NASA's Interface Region Imaging Spectrograph (IRIS) provides high resolution observations of the solar atmosphere through UV spectroscopy and imaging. Since the launch of IRIS in June 2013, we have conducted systematic observation campaigns in coordination with the Swedish 1-m Solar Telescope (SST) on La Palma. The SST provides complementary high-resolution observations of the photosphere and chromosphere. The SST observations include spectro-polarimetric imaging in photospheric Fe I lines and spectrally-resolved imaging in the chromospheric Ca II 8542 A, H-alpha, and Ca II K lines. We present a database of co-aligned IRIS and SST datasets that is open for analysis to the scientific community. The database covers a variety of targets including active regions, sunspots, plage, quiet Sun, and coronal holes.

We present a data pipeline for the newly installed SST/CHROMIS imaging spectrometer, as well as for the older SST/CRISP spectropolarimeter. The aim is to provide observers with a user-friendly data pipeline, that delivers science-ready data with the metadata needed for archival. We generalized the CRISPRED data pipeline for multiple instruments and added metadata according to recommendations worked out as part of the SOLARNET project. We made improvements to several steps in the pipeline, including the MOMFBD image restoration. A part of that is a new fork of the MOMFBD program called REDUX, with several new features that are needed in the new pipeline. The CRISPEX data viewer has been updated to accommodate data cubes stored in this format. The pipeline code, as well as REDUX and CRISPEX are all freely available through git repositories or web download. We derive expressions for combining statistics of individual frames into statistics for a set of frames. We define a new extension to the World Coordinate System, that allow us to specify cavity errors as distortions to the spectral coordinate.

We demonstrate that for data recorded with a solar telescope that uses adaptive optics and/or post-processing to compensate for many low- and high-order aberrations, the RMS granulation contrast is directly proportional to the Strehl ratio calculated from the residual (small-scale) wavefront error. We demonstrate that the wings of the high-order compensated PSF for SST are likely to extend to a radius of not more than about 2 arcsec, consistent with earlier conclusions drawn from straylight compensation of sunspot images. We report on simultaneous measurements of seeing and solar granulation contrast averaged over 2 sec time intervals at several wavelengths from 525 nm to 853.6 nm on the red-beam (CRISP beam) and wavelengths from 395 nm to 484 nm on the blue-beam (CHROMIS beam). These data were recorded with the Swedish 1-m Solar Telescope (SST) that has been revamped with an 85-electrode adaptive mirror and a new tip-tilt mirror, both of which were polished to exceptionally high optical quality. The highest 2-sec average image contrast measured in April 2015 through 0.3-0.9 nm interference filters at 525 nm, 557 nm, 630 nm and 853.5 nm with compensation only for the diffraction limited point spread function of SST is 11.8%, 11.8%, 10.2% and 7.2% respectively. Similarly, the highest 2-sec contrast measured at 395 nm, 400 nm and 484 nm in May 2016 through 0.37-1.3 nm filters is 16%, 16% and 12.5% respectively. The granulation contrast observed with SST compares favorably with that of other telescopes. Simultaneously with the above wideband red-beam data, we also recorded narrow-band continuum images with the CRISP imaging spectropolarimeter. We find that contrasts measured with CRISP are entirely consistent with the corresponding wide-band contrasts, demonstrating that any additional image degradation by the CRISP etalons and telecentric optical system is marginal or even insignificant.

Ellerman bombs and UV bursts are transient brightenings that are ubiquitously observed in the lower atmospheres of active and emerging flux regions. Here we present inversion results of SST/CRISP and CHROMIS, as well as IRIS data of such transient events. Combining information from the Mg II h & k, Si IV and Ca II 8542A and Ca II H & K lines, we aim to characterise their temperature and velocity stratification, as well as their magnetic field configuration. We find average temperature enhancements of a few thousand kelvin close to the classical temperature minimum, but localised peak temperatures of up to 10,000-15,000 K from Ca II inversions. Including Mg II generally dampens these temperature enhancements to below 8000 K, while Si IV requires temperatures in excess of 10,000 K at low heights, but may also be reproduced with secondary temperature enhancements of 35,000-60,000 K higher up. However, reproducing Si IV comes at the expense of overestimating the Mg II emission. The line-of-sight velocity maps show clear bi-directional jet signatures and strong correlation with substructure in the intensity images, with slightly larger velocities towards the observer than away. The magnetic field parameters show an enhancement of the horizontal field co-located with the brightenings at similar heights as the temperature increase. We are thus able to largely reproduce the observational properties of Ellerman bombs with UV burst signature with temperature stratifications peaking close to the classical temperature minimum. Correctly modelling the Si IV emission in agreement with all other diagnostics is, however, an outstanding issue. Accounting for resolution differences, fitting localised temperature enhancements and/or performing spatially-coupled inversions is likely necessary to obtain better agreement between all considered diagnostics.

We present data of 10 penumbral microjets (PMJs) observed in H\alpha, Ca II 8542 \AA, and Fe I 6302 \AA line pair with the Swedish 1 m Solar Telescope (SST) with CRISP and Ca II K with SST/CHROMIS in active region NOAA 12599 on the 12th October 2016 at \mu=0.68. All four Stokes parameters of the Ca II 8542 \AA and Fe I 6302 \AA lines were observed and a series of test pixels was inverted using the Stockholm inversion code. Our analysis revealed for the first time that PMJs are visible in H\alpha, where they appear as dark features with average line-of-sight (LOS) upflows of 1.1\pm0.6 km/s, matching the LOS velocities from the inversions. Based on the H\alpha observations we extend the previous average length and lifetime of PMJs to 2815\pm530 km and 163\pm25 s, respectively. The plane-of-sky (POS) velocities of our PMJs of up to 17 km/s tend to give increased velocities with distance travelled. Furthermore, two of our PMJs with significant Stokes V signal indicate that the PMJs possess an increased LOS magnetic field of up to 100 G compared to the local pre-/post- PMJ magnetic field, which propagates as quickly as the PMJs' POS velocities. Finally, we present evidence that PMJs display an on average 1 minute gradual precursory brightening that only manifests itself in the cores of the Ca II lines. We conclude that PMJs are not ordinary jets but likely are manifestations of heat fronts that propagate at the local Alfven velocity.

We study the polarization signals and physical parameters of penumbral microjets (PMJs) by using high spatial resolution data taken in the Fe I 630 nm pair, Ca II 854.2 nm and Ca II K lines with the CRISP and CHROMIS instruments at the Swedish 1-m Solar Telescope. We infer their physical parameters, such as physical observables in the photosphere and chromospheric velocity diagnostics, by different methods, including inversions of the observed Stokes profiles with the STiC code. PMJs harbor overall brighter Ca II K line profiles and conspicuous polarization signals in Ca II 854.2 nm, specifically in circular polarization that often shows multiple lobes mainly due to the shape of Stokes I. They usually overlap photospheric regions with sheared magnetic field configuration, suggesting that magnetic reconnections could play an important role in the origin of PMJs. The discrepancy between their low LOS velocities and the high apparent speeds reported on earlier, as well as the existence of different vertical velocity gradients in the chromosphere, indicate that PMJs might not be entirely related to mass motions. Instead, PMJs could be due to perturbation fronts induced by magnetic reconnections occurring in the deep photosphere that propagate through the chromosphere. This reconnection may be associated with current heating that produces temperature enhancements from the temperature minimum region. Furthermore, enhanced collisions with electrons could also increase the coupling to the local conditions at higher layers during the PMJ phase, giving a possible explanation for the enhanced emission in the overall Ca II K profiles emerging from these transients.

CHROMIS, a new imaging spectrometer at the Swedish 1-m Solar Telescope (SST), can observe the chromosphere in the H and K lines of Ca II at high spatial and spectral resolution. Accurate modeling as well as an understanding of the formation of these lines are needed to interpret the SST/CHROMIS observations. Such modeling is computationally challenging because these lines are influenced by strong departures from local thermodynamic equilibrium, three-dimensional radiative transfer, and partially coherent resonance scattering of photons. We aim to model the CaII H&K lines in 3D model atmospheres to understand their formation and to investigate their diagnostic potential for probing the chromosphere. We model the synthetic spectrum of Ca II using the radiative transfer code Multi3D in three different radiation-magnetohydrodynamic model atmospheres computed with the Bifrost code. We classify synthetic intensity profiles according to their shapes and study how their features are related to the physical properties in the model atmospheres. We investigate whether the synthetic data reproduce the observed spatially-averaged line shapes, center-to-limb variation and compare with SST/CHROMIS images. The spatially-averaged synthetic line profiles show too low central emission peaks, and too small separation between the peaks. The trends of the observed center-to-limb variation of the profiles properties are reproduced by the models. The Ca II H&K line profiles provide a temperature diagnostic of the temperature minimum and the temperature at the formation height of the emission peaks. The Doppler shift of the central depression is an excellent probe of the velocity in the upper chromosphere.

Context. The radiative losses in the solar chromosphere vary from 4~kW~m$^{-2}$ in the quiet Sun, to 20~kW~m$^{-2}$ in active regions. The mechanisms that transport non-thermal energy to and deposit it in the chromosphere are still not understood. Aims. We aim to investigate the atmospheric structure and heating of the solar chromosphere in an emerging flux region. Methods. We use observations taken with the CHROMIS and CRISP instruments on the Swedish 1-m Solar Telescope in the Ca II K, Ca II 854.2 nm, H$\alpha$, and Fe I 630.1 nm and 630.2 nm lines. We analyse the various line profiles and in addition perform multi-line, multi-species, non-Local Thermodynamic Equilibrium (non-LTE) inversions to estimate the spatial and temporal variation of the chromospheric structure. Results. We investigate which spectral features of Ca II K contribute to the frequency-integrated Ca II K brightness, which we use as a tracer of chromospheric radiative losses. The majority of the radiative losses are not associated with localized high-Ca II K-brightness events, but instead with a more gentle, spatially extended, and persistent heating. The frequency-integrated Ca II K brightness correlates strongly with the total linear polarization in the Ca II 854.2 nm line, while the Ca II K profile shapes indicate that the bulk of the radiative losses occur in the lower chromosphere. Non-LTE inversions indicate a transition from heating concentrated around photospheric magnetic elements below $\log{\tau_{500}} =-3$ to a more space-filling and time-persistent heating above $\log{\tau_{500}} =-4$. The inferred gas temperature at $\log{\tau_{500}} =-3.8$ correlates strongly with the total linear polarization in the Ca II 854.2 nm line, suggesting that that the heating rate correlates with the strength of the horizontal magnetic field in the low chromosphere.

Magnetic reconnection is thought to drive a wide variety of dynamic phenomena in the solar atmosphere. Yet the detailed physical mechanisms driving reconnection are difficult to discern in the remote sensing observations that are used to study the solar atmosphere. In this paper we exploit the high-resolution instruments Interface Region Imaging Spectrograph (IRIS) and the new CHROMIS Fabry-Perot instrument at the Swedish 1-m Solar Telescope (SST) to identify the intermittency of magnetic reconnection and its association with the formation of plasmoids in so-called UV bursts in the low solar atmosphere. The Si IV 1403A UV burst spectra from the transition region show evidence of highly broadened line profiles with often non-Gaussian and triangular shapes, in addition to signatures of bidirectional flows. Such profiles had previously been linked, in idealized numerical simulations, to magnetic reconnection driven by the plasmoid instability. Simultaneous CHROMIS images in the chromospheric Ca II K 3934A line now provide compelling evidence for the presence of plasmoids, by revealing highly dynamic and rapidly moving brightenings that are smaller than 0.2 arcsec and that evolve on timescales of order seconds. Our interpretation of the observations is supported by detailed comparisons with synthetic observables from advanced numerical simulations of magnetic reconnection and associated plasmoids in the chromosphere. Our results highlight how subarcsecond imaging spectroscopy sensitive to a wide range of temperatures combined with advanced numerical simulations that are realistic enough to compare with observations can directly reveal the small-scale physical processes that drive the wide range of phenomena in the solar atmosphere.

We discuss a recent paper by Schlichenmaier & Franz (SF; 2013, A&A, 555, A84), in which the claim is made that the penumbral dark downflows detected for the first time with the Swedish 1-m Solar Telescope (SST) by Scharmer et al. and Joshi et al. could be produced by overcompensation for straylight. We show that the analysis of SF is fundamentally flawed, because it ignores the constraints on the strength of such straylight from 3D convection simulations and on the spatial extent of the straylight point spread function from the measured minimum intensity in the sunspot umbra. Furthermore, we show that the claim made by SF, that the spatial straylight of Hinode is less than 10%, is false. We conclude that the analysis of SF is of no relevance in relation to the straylight compensation method applied to the SST data.

We discuss NICOLE inversions of Fe I 630.15 nm and 630.25 nm Stokes spectra from a sunspot penumbra recorded with the CRISP imaging spectropolarimeter on the Swedish 1-m Solar Telescope at a spatial resolution close to 0.15". We report on narrow radially extended lanes of opposite polarity field, located at the boundaries between areas of relatively horizontal magnetic field (the intra-spines) and much more vertical field (the spines). These lanes harbor convective downflows of about 1 km/s. The locations of these downflows close to the spines agree with predictions from the convective gap model (the "gappy penumbra") proposed six years ago, and more recent 3D MHD simulations. We also confirm the existence of strong convective flows throughout the entire penumbra, showing the expected correlation between temperature and vertical velocity, and having vertical RMS velocities of about 1.2 km/s.

Context. Recent discoveries of intensity correlated downflows in the interior of a sunspot penumbra provide direct evidence for overturning convection, adding to earlier strong indications of convection from filament dynamics observed far from solar disk center, and supporting recent simulations of sunspots. Aims. Using spectropolarimetric observations obtained at a spatial resolution approaching 0'.'1 with the Swedish 1-m Solar Telescope (SST) and its spectropolarimeter CRISP, we investigate whether the convective downflows recently discovered in the C i line at 538.03 nm can also be detected in the wings of the Fe i line at 630.15 nm Methods. We make azimuthal fits of the measured LOS velocities in the core and wings of the 538 nm and 630 nm lines to disentangle the vertical and horizontal flows. To investigate how these depend on the continuum intensity, the azimuthal fits are made separately for each intensity bin. By using spatially high-pass filtered measurements of the LOS component of the magnetic field, the flow properties are determined separately for magnetic spines (relatively strong and vertical field) and inter-spines (weaker and more horizontal field). Results. The dark convective downflows discovered recently in the 538.03 nm line are evident also in the 630.15 nm line, and have similar strength. This convective signature is the same in spines and inter-spines. However, the strong radial (Evershed) outflows are found only in the inter-spines. Conclusions. At the spatial resolution of the present SST/CRISP data, the small-scale intensity pattern seen in continuum images is strongly related to a convective up/down flow pattern that exists everywhere in the penumbra. Earlier failures to detect the dark convective downflows in the interior penumbra can be explained by inadequate spatial resolution in the observed data.

Recently measured straylight PSFs in Hinode/SOT make granulation contrast in observed data and synthetic MHD data consistent. Data from earthbound telescopes also need accurate correction for straylight and fixed optical aberrations. We aim to develop a method for measuring straylight in the post-focus imaging optics of the SST. We removed any influence from atmospheric turbulence and scattering by using an artificial target. We measured integrated straylight from three different sources in the same data: ghost images caused by reflections in the near-detector optics, PSFs corresponding to wavefront aberrations in the optics by using phase diversity, and extended scattering PSF wings of unknown origin by fitting to a number of different kernels. We performed the analysis separately in the red and blue beams. Wavefront aberrations, which possibly originate in the bimorph mirror of the adaptive optics, are responsible for a wavelength-dependent straylight of 20-30% of the intensity in the form of PSFs with 90% of the energy contained within a radius of 0.6". There are ghost images that contribute at the most a few % of straylight. The fraction of other sources of scattered light from the post-focus instrumentation of the SST is only \sim10^-3 of the recorded intensity. This contribution has wide wings with FWHM \sim16" in the blue and \sim34" in the red. The present method seems to work well for separately estimating wavefront aberrations and the scattering kernel shape and fraction. Ghost images can be expected at the same level for solar observations. The high-order wavefront aberrations possibly caused by the AO bimorph mirror dominate the measured straylight but are likely to change when imaging the Sun. We can therefore make no firm statements about the origin of straylight in SST data, but strongly suspect wavefront aberrations to be the dominant source.

Observations with the 1-m Swedish Solar Telescope of the flows seen in penumbral filaments are presented. Time sequences of bright filaments show overturning motions strikingly similar to those seen along the walls of small isolated structures in the active regions. The filaments show outward propagating striations with inclination angles suggesting that they are aligned with the local magnetic field. We interpret it as the equivalent of the striations seen in the walls of small isolated magnetic structures. Their origin is then a corrugation of the boundary between an overturning convective flow inside the filament and the magnetic field wrapping around it. The outward propagation is a combination of a pattern motion due to the downflow observed along the sides of bright filaments, and the Evershed flow. The observed short wavelength of the striation argues against the existence of a dynamically significant horizontal field inside the bright filaments. Its intensity contrast is explained by the same physical effect that causes the dark cores of filaments, light bridges and `canals'. In this way striation represents an important clue to the physics of penumbral structure and its relation with other magnetic structures on the solar surface. We put this in perspective with results from the recent 3-D radiative hydrodynamic simulations.

In this work, we study and quantify properties of strong-field small-scale convection and compare observed properties with those predicted by numerical simulations. We analyze spectropolarimetric 630.25 nm data from a unipolar ephemeral region near sun center. We use line-of-sight velocities and magnetic field measurements obtained with Milne-Eddington inversion techniques along with measured continuum intensities and Stokes V amplitude asymmetry at a spatial resolution of 0.15 arcseconds to establish statistical relations between the measured quantities. We also study these properties for different types of distinct magnetic features, such as micropores, bright points, ribbons, flowers and strings. We present the first direct observations of a small-scale granular magneto-convection pattern within extended regions of strong (more than 600 G average) magnetic field. Along the boundaries of the flux concentrations we see mostly downflows and asymmetric Stokes V profiles, consistent with synthetic line profiles calculated from MHD simulations. We note the frequent occurrence of bright downflows along these boundaries. In the interior of the flux concentrations, we observe an up/down flow pattern that we identify as small-scale magnetoconvection, appearing similar to that of field-free granulation but with scales 4 times smaller. Measured RMS velocities are 70% of those of nearby field-free granulation, even though the average radiative flux is not reduced. The interiors of these flux concentrations are dominated by upflows.

Context. For accurately measuring intensities and determining magnetic field strengths of small-scale solar (magnetic) structure, knowledge of and compensation for the point spread function is crucial. For images recorded with the Swedish 1-meter Solar Telescope, restoration with Multi-Frame Blind Deconvolution and Joint Phase Diverse Speckle methods lead to remarkable improvements in image quality but granulation contrasts that are too low, indicating additional stray light. Aims. We propose a method to compensate for stray light from high-order atmospheric aberrations not included in MFBD and JPDS processing. Methods. To compensate for uncorrected aberrations, a reformulation of the image restoration process is proposed that allows the average effect of hundreds of high-order modes to be compensated for by relying on Kolmogorov statistics for these modes. The applicability of the method requires simultaneous measurements of Fried's parameter r0. The method is tested with simulations as well as real data and extended to include compensation for conventional stray light. Results. We find that only part of the reduction of granulation contrast in SST images is due to uncompensated high-order aberrations. The remainder is still unaccounted for and attributed to stray light from the atmosphere, the telescope with its re-imaging system and to various high-altitude seeing effects. Conclusions. We conclude that statistical compensation of high-order modes is a viable method to reduce the loss of contrast occurring when a limited number of aberrations is explicitly compensated for with MFBD and JPDS processing. We show that good such compensation is possible with only 10 recorded frames. The main limitation of the method is that already MFBD and JPDS processing introduces high-order compensation that, if not taken into account, can lead to over-compensation.

To evaluate site quality and to develop multi-conjugative adaptive optics systems for future large solar telescopes, characterization of contributions to seeing from heights up to at least 12 km above the telescope is needed. We describe a method for evaluating contributions to seeing from different layers along the line-of-sight to the Sun. The method is based on Shack Hartmann wavefront sensor data recorded over a large field-of-view with solar granulation and uses only measurements of differential image displacements from individual exposures, such that the measurements are not degraded by residual tip-tilt errors. We conclude that the proposed method allows good measurements when Fried's parameter r_0 is larger than about 7.5 cm for the ground layer and that these measurements should provide valuable information for site selection and multi-conjugate development for the future European Solar Telescope. A major limitation is the large field of view presently used for wavefront sensing, leading to uncomfortably large uncertainties in r_0 at 30 km distance.

Numerical simulations have by now revealed that the fine scale structure of the penumbra in general and the Evershed effect in particular is due to overturning convection, mainly confined to gaps with strongly reduced magnetic field strength. The Evershed flow is the radial component of the overturning convective flow visible at the surface. It is directed outwards -- away from the umbra -- because of the broken symmetry due to the inclined magnetic field. The dark penumbral filament cores visible at high resolution are caused by the 'cusps' in the magnetic field that form above the gaps. Still remaining to be established are the details of what determines the average luminosity of penumbrae, the widths, lengths, and filling factors of penumbral filaments, and the amplitudes and filling factors of the Evershed flow. These are likely to depend at least partially also on numerical aspects such as limited resolution and model size, but mainly on physical properties that have not yet been adequately determined or calibrated, such as the plasma beta profile inside sunspots at depth and its horizontal profile, the entropy of ascending flows in the penumbra, etc.

Whereas penumbral models during the last 15 years have been successful in explaining Evershed flows and magnetic field inclination variations in terms of flux tubes, the lack of contact between these models and a convective process needed to explain the penumbral radiative heat flux has been disturbing. We report on recent observational and theoretical evidence that challenge flux tube interpretations and conclude that the origin of penumbral filamentary structure is overturning convection.

We discuss penumbral fine structure in a small part of a pore, observed with the CRISP imaging spectropolarimeter at the Swedish 1-m Solar Telescope (SST), close to its diffraction limit of 0.16 arcsec. Milne-Eddington inversions applied to these Stokes data reveal large variations of field strength and inclination angle over dark-cored penumbral intrusions and a dark-cored light bridge. The mid-outer part of this penumbra structure shows 0.3 arcsec wide spines, separated by 1.6 arcsec (1200 km) and associated with 30 deg inclination variations. Between these spines, there are no small-scale magnetic structures that easily can be be identified with individual flux tubes. A structure with nearly 10 deg more vertical and weaker magnetic field is seen midways between two spines. This structure is co-spatial with the brightest penumbral filament, possibly indicating the location of a convective upflow from below.

We discuss a numerical 3D radiation-MHD simulation of penumbral fine structure in a small sunspot. This simulation shows the development of short filamentary structures with horizontal flows, similar to observed Evershed flows, and an inward propagation of these structures at a speed compatible with observations. We conclude that the Evershed flow represents the horizontal flow component of overturning convection in gaps with strongly reduced field strength. The top of the flow is always directed outward--away from the umbra-- because of the broken symmetry due to the inclined magnetic field. Upflows occur in the inner parts of the gaps and most of the gas turns over radially (outwards and sideways), and descends back down again. The ascending, cooling and overturning flow tends to bend magnetic field lines down, forcing a weakening of the field that makes it easier for gas located in an adjacent layer--further in--to initiate a similar sequence of motion, aided by lateral heating, thus causing the inward propagation of the filament.

We present results of numerical 3D MHD simulations with radiative energy transfer of fine structure in a small sunspot of about 4 Mm width. The simulations show the development of filamentary structures and flow patterns that are, except for the lengths of the filaments, very similar to those observed. The filamentary structures consist of gaps with reduced field strength relative to their surroundings. Calculated synthetic images show dark cores like those seen in the observations; the dark cores are the result of a locally elevated $\tau=1$ surface. The magnetic field in these cores is weaker and more horizontal than for adjacent brighter structures, and the core support a systematic outflow. Movies show migration of the dark-cored structures towards the umbra, and fragments of magnetic flux that are carried away from the spot by a large scale `moat flow'. We conclude that the simulations are in qualitative agreement with observed penumbra filamentary structures, Evershed flows and moving magnetic features.

We present numerical 2D magnetostatic models for sunspot penumbrae consisting of radially aligned field-free gaps in a potential magnetic field, as proposed by Spruit and Scharmer (2006). The shape of the gaps and the field configurations around them are computed consistently from the condition of magnetostatic pressure balance between the gap and the magnetic field. The results show that field-free gaps in the {\it inner} penumbra are cusp-shaped and bounded by a magnetic field inclined by about $70^\circ$ from the vertical. Here, the magnetic component has a Wilson depression on the order 200--300 km relative to the top of the field-free gap; the gaps should thus appear as noticeably elevated features. This structure explains the large variations in field strength in the inner penumbra inferred from magnetograms and two-component inversions, and the varying appearance of the inner penumbra with viewing angle. In the {\it outer} penumbra, on the other hand, the gaps are flat-topped with a horizontal magnetic field above the middle of the gap. The magnetic field has large inclination variations horizontally, but only small fluctuations in field strength, in agreement with observations.

We interpret penumbral filaments as due to convection in field-free, radially aligned gaps just below the visible surface of the penumbra, intruding into a nearly potential field above. This solves the classical discrepancy between the large heat flux and the low vertical velocities observed in the penumbra. The presence of the gaps causes strong small-scale fluctuations in inclination, azimuth angle and field strength, but without strong forces acting on the gas. The field is nearly horizontal in a region around the cusp-shaped top of the gap, thereby providing an environment for Evershed flows. We identify this region with the recently discovered dark penumbral cores. Its darkness has the same cause as the dark lanes in umbral light-bridges, reproduced in numerical simulations by Nordlund and Stein (2005). We predict that the large vertical and horizontal gradients of the magnetic field inclination and azimuth in the potential field model will produce the net circular polarization seen in observations. The model also explains the significant elevation of bright filaments above their surroundings. It predicts that dark areas in the penumbra are of two different kinds: dark filament cores containing the most inclined (horizontal) fields, and regions between bright filaments, containing the least inclined field lines.

We present the first center-to-limb G-band images synthesized from high resolution simulations of solar magneto-convection. Towards the limb the simulations show "hilly" granulation with dark bands on the far side, bright granulation walls and striated faculae, similar to observations. At disk center G-band bright points are flanked by dark lanes. The increased brightness in magnetic elements is due to their lower density compared with the surrounding intergranular medium. One thus sees deeper layers where the temperature is higher. At a given geometric height, the magnetic elements are cooler than the surrounding medium. In the G-band, the contrast is further increased by the destruction of CH in the low density magnetic elements. The optical depth unity surface is very corrugated. Bright granules have their continuum optical depth unity 80 km above the mean surface, the magnetic elements 200-300 km below. The horizontal temperature gradient is especially large next to flux concentrations. When viewed at an angle, the deep magnetic elements optical surface is hidden by the granules and the bright points are no longer visible, except where the "magnetic valleys" are aligned with the line of sight. Towards the limb, the low density in the strong magnetic elements causes unit line-of-sight optical depth to occur deeper in the granule walls behind than for rays not going through magnetic elements and variations in the field strength produce a striated appearance in the bright granule walls.