Cuspy no more: how outflows affect the central dark matter and baryon distribution in Λ cold dark matter galaxies

Monthly Notices of the Royal Astronomical Society, May 2012

We examine the evolution of the inner dark matter (DM) and baryonic density profile of a new sample of simulated field galaxies using fully cosmological, Λ cold dark matter (ΛCDM), high-resolution SPH+N-Body simulations. These simulations include explicit H2 and metal cooling, star formation (SF) and supernovae-driven gas outflows. Starting at high redshift, rapid, repeated gas outflows following bursty SF transfer energy to the DM component and significantly flatten the originally ‘cuspy’ central DM mass profile of galaxies with present-day stellar masses in the 104.5–109.8 M⊙ range. At z= 0, the central slope of the DM density profile of our galaxies (measured between 0.3 and 0.7 kpc from their centre) is well fitted by ρDM ∝ rα with α≃−0.5 + 0.35 log10(M★/108 M⊙), where M★ is the stellar mass of the galaxy and 4 < log M★ < 9.4. These values imply DM profiles flatter than those obtained in DM-only simulations and in close agreement with those inferred in galaxies from the THINGS and LITTLE THINGS surveys. Only in very small haloes, where by z= 0 SF has converted less than ∼0.03 per cent of the original baryon abundance into stars, outflows do not flatten the original cuspy DM profile out to radii resolved by our simulations. The mass (DM and baryonic) measured within the inner 500 pc of each simulated galaxy remains nearly constant over 4 orders of magnitudes in stellar mass for M★ < 109 M⊙. This finding is consistent with estimates for faint Local Group dwarfs and field galaxies. These results address one of the outstanding problems faced by the CDM model, namely the strong discrepancy between the original predictions of cuspy DM profiles and the shallower central DM distribution observed in galaxies.

A PDF file should load here. If you do not see its contents the file may be temporarily unavailable at the journal website or you do not have a PDF plug-in installed and enabled in your browser.

Alternatively, you can download the file locally and open with any standalone PDF reader:

Cuspy no more: how outflows affect the central dark matter and baryon distribution in Λ cold dark matter galaxies

Cuspy no more: how outflows affect the central dark matter and baryon distribution in cold dark matter galaxies F. Governato 4 A. Zolotov 3 A. Pontzen 0 1 C. Christensen 2 8 S. H. Oh 6 7 A. M. Brooks 5 11 T. Quinn 4 S. Shen 10 J. Wadsley 9 0 James Martin Research Fellow 1 Oxford Astrophysics, University of Oxford , Denys Wilkinson Building, Keble Road, Oxford OX1 3RH 2 Theory Fellow 3 Racah Institute of Physics, Hebrew University , Jerusalem 91904 , Israel 4 Astronomy Department, University of Washington , Box 351580, Seattle, WA 98195-1580 , USA 5 Department of Astronomy, University of Madison-Wisconsin , 2535 Sterling Hall, 475 N. Charter Street, Madison, WI 53076 , USA 6 School of Physics, University of Sydney , NSW 2006 , Australia 7 International Centre for Radio Astronomy Research (ICRAR), University of Western Australia , 35 Stirling Highway, Perth, WA 6009 , Australia 8 Department of Astronomy/Steward Observatory , 933 North Cherry Avenue, Tucson, AZ 85721-0065 , USA 9 Department of Physics and Astronomy, McMaster University , Hamilton, Ontario L88 4M1 , Canada 10 Department of Astronomy and Astrophysics, University of California , Santa Cruz, CA 95064 , USA 11 Grainger Postdoctoral Fellow A B S T R A C T We examine the evolution of the inner dark matter (DM) and baryonic density profile of a new sample of simulated field galaxies using fully cosmological, cold dark matter ( CDM), high-resolution SPH+N-Body simulations. These simulations include explicit H2 and metal cooling, star formation (SF) and supernovae-driven gas outflows. Starting at high redshift, rapid, repeated gas outflows following bursty SF transfer energy to the DM component and significantly flatten the originally 'cuspy' central DM mass profile of galaxies with presentday stellar masses in the 104.5-109.8 M range. At z = 0, the central slope of the DM density profile of our galaxies (measured between 0.3 and 0.7 kpc from their centre) is well fitted by ρDM ∝ rα with α −0.5 + 0.35 log10(M /108 M ), where M is the stellar mass of the galaxy and 4 < log M < 9.4. These values imply DM profiles flatter than those obtained in DM-only simulations and in close agreement with those inferred in galaxies from the THINGS and LITTLE THINGS surveys. Only in very small haloes, where by z = 0 SF has converted less than ∼0.03 per cent of the original baryon abundance into stars, outflows do not flatten the original cuspy DM profile out to radii resolved by our simulations. The mass (DM and baryonic) measured within the inner 500 pc of each simulated galaxy remains nearly constant over 4 orders of magnitudes in stellar mass for M < 109 M . This finding is consistent with estimates for faint Local Group dwarfs and field galaxies. These results address one of the outstanding problems faced by the CDM model, namely the strong discrepancy between the original predictions of cuspy DM profiles and the shallower central DM distribution observed in galaxies. hydrodynamics - galaxies; evolution - galaxies; formation - galaxies; star for- - 1 I N T R O D U C T I O N The predictions of the cold dark matter ( CDM) cosmological model are in excellent agreement with observations of the assembly of cosmic structures on large scales (Eke, Cole & Frenk 1996; Riess et al. 1998; Spergel et al. 2007) . The observed properties of dwarf galaxies, however, have presented strong challenges to the model at galactic scales. Dark matter (DM) only simulations predict that DM haloes should follow a quasi-universal Einasto profile (Navarro, Frenk & White 1996b; Moore et al. 1999; Reed et al. 2005; Macci o`, Kang & Moore 2009; Stadel et al. 2009; Navarro et al. 2010) , defined by a power-law density profile ρ ∝ rα with −1.5 < α < −1 in their central regions. Such simulations therefore predict steep (or ‘cuspy’) inner density profiles. Observations of small galaxies (Vpeak ∼ 30–60 km s−1), however, have repeatedly shown that their central DM profiles are shallower than the predictions of DM-only simulations at scales of 0.1–1 kpc (e.g. Flores & Primack 1994; Moore et al. 1999; Swaters et al. 2003; Simon et al. 2005; de Blok et al. 2008; Donato et al. 2009; Primack 2009; Elson, de Blok & Kraan-Korteweg 2010; Kuzio de Naray et al. 2010; Oh et al. 2011b) . In these observational works, the measured α slopes range between 0 and −1. [Note: in this paper we will identify haloes with DM profiles shallower than Navarro–Frenk–White (NFW) or Einasto as ‘cored’.] These observational results provided a longstanding challenge to CDM at scales that cannot probed by cosmic microwave background or Lyα experiments (Croft et al. 2002) . They also prompted the development of several alternative DM models such as warm DM (WDM), meta-DM and self-interacting DM (Spergel & Steinhardt 2000; Dave´ et al. 2001; Knebe et al. 2002; Ahn & Shapiro 2005; Strigari, Kaplinghat & Bullock 2007; Col´ın, Valenzuela & Avila-Reese 2008; Martinez et al. 2009; Abdo et al. 2010; Koda & Shapiro 2011; Loeb & Weiner 2011; Lovell et al. 2012; Vogelsberger, Zavala & Loeb 2012) and alternative gravity theories such as MOND (McGaugh 2005; Gentile, Famaey & de Blok 2011) . There is emerging evidence that our poor understanding of the baryonic processes involved in galaxy formation is the source of the inconsistency between the observations of dwarf galaxies and the predictions of CDM. Models of the effect of feedback on the structure of galaxies and the efficiency of star formation (SF) were originally motivated by the discrepancy between the observed number of dwarf galaxies and the much higher number density of small DM haloes predicted by CDM models (Dekel & Silk 1986; Moore et al. 1998; Bower et al. 2006) . Without requiring a major change to the CDM scenario, several models have been presented in the past advocating the evolution of originally cuspy CDM haloes into haloes with ‘cored’ density profiles through supernova (SN) feedback or dynamical friction (Navarro, Eke & Frenk 1996a; Dekel, Devor & Hetzroni 2003b; Dekel et al. 2003a; El-Zant et al. 2004; Mo & Mao 2004; Read & Gilmore 2005; Goerdt et al. 2006, 2010; Mashchenko, Couchman & Wadsley 2006; Tonini, Lapi & Salucci 2006; Del Popolo 2009; Romano-D´ıaz et al. 2009; Lackner & Ostriker 2010; de Souza et al. 2011; Pontzen & Governato 2011) . In Governato et al. (2010, hereafter G10), we presented self-consistent, DM + gas dynamic simulations where shallow DM cores arise naturally in a CDM cosmology (see also Maccio` et al. 2012) . In these simulations, energy feedback from SNe in star-forming regions generates repeated, fast gas outflows. These outflows efficiently remove gas from the inner kpc of protogalaxies (Brook et al. 2011) and, in smaller galaxies, are able to expel a large fraction of gas from the galaxy altogether. In Pontzen & Governato (2011) , we presented a coherent analytical model for core formation that correctly matches results from our simulations. In this model, multiple rapid gas outflows transfer energy to the collisionless DM and create DM cores of about 1 kpc in size in haloes of total mass 2–3 × 1010 M . As also described in detail in G10, a crucial ingredient is the spatial distribution of the SF events. If SF is allowed in low-density gas ρ ∼ 0.1–1 amu cm−3 (typical in most previous, lower resolution simulations), outflows are weaker and do not generate cores, even if the feedback scheme remains the same. The necessity of simulations to resolve dense gas regions calls for simulations of very high mass and spatial resolution (Mashchenko, Wadsley & Couchman 2008; Saitoh et al. 2008; Ceverino & Klypin 2009) . As a key step towards understanding the properties of DM haloes at very small masses, many faint Milky Way (MW) dwarf satellites have recently been discovered, some with less than one millionth of the MW’s luminosity (e.g. Willman et al. 2005; Belokurov et al. 2007) . The first kinematic studies of ‘ultra-faint’ dwarf galaxies have measured mass-to-light ratios (M/L) as low as 1000 (Simon & Geha 2007; Geha et al. 2009) . These results suggest that the smallest cosmic structures where SF took place might have been identified. It has been argued that the observed population of faint galaxy satellites should be hosted inside cuspy haloes, as small cored haloes would easily be destroyed by the MW tidal field (Pen˜arrubia et al. 2010) . However, estimates of the mass profiles of ultra-faint dwarfs are uncertain and critically depend on the detailed kinematic of their stars, which are used as dynamical tracers (Walker & Pe n˜arrubia 2011) . Unlike most field dwarfs, where the evidence for DM cores is robust, current observations have not securely determined if lowmass satellites of the MW are hosted inside cored or cuspy DM haloes, although Walker & Pen˜arrubia (2011) and Jardel & Gebhardt (2011) reported that the Sculptor and Fornax dwarf galaxies might indeed have cores. Surprisingly, while the MW low-luminosity dwarfs span 5 orders of magnitude in luminosity, dynamical studies estimate that these galaxies contain the same total (baryons+DM) mass in their innermost part, about 107 M within the inner 300 pc (Strigari et al. 2008) . This result provides a useful constraint for all models of galaxy formation and can be interpreted assuming a truncation in the halo mass function as predicted in WDM models (AvilaReese et al. 2001; Bode, Ostriker & Turok 2001 ; Boylan-Kolchin, Bullock & Kaplinghat 2011b; Schneider et al. 2011; Parry et al. 2012) or to the suppression of SF in haloes with virial temperature less than the cosmic UV background (Quinn, Katz & Efstathiou 1996; Benson et al. 2002; Stringer, Cole & Frenk 2010) . Both processes would set a minimum halo scale for SF, hence a common mass at a fixed radius. Feedback and gas heating from cosmic sources have been shown (Governato et al. 2007; Font et al. 2011, among many) to alleviate the above mentioned overabundance of satellites in CDM, while WDM faces already significant observational constraints (Viel et al. 2005; Kuzio de Naray et al. 2010; Lovell et al. 2012) and several works presented evidence that WDM haloes would form cores much smaller than observed, (Dalcanton & Hogan 2001; Knebe et al. 2002; Strigari et al. 2007) , making this model less attractive. A few models have tried to explain the observed flat stellar mass– central mass relation measured in Strigari et al. (2008) , while assuming cuspy DM halo profiles (Rashkov et al. 2012) . Some (Li et al. 2009; Macci o` et al. 2009) invoke a large scatter in star formation histories (SFHs) and halo assembly histories of dwarf galaxies to explain the observed flat luminosity–central mass relation over a large range of halo masses. These results, however, are based on DM-only simulations paired to semi-analytical models. They therefore neglect important interactions between the baryonic and the DM component of haloes, as well as the effect of H2 cooling on SF in dense regions (Gnedin, Tassis & Kravtsov 2009). Ongoing surveys such as ANGST (Dalcanton et al. 2009) , THINGS (Walter et al. 2008) and LITTLE THINGS (Zhang et al. 2011) are providing data on the central mass distribution of field galaxies with Vpeak < 60 km s−1 and provide new, strong constraints on the central mass distribution of galaxies (Oh et al. 2011a, hereafter OH11) that need to be properly taken into account. In particular, LITTLE THINGS and THINGS are able to measure DM profiles over a range of galaxy stellar masses and to evaluate the total mass within the central region of galaxies more massive than those in the original ‘Strigari relation’. To summarize, observational measurements of the central mass distribution of galaxies and of the central slope of the underlying CDM DM profile as a function of a galaxy stellar mass, may shed light on the nature of DM and of baryon–DM interactions at scales much smaller than those probed by cosmological test. In this work we focus on field galaxies and their structural properties, and compare the highest resolution set of simulations of small galaxies to date with measurements of DM slopes obtained from new data from the THINGS and LITTLE THINGS surveys. Focusing on field galaxies allows us to separate the effects of gas outflows from those of tidal interactions (Mayer et al. 2001) . The goal of the analysis presented here is to evaluate if galaxies formed in a CDM cosmology have central DM and baryon distributions consistent with observations once the effect of realistic gas outflows is evaluated in a full cosmological context. To achieve this goal, we study a new set of high-resolution galaxies formed in a full CDM cosmological context. The simulations presented here include a consistent implementation of metal line cooling and H2 physics (Christensen et al. 2011, hereafter CH11) , essential to correctly model SF in lowmetallicity gas and in low-mass haloes (Li et al. 2009; Krumholz & Dekel 2011) . The highest resolution simulations in our sample resolve individual SF regions as small as a few 104 M . The simulations have not been tuned to produce cores, but rather to form a realistic amount of stars and to reproduce the stellar mass–halo mass relation (Conroy, Wechsler & Kravtsov 2006; Tollerud et al. 2011; Munshi et al., in preparation) . As a test of our predictions for the central mass distribution of galaxies, we also show that simulations naturally reproduce an updated flat stellar mass–central mass relation obtained combining MW and field dwarf galaxies. We describe the simulations and the THINGS data in Section 2, the results on the shape of the DM profiles in Section 3 and the results on the central mass–stellar mass relation in Section 4. In Section 5 we conclude. 2 S I M U L AT I O N S A N D O B S E RVAT I O N A L D ATA The simulations used in this work were run with the N-Body + smoothed particle hydrodynamics (SPH) code GASOLINE (Wadsley, Stadel & Quinn 2004; Stinson et al. 2006) in a fully cosmological CDM context: 0 = 0.26, = 0.74, h = 0.73, σ 8 = 0.77, n = 0.96. The galaxies were originally selected from two uniform DM-only simulations of 25 and 50 Mpc per side. From these volumes, five field-like regions where selected, each centred on a galaxy-sized halo of different mass (3 × 1011, 2 × 1011, 3 × 1010, 2 × 1010 and 1010 M ). Each region was then re-simulated at higher resolution and with baryons using the ‘zoomed-in’ volume renormalization technique (Katz & White 1993; Brooks et al. 2007; Pontzen et al. 2008) , while fully preserving the surrounding large-scale structure. This technique allows for significantly higher resolution while capturing the effect of large-scale torques that deliver angular momentum to the galaxy (Barnes & Efstathiou 1987) . The mass overdensity δρ/ρ for each chosen field ranges from −0.07 to 0.58 when measured on a scale of 4 h−1 Mpc (see Table 1). To simulate even very small haloes with millions of resolution elements, the mass and spatial resolution of each zoomed region are inversely proportional to the mass of the largest halo in each one of them (see Table 1). The force spline softening ranges between 64 and 170 pc in all runs and it is kept fixed at z < 10. Star particles are formed with a mass of 400–8000 M . The gas smoothing length is allowed to shrink to 0.1 times the gravitational softening in very dense regions (resulting values around 0.5 are typical) to ensure that hydro forces dominate at very small scales. The main galaxy in every zoomed region contains several million particles within its virial radius (Rvir, defined as the radius at which the average halo density = 100 × ρcrit). Every zoomed region also contains several smaller galaxies with identical mass and spatial resolution, which we include in our analysis. With this approach, the total highresolution sample contains 15 field galaxies with between several million and 50 000 DM particles within Rvir covering a halo mass range of 3 orders of magnitude. Galaxies and their parent haloes were identified using AHF1 (Knollmann & Knebe 2009) . Similar to criteria used in previous works, one galaxy undergoing strong interactions at z = 0 was excluded from the sample. In this work we follow up on G10 and OH11, where we studied the stellar and DM content of haloes M ∼ 1010 M and extend our predictions on a larger range in halo masses (from a few times 108 to 3 × 1011 M ), peak velocities Vpeak (10–100 km s−1) and stellar masses M (from 104 to almost 1010 M ), spanning a range of halo spin values and accretion histories. In observational terms, this range covers faint dwarfs to normal field disc galaxies (Geha et al. 2006) . No haloes in our sample have been contaminated by particles from the lower resolution volumes and the zoomed-in regions are large enough (a few Mpc in each case) to contain all the gas expelled from galaxies by outflows (that can reach as far as several times the virial radius of the parent galaxy). This approach allows us to test for the effects of resolution over a large range in halo masses (i.e. Figs 1 and 2 contain haloes of overlapping mass but ran at different resolutions). Given the high resolution of our simulations, star-forming regions as small as 105 M are identified by hundreds of gas and star particles. As a significant improvement over most cosmological simulations carried out to the present time, these new simulations include both metal lines and H2 cooling (Shen, Wadsley & Stinson 2010; CH11) . The simulations include a dust-dependent description of H2 creation and destruction by Lyman–Werner radiation (Gnedin et al. 2009; CH11) . Metal lines cooling, the H2 fraction and self-shielding of high-density gas from local radiation play an important role 1 AMIGA Halo Finder, available for download at http://popia. in determining the structure of the interstellar medium and where SF can occur (Kennicutt 1998; Krumholz & McKee 2005; Bigiel et al. 2008; Elmegreen, Bournaud & Elmegreen 2008; Gnedin et al. 2009;Feldmann, Gnedin & Kravtsov 2011) .With this approach, the local SF efficiency is linked directly to the local H2 abundance, as regulated by the gas metallicity and local radiation from young stars. As a result, in our simulations stars naturally form in highdensity regions around 10–100 amu cm−3 without having to resort to simplified approaches based on a fixed local gas density threshold (Saitoh et al. 2008; G10; Guedes et al. 2011; Kuhlen et al. 2011) . A Kroupa (1993) initial mass function and relative yields are assumed. We include a gas heating spatially uniform, time evolving UV cosmic background following an updated model of Haardt & Madau (1996). Gas heating from UV radiation progressively suppresses SF in galaxies below 1010 M , making small DM haloes completely void of stars (Benson et al. 2002) and reducing the overabundance of dwarf satellite galaxies (Moore et al. 1998) . The smallest galaxies in our sample have some SF occurring before reionization (z ∼ 9 in our model) likely associated to H2 cooling and then in small sparse bursts thereafter. The full details of our physically motivated SN feedback implementation and its applications have been described in several papers and shown to reproduce many galaxy properties over a range of redshifts (Stinson et al. 2006; Brooks et al. 2007; Governato et al. 2007, 2009; Pontzen et al. 2008; Zolotov et al. 2009; Pontzen et al. 2010; Brook et al. 2011; Brooks et al. 2011; Guedes et al. 2011) . As in G10, the star formation rate (SFR) in our simulations is set by the local gas density (ρgas)1.5 and a SF efficiency parameter, c∗ = 0.1, to give the correct normalization of the Kennicutt–Schmidt relation (the SF efficiency for each star-forming region is much lower than the implied 10 per cent, as only a few star particles are formed before gas is disrupted by SN winds). The maximum temperature for gas to turn into stars is set to 3000 K and the efficiency of SF is then further multiplied by the H2 fraction, which effectively drops to zero in warm gas with T >; 10 000 K. As massive stars evolve into SN, mass, thermal energy and metals are deposited into nearby gas particles. Gas cooling is turned off until the end of the snowplough phase as described by the Sedov–Taylor solution, typically a few million years. The amount of energy deposited amongst those neighbours is 1051 erg per SN event. Energy deposition from SN feedback leads to enhanced gas outflows that remove low angular momentum gas from the central regions of galaxies (Brook et al. 2011) . We have verified that in this set of simulations the ‘loading factor’ of the winds, i.e. the amount of baryons removed, is typically a few times the current SFR, similar to what is observed in real galaxies over a range of redshifts (Martin 1999; Shapley et al. 2003; Kirby, Martin & Finlator 2011; van der Wel 2011) . As SF is limited by the local H2 abundance, stars form only in high-density peaks sufficiently shielded from radiation from hot stars and the SFHs of the galaxies in our simulated sample are bursty over a significant fraction of the Hubble time, but especially at high redshift, where each galaxy is still divided into individual progenitors. Bursty phases typically last 10–100 Myr with SFR variations on shorter time-scales and SF enhanced by a factor of ∼4–20, similar to what measured in Local Group dwarfs (McQuinn et al. 2010) . As discussed in Pontzen & Governato (2011) , a bursty SF is necessary to create the fast outflows able to transfer energy to the DM component. Outflows also decrease the SF efficiency in haloes with total mass smaller than a few 1010 M . In our set of simulations, outflows are predominant at high-z2 when SF peaks and galaxy interactions are common. These outflows affect the haloes that will subsequently merge to form the central regions of the final, present-day galaxies. 2 In the mass range explored by our simulation (up to haloes with Mvir = 3 × 1011), the ratio of stellar mass to halo mass (the SF efficiency) is a strong function of halo mass, roughly scaling as Mstars ∝ Mv2ir. In this halo range, SF becomes substantially less efficient in smaller galaxies. In haloes with total mass smaller than 109 M (also equivalent to a virial temperature Tvir < 104 K), SN feedback and the cosmic UV background strongly suppress SF. Furthermore, in haloes this small stars only form when H2 cooling is introduced, as gas can cool below Tvir. As a result, in the smallest haloes only a very small fraction of baryons is then turned into stars. The more massive galaxies in our sample turn only ∼10 per cent of their primordial baryon content into stars, after having expelled about 30 per cent of their gas outside Rvir. Typical dwarfs in our sample turn a few per cent of their primordial gas fraction into stars, and the smallest galaxies ∼0.01 per cent. In OH11 (see their fig. 5), we verified that galaxies with Vpeak < 60 km s−1 form the correct amount of stars when compared with a local sample with resolved photometric and kinematic data. In a future paper, we will show how our sample closely matches the stellar mass–halo mass relation inferred using halo occupation methods (Moster et al. 2010; Munshi et al. in preparation) . As a reference and a resolution test of the simulations, most of the above runs have been repeated including only the collisionless CDM component. The DM and baryonic mass distribution of the simulated galaxies will be compared with those measured from extensive H I data from a sample of nearby dwarf galaxies from THINGS (Walter et al. 2008) and LITTLE THINGS (Hunter et al., in preparation) surveys which focused on field galaxies. The high-resolution H I data (∼6 arcsec angular; ≤5.2 km s−1 velocity resolution) combined with Spitzer IRAC 3.6 μm and ancillary optical images significantly reduce various observational systematic effects inherent in lower resolution data, such as beam smearing, dynamical centre offset and non-circular motions, and thus enable us to derive more reliable mass models of the galaxies. For a comparison with our simulations, we select a sample of 22 dwarf galaxies (seven from THINGS and 15 from LITTLE THINGS) that show a clear rotation pattern in their velocity fields. These high-quality multiwavelength data allow us to measure the enclosed amount of mass and the inner slope of the DM density profile at 500 pc of the galaxies with good accuracy. 3 T H E E VO L U T I O N O F D M C O R E S A S A F U N C T I O N O F H A L O M A S S A N D R E D S H I F T With the goal of measuring when and how much gas outflows affect the underlying DM profiles in CDM galaxies, in this section we focus on how the central DM density profiles differ from the simple predictions of DM-only runs once cooling, SF processes and gas outflows are introduced. To do this, we measure the slope α of the DM density profile at 500 pc for all the well-resolved galaxies formed in our hydrodynamical simulations, and then compare them with observational data as well as predictions from DM-only simulations. The α value of the central DM density profile is obtained by spherically averaging the density and fitting the density profile with ρDM ∝ rα between 300 and 700 pc. α is then formally defined at 500pc. In this section we exclusively study field galaxies to avoid the effects that satellite – main halo interactions might have on the density profiles (Mayer et al. 2001; Stoehr et al. 2002; Kazantzidis et al. 2004) . Figs 1 and 2 show the value of α as a function of galaxy stellar mass and virial mass. The zoomed-in runs approach allows us cover a large range of galaxy stellar masses, almost 6 orders of magnitude. Both figures clearly show a trend with increasing galaxy stellar (or total) mass showing a central DM profile significantly flatter than the one predicted by CDM simulations that only included a DM component (solid line, showing results from Maccio` et al. 2007) . In Fig. 1 the DM-only predictions are mapped on to the x-axis by assuming the same stellar mass–halo mass relation as in our runs. The DM profiles become progressively flatter up to the most massive systems probed by our simulations, having peak velocities of about 100 km s−1. This result supports the model where SN originated outflows are able to transfer significant amounts of energy to the DM at the centre of each galaxy, lowering the DM central density (Pontzen & Governato 2011) . At z = 0, our simulations predict that the central slope of the DM density profile (again, measured between 0.3 and 0.7 kpc) as a function of stellar mass is well fitted over a mass range 4 < log M < 9.4 by ρDM ∝ rα with α −0.5 + 0.35 log10 M /108 M . (1) The α trend in the DM-only haloes in Figs 1 and 2, with less negative (flatter) values for increasing stellar and halo masses, is partly driven by the constant radius used to compare with observations over a large range in galaxy masses, as in DM Einasto-like density profiles α gets steeper in their outer parts (i.e. for larger R/Rvir values), with α rolling from ∼−1 close to the centre to −3 at Rvir. For larger haloes, 500 pc represents a smaller fraction of the virial radius, hence the DM profile is flatter (less negative α values) than for smaller haloes, where at 500 pc we are instead starting to measure the outer part of the DM profile. This effect is present also in the runs with SF-induced outflows, but with a significant shift to flatter profiles and less negative α values over a large range in stellar masses. As a result, central DM profiles in runs with outflows become rapidly flatter at halo masses larger than 109 M (circles in Fig. 2, see also Fig. 3). The DM-only runs in our sample follow closely the results from a set of high-resolution haloes presented by Maccio` et al. (2007) . Scatter at a similar mass is small and there is no trend with resolution (small versus large crosses). However, in the lowest mass dwarfs in our sample, when SF converts less than 1 M of baryons per 5 × 104 M of DM (corresponding to Mvir < 5 × 109, or also Vpeak < 20 km s−1), outflows are too weak to remove sufficient amount gas and cause DM heating through energy transfer. These haloes maintain their original cuspy DM profile. Note that haloes formed in simulations that include outflows accrete slightly less mass over a Hubble time compared to DM-only runs (Fig. 2). In Pontzen & Governato (2011) , we presented an analytical model of the cusp flattening process. In the galaxies in our sample, this process commonly starts at high redshift (z ∼ 4) and continues down to z ∼ 1, when the SFR declines. Fig. 3 shows the DM (spherically averaged) density profile as a function of radius and redshift for a dwarf with final total mass ∼1010 M , showing a rapid flattening at high-z. Around z = 3, as the SFR peaks, the galaxy is undergoing several starbursts associated with a rapid accretion and mergers. The DM density profile flattens and reaches a relatively stable profile by z = 1, when the SFR declines. The last small starbursts have a relatively small effect. By z = 0 the DM profile shows a flattened profile within the central 1 kpc and its central density is almost an order of magnitude lower than in the DM-only case. An interesting result of this work is that massive blowouts that could potentially disrupt the formation of gaseous and stellar discs observed in field dwarfs are not required for the formation of cores. Each outflow typically removes gas only from a small area (typically <1 kpc) around the galaxy centre and discs remain fairly thin (Sa´nchez-Janssen, Me´ndez-Abreu & Aguerri 2010) . This model is similar to that proposed in Read & Gilmore (2005) , where the need for repeated outflows was stressed, and Mashchenko et al. (2008) , where ongoing SF was associated with cusp-flattening. In fact, some studies were not able to create cores using single, massive blowouts (Gnedin & Zhao 2002; Boylan-Kolchin, Bullock & Kaplinghat 2011a) . Our results suggest that those models fail as the amount of baryons removed from the central kpc is small over a single outflow event, irrespective of the fact that the whole galaxy gas reservoir can be blown away. Consequently, the energy transfer to the central DM over a single outflow event is small compared to the binding energy of the DM cusp and multiple events become necessary. In the scenario proposed in this work, only very small galaxies (Vpeak < 20), where SF is extremely inefficient, retain the cuspy CDM profile that was originally predicted by DM-only simulations. In very small objects, SFRs are too low to generate sufficient energy to modify the CDM cusp. This sets an interesting scale (Willman et al. 2005; Tollerud et al. 2008) where primordial cuspy profiles would still be observable. These observations would be able to discriminate between CDM and alternative DM models, where cores are generated by primordial effects and would then be present even in the lowest stellar mass objects. The comparison with the THINGS and LITTLE THINGS data set shown in Fig. 1 [more details in Oh et al. (in preparation)] shows an extremely good agreement between the simulations and the real galaxies. The observational data set includes 22 galaxies. The simulations reproduce both the absolute values of α and the trend with stellar mass. Using the following formula (de Blok et al. 2001) , we first converted the DM rotation curves derived subtracting the baryons from the total kinematics of the 22 dwarf galaxies to the DM density (2) profiles (see OH11; Oh et al. 2011b, for more details) : 1 ρ(R) = 4πG 2 V ∂V R ∂R + V R 2 , where V is the rotation velocity observed at radius R, and G is the gravitational constant. We then measured the logarithmic inner slopes α of the derived DM density profiles assuming a power law (ρ ∼ rα). After determining a break radius (<1 kpc) where the slope changes most rapidly, we measured the inner slope by performing a least-squares fit to the data points of a given profile within the break radius. A similar analysis had been performed by OH11 on two simulated dwarfs that were compared to galaxies from the THINGS survey. Their analysis compared estimates of α obtained from the observed mass distributions and from artificial observations (H I data cubes paired with artificial photometric images) of the simulations. The two methods showed good agreement. Relevant to the results presented here, OH11 also showed that observations can correctly recover the DM profile of galaxies with no significant biases. Those tests support the analysis and the results presented here. In a future work (Oh et al., in preparation), the observational data set will be presented in more detail and it will be compared with results obtained by artificial observations of the new simulations. We have verified that our results are robust versus unwanted numerical effects. In particular, we find that the graininess of the DM potential (due to a finite number of particles) does not substantially affect its response to gas outflows. To this aim we increased the number of DM particles by a factor of 8 for all the haloes in field 3, finding that the measured DM slopes remain substantially unchanged. In G10 we verified that gas resolution effects dominate over pure DM resolution effects as poorer resolution creates more cuspy cores, likely as outflows become poorly resolved and artificial viscosity brings more gas to a galaxy centre. As most haloes were re-run including only the DM component, we also verified that the central DM profiles in the absence of baryons were as cuspy as those in the literature (e.g. Reed et al. 2005; Maccio` et al. 2007) . Fig. 2 shows good agreement between the slopes measured from our simulations and the predictions inferred from Macci o` et al. (2007) over the whole sample range in mass and resolution. This test, combined with the existing data points in the halo range 1010–1011 that span a range in mass and force resolution (Table 1), show that core formation and sizes are stable quantities over the resolution range explored in this work. The analytical model of core formation presented in Pontzen & Governato (2011) shows that the creation of DM cores should be a generic property of fast, repeated gas (out)flows. Core creation should then be a common outcome of any feedback scheme that can create such flows (Springel, Di Matteo & Hernquist 2005; Keresˇ et al. 2009; Oppenheimer et al. 2010; Choi & Nagamine 2011; Hopkins, Quataert & Murray 2011) as long as sufficient spatial resolution and high SF surface densities are achieved. However, as some traditional SPH implementations provide a poor description of Rayleigh–Taylor and Kelvin–Helmholtz gas instabilities (Agertz et al. 2007) , it will be important to evaluate how potential improvements recently introduced by different gas-dynamical treatments (Bauer & Springel 2011; Keres et al. 2011; Read & Hayfield 2011) affect the outcome of baryon energy injection from different astrophysical processes. Encouragingly, results showing the formation of DM cores due to repeated energy injection (from supermassive black holes) into the gas component were also recently obtained using an adaptive mesh refinements (AMR) code (Martizzi et al. 2011) . CDM The extremely good agreement between observations and simulations strongly supports that baryon–DM interactions, and specifically repeated baryonic outflows, are able to lower the DM density at the centre of galaxies, creating DM ‘cores’. This result resolves one of the outstanding problems faced by the CDM model of galaxy formation, namely the strong discrepancy between the original model predictions and the observed DM distribution in galaxies. In this section we compare the total mass contained in the central regions of our simulated galaxy sample with the existing constraints coming from 12 galaxies in the LITTLE THINGS and THINGS samples and from a sample of local dwarf spheroidals (dSph; Walker et al. 2009) . This comparison is necessary to test if the distribution of stars, gas and DM of the simulated galaxies is realistic over a range of galaxy masses, hence making our predictions on the shape of the DM density profiles of galaxy haloes robust. With this goal we will focus on a recent observed relation that has placed strong constraints on the CDM model, namely the central mass–luminosity relation (Strigari et al. 2008, hereafter S08) . As our simulations have not been tuned to be a good match to these observational data, they will provide a good test of the efficiency of SF as a function of stellar mass and of the relative effect of outflows as a function of halo mass. Fig. 4 shows the total mass within 500 pc for the simulated field galaxies in our sample, as a function of their total stellar mass. The figure also shows similar estimates from observed galaxies. For the observational sample, stellar masses are obtained assuming a M/L ∼ 2 for the Walker et al. sample (Walker, private communication) and from the optical and Spitzer IRAC 3.6-μm photometry for the THINGS and LITTLE THINGS samples (Oh et al., in preparation). In particular, for the THINGS and LITTLE THINGS samples, we obtain the 3.6 μm M/L values from optical colours based on Bruzual & Charlot (2003) stellar population synthesis models (Bell & de Jong 2001; Oh et al. 2008) . In addition, assuming a spherical potential, we calculate the total masses (DM + baryons) within 500 pc of the THINGS and LITTLE THINGS samples from their total rotation curves. Measurements from simulations are in extremely good agreement with the observed estimates, which with the addition of the THINGS and LITTLE THINGS data points extend the results in S08 by almost 2 orders of magnitude to larger galaxy masses. It is also remarkable that the two observational samples agree with each other very well, being one comprised of MW satellites and the other of field, rotationally supported galaxies. Our simulations predict that the flat central mass–stellar mass relation will start an upward trend for galaxies with stellar masses >109 M . At that mass scale, our simulated galaxies start having a small bulge component and the total mass within 500 pc starts growing. The flat central mass–stellar mass relation shown by our simulated CDM galaxies differs from the naive expectations from DM-only runs, which show a clear correlation between halo peak velocity (or halo mass) and the mass within a fixed radius; in those models, more massive haloes contain more mass (Li et al. 2009) within a fixed radius. Our analysis, focusing on simulated field haloes, also removes the complications introduced by the dynamical interactions with the host galaxy potential as for the satellites of the MW (Stoehr et al. 2002) . Fig. 5 shows the circular velocity V c (defined as √(M /r ) ) for the galaxies in our sample, where M is the total mass within r. This plot supports the finding that the total amount of mass within the central kpc is a weak function of the galaxy peak velocity. These findings suggest that in our simulations a flat central mass–stellar mass relation originates from having a large range in luminosities over a relatively small range in halo masses and is possibly further helped by the flattening of the central DM mass profile in the more massive galaxies. The continuous and dashed lines in Fig. 4 show the resulting central mass–stellar mass relation if a different mapping between halo and stellar mass is adopted (normalized at 109 M ). The continuous line is for M ∝ M2Vir as in our simulations. The dashed line assumes a linear relation M ∝ MVir that would be created by having less efficient feedback at smaller masses, resulting in more stars. In this second case, there is a clear trend between stellar mass and total central mass, as a much larger range in halo masses is mapped over the same range in stellar masses. As a consequence, if simulations had a weaker feedback they would not match the central mass–stellar mass relation. Note that we only plot the subsample of Strigari’s sample that could be safely extrapolated out to 500 pc. As shown in S08, the full sample continues to follow a flat central mass–stellar mass relation down to ∼103 M (in stars) when measured at 300 pc. We decided against extrapolating to 500 pc the stellar and total masses of the smallest galaxies in the observed sample. This extrapolation becomes less reliable going to very faint/small galaxies with very small sizes (Walker, private communication). On the other hand, our simulations would become less reliable at smaller radii. We plan to study the mass distribution in ultra-faint dwarfs and in satellites of MW-like systems in future, higher resolution simulations that will allow robust predictions at smaller radii. The scatter in accretion histories and SF truncation times of the MW satellites due to ram pressure (Mayer et al. 2007) has often been cited as a possible origin of the central mass–stellar mass relation (Li et al. 2009; Maccio` et al. 2009; Parry et al. 2012) , as it would cause a large scatter in the amount of formed stars at a given halo mass. This mechanism cannot be responsible for the flat trend in Fig. 4, where the THINGS galaxies and all the simulated galaxies are field galaxies with a prolonged, untruncated SFH. We plan a more detailed comparison using a larger ensemble of simulated satellites of MW analogues (Brooks et al., in preparation). These results are very encouraging, as the ability of creating realistic galaxies (see also G10) gives further support to gas outflows as the origin of the flattening of CDM density profiles. In this work we used fully cosmological hydrodynamical simulations to show that once baryonic processes are correctly taken into account, shallow central DM profiles are a common property of field galaxies formed within the CDM model. Our predictions are in excellent agreement with observational estimates of the DM distribution in galaxies from the THINGS and LITTLE THINGS data samples and extend results from G10 and OH11. The introduction of gas outflows in high-resolution simulations resolves the longstanding tension between the observed ‘cored’ DM distribution at the centre of small galaxies and the dense cuspy DM distribution predicted in (C)DM-only simulations. SN feedback is shown to be a vital ingredient in galaxy formation models, where the removal of low angular momentum gas creates at the same time galaxies with DM cores and bulgeless dwarfs (Fall 1983, Bullock et al. 2001; van den Bosch, Burkert & Swaters 2001; Mashchenko, Wadsley & Couchman 2008; G10; Brook et al. 2011; Pontzen & Governato 2011) . This work highlights the fundamental role that baryon and DM interactions play in shaping galaxy properties as fundamental as their central DM and baryon distribution. Predictions on the detailed properties of galaxies based on DM-only simulations or methods where baryon dynamics are not fully coupled to the DM need to be viewed with some caution. The simulated galaxies described in this work covered more than 5 orders of magnitude in stellar mass and circular velocities from 10 to 100 km s−1. These simulations were carried to z = 0, included metal cooling and H2 related processes with the spatial and mass resolution to identify individual star-forming regions. In these simulations, bursty SF limited to dense, H2-rich regions creates repeated, fast outflows which break the adiabatic approximation. Over several Gyr, these fast and repeated outflows progressively lower the central DM density of galaxy haloes and turn CDM central ‘cuspy’ profiles into much shallower ‘cores’. With the combination of SN feedback and cosmic UV background adopted in this work, the DM distribution remains cuspy (or at least with cores smaller than our current spatial resolution) only in dwarfs with M less than 105 M , where less than 0.03 per cent M of the original baryon fraction was turned into stars. These findings strongly support the analytical model presented in Pontzen & Governato (2011) and other numerical works on the formation of cores (G10; Martizzi et al. 2011; Maccio` et al. 2012) , while making detailed observable predictions. As an important consistency check of our model of baryon–DM interactions, we compared our simulations to the observed central mass–luminosity relation for dwarf, extended to a sample containing also field galaxies from the THINGS and LITTLE THINGS surveys (Oh et al., in preparation; S08). Our simulations reproduce the almost constant total mass within the central 500 pc of dwarf galaxies over a wide range of masses, up to a few times 108 M . In our framework, this result is caused by rapidly decreasing SF efficiency at decreasing halo masses and SN feedback simultaneously lowering the central DM density in more massive galaxies. These two effects cause galaxies over a large range in luminosities to inhabit haloes with a relatively small mass range within the central kpc. The correct modelling of baryon–DM interactions in galaxy formation simulations is still in its early stages and the effects of different feedback schemes are possibly even larger than the discrepancies still existing between different gas-dynamical codes (Scannapieco et al. 2011) . It is remarkable that feedback processes similar to those commonly observed in galaxies can simultaneously improve on at least three fundamental problems in galaxy formation: (i) the substructure overabundance problem by rapidly decreasing SF efficiency at smaller halo masses; (ii) the formation of bulgeless galaxies in hierarchical models by the selective removal of low angular momentum gas and (iii) the existence of DM cores in CDM cosmologies by allowing energy transfer from baryons to the DM matter. While it is possible that other physical processes are involved in each of the above problems, numerical and analytical models point to the ubiquitous role of energy feedback. Further comparisons with observational data, especially the constraints coming from the MW satellites, will involve understanding in detail how the properties of the mass distribution of faint dwarf galaxies can be inferred from their stellar kinematics (Boylan-Kolchin, Bullock & Kaplinghat 2011a; Evans, An & Deason 2011; Koposov et al. 2011; Adams et al. 2012) and how interactions with the main galaxy affect the DM, SFHs and final stellar distribution of galaxy satellites. We expect these comparisons to add further constraints on the effects of SN feedback as a function of halo mass and to guide predictions for DM direct detection experiments (Dalal & Kochanek 2002). In the near future, we will also extend our analysis to higher mass systems as ∼L galaxies, where the presence of DM cores is still debated (Chemin, de Blok & Mamon 2011; Swaters et al. 2011) . Measurements of the mass distribution in very faint galaxies will be able to strongly distinguish between baryon–DM interaction models from those invoking alternative DM models to explain the observed central distribution of galaxies. AC K N OW L E D G M E N T S FG and TQ were funded by NSF grant AST-0908499. FG acknowledges support from NSF grant AST-0607819 and NASA ATP NNX08AG84G. AZ acknowledges support from NSF grant AST0908446, ISF grant 6/08 and GIF grant G-1052-104.7/2009. AMB gratefully acknowledges support from The Grainger Foundation. Some of SHO’s research was carried out at ‘The Centre for All-sky Astrophysics’, which is an Australian Research Council Centre of Excellence, funded by grant CE11E0090. Simulations were run at TACC and NAS. We thank Matthew Walker for sharing his data and Oleg and Nick Gnedin, Avishai Dekel, Piero Madau, Mike Boylan-Kolchin and Jorge Pen˜arrubia for useful discussions. R E F E R E N C E S This paper has been typeset from a TEX/LATEX file prepared by the author. Abdo A. A. et al., 2010 , ApJ, 712 , 147 Adams J. J. , Gebhardt K. , Blanc G. A. , Fabricius M. H. , Hill G. J. , Murphy J. D. , van den Bosch R. C. E., van de Ven G., 2012 , ApJ, 745 , 92 Agertz O. et al., 2007 , MNRAS, 380 , 963 Ahn K. , Shapiro P. R. , 2005 , MNRAS, 363 , 1092 Avila Reese V. , Col´ın P. , Valenzuela O. , D'Onghia E. , Firmani C. , 2001 , ApJ, 559 , 516 Barnes J. , Efstathiou G. , 1987 , ApJ, 319 , 575 Bauer A. , Springel V. , 2011 , preprint (arXiv: 1109 .4413) Bell E. F. , de Jong R. S., 2001 , ApJ, 550 , 212 Belokurov V. et al., 2007 , ApJ, 654 , 897 Benson A. J. , Frenk C. S. , Lacey C. G. , Baugh C. M. , Cole S. , 2002 , MNRAS, 333 , 177 Bigiel F. , Leroy A. , Walter F. , Brinks E ., de Blok W. J. G. , Madore B. , Thornley M. D. , 2008 , AJ, 136 , 2846 Bode P. , Ostriker J. P. , Turok N. , 2001 , ApJ, 556 , 93 Bower R. G. , Benson A. J. , Malbon R. , Helly J. C. , Frenk C. S. , Baugh C. M. , Cole S. , Lacey C. G. , 2006 , MNRAS, 370 , 645 Boylan-Kolchin M. , Bullock J. S. , Kaplinghat M. , 2011a, preprint (arXiv e-prints) Boylan-Kolchin M. , Bullock J. S. , Kaplinghat M. , 2011b , MNRAS, 415 , L40 Brook C. B . et al., 2011 , MNRAS, 415 , 1051 Brooks A. M. , Governato F. , Booth C. M. , Willman B. , Gardner J. P. , Wadsley J. , Stinson G. , Quinn T. , 2007 , ApJ, 655 , L17 Brooks A. M. et al., 2011 , ApJ, 728 , 51 Bruzual G. , Charlot S. , 2003 , MNRAS, 344 , 1000 Bullock J. S. , Dekel A. , Kolatt T. S. , Kravtsov A. V. , Klypin A. A. , Porciani C. , Primack J. R. , 2001 , ApJ, 555 , 240 Ceverino D. , Klypin A. , 2009 , ApJ, 695 , 292 Chemin L. , de Blok W. J. G. , Mamon G. A. , 2011 , AJ, 142 , 109 Choi J.-H. , Nagamine K. , 2011 , MNRAS, 410 , 2579 Christensen C. , Quinn T. , Governato F. , Stilp A. , Shen S. , Wadsley J. , 2011 , MNRAS, submitted Col´ın P. , Valenzuela O. , Avila Reese V., 2008 , ApJ, 673 , 203 Conroy C. , Wechsler R. H. , Kravtsov A. V. , 2006 , ApJ, 647 , 201 Croft R. A. C. , Weinberg D. H. , Bolte M. , Burles S. , Hernquist L. , Katz N. , Kirkman D. , Tytler D. , 2002 , ApJ, 581 , 20 Dalal N. , Kochanek C. S. , 2002 , ApJ, 572 , 25 Dalcanton J. J. , Hogan C. J. , 2001 , ApJ, 561 , 35 Dalcanton J. J . et al., 2009 , ApJS, 183 , 67 Dave ´ R., Spergel D. N. , Steinhardt P. J. , Wandelt B. D. , 2001 , ApJ, 547 , 574 de Blok W. J. G. , McGaugh S. S. , Bosma A. , Rubin V. C. , 2001 , ApJ, 552 , L23 de Blok W. J. G. , Walter F. , Brinks E. , Trachternach C. , Oh S. , Kennicutt R. C. , 2008 , AJ, 136 , 2648 de Souza R . S., Rodrigues L. F. S. , Ishida E. E. O. , Opher R. , 2011 , MNRAS, 415 , 2969 Dekel A. , Silk J. , 1986 , ApJ, 303 , 39 Dekel A. , Arad I. , Devor J. , Birnboim Y. , 2003a, ApJ, 588 , 680 Dekel A. , Devor J. , Hetzroni G. , 2003b , MNRAS, 341 , 326 Del Popolo A. , 2009 , ApJ, 698 , 2093 Donato F . et al., 2009 , MNRAS, 397 , 1169 Eke V. R. , Cole S. , Frenk C. S. , 1996 , MNRAS, 282 , 263 Elmegreen B. G. , Bournaud F. , Elmegreen D. M. , 2008 , ApJ, 688 , 67 Elson E. C. , de Blok W. J. G. , Kraan-Korteweg R . C., 2010 , MNRAS, 404 , 2061 El-Zant A. A. , Hoffman Y. , Primack J. , Combes F. , Shlosman I. , 2004 , ApJ, 607 , L75 Evans N. W. , An J. , Deason A. J. , 2011 , ApJ, 730 , L26 Fall S. M. , 1983 , in Athanassoula E., ed., IAU Symp . Vol. 100 , Internal Kinematics and Dynamics of Galaxies. Reidel, Dordrecht, p. 391 Feldmann R. , Gnedin N. Y. , Kravtsov A. V. , 2011 , ApJ, 732 , 115 Flores R. A. , Primack J. R. , 1994 , ApJ, 427 , L1 Font A. S. et al., 2011 , MNRAS, 417 , 1260 Geha M. , Blanton M. R. , Masjedi M. , West A. A. , 2006 , ApJ, 653 , 240 Geha M. , Willman B. , Simon J. D. , Strigari L. E. , Kirby E. N. , Law D. R. , Strader J. , 2009 , ApJ, 692 , 1464 Gentile G. , Famaey B ., de Blok W. J. G. , 2011 , A &A, 527 , A76 Gnedin O. Y. , Zhao H. , 2002 , MNRAS, 333 , 299 Gnedin N. Y. , Tassis K. , Kravtsov A. V. , 2009 , ApJ, 697 , 55 Goerdt T. , Moore B. , Read J. I. , Stadel J. , Zemp M. , 2006 , MNRAS, 368 , 1073 Goerdt T. , Moore B. , Read J. I. , Stadel J. , 2010 , ApJ, 725 , 1707 Governato F. , Willman B. , Mayer L. , Brooks A. , Stinson G. , Valenzuela O. , Wadsley J. , Quinn T. , 2007 , MNRAS, 374 , 1479 Governato F . et al., 2009 , MNRAS, 398 , 312 Governato F . et al., 2010 , Nat, 463 , 203 ( G10 ) Guedes J. , Callegari S. , Madau P. , Mayer L. , 2011 , ApJ, 742 , 76 Haardt F. , Madau P. , 1996 , ApJ, 461 , 20 Hopkins P. F. , Quataert E. , Murray N. , 2011 , MNRAS, 417 , 950 Jardel J. , Gebhardt K. , 2011 , preprint (arXiv e-prints) Katz N. , White S. D. M. , 1993 , ApJ, 412 , 455 Kazantzidis S. , Mayer L. , Mastropietro C. , Diemand J. , Stadel J. , Moore B. , 2004 , ApJ, 608 , 663 Kennicutt R. C. , 1998 , ApJ, 498 , 541 Keresˇ D. , Katz N. , Dave ´ R., Fardal M. , Weinberg D. H. , 2009 , MNRAS, 396 , 2332 Keres D. , Vogelsberger M. , Sijacki D. , Springel V. , Hernquist L. , 2011 , preprint (arXiv e-prints) Kirby E. N. , Martin C. L. , Finlator K. , 2011 , ApJ, 742 , L25 Knebe A. , Devriendt J. E. G. , Mahmood A. , Silk J. , 2002 , MNRAS, 329 , 813 Knollmann S. R. , Knebe A. , 2009 , ApJS, 182 , 608 Koda J. , Shapiro P. R. , 2011 , MNRAS, 415 , 1125 Koposov S. E . et al., 2011 , ApJ, 736 , 146 Krumholz M. R. , Dekel A. , 2011 , preprint (arXiv: 1106 .0301) Krumholz M. R. , McKee C. F. , 2005 , ApJ, 630 , 250 Kuhlen M. , Krumholz M. , Madau P. , Smith B. , Wise J. , 2011 , preprint (arXiv e-prints) Kuzio de Naray R ., Martinez G. D. , Bullock J. S. , Kaplinghat M. , 2010 , ApJ, 710 , L161 Lackner C. N. , Ostriker J. P. , 2010 , ApJ, 712 , 88 Li Y. , Helmi A. , De Lucia G., Stoehr F., 2009 , MNRAS, 397 , L87 Loeb A. , Weiner N. , 2011 , Phys. Rev. Lett., 106 , 171302 Lovell M. et al., 2012 , MNRAS, 420 , 2318 Maccio ` A. V. , Dutton A . A., van den Bosch F. C. , Moore B. , Potter D. , Stadel J. , 2007 , MNRAS, 378 , 55 Maccio ` A. V. , Kang X. , Moore B. , 2009 , ApJ, 692 , L109 Maccio ` A. V. , Stinson G. , Brook C. B. , Wadsley J. , Couchman H. M. P. , Shen S. , Gibson B. K. , Quinn T. , 2012 , ApJ, 744 , L9 McGaugh S. S. , 2005 , ApJ, 632 , 859 McQuinn K. B. W . et al., 2010 , ApJ, 721 , 297 Martin C. L. , 1999 , ApJ, 513 , 156 Martinez G. D. , Bullock J. S. , Kaplinghat M. , Strigari L. E. , Trotta R. , 2009 , J. Cosmol . Astropart. Phys. , 6 , 14 Martizzi D. , Teyssier R. , Moore B. , Wentz T. , 2011 , preprint (arXiv: 1112 .2752) Mashchenko S. , Couchman H. M. P. , Wadsley J. , 2006 , Nat, 442 , 539 Mashchenko S. , Wadsley J. , Couchman H. M. P. , 2008 , Sci, 319 , 174 Mayer L. , Governato F. , Colpi M. , Moore B. , Quinn T. , Wadsley J. , Stadel J. , Lake G. , 2001 , ApJ, 559 , 754 Mayer L. , Kazantzidis S. , Mastropietro C. , Wadsley J. , 2007 , Nat, 445 , 738 Mo H. J. , Mao S. , 2004 , MNRAS, 353 , 829 Moore B. , Governato F. , Quinn T. , Stadel J. , Lake G. , 1998 , ApJ, 499 , L5 Moore B. , Quinn T. , Governato F. , Stadel J. , Lake G. , 1999 , MNRAS, 310 , 1147 Moster B. P. , Somerville R. S. , Maulbetsch C., van den Bosch F. C. , Maccio ` A. V. , Naab T. , Oser L. , 2010 , ApJ, 710 , 903 Navarro J. F. , Eke V. R. , Frenk C. S. , 1996a , MNRAS, 283 , L72 Navarro J. F. , Frenk C. S. , White S. D. M. , 1996b, ApJ, 462 , 563 Navarro J. F . et al., 2010 , MNRAS, 402 , 21 Oh S. , de Blok W. J. G. , Walter F. , Brinks E. , Kennicutt R. C. , 2008 , AJ, 136 , 2761 Oh S.-H. , Brook C. , Governato F. , Brinks E. , Mayer L., de Blok W. J. G. , Brooks A. , Walter F. , 2011a , AJ, 142 , 24 ( OH11 ) Oh S.-H ., de Blok W. J. G. , Brinks E. , Walter F. , Kennicutt R. C. , Jr , 2011b, AJ, 141 , 193 Oppenheimer B. D. , Dave´ R., Keresˇ D. , Fardal M. , Katz N. , Kollmeier J. A. , Weinberg D. H. , 2010 , MNRAS, 406 , 2325 Parry O. H. , Eke V. R. , Frenk C. S. , Okamoto T. , 2012 , MNRAS, 419 , 3304 Pen ˜arrubia J., Benson A. J. , Walker M. G. , Gilmore G. , McConnachie A. W. , Mayer L. , 2010 , MNRAS, 406 , 1290 Pontzen A. , Governato F. , 2011 , MNRAS, in press (arXiv e-prints) Pontzen A. et al., 2008 , MNRAS, 390 , 1349 Pontzen A. et al., 2010 , MNRAS, 402 , 1523 Primack J. R. , 2009 , New J. Phys., 11 , 105029 Quinn T. , Katz N. , Efstathiou G. , 1996 , MNRAS, 278 , L49 Rashkov V. , Madau P. , Kuhlen M. , Diemand J. , 2012 , ApJ, 745 , 142 Read J. I. , Gilmore G. , 2005 , MNRAS, 356 , 107 Read J. I. , Hayfield T. , 2011 , preprint (arXiv: 1111 .6985) Reed D. , Governato F. , Verde L. , Gardner J. , Quinn T. , Stadel J. , Merritt D. , Lake G. , 2005 , MNRAS, 357 , 82 Riess A. G. et al., 1998 , AJ, 116 , 1009 Romano-D´ ıaz E., Shlosman I. , Heller C. , Hoffman Y. , 2009 , ApJ, 702 , 1250 Saitoh T. R. , Daisaka H. , Kokubo E. , Makino J. , Okamoto T. , Tomisaka K. , Wada K. , Yoshida N. , 2008 , PASJ, 60 , 667 Sa´nchez-Janssen R. , Me´ ndez -Abreu J., Aguerri J. A. L. , 2010 , MNRAS, 406 , L65 Scannapieco C. et al., 2011 , preprint (arXiv: 1112 .0315) Schneider A. , Smith R. E. , Maccio A. V. , Moore B. , 2011 , preprint (arXiv: 1112 .0330) Shapley A. E. , Steidel C. C. , Pettini M. , Adelberger K. L. , 2003 , ApJ, 588 , 65 Shen S. , Wadsley J. , Stinson G. , 2010 , MNRAS, 407 , 1581 Simon J. D. , Geha M. , 2007 , ApJ, 670 , 313 Simon J. D. , Bolatto A. D. , Leroy A. , Blitz L. , Gates E. L. , 2005 , ApJ, 621 , 757 Spergel D. N. , Steinhardt P. J. , 2000 , Phys. Rev. Lett., 84 , 3760 Spergel D. N . et al., 2007 , ApJS, 170 , 377 Springel V. , Di Matteo T., Hernquist L. , 2005 , MNRAS, 361 , 776 Stadel J. , Potter D. , Moore B. , Diemand J. , Madau P. , Zemp M. , Kuhlen M. , Quilis V. , 2009 , MNRAS, 398 , L21 Stinson G. , Seth A. , Katz N. , Wadsley J. , Governato F. , Quinn T. , 2006 , MNRAS, 373 , 1074 Stoehr F. , White S. D. M. , Tormen G. , Springel V. , 2002 , MNRAS, 335 , L84 Strigari L. E. , Kaplinghat M. , Bullock J. S. , 2007 , Phys. Rev. D, 75 , 061303 Strigari L. E. , Bullock J. S. , Kaplinghat M. , Simon J. D. , Geha M. , Willman B. , Walker M. G. , 2008 , Nat, 454 , 1096 ( S08 ) Stringer M. , Cole S. , Frenk C. S. , 2010 , MNRAS, 404 , 1129 Swaters R. A. , Madore B. F., van den Bosch F. C. , Balcells M. , 2003 , ApJ, 583 , 732 Swaters R. A. , Sancisi R ., van Albada T. S., van der Hulst J. M. , 2011 , ApJ, 729 , 118 Tollerud E. J. , Bullock J. S. , Strigari L. E. , Willman B. , 2008 , ApJ, 688 , 277 Tollerud E. J. , Bullock J. S. , Graves G. J. , Wolf J. , 2011 , ApJ, 726 , 108 Tonini C. , Lapi A. , Salucci P. , 2006 , ApJ, 649 , 591 van den Bosch F. C. , Burkert A. , Swaters R. A. , 2001 , MNRAS, 326 , 1205 van der Wel A . et al., 2011 , ApJ, 730 , 38 Viel M. , Lesgourgues J. , Haehnelt M. G. , Matarrese S. , Riotto A. , 2005 , Phys. Rev. D, 71 , 063534 Vogelsberger M. , Zavala J. , Loeb A. , 2012 , MNRAS, submitted (arXiv: 1201 .5892) Wadsley J. W. , Stadel J. , Quinn T. , 2004 , New Astron., 9 , 137 Walker M. G. , Pen˜arrubia J., 2011 , ApJ, 742 , 20 Walker M. G. , Mateo M. , Olszewski E. W. , Pen˜arrubia J., Wyn Evans N. , Gilmore G. , 2009 , ApJ, 704 , 1274 Walter F. , Brinks E ., de Blok W. J. G. , Bigiel F. , Kennicutt R. C. , Jr , Thornley M. D. , Leroy A. , 2008 , AJ, 136 , 2563 Willman B. et al., 2005 , ApJ, 626 , L85 Zhang H. , Hunter D. , LITTLE THINGS Team , 2011 , in Treyer M., Wyder T. , Neill J. , Seibert M. , Lee J ., eds, ASP Conf. Ser . Vol. 440 , UP2010 : Have Observations Revealed a Variable Upper End of the Initial Mass Function? Astron . Soc. Pac., San Francisco, p. 247 Zolotov A. , Willman B. , Brooks A. M. , Governato F. , Brook C. B. , Hogg D. W. , Quinn T. , Stinson G. , 2009 , ApJ, 702 , 1058

This is a preview of a remote PDF:

F. Governato, A. Zolotov, A. Pontzen, C. Christensen, S. H. Oh, A. M. Brooks, T. Quinn, S. Shen, J. Wadsley. Cuspy no more: how outflows affect the central dark matter and baryon distribution in Λ cold dark matter galaxies, Monthly Notices of the Royal Astronomical Society, 2012, 1231-1240, DOI: 10.1111/j.1365-2966.2012.20696.x