Pending problems in QSOs

Quasars (Quasi Stellar Objects, abbreviated as QSOs) are still nowadays, close to half a century after their discovery, objects which are not completel y understood. In this brief review a description of the pending problems, inconsistencies and caveats in the QSO's research is presented. The standard paradigm model based on the existence of very massive black holes that are responsible for the QSO's huge luminosities, resulting from to their cosmological redshifts, leaves many facts without explanation. There are several observations which lack a clear explanation, for instance: the absence of bright QSOs at low redshifts, a mysterious evolution not properly understood; the inconsistencies of the absorption lines, such as the different structure of the clouds along the QSO's line of sight and their tangential directions; the spatial correlation between QSOs and galaxies; and many others.

gas emitting those spectral lines. Fast motions strongly would indicate a large mass. Since they cannot continue to feed at high rates for 10 billion years, after the accretion of the surrounding gas and dust is terminated, they would become ordinary galaxies, in few tens of Myr. Unified models were developed in which QSOs were classified as a particular kind of an active galaxy, like a Seyfert 1 but with higher luminosity due to the higher mass of the black hole and a general consensus emerged. In many cases it is simply based in the viewing angle that distinguishes them from other classes of active galaxies, such as Seyfert 2, blazars, radio galaxies.
There are plenty of books and reviews about these fascinating objects and the standard hypothesis to interpret them (e.g., Rees 1984;Antonucci 1993; Kembhavi & Narlikar 1999). In this review I pretend to do something different: rather than presenting the successes of the standard theory in our understanding of the QSOs, I want to show the dark side, the aspects of which are still not very clear and deserve further consideration, either for improving the present standard theory, or to modify it, and even to change it completely if it were necessary. This article is not a forum for the discussion of all the possible theoretical approaches, but deals with the observational facts which could affect the discussion of what is known and/or what is still unknown.
Quasars (Quasi Stellar Objects, abbreviated as QSOs) are still nowadays, close to half a century after their discovery, objects which are not completely understood. In this brief review a description of the pending problems, inconsistencies and caveats in the QSO's research is presented. The standard paradigm model based on the existence of very massive black holes that are responsible for the QSO's huge luminosities, resulting from their cosmological redshifts, leaves many facts without explanation.
Possibly not all of the work cited in the references of this review is correct. My task here is just to compile the bibliography on the pending problems, not to critically examine them. This review is not complete, there are indeed hundreds or thousands of references relevant to these questions, although I think the references hereby presented are quite representative. Nonetheless, given the sort of the material displayed in our references, one can get a general glimpse of what are the most relevant topics discussed nowadays pertaining to the nature of QSOs.

Very high luminosity at high redshift
As said previously, the most remarkable characteristic of QSOs is perhaps their very high luminosity. The luminosities of the brightest QSOs, using the standard interpretation, are as bright as several thousands of cD galaxies (the brightest galaxy in a cluster of galaxies at low redshift z). Only one QSO may be as bright as hundreds of large clusters of galaxies (with ∼ 1000 large galaxies in each of them) in a relatively compact region. The regions should be very compact in order to justify their strong variability in short times.
To reduce their luminosity, some prefer to think that there is a magnification due to gravitational lenses, but no evidence of that was found. Yamada et al. (2003) or Richards et al. (2004) examined the fields of some QSOs at z ∼ 6 and concluded that they are not gravitationally magnified, so the luminosity at that redshift must be very high. The anisotropy of the QSO radiation, observable only when the beam is pointed towards us, would also reduce the luminosity; but the number of sources would be much higher, a huge number, and we would then see a large number of them in the nearby Universe, even if their beams of maximum flux are not pointing towards us; local AGNs (Active Galactic Nuclei) would experience also this strong anisotropy radiation emission. It is not the case, we do not observe that. Therefore, the luminosity of the QSOs must be really huge, provided that their assumed distance is correct.
Since their discovery, long debates have taken place on whether the distance pointed by the redshift is real or not. After that period of debate in the 60s and the early 70s, the mainstream of astronomers adopted the consensus that the redshift of the QSOs is cosmological in origin and therefore the luminosity is intrinsically very high. Hence, it could be explained in terms of supermassive black holes (e.g., Kembhavi & Narlikar 1999, ch. 5;Djorgovski et al. 2008) while discarding other alternative interpretations.
Nonetheless, some apparent inconsistencies remain still within the standard explanation.
The variability of some QSOs is of the order of some hours. Using the expression for Eddington's luminosity, based on the characteristic scale of the Schwarzshild radius of a black hole, gives luminosities of the order of 10 44 erg/s (Kembhavi & Narlikar 1999, §15.3), which is much lower than what is observed.
The relativistic beaming of jets pointing nearly directly toward us could explain the values in the most extreme cases but this ad hoc solution is not clear yet. Also, the inverse-Compton catastrophe (Hoyle et al. 1966;Kembhavi & Narlikar 1999, §15.3) introduces some constraints on the energy production mechanism, and it is not understood how this problem is solved for the case of the huge luminosities of QSOs. In the case of ultramassive black holes of around 10 10 M ⊙ , necessary to explain the extremely high luminosities of high redshift QSOs, they would attract the surrounding material at relativistic speeds producing large tidal forces such that the infalling matter would be heated up during its compression and become strongly redshifted (Kundt 2009), something which is not observed. Some physical variables should be proportional to the distance of a source, such as the Faraday rotation or the time dilation factor, but they are not observed to be correlated to the redshift. The polarization of radio emission rotates as it passes through magnetized extragalactic plasmas. Such Faraday rotations in QSOs should increase (on average) with distance. If redshift indicates distance, then rotation and redshift should increase together. However, the mean Faraday rotation is less near z = 2 than near z = 1 (Arp 1988). Time dilation, which is observed in supernovae, should also be observed in QSOs, increasing the periods of variability with the distance, but it is not observed in QSOs against expectations (Hawkins 2001).
Moreover, the huge dispersion in the magnitude-redshift relation for QSOs (Hewitt & Burbidge 1987) makes impossible to derive a Hubble law for them. This is not a strong argument since the intrinsic dispersion of luminosities might be high itself, but it might be pointing out that something is wrong with the distance measurement.

Host galaxies
The luminosity of the host galaxies, which are supposed to be normal galaxies and whose luminosity come for these high luminosities is that we could be observing a very young populations of stars. Magain et al. (2005) reported on the observation of a quasar lying at the edge of a gas cloud, whose size is comparable to that of a small galaxy, but whose spectrum shows no evidence for stars. The gas cloud is excited by the quasar itself. Magain et al. (2005) could not see any host galaxy in it; if a host galaxy were present, it should be at least six times fainter than it would normally be expected to be for such a bright quasar. This tells us that the host galaxies, although they are normally brighter than normal galaxies, in some cases are much fainter or inexistent. We do not know the reason.
Other problems to solve in the host galaxies remain: the dynamical mass of molecular gas of a case at z = 6.4 (∼ 5.5 × 10 10 M ⊙ ) is too high to leave room for other kinds of matter, and it cannot accommodate the predicted 10 12 M ⊙ stellar bulge necessary for its massive black hole (Walter et al. 2007). There are also unexpected non-detections of cold neutral gas in the host galaxies of high redshift QSOs (at > 10 23 W/Hz; Curran et al. 2008).

Age and metallicity of high redshift QSOs
Some QSOs are apparently somewhat older than the Universe at their corresponding redshift. For instance, there is a QSO at redshift z = 3.91 that has an age of 2-3 Gyr, which constrains Ω m to be less than 0.21 (Jain & Dev 2006), lower than the accepted values for the standard cosmology nowadays.
Possibly the age measurement is somewhat overestimated and this would explain the inconsistency, but it is important to bear in mind that there are pending cases like this to be solved.
Big Bang requires that stars, QSOs and galaxies in the early universe be "primitive", meaning mostly metal-free, because it requires many generations of supernovae to build up metal content in stars. But the observations show the existence of even higher than solar metallicities in the "earliest" QSOs and galaxies , Becker et al. 2001, Constantin et al. 2002, Simon et al. 2007). The iron to magnesium ratio increases at higher redshifts (Iwamuro et al. 2002). And what is even more amazing: there is no evolution of some line ratios, including iron abundance (Dietrich et al. 2003, Freudling et al. 2003, Maiolino et al. 2003, Barth et al. 2003) between z = 0 and z = 6.5, iron abundance at z ∼ 6 QSOs is similar to its abundance in local QSOs. The amount of dust in high redshift galaxies and QSOs is also much higher than expected (Dunne et al. 2003). In view of these evidences, orthodox cosmologists claim now that the star formation began very early and produced metals up to the solar abundance quickly, in roughly half Gyr. However, it is not enough to come up with such a surprising claim, it needs to be demonstrated, and I do not see any evidence in favor of such a quick evolution in the local galaxies.

Evolution or non-evolution of QSOs
There is another remarkable fact about the luminosity of QSOs. They are extremely bright at high redshift, but QSOs at low redshift have got a much lower luminosity. From the analysis of the bolometric luminosity function of QSOs at different redshift (Hopkins et al. 2007), it is clear that the relative abundance of high luminosity QSOs decreases quickly at low redshift. In visible, below z = 0.3, the rate of luminosity decrease begins to slow down and below z = 0.1 the luminosity begins to increase again (Bell 2007). At SDSS survey, all QSOs at z < 0.4 are fainter than M B = −26 (with K-corrections) while there are plenty of QSOs tens of times brighter than this limit at higher redshifts. Also in other wavelengths this fact is observed clearly: in the X-ray region it is particularly strong the evolution at low redshift (Shen et al. 2006); or in the radio regime (Bridle & Perley 1984;Bell 2006, Figs. 9, 10). A strong density and luminosity evolution is required. It seems that we live in the era in which the bright QSOs have disappeared.
It is usual to claim that evolution is the wild card which solves this kind of problems. Something very different should have happened at high redshift with respect to the low redshift Universe to obtain this different level of luminosity. However, no visible signs of this evolution are observed. There is no indication of any significant evolution in the X-ray properties of quasars between redshifts 0 and 6, apart from the intrinsic luminosity, suggesting that the physical processes of accretion onto massive black holes have not changed over the bulk of cosmic time (Vignali et al. 2005). Also, the spectral features of low and high redshift QSOs are very similar (Segal & Nicoll 1998). There are not either clear signs of the variation of the ratio between bolometric and Eddington luminosities (Gutiérrez & López-Corredoira 2009). Therefore, the situation is that QSOs have a strong evolution in the values of their luminosity but not significant change in other properties, and it is not well understood which is the cause of the luminosity evolution. Possibly the environment might change. Nonetheless, do we know the connection between the triggering of activity and the environment?

Triggering of activity
It is usually suggested that the interactions with the companion galaxies are related to the mechanism of feeding the black hole of the QSO (Stockton 1982;Canalizo & Stockton 2001). Horst & Duschl (2008) presented the results of an extremely simple cosmological model combined with an evolutionary scenario in which both the formation of the black hole as well as the gas accretion onto it are triggered by major mergers of gas-rich galaxies. Despite the very generous number of approximations their model reproduces the quasar density evolution in remarkable agreement with some observations. However, other authors There is evidence for a significant post-starburst population in many luminous AGNs, and that a direct, causal link might exist between star formation and black hole accretion (Ho 2005). The detection of large amounts of warm, extended, molecular gas also points that QSOs have vigorous star formation (Walter et al. 2007). However, it is also common nowadays the proposal that AGN host galaxies are a transition population, being the AGNs the mechanism for star-formation quenching, where the black hole blows out the gas. Therefore, to sum up this section, we have no idea of the mechanism which triggers the activity in galaxies, and the different observations point to different directions within the actual proposed scenarios.

Superluminal motions
Superluminal motions of sources at high distance (D) are observed, i.e. angular speeds ω between two radio emitting blobs which imply linear velocities v = Dω greater than the speed light (Cohen 1986). There are some explanations. The so called relativistic beaming model (Rees 1967) assumes that there is one blob A which is fixed while blob B is traveling almost directly towards the observer with speed

Periodicity of redshifts
Another problem with QSOs which has a long history and without a clear agreement is the periodicity of redshifts. In a homogeneous and isotropic universe we expect the redshift distribution of extragalactic objects to approximate a continuous and aperiodic distribution. However, a periodicity with ∆z = 0.031 The standard consensus is that all these cases are just random projections of background/foreground objects rather than the real associations of objects with different redshifts. This might be true in many cases, but the statistics still shows an excess number compared to the expected values for random projections. Hence, it remains difficult to explain these results in terms of random projections. Typical rebuffs such as "it is an a posteriori statistical calculation" or other considerations such as a bias, incompleteness, gravitational lensing, do not solve the anomalies in general (López-Corredoira & Gutiérrez 2006; López-Corredoira 2009). On the other hand, the main supporters of the hypothesis of non-cosmological redshifts continue to produce tens of analyses of cases in favor of their ideas without too much care, pictures without rigorous statistical calculations in many cases, or with wrong identifications, underestimated probabilities, biases, use of incomplete surveys for statistics, etc., in many other cases. Some cases which were claimed to be anomalous in the past have found an explanation in standard terms (López-Corredoira 2009). There are, however, many papers in which no objections are found in the arguments and they present quite controversial objects, but due to the bad reputation of the topic, the community 8 simply ignores them. This has become a topic in which everybody has an opinion without having read the papers or knowing the details of the problem, because some leading cosmologists have said it is bogus. Therefore, despite the many efforts by most cosmologists to forget this old problem encountered with QSOs, the unexplained data are still there pointing out to us that we do not understand these phenomena completely. I maintain a neutral position, neither in favor of nor against non-cosmological redshifts.

Emission lines
The standard model assumes that QSOs are the same type of objects as Seyfert 1 galaxies but much brighter. Both of them present the characteristic broad emission lines for hydrogen, carbon and other elements, together with a narrow emission of "forbidden" lines in the case of elements like oxygen, nitrogen, sulfur, etc. According to the standard black hole scenario and its accretion disc (e.g., Kembhavi & Narlikar 1999, ch. 5), the broad lines stem from the inner region with a strong velocity dispersion, while the forbidden lines would be generated in the outer regions. These lines depend on the physical conditions of the clouds and the spectral energy distribution that photo-ionizes the clouds. Although this scenario explains the main basic spectral features, it remains to clarify some detailed observations. For instance, some analyses of the spectra are given by Sulentic (2006), who believes that the double-peaked Balmer emission lines are better fitted with the bi-cone outflow model rather than the model of accretion disks. Indeed, the double-peaked optical emission lines are present only in ∼ 5% of the AGNs, which raises another problem in supporting the black holes hypothesis as the engines of the activity.
According to the unification model, the differences between narrow-line AGNs (Seyfert 2) and broadline ones (Seyfert 1) stem from the orientation of the toroidal regions of large extinction in the same type of galaxies. But the relative ratios of narrow and broad line AGNs cannot be reconciled with this simple model of unification in which the tori have the same optical depth and their opening angle is found to be independent of the luminosity (Lawrence 1987); it requires a modification of the tori's extinction and/or cone angles for objects of different luminosity. Among the spiral AGNs with close satellite companions, only a 2.6% (1/39) of them are Seyfert 1 (Dultzin et al. 2008); this cannot be explained either by a simple unification model and requires modifications in terms of extra extinctions in presence of companions. Other observations, pointing out that the differences between Seyfert 2 and Seyfert 1 cannot be due entirely to different orientations of the same object, can also be found in the literature (e.g., Kembhavi & Narlikar 1999, §12.6.5). Hence, it is required, in general, a revision of the simple unification model. This does not mean that the main aspects of the unification model are wrong, but there are many observations which do not fit its predictions unless the model is made more complex with the introduction of more ad hoc terms.

9
According to the standard interpretation, the absorption lines in QSOs are produced by the footprint in the intergalactic medium that the paths of the photons take. The redshift is associated to the path of the photons taken across the cloud. However, some observations are not always consistent with this scenario.
For instance, the HST NICMOS spectrograph has searched for objects associated to the absorption lines of damped Lyα systems (DLAs) of some QSOs directly in the infrared, but failed for the most part to detect them (Colbert & Malkan 2002). Moreover, the relative abundances of DLAs have a surprising uniformity, unexplained in the standard model (Prochaska & Wolfe 2002), except for the clouds which have a velocity difference less than 6000 km/s from a QSOs, where the excess density by a factor 2 (at 3.5σ) (Russell et al. 2006 2007) conducted a similar study using CIV absorbers with GRB systems and their column density distribution and number density of this sample do not show any statistical differences with the same quantities measured in the QSO spectra. Maybe the discrepancy stems from a higher dust extinction in the strong MgII QSO samples studied up to now (Sudilovsky et al. 2007). Frank et al. (2007) propose that the solution is that the QSO beam size is 2 times larger than the GRB beam sizes on average. Porciani et al. (2007) have argued that the combined action of some effects can substantially reduce the statistical significance of the discrepancy. Possibly this discrepancy can be solved in standard terms but it is not certain.

Conclusions
Irrespective of who is right or wrong, either the researchers who are following the standard interpretation of QSOs or the others who are following a different one, the general impression which emerges from all of these problems is that we do not yet understand very well many aspects of these objects and further research is still necessary.
If I had to express my particular opinion, I would say that the three most puzzling points are: 1) the total absence of bright QSOs at low redshift, a mysterious evolution not properly understood; 2) the inconsistencies of the absorption lines, such as the different structure of the clouds, when performing comparisons between measurements along the tangential and the line of sight of QSOs; 3) the spatial correlations among QSOs and nearby galaxies.
Nonetheless, one must not forget that there are also good reasons to support the standard scenario of QSOs, particularly the results about the large distances and luminosities. Just to select three among them: 1) the association of host galaxies with their QSOs shows that the luminosity of the central part of the object is much higher than the rest of the galaxy, and the hosts have angular sizes decreasing with redshift; 2) the absorption lines in many cases have a successful interpretation in terms of gas or galaxies intervening along the line of sight; 3) cases involving gravitational lensing indicate that the redshift of QSOs is much higher than the redshift of the lensing galaxy. And, within the large distance/luminosity assumption, there are also good reasons to support the standard paradigm model based on the existence of very massive black holes with their accretion discs.
It would be desirable that we could proclaim that we understand everything related to these fascinating objects before other 50 years went by. But up to now we should leave at least some room for more discussions and even having an open mind to embrace novel hypotheses in the interpretation of QSOs.