Intrinsic Shapes Of Galaxies Research Articles

Kinematic lensing inference – I. Characterizing shape noise with simulated analyses

ABSTRACT The unknown intrinsic shape of source galaxies is one of the largest uncertainties of weak gravitational lensing (WL). It results in the so-called shape noise at the level of $\sigma _\epsilon ^{\mathrm{WL}} \approx 0.26$, whereas the shear effect of interest is of the order of per cent. Kinematic lensing (KL) is a new technique that combines photometric shape measurements with resolved spectroscopic observations to infer the intrinsic galaxy shape and directly estimate the gravitational shear. This paper presents a KL inference pipeline that jointly forward-models galaxy imaging and slit spectroscopy to extract the shear signal. We build a set of realistic mock observations and show that the KL inference pipeline can robustly recover the input shear. To quantify the shear measurement uncertainty for KL, we average the shape noise over a population of randomly oriented disc galaxies and estimate it to be $\sigma _\epsilon ^{\mathrm{KL}}\approx 0.022\!-\!0.038$ depending on emission-line signal-to-noise ratio. This order of magnitude improvement over traditional WL makes a KL observational programme feasible with existing spectroscopic instruments. To this end, we characterize the dependence of KL shape noise on observational factors and discuss implications for the survey strategy of future KL observations. In particular, we find that prioritizing quality spectra of low-inclination galaxies is more advantageous than maximizing the overall number density.

The MillenniumTNG project: intrinsic alignments of galaxies and haloes

ABSTRACT The intrinsic alignment (IA) of observed galaxy shapes with the underlying cosmic web is a source of contamination in weak lensing surveys. Sensitive methods to identify the IA signal will therefore need to be included in the upcoming weak lensing analysis pipelines. Hydrodynamical cosmological simulations allow us to directly measure the intrinsic ellipticities of galaxies, and thus provide a powerful approach to predict and understand the IA signal. Here we employ the novel, large-volume hydrodynamical simulation MTNG740, a product of the MillenniumTNG (MTNG) project, to study the IA of galaxies. We measure the projected correlation functions between the intrinsic shape/shear of galaxies and various tracers of large-scale structure, w+g, w+m, w++ over the radial range $r_{\rm p} \in [0.02 , 200]\, h^{-1}{\rm Mpc}$ and at redshifts z = 0.0, 0.5, and 1.0. We detect significant signal-to-noise IA signals with the density field for both elliptical and spiral galaxies. We also find significant intrinsic shear–shear correlations for ellipticals. We further examine correlations of the intrinsic shape of galaxies with the local tidal field. Here we find a significant IA signal for elliptical galaxies assuming a linear model. We also detect a weak IA signal for spiral galaxies under a quadratic tidal torquing model. Lastly, we measure the alignment between central galaxies and their host dark-matter haloes, finding small to moderate misalignments between their principal axes that decline with halo mass.

Relativistic distortions in galaxy density–ellipticity correlations: gravitational redshift and peculiar velocity effects

ABSTRACT We study relativistic effects, arising from the light propagation in an inhomogeneous universe. We particularly investigate the effects imprinted in a cross-correlation function between galaxy positions and intrinsic galaxy shapes (GI correlation). Considering the Doppler and gravitational redshift effects as major relativistic effects, we present an analytical model of the GI correlation function, from which we find that the relativistic effects induce non-vanishing odd multipole anisotropies. Focusing particularly on the dipole anisotropy, we show that the Doppler effect dominates at large scales, while the gravitational redshift effect originated from the halo potential dominates at the scales below 10–$30\, {\rm Mpc}\, h^{-1}$, with the amplitude of the dipole GI correlation being positive over all the scales. Also, we newly derive the covariance matrix for the modelled GI dipole. Taking into account the full covariance, we estimate the signal-to-noise ratio and show that the GI dipole induced by the relativistic effects is detectable in future large-volume galaxy surveys. We discuss how the measurement of dipole GI correlation could be helpful to detect relativistic effects in combination with the conventional galaxy–galaxy cross-correlation.

The impact of self-interacting dark matter on the intrinsic alignments of galaxies

ABSTRACT The formation and evolution of galaxies is known to be sensitive to tidal processes leading to intrinsic correlations between their shapes and orientations. Such correlations can be measured to high significance today, suggesting that cosmological information can be extracted from them. Among the most pressing questions in particle physics and cosmology is the nature of dark matter. If dark matter is self-interacting, it can leave an imprint on galaxy shapes. In this work, we investigate whether self-interactions can produce a long-lasting imprint on intrinsic galaxy shape correlations. We investigate this observable at low redshift (z < 0.4) using a state-of-the-art suite of cosmological hydro-dynamical simulations where the dark matter model is varied. We find that dark matter self-interactions induce a mass-dependent suppression in the intrinsic alignment signal by up to 50 per cent out to tens of mega-parsecs, showing that self-interactions can impact structure outside the very core of clusters. We find evidence that self-interactions have a scale-dependent impact on the intrinsic alignment signal that is sufficiently different from signatures introduced by differing baryonic physics prescriptions, suggesting that it is detectable with upcoming all-sky surveys.

Radio–optical galaxy shape and shear correlations in the COSMOS field using 3 GHz VLA observations

Abstract We present a weak-lensing analysis of the 3 GHz Very Large Array radio survey of the COSMOS field, which we correlate with overlapping Hubble Space Telescope-Advanced Camera for Survey optical observations using both intrinsic galaxy shape and cosmic shear correlation statistics. After cross-matching sources between the two catalogues, we measure the correlations of galaxy position angles and find a Pearson correlation coefficient of 0.14 ± 0.03. This is a marked improvement from previous studies which found very weak, or non-existent correlations, and gives insight into the emission processes of radio and optical galaxies. We also extract power spectra of averaged galaxy ellipticities (the primary observable for cosmic shear) from the two catalogues, and produce optical–optical, radio–optical, and radio–radio spectra. The optical–optical autopower spectrum was measured to a detection significance of 9.80σ and is consistent with previous observations of the same field. For radio spectra (which we do not calibrate, given the unknown nature of their systematics), although we do not detect significant radio–optical (1.50σ) or radio–radio (1.45σ) E-mode power spectra, we do find the E-mode spectra to be more consistent with the shear signal expected from previous studies than with a null signal, and vice versa for B-mode and EB cross-correlation spectra. Our results give promise that future radio weak-lensing surveys with larger source number densities over larger areas will have the capability to measure significant weak-lensing signals.

Prospects for recovering galaxy intrinsic shapes from projected quantities

The distribution of three dimensional intrinsic galaxy shapes has been a longstanding open question. The difficulty stems from projection effects meaning one must rely on statistical methods applied to galaxy samples to infer intrinsic shape distributions. Theoretical work using analytical galaxy potentials suggests a relationship between galaxy intrinsic shape (as defined by its "triaxiality", in practice a proxy for how prolate a galaxy is) and the intrinsic misalignment angle between kinematic and morphological axes ($\Psi_{\rm int}$). This relationship reduces the number of unknowns, providing more reliable inferred intrinsic shape distributions than methods using photometry alone. Here we explore the connection between intrinsic shape and stellar kinematics using cosmological hydrodynamical simulations from the Illustris project. The strongest relationship we find is that galaxy intrinsic flattening is correlated with specific angular momentum (j) with high j galaxies being flatter than galaxies with low specific angular momentum. Our analysis shows that, although the majority of kinematically misaligned galaxies exhibit prolate shapes, examples of kinematically aligned prolate galaxies are also present. Clearly a direct correspondence between prolate shape and minor-axis rotation (often referred to as "prolate rotation") is not present in Illustris. Thus, we demonstrate that the assumption of a simple relationship between $\Psi_{\rm int}$ and intrinsic shape commonly employed in shape recovery studies is not valid for Illustris galaxies. We suggest improvements on the method as well as some alternative methods for future work in this area.

The SAMI Galaxy Survey: the intrinsic shape of kinematically selected galaxies

Using the stellar kinematic maps and ancillary imaging data from the Sydney AAO Multi Integral field (SAMI) Galaxy Survey, the intrinsic shape of kinematically-selected samples of galaxies is inferred. We implement an efficient and optimised algorithm to fit the intrinsic shape of galaxies using an established method to simultaneously invert the distributions of apparent ellipticities and kinematic misalignments. The algorithm output compares favourably with previous studies of the intrinsic shape of galaxies based on imaging alone and our re-analysis of the ATLAS3D data. Our results indicate that most galaxies are oblate axisymmetric. We show empirically that the intrinsic shape of galaxies varies as a function of their rotational support as measured by the "spin" parameter proxy Lambda_Re. In particular, low spin systems have a higher occurrence of triaxiality, while high spin systems are more intrinsically flattened and axisymmetric. The intrinsic shape of galaxies is linked to their formation and merger histories. Galaxies with high spin values have intrinsic shapes consistent with dissipational minor mergers, while the intrinsic shape of low-spin systems is consistent with dissipationless multi-merger assembly histories. This range in assembly histories inferred from intrinsic shapes is broadly consistent with expectations from cosmological simulations.

Multitracing anisotropic non-Gaussianity with galaxy shapes

Correlations between intrinsic galaxy shapes on large-scales arise due to the effect of the tidal field of the large-scale structure. Anisotropic primordial non-Gaussianity induces a distinct scale-dependent imprint in these tidal alignments on large scales. Motivated by the observational finding that the alignment strength of luminous red galaxies depends on how galaxy shapes are measured, we study the use of two different shape estimators as a multi-tracer probe of intrinsic alignments. We show, by means of a Fisher analysis, that this technique promises a significant improvement on anisotropic non-Gaussianity constraints over a single-tracer method. For future weak lensing surveys, the uncertainty in the anisotropic non-Gaussianity parameter, $A_2$, is forecast to be $\sigma(A_2)\approx 50$, $\sim 40\%$ smaller than currently available constraints from the bispectrum of the Cosmic Microwave Background. This corresponds to an improvement of a factor of $4-5$ over the uncertainty from a single-tracer analysis.

Redshift and luminosity evolution of the intrinsic alignments of galaxies in Horizon-AGN

Intrinsic galaxy shape and angular momentum alignments can arise in cosmological large-scale structure due to tidal interactions or galaxy formation processes. Cosmological hydrodynamical simulations have recently come of age as a tool to study these alignments and their contamination to weak gravitational lensing. We probe the redshift and luminosity evolution of intrinsic alignments in Horizon-AGN between z=0 and z=3 for galaxies with an r-band absolute magnitude of <-20. Alignments transition from being radial at low redshifts and high luminosities, dominated by the contribution of ellipticals, to being tangential at high redshift and low luminosities, where discs dominate the signal. This cannot be explained by the evolution of the fraction of ellipticals and discs alone: intrinsic evolution in the amplitude of alignments is necessary. The alignment amplitude of elliptical galaxies alone is smaller in amplitude by a factor of ~2, but has similar luminosity and redshift evolution as in current observations and in the nonlinear tidal alignment model at projected separations of > 1 Mpc. Alignments of discs are null in projection and consistent with current low redshift observations. The combination of the two populations yields an overall amplitude a factor of ~4 lower than observed alignments of luminous red galaxies with a steeper luminosity dependence. The restriction on accurate galaxy shapes implies that the galaxy population in the simulation is complete only to an r-band absolute magnitude of <-20. Higher resolution simulations will be necessary to avoid extrapolation of the intrinsic alignment predictions to the range of luminosities probed by future surveys.

The effects of environment on the intrinsic shape of galaxies

We measure the effect of the environment on the intrinsic shapes of spiral and elliptical galaxies by finding the 3D shape distribution and dust extinction that fits better the projected shape of galaxies in different environment. We find that spiral galaxies in groups are very similar to field spirals with similar intrinsic properties (magnitudes, sizes and colours). But for spirals in groups, those in denser environments or closer to the centre of the group tend to have a more circular disc than similar galaxies in less dense environments or far from the group centres. Also we find that central spiral galaxies in their groups tend to be thinner than other similar spirals. For ellipticals, we do not find any important dependence of their shape on their position in a group or on the local density. However, we find that elliptical galaxies in groups tend to be more spherical than field ellipticals with similar intrinsic properties. We find that, once in groups, the shape of member galaxies do not depend on group mass, regardless of their morphological type.

Detection of universality of dark matter profile from Subaru weak lensing measurements of 50 massive clusters

Abstract We develop a novel method of measuring the lensing distortion profiles of clusters by stacking the “scaled” amplitudes of background galaxy ellipticities as a function of the “scaled” centric radius according to the Navarro–Frenk–White (NFW) prediction of each cluster, based on the assumption that the different clusters in a sample follow the universalNFW profile. First we demonstrate the feasibility of this method using both the analytical NFW model and simulated halos in a suite of high-resolution N-body simulations. We then apply, as a proof of concept, this method to the Subaru weak lensing data and the XMM/Chandra X-ray observables for a sample of 50 massive clusters in the redshift range 0.15 ≤ z ≤ 0.3, where their halo masses differ from each other by up to a factor of 10. To estimate the NFW parameters of each cluster, we use the halo mass proxy relation of X-ray observables, based on either the hydrostatic equilibrium or the gas mass, and then infer the halo concentration from the model scaling relation of halo concentration with halo mass. We evaluate the performance of the NFW scaling analysis by measuring the scatters of 50 cluster lensing profiles relative to the NFW predictions over a range of radii, 0.14 ≤ R/[h−1 Mpc] ≤ 2.8. We found 4–6 σ-level evidence of the universal NFW profile in 50 clusters, for both the X-ray halo mass proxy relations, although the gas mass appears to be a better proxy of the underlying true mass. By comparing the measurements with the simulations of cluster lensing profiles taking into account the statistical errors of intrinsic galaxy shapes in the Subaru data, we argue that additional halo mass errors or intrinsic scatters of σ(M500c)/M500c ∼ 0.2–0.3 could reconcile the difference between measurements and simulations. This method allows us to some extent to preserve characteristics of individual clusters in the statistical weak lensing analysis, thereby yielding a new means of exploiting the underlying genuine form of the halo mass profile and the halo mass proxy relations via weak lensing information, under the assumption of the existence of the universal profile.

THE THIRD GRAVITATIONAL LENSING ACCURACY TESTING (GREAT3) CHALLENGE HANDBOOK

The GRavitational lEnsing Accuracy Testing 3 (GREAT3) challenge is the third in a series of image analysis challenges, with a goal of testing and facilitating the development of methods for analyzing astronomical images that will be used to measure weak gravitational lensing. This measurement requires extremely precise estimation of very small galaxy shape distortions, in the presence of far larger intrinsic galaxy shapes and distortions due to the blurring kernel caused by the atmosphere, telescope optics, and instrumental effects. The GREAT3 challenge is posed to the astronomy, machine learning, and statistics communities, and includes tests of three specific effects that are of immediate relevance to upcoming weak lensing surveys, two of which have never been tested in a community challenge before. These effects include realistically complex galaxy models based on high-resolution imaging from space; spatially varying, physically-motivated blurring kernel; and combination of multiple different exposures. To facilitate entry by people new to the field, and for use as a diagnostic tool, the simulation software for the challenge is publicly available, though the exact parameters used for the challenge are blinded. Sample scripts to analyze the challenge data using existing methods will also be provided. See http://great3challenge.info and http://great3.projects.phys.ucl.ac.uk/leaderboard/ for more information.

Intrinsic galaxy shapes and alignments – II. Modelling the intrinsic alignment contamination of weak lensing surveys

Intrinsic galaxy alignments constitute the major astrophysical systematic of forthcoming weak gravitational lensing surveys but also yield unique insights into galaxy formation and evolution. We build analytic models for the distribution of galaxy shapes based on halo properties extracted from the Millennium Simulation, differentiating between early- and late-type galaxies as well as central galaxies and satellites. The resulting ellipticity correlations are investigated for their physical properties and compared to a suite of current observations. The best-faring model is then used to predict the intrinsic alignment contamination of planned weak lensing surveys. We find that late-type galaxy models generally have weak intrinsic ellipticity correlations, marginally increasing towards smaller galaxy separation and higher redshift. The signal for early-type models at fixed halo mass strongly increases by three orders of magnitude over two decades in galaxy separation, and by one order of magnitude from z=0 to z=2. The intrinsic alignment strength also depends strongly on halo mass, but not on galaxy luminosity at fixed mass, or galaxy number density in the environment. We identify models that are in good agreement with all observational data, except that all models over-predict alignments of faint early-type galaxies. The best model yields an intrinsic alignment contamination of a Euclid-like survey between 0.5-10% at z>0.6 and on angular scales larger than a few arcminutes. Cutting 20% of red foreground galaxies using observer-frame colours can suppress this contamination by up to a factor of two.

Flexion measurement in simulations of Hubble Space Telescope data

We present a simulation analysis of weak gravitational lensing flexion and shear measurement using shapelet decomposition, and identify differences between flexion and shear measurement noise in deep survey data. Taking models of galaxies from the Hubble Space Telescope Ultra Deep Field (HUDF) and applying a correction for the HUDF point spread function we generate lensed simulations of deep, optical imaging data from Hubble's Advanced Camera for Surveys (ACS), with realistic galaxy morphologies. We find that flexion and shear estimates differ in our measurement pipeline: whereas intrinsic galaxy shape is typically the dominant contribution to noise in shear estimates, pixel noise due to finite photon counts and detector read noise is a major contributor to uncertainty in flexion estimates, across a broad range of galaxy signal-to-noise. This pixel noise also increases more rapidly as galaxy signal-to-noise decreases than is found for shear estimates. We provide simple power law fitting functions for this behaviour, for both flexion and shear, allowing the effect to be properly accounted for in future forecasts for flexion measurement. Using the simulations we also quantify the systematic biases of our shapelet flexion and shear measurement pipeline for deep Hubble data sets such as Galaxy Evolution from Morphology and SEDs, Space Telescope A901/902 Galaxy Evolution Survey or the Cosmic Evolution Survey. Flexion measurement biases are found to be significant but consistent with previous studies.

The intrinsic shape of galaxies in SDSS/Galaxy Zoo

By modelling the axis ratio distribution of SDSS DR8 galaxies we find the intrinsic 3D shapes of spirals and ellipticals. We use morphological information from the Galaxy Zoo project and assume a non-parametric distribution intrinsic of shapes, while taking into account dust extinction. We measure the dust extinction of the full sample of spiral galaxies and find a smaller value than previous estimations, with an edge-on extinction of $E_0 = 0.284^{+0.015}_{-0.026}$ in the SDSS r band. We also find that the distribution of minor to major axis ratio has a mean value of $ 0.267 \pm 0.009 $, slightly larger than previous estimates mainly due to the lower extinction used; the same affects the circularity of galactic discs, which are found to be less round in shape than in previous studies, with a mean ellipticity of $0.215 \pm 0.013$. For elliptical galaxies, we find that the minor to major axis ratio, with a mean value of $0.584 \pm 0.006$, is larger than previous estimations due to the removal of spiral interlopers present in samples with morphological information from photometric profiles. These interlopers are removed when selecting ellipticals using Galaxy Zoo data. We find that the intrinsic shapes of galaxies and their dust extinction vary with absolute magnitude, colour and physical size. We find that bright elliptical galaxies are more spherical than faint ones, a trend that is also present with galaxy size, and that there is no dependence of elliptical galaxy shape with colour. For spiral galaxies we find that the reddest ones have higher dust extinction as expected, due to the fact that this reddening is mainly due to dust. We also find that the thickness of discs increases with luminosity and size, and that brighter, smaller and redder galaxies have less round discs.

Intrinsic galaxy shapes and alignments – I. Measuring and modelling COSMOS intrinsic galaxy ellipticities

The statistical properties of the ellipticities of galaxy images depend on how galaxies form and evolve, and therefore constrain models of galaxy morphology, which are key to the removal of the intrinsic alignment contamination of cosmological weak lensing surveys, as well as to the calibration of weak lensing shape measurements. We construct such models based on the halo properties of the Millennium Simulation and confront them with a sample of 90,000 galaxies from the COSMOS Survey, covering three decades in luminosity and redshifts out to z=2. The ellipticity measurements are corrected for effects of point spread function smearing, spurious image distortions, and measurement noise. Dividing galaxies into early, late, and irregular types, we find that early-type galaxies have up to a factor of two lower intrinsic ellipticity dispersion than late-type galaxies. None of the samples shows evidence for redshift evolution, while the ellipticity dispersion for late-type galaxies scales strongly with absolute magnitude at the bright end. The simulation-based models reproduce the main characteristics of the intrinsic ellipticity distributions although which model fares best depends on the selection criteria of the galaxy sample. We observe fewer close-to-circular late-type galaxy images in COSMOS than expected for a sample of randomly oriented circular thick disks and discuss possible explanations for this deficit.

Constraints on intrinsic alignment contamination of weak lensing surveys using the MegaZ-LRG sample

Correlations between the intrinsic shapes of galaxies and the large-scale galaxy density field provide an important tool to investigate galaxy intrinsic alignments, which constitute a major astrophysical systematic in cosmological weak lensing (cosmic shear) surveys, but also yield insight into the formation and evolution of galaxies. We measure galaxy position-shape correlations in the MegaZ-LRG sample for more than 800,000 luminous red galaxies, making the first such measurement with a photometric redshift sample. In combination with a re-analysis of several spectroscopic SDSS samples, we constrain an intrinsic alignment model for early-type galaxies over long baselines in redshift (z ~ 0.7) and luminosity (4mag). We develop and test the formalism to incorporate photometric redshift scatter in the modelling. For r_p > 6 Mpc/h, the fits to galaxy position-shape correlation functions are consistent with the scaling with r_p and redshift of a revised, nonlinear version of the linear alignment model for all samples. An extra redshift dependence proportional to (1+z)^n is constrained to n=-0.3+/-0.8 (1sigma). To obtain consistent amplitudes for all data, an additional dependence on galaxy luminosity proportional to L^b with b=1.1+0.3-0.2 is required. The normalisation of the intrinsic alignment power spectrum is found to be (0.077 +/- 0.008)/rho_{cr} for galaxies at redshift 0.3 and r band magnitude of -22 (k- and evolution-corrected to z=0). Assuming zero intrinsic alignments for blue galaxies, we assess the bias on cosmological parameters for a tomographic CFHTLS-like lensing survey. Both the resulting mean bias and its uncertainty are smaller than the 1sigma statistical errors when using the constraints from all samples combined. The addition of MegaZ-LRG data reduces the uncertainty in intrinsic alignment bias on cosmological parameters by factors of three to seven. (abridged)

The WiggleZ Dark Energy Survey: direct constraints on blue galaxy intrinsic alignments at intermediate redshifts

Correlations between the intrinsic shapes of galaxy pairs, and between the intrinsic shapes of galaxies and the large-scale density field, may be induced by tidal fields. These correlations, which have been detected at low redshifts (z<0.35) for bright red galaxies in the Sloan Digital Sky Survey (SDSS), and for which upper limits exist for blue galaxies at z~0.1, provide a window into galaxy formation and evolution, and are also an important contaminant for current and future weak lensing surveys. Measurements of these alignments at intermediate redshifts (z~0.6) that are more relevant for cosmic shear observations are very important for understanding the origin and redshift evolution of these alignments, and for minimising their impact on weak lensing measurements. We present the first such intermediate-redshift measurement for blue galaxies, using galaxy shape measurements from SDSS and spectroscopic redshifts from the WiggleZ Dark Energy Survey. Our null detection allows us to place upper limits on the contamination of weak lensing measurements by blue galaxy intrinsic alignments that, for the first time, do not require significant model-dependent extrapolation from the z~0.1 SDSS observations. Also, combining the SDSS and WiggleZ constraints gives us a long redshift baseline with which to constrain intrinsic alignment models and contamination of the cosmic shear power spectrum. Assuming that the alignments can be explained by linear alignment with the smoothed local density field, we find that a measurement of \sigma_8 in a blue-galaxy dominated, CFHTLS-like survey would be contaminated by at most +/-0.02 (95% confidence level, SDSS and WiggleZ) or +/-0.03 (WiggleZ alone) due to intrinsic alignments. [Abridged]

Theoretical estimates of intrinsic galaxy alignment

ABSTRA C T It has recently been argued that the observed ellipticities of galaxies may be determined at least in part by the primordial tidal gravitational field in which the galaxy formed. Long-range correlations in the tidal field could thus lead to an ellipticity ‐ ellipticity correlation for widely separated galaxies. We present a new model relating ellipticity to angular momentum, which can be calculated in linear theory. We use this model to calculate the angular power spectrum of intrinsic galaxy shape correlations. We show that, for low-redshift galaxy surveys, our model predicts that intrinsic correlations will dominate correlations induced by weak lensing, in good agreement with previous theoretical work and observations. We find that our model produces ‘E-mode’ correlations enhanced by a factor of 3.5 over B-modes on small scales, making it harder to disentangle intrinsic correlations from those induced by weak gravitational lensing.

Intrinsic correlation of galaxy shapes: implications for weak lensing measurements

Weak gravitational lensing is now established as a powerful method to measure mass fluctuations in the universe. It relies on the measurement of small coherent distortions of the images of background galaxies. Even low-level correlations in the intrinsic shapes of galaxies could however produce a significant spurious lensing signal. These correlations are also interesting in their own right, since their detection would constrain models of galaxy formation. Using haloes found in N-body simulations, we compute the correlation functions of the intrinsic ellipticity of spiral galaxies assuming that the disc is perpendicular to the angular momentum of the dark matter halo. We also consider a simple model for elliptical galaxies, in which the shape of the dark matter halo is assumed to be the same as that of the light. For deep lensing surveys with median redshifts ∼1, we find that intrinsic correlations of ∼10−4 on angular scales are generally below the expected lensing signal, and contribute only a small fraction of the excess signals reported on these scales. On larger scales we find limits to the intrinsic correlation function at a level ∼10−5, which gives a (model-dependent) range of separations for which the intrinsic signal is about an order of magnitude below the ellipticity correlation function expected from weak lensing. Intrinsic correlations are thus negligible on these scales for dedicated weak lensing surveys. For wider but shallower surveys such as SuperCOSMOS, APM and SDSS, we cannot exclude the possibility that intrinsic correlations could dominate the lensing signal. We discuss how such surveys could be used to calibrate the importance of this effect, as well as study spin–spin correlations of spiral galaxies.