Deep Near-Infrared Survey of the southern sky

Status report as of May 15, 1998

DENIS image of the galactic center

The front picture is a three colour image of a typical DENIS field (12×12 arcminutes) roughly towards the center of our Galaxy. Each of the approximately 4000 dots is a star and most of them should be close to the actual center of the Galaxy. The colour of the dot shows how red the star is by itself or how much it gets red by intervening clouds. The colours of the stars in the left side of the field are all much redder because of the presence of such a cloud; in the lower right quarter a very dark cloud obscures a small area. Photographs in visual light of the same area contain only a very small fraction of all these stars: the light is extinguished by clouds because at visual wavelengths the extinction is so much stronger. This result is a standard product from the DENIS survey.

Progress report on Denis,
prepared by H. Habing & N. Epchtein
May 15, 1998

1. Goals of DENIS

DENIS is a project to survey the all-southern sky in three wavelength bands (Gunn-i 0.82 µm; J, 1.25 µm; and Ks, 2.15 µm) with limiting magnitudes 18.5, 16.5 and 14.0, respectively. The observations are performed with the 1m-ESO telescope at La Silla (Chile). The DENIS instrument is made up of a 3-channel camera built of commercially available detector arrays by the Observatoire de Paris and with major contributions from other European Institutes, notably: the IAS in Frascati, the Observatoire de Grenoble, the University of Innsbruck, the Observatoire de Lyon, and the IAC in Tenerife. The survey is carried out by observing strips of 30 in declination and 12 arcminutes in Right Ascension with an overlap of 2 arcminutes between consecutive strips. The observations started at the end of 1995 and will be completed by the summer of 2001. The data are reduced in two consecutive steps, the first at the Institut d'Astrophysique de Paris and the second at the Leiden Observatory. The position of a general extracted point source is provided with an accuracy better than 1 arcsec and its magnitude to better than 0.1 mag. The data will be made publicly available as soon as possible after completing the data reduction. The Centre de Données Astronomiques de Strasbourg (CDS) is implementing the final databases and will eventually provide access of the processed and calibrated data to the worldwide community. It is expected that DENIS data will be of great help in determining projects for the VLT and future space missions.

The principal investigator of the DENIS project is N. Epchtein (Observatoire de la Côte d'Azur); scientists and engineers from seven European countries and from Brasil are involved. The scientific exploitation of the already existing data is carried out by several of the different groups involved. These special programs allow a continuous monitoring of the quality of the survey data and therefore give benefit to the project. The programs also give motivation to the many groups that voluntarily spend time and energy to make a successful project.

2. The survey

2.1 Observations already completed

The survey observations with all 3 cameras started in routine mode in December 1995 and continued successfully until mid 1997 when a series of technical failures began. Because of insufficient technical support caused by financial problems the survey was interrupted in May 1997, due to electronics problems in the J–camera, then again from November 1997 until January 1998, due to electronics problems in the K–camera, although during this second interruption observations were carried on with I– and J–cameras, mainly on the Magellanic Clouds. Bad weather in 1997 due to El Niño also contributed to make 1997 considerably less successful than 1996. The updated overview of the survey status shows that about 35% of the southern hemisphere has been covered, which makes a database of about 1,000,000 images. More than 80% of these images has been judged of good or acceptable quality. Low-quality measurements are routinely detected by the data reduction software; they were caused mostly by poor focus or seeing. We expect that the survey will be completed during 2001, under the assumption that we will not suffer again from long periods of instrumental breakdowns.

2.2 Data reduction; availability of data

All the images received from La Silla have been processed at the PDAC, the Paris Data Analysis Center, and will be made available on high-density data tapes. The LDAC, the Leiden Data Analysis Center, started the point source extraction in mid 1997. The source extraction software implemented especially by Emmanuel Bertin and Erik Deul for DENIS is now used also for the newly started ESO imaging survey (EIS). Because the data reduction takes less time than their measurement the existing backlog is reduced and will have disappeared within a few months. About one Tbyte of data is now available to the consortium members.

2.3 Relation to the 2MASS-survey

The 2MASS or ``Two Micron All Sky Survey'' will survey the Northern and the Southern sky each with a 1.3m dedicated telescope in the J–, H–, and K– photometric bands with limiting magnitudes 17.1, 16.4, 15.6, respectively. The project is funded by NASA and brings together astronomers from six institutes in the USA; the principal investigator is M.F. Skrutskie from University of Massachusetts in Amherst. The survey of the North hemisphere started in April 1997, that in the South very recently. Experience in the first year of the survey shows that the project will achieve its goals, and the nominal completion time is 2001. Poor and unpredictable wheather conditions seriously slowed down DENIS progress, especially in Winter 1997.

The overlap between DENIS and 2MASS is large, but there is one essential difference: DENIS surveys the sky also in the I–band at 0.81 µm whereas the shortest wavelength for 2MASS is the J–band at 1.2 µm. In spite of the overlap the two surveys are therefore complementary. Experience shows that R–band photometry at a wavelength of 0.62 µm often provides the data at the longest wavelengths. For many astrophysical objects the gap between the R– and J–bands is often too large to arrive at solid statistical conclusions. I–band photometry narrows this gap significantly.

3. First scientific results

3.1 Faint stars in the solar neighbourhood: red and brown dwarfs

The range of DENIS wavelengths is large (from the I-band to the K-band is a factor 2.5 in wavelength) and this makes the survey data well suited for the detection of red dwarfs of the latest spectral types (M7 or later) about which little is known. The analysis of 220 square degrees of high galactic data led to the identification of 25 dwarfs later than M7V, doubling the known inventory of these very cool objects (Delfosse et al., 1998), and 3 brown dwarf candidates. One of these three is DENIS-P J1228-1547 (Delfosse et al., 1997); Keck spectroscopy showed the presence of a strong lithium absorption line at 680.7 nm (Martín et al., 1997; Tinney et al., 1997); in such fully convective objects lithium nuclear burning never took place and thus there has been no hydrogen burning either. Interestingly, one of the other two candidates DENIS-P J0205-1159 has no lithium (Martín, private communication) in spite of a significantly lower effective temperature.

Very few of these enigmatic objects are known. The obvious priority now is to extend this analysis to a much larger sky area. Extrapolation from the above results shows that the already observed data contain 300 extreme M dwarfs and 30 brown dwarfs. Together these will provide a first statistically significant sample of these objects, sufficiently large to determine their local density (i.e. the disk luminosity function) and to form the basis for the calibration of the observable properties of brown dwarfs (flux calibrated spectra and distances) into more fundamental physical quantities (effective temperature, luminosity, mass).

3.2 The interstellar medium; extinction

Detailed maps of the extinction in nearby molecular clouds (in Chameleon, Ophiuchus, Orion and Serpens) can be obtained via starcounts in the DENIS data. An example is the study by L. Cambrésy et al. (1997) of the nearby Cham I cloud that has produced an extinction map of much higher angular resolution than before (see the enclosed plate that shows the extinction map derived from the DENIS star count in the J band and translated into visual extinction scale. There is an excellent coincidence of the 'cold' dust emission represented by the isophots in white as measured by IRAS at 100-micron) . The DENIS survey observations will make possible to obtain similar maps of other well–known and nearby clouds, e.g. those in Ophiuchus and Serpens, but also in small globules. When these extinction maps are combined with maps in the CO millimeter line emission, with 60 and 100 µm maps obtained by IRAS the distribution of gas, dust and of the temperature will be known much better than so far. Counting stars in the whole DENIS data base will eventually provide the first deep maps of interstellar extinction throughout the Galaxy.

3.3 Young stars in dark clouds

The exploration of DENIS data of dense molecular clouds will lead to the detection of a large number of stars in the process of formation (young stellar objects or YSOs, protostars, etc.). E. Copet (thesis, 1996) has studied DENIS data of the Orion molecular cloud and discovered many new faint YSOs which are likely to be T Tau stars. From an analysis of similar data Cambrésy et al. (1998) detected YSOs that probably form a new population of low-luminosity T Tau stars: a full survey of the clouds in Chameleon (an area of several tens of square degrees) and in other similar regions will improve our knowledge of the low-end of the luminosity function. Because young stars sometimes are strong X-ray emitters the DENIS data are compared with ROSAT surveys results; this work is in collaboration with R. Neuhauser (MPE Garching).

3.4 The distribution of stars in the inner disk and in the bulge of our Galaxy

At visual wavelength the search for stars in the inner galactic disk is limited to distances smaller than one or two kiloparsecs, while at 2 µm red giants and supergiants are detected up to galactic center distances: the extinction in the infrared is much smaller (AK~AV /10). DENIS will make the first large-scale digital survey in the infrared; in addition it has homogeneous magnitude limits in its photometric bands, with the exception of a modest variation (0.6 mag) in the early data in the K-band. Thus new studies are possible of the distribution of dust, red giants and red supergiants in our Galaxy; these studies tackle the question what are the distributions of different kinds of stars in our Galaxy? The newly detected supergiants will give the distribution of the star formation rate and the red giants the distribution and evolution of older stellar populations in the galactic plane. Detailed new information will be obtained about the inner spiral arms, the ring, the bar and the bulge of the Galaxy. The COBE/DIRBE data have been important in finding such large scale structures; the much higher angular resolution of DENIS (a few arcseconds instead of a few degrees with COBE) will give more detailed information and will solve issues that could not be solved with the COBE/DIRBE data. The distribution of different kinds of stars in our Galaxy, that has been a major topic of research earlier in this century, is becoming an important topic again, mainly due to new and larger surveys, of which DENIS is an example, but also the surveys that search for the microlensing effects (MACHO, EROS etc.)

The DENIS data will be used to interpret the data on the inner Galaxy at 7 and 15 µm obtained through the ISOGAL program, one of the longest ISO programs. The DENIS and ISO teams of Co-investigators have many astronomers in common, e.g. the Principal Investigators (N. Epchtein for DENIS and A. Omont for ISOGAL). Several other surveys of the inner Galaxy exist already, e.g. radiocontinuum studies, and we expect that the additional information from the DENIS and ISOGAL survey on the stellar content will be important, perhaps essential to obtain a comprehensive model for the inner Galaxy.

The existing survey data have already significantly constrained models for the large scale structure of the Galaxy (Ruphy, 1996, Ruphy et al, 1997). The scale length in the galactic plane has been determined as well as the outer edge of the stellar disc, the position of the Sun relative to the galactic plane and the shape and density of some spiral arms and of the 3 kpc ring; we look forward to cross-identify DENIS data on the ring and the bulge with those of ISOGAL. DENIS adds data that reveal the presence of stars that otherwise would have been missed.

There is a strong interest in DENIS survey data on the bulge and the bar of our Galaxy; again: there is considerable overlap between ISOGAL and DENIS survey. Schultheis and Hron (Vienna) analysed the DENIS measurements of 30,000 sources in the Palomar-Groningen field #3 (PG3); among these there are 36 previously known AGB variables. Carbon stars are missing, confirming earlier studies by V. and B. Blanco. Whitelock (Cape Town) is studying red giants in two areas where we see the near and the far end of the galactic bar.

Planetary nebulae (PNe) are an interesting category of objects as well. PNe emerge from AGB stars and represent the same galactic population. The study of PN properties thus complements that of the AGB stars; in fact, they are probably the best objects from this population to derive the abundances of the various elements and the variation of these abundances over the face of the Galaxy. Already 250 of the 750 known PNe in the southern sky have been detected by the DENIS survey. The survey will provide better surface photometry at its three wavelengths than previously known. It also will discover faint, red, background stars that often mess up photometry at visual wavelengths but that also may be used as distance indicators (Kimeswenger, Kerber, Roth et al., 1998). A further goal concerns the few thousand PNe in the inner Galaxy that must exist, but that have escaped detection so far because of interstellar extinction. Kimeswenger, Kerber, Roth et al. expect that DENIS will find many new PNe.

Even at medium or high latitude fields, where interstellar extinction at visual wavelengths is not a problem, the DENIS survey data will identify red giants that can be used to trace their distribution in the halo, well before optical surveys produce results (e.g. the SLOAN survey).

3.5 Stars in the Magellanic Clouds

The Large Magellanic Cloud has a distance modulus of (m–M)=18.6 and the SMC has (m–M)=19.1. Therefore DENIS records data of all stars with (V–I)> 3 that have a higher luminosity than about 2500 L or Mbol< –4.0 and thus all AGB stars except the few that are very red. Earlier photometric and even spectroscopic surveys have produced incomplete samples and the DENIS data will largely correct for this incompleteness. There are large differences in luminosities and element abundances between populations of AGB stars in different surroundings, e.g. between the SMC, the LMC, the outer disk of our Galaxy and our bulge. Most AGB stars in the SMC are carbon-rich whereas in the bulge the AGB stars are all of M–type (``oxygen-rich"). The reasons for these variations are not well understood at present and the DENIS samples may give the data for a breakthrough. Knowledge so obtained will be of great importance in the study of AGB stars in other Local Group dwarf galaxies, studies that are within the grasp of the VLT. Besides DENIS data reveal upper MS stars and SGs that together with RGB and AGB stars allow to investigate the important ratios for stellar population studies. A first goal is to publish a catalogue of point sources in both Clouds; the catalogue is planned to be publicly available around mid 1999. In several small fields in the LMC and SMC ISO detects the only group of AGB stars that DENIS will miss, redder objects that are believed to experience a great mass loss (Loup et al. 1997). A large consortium, started by C. Loup, is studying LMC and SMC data of DENIS and ISO.

3.6 Galaxies and cosmology

The DENIS survey makes it possible to construct large samples of galaxies that are not limited by uncertain interstellar extinction effects on the cut–off magnitude. Moreover, the near infrared domain provides galaxy samples that are more mass-weighted and less affected by recent star formation than galaxy samples in other wavebands; this will have several cosmological implications.

Vauglin and Paturel (both from Lyon) extract galaxies from the survey measurements in a homogeneous manner. The I-band data are the best suited to separate stars from galaxies and to determine the parameters of extended objects. The systematic routine to extract bright galaxies from the I-band images has begun in Lyon in February 1996 and has already led to a catalog of about 36,000 galaxies, of which roughly 2/3 are new. The comparison with another existing sample (that of Mathewson) shows that the uncertainty in DENIS I-magnitude is 0.13 mag.

A special project is the search for galaxies in the Zone of Avoidance in the direction of the ``Great Attractor". A systematic search is being made at Lyon. Routine analysis of the DENIS data base started in March 1997. More than 200 extended objects have already been listed; some of these are not galaxies but newly discovered globular clusters.

It is possible to go as faint as I = 16.5 and extract galaxies with very high (>95%) completeness and reliability, as shown by G. Mamon and colleagues. The galaxy counts in the I– and J– bands are Euclidean and thus require no very recent galaxy evolution (in contrast to counts performed in the optical). A catalogue with 900,000 galaxies in the I–band, 500,000 at J and 6,000 at K is expected. The homogeneity of the extraction is essential for statistical cosmological studies of galaxies such as the measurement of the angular two-point correlation function of galaxies, the derivation of the primordial density fluctuation spectrum, the building of catalogs of groups and clusters, and the study of the variation of galaxy colours with environment. So far, such studies have been based on photographic photometry that is less accurate and has a cut-off that is less homogeneous.

The galaxies discovered in the DENIS survey are expected to be the input for three spectroscopic surveys in the southern sky: one in which radial velocities will be measured (in 2001–2002) for roughly 100,000 galaxies using the UK Schmidt Telescope and a robotic multifibre spectroscope (6dF; PI's M. Colless and G. Mamon); the second (in 1999-2002; PI: G. Paturel) measuring the 21 cm line of roughly 5000 inclined spiral galaxies with the Nançay telescope (which is limited to δ> –39); and the third for roughly 20,000 inclined spirals as well as early type galaxies using the 6dF instrument at high spectral resolution (2003–2005; PI: G. Mamon).

4. Conclusions

The DENIS survey is well underway. Technical problems with the 3-colour camera have been overcome and the reduction software is operating routinely. The photometric and positional accuracy reached agree with the specifications before the start of the project. If financial support remains at its present level the survey is expected to be completed in mid 2001. In addition several projects are carried out that use and validate the survey measurements. The results will be important for their own sake and will also be very valuable for further studies with the VLT. The prospects of DENIS are great.


  • B. de Batz, Observatoire de Paris (bertrand.debatz -at-
  • M. Dennefeld, IAP (dennefeld -at-
  • E. Deul, Leiden Observatory (deul -at-
  • N. Epchtein, Observatoire de la Côte d'Azur (epchtein -at-
  • T. Forveille, Observatoire de Grenoble (forveille -at-
  • P. Fouqué, Observatoire de Paris and ESO, La Silla (pfouque -at-
  • F. Garzón, Instituto de Astrofisica de Canarias, Tenerife (fgl -at-
  • H. J. Habing, Leiden Observatory (habing -at-
  • J. Hron, Vienna Observatory (HRON -at-
  • S. Kimeswenger, Innsbruck University (Stefan.Kimeswenger -at-
  • F. Lacombe, Observatoire de Paris (francois.lacombe -at-
  • T. Le Bertre, Observatoire de Paris (thibaut.lebertre -at-
  • C. Loup, IAP and ESO (loup -at-
  • G. Mamon, IAP and Observatoire de Paris (gam -at-
  • A. Omont, IAP. Paris (omont -at-
  • G. Paturel, Observatoire de Lyon (patu -at-
  • P. Persi, Istituto Astrofisica Spaziale, Frascati, Roma (persi -at-
  • A. Robin, Observatoire de Besançon (annie -at-
  • D. Rouan, Observatoire de Paris (daniel.rouan -at-
  • G. Simon, Observatoire de Paris (guy.simon -at-
  • D. Tiphène, Observatoire de Paris (tiphene -at-
  • I. Vauglin, Observatoire de Lyon (vauglin -at-
  • S. Wagner, Landessternwarte Heidelberg (S.Wagner -at-


  • 1997A&A...324L...5C– L. Cambrésy, N. Epchtein, E. Copet, B. de Batz, S. Kimeswenger, T. Le Bertre, D. Rouan, D. Tiphène, 1997, Astron. Astrophys.,324, L5: ``Extinction map of Chameleon I molecular cloud with DENIS star counts''
  • 1996A&A...313L..21R– S. Ruphy, A.C. Robin, N. Epchtein, E. Copet, E. Bertin, P. Fouqué and F. Guglielmo, 1996, Astron. Astrophys.313, L21 : ''New determination of the disc scale length and the radial cut-off in the anticenter with DENIS data ''
  • 1997A&A...326..597R– S. Ruphy, N. Epchtein , M. Cohen, E. Copet, B. de Batz, J. Borsenberger, P. Fouqué, S. Kimeswenger, F. Lacombe, T. Le Bertre, D. Rouan, D. Tiphène, 1997, Astron. Astrophys. 326, 597 : ''Stellar populations and inhomogeneities of the galactic plane from DENIS star counts''
  • – E. Copet, N. Epchtein, D. Rouan, B. de Batz, D. Tiphène, F. Lacombe, S. Pau, P. Fouqué, T. Le Bertre, T. Forveille, S. Kimeswenger, J.C. Renault, L. Capoani, 1998, Astron. Astrophys., submitted, ''DENIS : Deep near infrared survey of the southern sky. I : the focal instrument''
  • 1997A&A...327L..25D– X. Delfosse, C. Tinney, T. Forveille, N. Epchtein, E. Bertin, J. Borsenberger, E. Copet, B. de Batz, P. Fouqué, S. Kimeswenger, T. Le Bertre, F. Lacombe, D. Rouan, D. Tiphène, 1997, Astron. Astrophys. 327, L25, ``Field brown dwarfs found by DENIS''
  • 1997A&A...327L..29M – E. Martín, G. Basri, X. Delfosse, T. Forveille, 1997, Astron. Astrophys. Lett. 327, L29, ``Keck HIRES spectra of the brown dwarf DENIS PJ1228-1547''
  • 1997ApJ...490L..95T – C.G. Tinney, X. Delfosse, T. Forveille, 1997, ApJ. Lett., 490, L95, ``DENIS-P J1228.2-1547– A New Benchmark Brown Dwarf''
  • 1996A&A...315L.165P – M. Pérault, A. Omont, G. Simon (+ 21 other authors), 1996, Astron. Astrophys., 315, L165 , ``First ISOCAM images of the Milky Way "
  • 1998A&A...332..300K – S. Kimeswenger, F. Kerber, M. Roth, M. Dennefeld, B. de Batz, J. Borsenberger, E. Copet, E. Deul, N. Epchtein, T. Forveille, P. Fouqué, J. Hron, F. Lacombe, T. Le Bertre, D. Rouan, M. Schultheis, D. Tiphène, 1998, Astron. Astrophys., 332, 300 , ``Planetary Nebulae with DENIS. Capabilities of Imaging nebulae''
  • 1998MNRAS.295..119U – M. Unavane, G. Gilmore, N. Epchtein, G. Simon, D. Tiphène, B. de Batz, 1998, MNRAS, 295, 119, ``The inner Galaxy resolved at IJK using DENIS data''
  • 1998A&A...338..977C – L. Cambrésy, E. Copet, N. Epchtein, B. de Batz, J. Borsenberger, P. Fouqué, S. Kimeswenger, D. Tiphène, 1998, ``New Young Stellar Object candidates in the Chamaeleon molecular cloud discovered by DENIS'', Astron. Astrophys., 338, 977
  • 1998A&A...338.1066T – C.G. Tinney, X. Delfosse, T. Forveille, F. Allard, 1998, Astron. Astrophys. 338, 1066, Optical spectroscopy of DENIS mini-survey brown dwarf candidates.
  • – X. Delfosse, C.G. Tinney, T. Forveille, N. Epchtein, J. Borsenberger, P. Fouqué, S. Kimeswenger, D. Tiphène, 1998, ``Searching for Very Low–mass Stars and Brown Dwarfs with DENIS'', Astron. Astrophys., submitted
  • – S. Kimeswenger, C. Kienel, J. Hron, B. de Batz, E. Bertin, J. Borsenberger, L. Capoani, E. Copet, E. Deul, P.A. Duc, N.Epchtein, T. Forveille, P. Fouque, F. Lacombe, T. Le Bertre, S. Pau, J.C. Renault, D. Rouan, D. Tiphène, 1998, ``DENIS: Deep Near Infrared Southern Sky Survey. II: Data reduction'', Astron. Astrophys., submitted


  • – N. Epchtein, A. Omont, W.B. Burton, P. Persi, (eds.), 1994, ``Science with astronomical Near-Infrared Survey'' , Proceedings of a Conference held at Les Houches, France, September 1993, reprinted in Astr. Sp. Sci., vol. 217
  • – P. Persi, W.B. Burton, N.Epchtein, A. Omont (eds.), 1995, Proceedings of the first Euroconference on ``Near-infrared surveys'', San Miniato, Italy, Jan. 1995, reprinted in Mem. S.A. It., 66, 3
  • – F. Garzon, N. Epchtein, A. Omont, W.B. Burton, P. Persi (eds.), 1997, Proceedings of the second Euroconference on ``The impact of large-scale near-infrared sky surveys'', Puerto de la Cruz, Spain, ASSL series vol. 210, Kluwer Acad. Pub., Dordrecht
  • – Epchtein, (ed.), 1998, Proceedings of the 3rd Euroconference on ``The impact of near infrared surveys on galactic and extragalactic astronomy'', Meudon juin 1997, Kluwer ASSL series (in press)

    Over the last 5 years the DENIS PI, the CoI-s and their associates have presented contributions concerning the progress of the survey and first scientific conclusions at more than 15 other workshops, conferences, symposia.


    Theses already defended:

  • – E. Copet, April 1996, Paris 6, (instrumentation and star formation), supervisor: D. Rouan
  • – S. Ruphy, December 1996, Paris 6, (galactic structure), supervisor: N.Epchtein,
  • – X. Delfosse, October 1997, Grenoble, (brown dwarfs), supervisor: T. Forveille.

    Theses research in progress:

  • – L. Cambrésy, Paris 7, (1999), (interstellar extinction and star forming rates in our Galaxy), supervisor: N. Epchtein,
  • – S. Derriere, Strasbourg (2000), (galactic structure), supervisor: D. Egret and A. Robin,
  • – A. Baudrand, Observatoire de Paris (2001), (asteroids), supervisor: M.A. Barucci,
  • – M.R. Cioni, Leiden (2001), (AGB stars in the Magellanic Clouds), supervisor: H.J. Habing,
  • – M. Schultheis, Vienna (1998), (AGB stars in the bulge), supervisor: J. Hron,
  • – R. Vavrek, Paris 7 (2001), (advanced statistical methods applied to DENIS data in dark clouds and SFR); supervisors: L.J. Balazs and N. Epchtein.

    The photograph at the backside has been taken with the Danish 1.54 m telescope at ESO, La Silla, Chile at a wavelength of 0.82µm. It covers an area of 10.6×10.6 arcminute, that is about one millionth of the sky. Each dot is a star. The faint star somewhat NE of the center and marked by a cross is a brown dwarf, an object too small to start nuclear fusion: it can not heat itself. The existence of brown dwarfs has been hypothesized by theoreticians for over 30 years. DENIS found this one brown dwarf among millions of other run-of-the-mill stars; although a few dwarfs had been found before, this was the first isolated dwarf.

    DENIS brown dwarf