Our Galaxy is pervaded by strong magnetic fields, which, together withthe ordinary matter (made of gas and dust) and the cosmic rays,form the three basic constituents of the interstellar medium.After briefly reviewing our current observational knowledge aboutGalactic magnetic fields, I will address, from a theoretical point of view, the fundamental questionsof their origin and their impact on the interstellar medium.


Total magnetic fields in spiral galaxies, as observed through their total synchrotronemission, are strongest (up to ≲30 µG) in the spiral arms. The degree of radio polarization is low; the field in the arms must be mostly turbulent or tangled. Polarized synchrotronemission shows that the resolved regular fields are generally strongest in the interarm regions (up to ≲15 µG), sometimes forming “magnetic arms” parallel to the optical arms. The field structure is spiral in almost every galaxy, even in flocculent and bright irregular types which lack spiralarms. The observed large-scale patterns of Faraday rotation in several massive spiral galaxies reveal coherent regular fields, as predicted by dynamo models. However, in most galaxies observed so far no simple patterns of Faraday rotation could be found. Either many dynamo modes are superimposed and cannot be resolved by present-day telescopes, or most of the apparently regular field is in fact anisotropic random, with frequent reversals, due to shearing and compressing gas flows. In galaxies with massive bars, the polarization pattern follows the gas flow. However, around strong shocks in bars, the compression of theregular field is much lower than that of the gas; the regular field decouples from the cold gas and is strong enough to affect the flow of the diffuse warm gas. – The average strength of the total magnetic field in the Milky Way is 6µG near the sun and increases to 20–40 μG in the Galactic center region.The Galactic field is mostly parallel to the plane, except in the center region. Rotation measure data from pulsars indicate several field reversals, unlike external galaxies, but some reversals could be due to distortions of thenearby field.

Observations of magnetic fields in molecular clouds are essential for understanding their role in the evolution of dense clouds and in the star formation process. The two extreme-case models of star formation are the weak magnetic field scenario, in which turbulence drives molecular cloud evolution and therefore star formation, and the strong magnetic field scenario, with magnetic fields governing molecular cloud formation and ambipolar diffusion ultimately driving star formation. Continuum and spectral line polarization observations make it possible to infer magnetic field strengths and morphologies, and therefore to test which of the two star formation scenarios dominates. The observational techniques available for such observations – the Zeeman effect, mapping of polarized dust emission, and mapping of linearly polarized line emission – are briefly described, together with how these observations can test star formation theory. Examples of the existing observations and the current state of the testing of magnetically controlled star formation are presented. Finally, the prospects for the future of study of magnetic fields in molecular clouds and their role in the star formation process is discussed.

The diffuse (non self-gravitating) interstellar medium of the Galaxy is almost impossibly complex and diverse. Temperatures and densities range over many orders of magnitude. The magnetic field links all regimes of the gas, including cosmic rays. Shocks from supernovae and other sources buffet the medium. Into this maelstrom, theorists have ventured. Some models emphasize time-independent thermal equilibrium between hot and cold phases of the gas. Other models stress frequent dynamical events that throw much of the medium out of thermal equilibrium. Some models include magnetic field effects, others do not. Here we report on the nature of diffuse atomic gas, emphasizing observational results from a recent, extensive survey of HI emission and absorption along random lines of sight in the local diffuse medium. We find much of the warmer diffuse atomic gas is out of thermal equilibrium, yet the medium retains a clear dichotomy between warmer and cooler phases. The warmer phase comprises about half of the total diffuse atomic hydrogen gas. Magnetic fields have a median value of about 6 μG in the cold gas, insuring that their dynamical effects cannot be ignored. The conundrum of similar magnetic field strengths in diffuse gas at widely disparate densities remains as an observational fact and a challenge to explain theoretically.

The cold diffuse interstellar medium (ISM) is turbulent withsupersonic, most likely transAlfvénic motions. All its tracers,including velocity, are structured down to the smallest scalesaccessible to observations. This structure, though, is not random andcan be quantified using the statistics of its Fourier phaseincrements. The quantity of structure of interstellar clouds is largerthan that of fractal Brownian motion (fBm) fields and of mildlycompressible turbulence. We propose that this structure is a responseof the medium to a general property of turbulence, the space-timeintermittency of its dissipation. The bursts of turbulentdissipation, because they temporarily (over a few hundred years) andlocally heat the diffuse medium up to ~103 K, irreversiblymodify the subsequent thermal and chemical evolution of the gas.Preliminary models of such bursts predict chemical abundances andexcitation of a subset of molecules that are in agreement with theobservations. Timescales of the velocity fluctuations are shortenough to locally decouple the neutrals from the ionized species inthe diffuse medium and generate helical magnetic fields. Small scalefluctuations of the polarization properties of the diffuse ISM aretherefore expected.

So far our understanding of the evolution of the ISM has beenscanty, because of the inherent nonlinearity of all the processesinvolved. Modelling the interstellar medium in galaxies in aself-consistent way requires an approach that must take into accountthe relevant scales, which may cover several orders of magnitude(e.g., from kpc to less than 1 pc). This is a difficult task thatrequires the use of sophisticated numerical codes, adequatecomputing power, and precision input data from observations. Thelarge range of scales can be resolved by means of adaptive meshrefinement (AMR), that is, the grids reproducing the computationaldomain are refined on the fly such that a minimum number of cells isneeded to provide the most complete description of the flow.
In this paper we review the most recent results on 3D modeling ofthe interstellar medium and disk-halo interaction in a section ofthe Milky Way that includes the Galactic magnetic field, backgroundheating due to starlight, self-gravity and allows for theestablishment of the duty-cycle between the disk and halo (commonlyknown as galactic fountain) by using a grid that extends up to 10kpc on either side of the midplane. Our simulations capture both thelargest structures (e.g., superbubbles) together with the smallerones (e.g., filaments and eddies) down to 0.625 pc. We investigate,among other things, the variability of the magnetic field in theGalactic disk and its correlation with the density, the rôle ofram pressure in the dynamics of disk gas and the relative weight ofthe ram, thermal and magnetic pressures, the mass distribution andthe volume filling factors of the different temperature regimes inthe ISM, as well as the scales at which energy is injected into theinterstellar turbulence and we give an estimate for the dimension ofthe most dissipative structures.

Recent radio polarization observations have revealed a plethora ofunexpected features in the polarized Galactic radio background thatarise from propagation effects in the random (turbulent) interstellarmedium. The canals are especially striking among them, a random networkof very dark, narrow regions clearly visible in many directions againsta bright polarized Galactic synchrotron background. There are no obviousphysical structures in the ISM that may have caused the canals, and sothey have been called Faraday ghosts. They evidently carry informationabout interstellar turbulence but only now is it becoming clear how thisinformation can be extracted. Two theories for the origin of the canalshave been proposed; both attribute the canals to Faraday rotation, butone invokes strong gradients in Faraday rotation in the sky plane(specifically, in a foreground Faraday screen) and the other only relieson line-of-sight effects (differential Faraday rotation). In this reviewwe discuss the physical nature of the canals and how they can be used toexplore statistical properties of interstellar turbulence. This opensstudies of magnetized interstellar turbulence to new methods ofanalysis, such as contour statistics and related techniques ofcomputational geometry and topology. In particular, we can hope tomeasure such elusive quantities as the Taylor microscale and theeffective magnetic Reynolds number of interstellar MHD turbulence.

After a quick tour of polarization basics, I review interstellar polarization starting with its galactic distribution, and then going into molecular clouds. The properties of interstellar polarization are discussed, its wavelength dependence and the 3 parameters used to specify it. Then more recent work since approximately the last 15 years which show significant deviations from the usual behavior in the densest regions of molecular clouds is presented. Recent results for two specific regions are presented, and finally I conclude by summarizing the main points.


Observations of far-infrared (FIR) and submillimeter (SMM) polarizedemission are used to study magnetic fields and dust grains in denseregions on the interstellar medium (ISM). These observations placeconstraints on models of molecular clouds, star-formation, grainalignment mechanisms, and grain size, shape, and composition.
The FIR/SMM polarization is strongly dependent onwavelength. We have attributed this wavelength dependence to samplingdifferent grain populations at different temperatures. To date, mostobservations of polarized emission have been inthe densest regions of the ISM. Extending these observations toregions of the diffuse ISM, and to microwave frequencies, will provideadditional tests of grain and alignment models.
An understanding of polarized microwave emission from dust is key toan accurate measurement of the polarization of the cosmic microwavebackground. The microwave polarization spectrum will put limits on thecontributions to polarized emission from spinning dust and vibratingmagnetic dust.


Several of the current and next-generation cosmic microwave background(CMB) experiments have polarimetric capability, promising to add tothe finesse of precision cosmology. One of the contaminating Galacticforegrounds is thermal emission by dust. Since optical interstellarpolarization is commonly seen, from differential extinction by alignedaspherical dust particles, it is expected that thermal emission fromthese grains will be polarized. Indeed, in the Galactic plane and indark (molecular) clouds, dust emission in the infrared andsubmillimetre has been measured to be polarized. It seems likely thatthe faint diffuse cirrus emission, of more relevance to CMBexperiments, will be polarized too. We discuss how well the amount ofpolarization of this component can be predicted, making use of what isknown about optical (and infrared and ultraviolet) interstellarpolarization and extinction. Some constraints on the alignment of thecarrier of the dust-correlated anomalous microwave emission can bemade as well.


Future cosmology space missions will concentrate on measuring thepolarization of the Cosmic Microwave Background (CMB), which potentiallycarries invaluable information about the earliest phases of the evolutionof our universe. Such ambitious projects will ultimately be limited bythe sensitivity of the instrument and by the accuracy at whichpolarized foreground emission from our own Galaxy can be subtractedout. We present the PILOT balloon project which will aim atcharacterizing one of these foreground sources, the polarization ofthe dust continuum emission in the diffuse interstellar medium. ThePILOT experiment will also constitute a test-bed for usingmultiplexed bolometer arrays for polarization measurements.


In this short review, I try to give a flavor of what are the motivationsand difficulties of observing the Cosmic MicrowaveBackground (CMB)polarization. After describing briefly the physics of CMB polarization,I give a snapshot of the observational status, as well as the dataprocessing techniques developed so far. I end up describing how polarizationinformation should constrain some key aspects of the early universephysics.


This contribution reviews the most recent experimental resultsobtained on the polarization anisotropy of the Cosmic MicrowaveBackground and those anticipated in the near future. Some basicarguments for a dedicated satellite are also presented, as well as theneed for a better characterization of polarized Galactic foregrounds.


Planck is designed to image the anisotropies of the Cosmic MicrowaveBackground (CMB) over the whole sky, with unprecedented sensitivity (ΔT/T ~ 2 × 10-6) and angular resolution (~5 arcmin). Planckwill provide a major source of information relevant to several cosmologicaland astrophysical issues, such as testing theories of the early universe andthe origin of cosmic structure. The ability to measure to high accuracy theangular power spectrum of the CMB fluctuations will allow the determination offundamental cosmological parameters with an uncertainty of order a fewpercent. In addition to the main cosmological goals of the mission, the Plancksky survey will be used to study in detail the very sources of emission which“contaminate” the signal due to the CMB. This will result in a wealth ofinformation on the properties of extragalactic sources, and on the dust andgas in our own galaxy. One specific notable result will be the measurement ofthe polarized emission from our own Galaxy. The ability of Planck to measurepolarization across a wide frequency range (30–350 GHz), with high precisionand accuracy, and over the whole sky, will provide unique insight into theproperties of the interstellar medium. An overview is presented of the Planckmission, its scientific objectives, the key elements of its technical design,and its current status.