observations of nova V574 Puppis. From the V band light curve the distance to the nova is estimated to be ∼ µ m in the near-IR spectra. The nova continuum is modeledand found to evolve from a λ − . dependence to a free-freeemission during the period of our observations. A recombi-nation analysis of the HI lines is presented. We estimate themass of the ionized gas in the ejecta and show it to lie inthe range 10 − M ⊙ and 10 − M ⊙ . ACKNOWLEDGMENTS
We wish to thank the referee Prof. A Evans for his usefulsuggestions on the paper. The research work at PhysicalResearch Laboratory is funded by the Department of Space,Government of India. We thank George Koshy for help withsome of the observations. We acknowledge with thanks thevariable star observations from the AAVSO InternationalDatabase, contributed by observers worldwide, and used inthis research. This research has made use of the AFOEVdatabase, operated at CDS, France.
REFERENCES
Ashok, N. M., Banerjee, D. P. K., 2004, IAU Circ., 8447, 4Ayani, K. 2004, IAU Circ., 8443, 2Banerjee, D. P. K., Das, R. K., Ashok, N. M., 2009, MNRAS,399, 357Das, R. K., Banerjee, D. P. K., Ashok, N. M., Chesneau, O., 2008,MNRAS, 391, 1874Das, R. K., Banerjee, D. P. K., Ashok, N. M., 2009, MNRAS,398, 375della Valle, M., Livio, M., 1995, ApJ, 452, 704Ennis, D., Beckwith, S., Gatley, I., Matthews, K., Becklin, E. E.,Elias, J., Neugebauer, G., Willner, S. P., 1977, ApJ, 214, 478Gehrz, R. D., 1988, ARA&A, 26, 377Lynch D.K., Rudy R.J., Mazuk S., Puetter R.C., 2000, ApJ, 541,791Lynch, D. K., Wilson, J. C., Rudy, R. J., Venturini, C., Mazuk,S., Miller, N. A., Puetter, R. C., 2004, AJ, 127, 1089Lynch, D. K., Rudy, R. J., Prater, T. R., Gilbert, A. M., Mazuk,S., Perry, R. B., Puetter, R. C., 2007, IAU Circ., 8906, 1Naik, S., Banerjee, D. P. K., Ashok, N. M., 2009, MNRAS, 394,1551Nakano, S., Tago, A., Sakurai, Y., Kushida, R., Kadota, K., 2004,IAU Circ., 8443, 1Ness, J.-U., Schwarz, G. J., Retter, A., Starrfield, S., Schmitt,J. H. M. M., Gehrels, N., Burrows, D., Osborne, J. P., 2007,ApJ, 663, 505Rudy, R. J., Lynch, D. K., Mazuk, S., Venturini, C. C., Puetter,R. C., Perry, R. B., 2002, AAS, 201, 4006 Rudy, R. J., Lynch, D. K., Mazuk, S., Venturini, C. C., Puetter,R. C., Perry, R. B., Walp, B., 2005, IAU Circ., 8643, 2Rudy, R. J. et al., 2006, AAS, 209, 906Samus, N. N., Kazarovets, E., 2004, IAU Circ., 8445, 2Siviero, A., Munari, U., Jones, A. F., 2005, IBVS, 5638, 1Storey, P. J., Hummer, D. G., 1995, MNRAS, 292, 41Warner, B., 1995,
Cataclysmic Variable Stars. Cambridge Astro-physics Series , Cambridge Univ. Press, Cambridge, New YorkWhitelock, P. A., Carter, B. S., Feast, M. W., Glass, I. S., Laney,D., Menzies, J. W., Walsh, J., Williams, P. M., 1984, MNRAS,211, 421Yaron, O., Prialnik, D., Shara, M. M., Kovetz, A., 2005, ApJ,623, 398 c (cid:13)000