Final Report Summary - HER-SFR (Star, gas and dust in star forming regions: studying Star Formation Rate tracers at small galactic scales)
The main goal of the project has been to perform a detailed multi-wavelength study of HII regions located in galaxies of different types, star formation rates (SFR) and environments. We want to understand better the relation between the stellar population and the dust and gas components within the star forming regions. As the amount of stars formed in a galaxy correlates with important properties of the galaxy, being able to estimate accurately the SFR is crucial to understand the evolution of galaxies. The study of the interstellar dust is improving considerably nowadays mainly due to the scientific contribution of three space missions: mid-far IR SPITZER (http://ssc.spitzer.caltech.edu) (IRAC; 3.6-8μm and MIPS 24-160μm), UV GALEX (Far-UV; 1350-1750AA and Near-UV; 1750-2750AA, and the most recent European missions HERSCHEL (PACS; 70-160μm and SPIRE; 250- 500μm) (http://www.rssd.esa.int/Herschel) and PLANCK (350μm-1.4mm). These satellite missions, together with observations in the visible from terrestrial telescopes, have produced the most complete spectral energy distributions (SEDs) of galaxies up to now allowing the study of the stellar, gas and dust components in the widest wavelength range and the best resolution ever.
The project has been carried out to a large extent in the nearby spiral galaxy, M33, which hosts a significant number of HII regions covering a wide range of luminosities and sizes and for which the required observations are available. It has been complemented with a study of two low-metallicity dwarf galaxies: NGC 1569 and NGC 4214 to cover a range in metallicities and environments. The project was divided into two primary goals. I briefly summarise them here and show the main results achieved during the project.
1.- Properties of the interstellar dust in a sample of HII regions with different morphologies and environments.
The new IR bands proposed as SFR calibrators are based on the strength of statistical correlations between the emissions in these bands and the Halpha emission, the classical SFR indicator. Considerable analysis has been carried out from a statistical point of view, but little has been done to compare the spatial distribution of the proposed SFR tracers and the location where the stars actually form within HII emitting knots (e.g. Relano & Kennicutt 2009). It is then necessary to study the spatial relation between the dust, gas and stars within the HII regions.
We have extracted the SED of a set of HII Regions in M33 covering a wide range of Halpha luminosities, sizes and different morphologies (compact, shells and mixed). We have looked for trends in the shape of the SED of HII regions and the morphology of the objects. We have found that the shells tend to emit less at the 24μm band and that the IR peak of the SED, which is related to the temperature of the dust within the HII region, is located at longer wavelengths for shells than for compact and mixed regions. Moreover, the logarithmic 100 μm/70 μm ratio for filled and mixed regions remains constant over one order of magnitude in Halpha and FUV surface brightness, while the shells and clear shells exhibit a wider range of values of almost two orders of magnitude. This could mean that for filled and mixed objects, the dust is so close to the stars within the regions that it is very efficiently heated and reaches a very well defined, narrow range of temperature, independently of the radiation field intensity. For shells and clear shells, other parameters may affect the dust-heating mechanism. It could probably be the location of the dust relative to the stars or the evolutionary state of the stellar population within the region that may lead to a dispersion in the correlation at low-intensity radiation fields.
We have modelled the individual SED of the HII regions in M33 with DUSTEM (Compiegne et al. 2011), which is an updated version of the classical dust model from Desert et al. (1990). The model includes three different type of grains: (i) large (radius 0.4 to 1 nm) polycyclic aromatic hydrocarbons (PAHs), (ii) carbonaceous nanoparticles (radius 1 to 10 nm) called very small grains (VSGs), and (iii) big grains (BGs, 10 to 100 nm). In the diffuse interstellar medium, small grains (PAHs and VSGs) undergo temperature spikes triggered by the absorption of stellar photons and cool by emission in the near and mid-IR range. Conversely, BGs, which have a longer cooling time and a shorter timescale between absorption of two photons because of their size, stay at constant temperature and emit like grey bodies. With DUSTEM we have obtained the mass fraction (relative to the hydrogen mass) of the different type of grains for each individual HII region in our sample. We found that the regions with shell morphology have a lower mass fraction of VSGs compared to mixed and compact regions. For these regions, shocks might be very important and break the BGs into VSGs very efficiently. A pixel-by-pixel SED fitting for the most luminous HII regions in M33 shows that the mass fraction of VSGs is higher in places close to the location of the stellar clusters, where the effect of shocks produced by the interaction of the stellar winds and the interstellar medium (ISM) can be very important.
2.- Effect of the leakage of ionising radiation in SFR measurements.
In the calibration used to convert the Halpha luminosity to the SFR it is generally assumed that every Lyman continuum photon emitted by the stars results in the ionisation of a hydrogen atom. However, if there is leakage of Lyman continuum photons from the star forming region, the observed Halpha flux would be lower than expected and the true SFR based on Halpha emission would be underestimated. Quantifying the fraction of ionising photons leaking from HII regions will be crucial in understanding better: a) the different results for the SFR measurements based on UV and Halpha indicators, and b) the heating of the dust associated with the diffuse ISM. Whether the bulk of the diffuse dust emission comes from dust heated by evolved stellar populations (and therefore not associated to current SFR) or by radiation escaped from star forming regions is the subject of a long debate (Lonsdale et al. 1987, Sauvage &Thuan 1992, Calzetti et al. 2010 and others) and represents one of the most important uncertainties in quantifying the star formation rate for whole galaxies. Therefore, disentangling the different heating sources of the dust is of particular importance to quantify the star formation both in nearby and distant galaxies.
We have found observational evidence that leakage of ionising photons from star-forming regions can affect the quantification of the star formation rate (SFR) in galaxies: leakage could decrease the SFR(Halpha)/SFR(FUV) ratio by up to 25 per cent. The evidence is based on the observation that the SFR(Halpha)/SFR(FUV) ratio is lower for objects showing a shell Halpha structure than for regions exhibiting a much more compact morphology. Although leakage cannot entirely explain the observed trend of SFR(Halpha)/SFR(FUV) ratios for systems with low SFR, we show that the mechanism can lead to different SFR estimates when using Halpha and FUV luminosities. Therefore, further study is needed to constrain the contribution of leakage to the low SFR(Halpha)/SFR(FUV) ratios observed in dwarf galaxies and its impact on the Halpha flux as an SFR indicator in such objects.
We have modelled the SED of the dwarf galaxy NGC 4214, separating the dust emission components: HII regions (plus their associated Photodissociation regions (PDRs) on pc scales) and the diffuse dust (on kpc scales). We analyse the full UV to FIR/submm SED of the galaxy using a radiation transfer model which self-consistently treats the dust emission from diffuse and star-forming complexes components, considering the illumination of diffuse dust both by the distributed stellar populations, and by escaping light from the HII regions. While maintaining consistency with the framework of this model we additionally use a model that provides a detailed description of the dust emission from the HII regions and their surrounding PDRs. We achieve a satisfactory fit for the emission from HII+PDR regions on pc scales, with the exception of the emission at 8 μm, which is underpredicted by the model. For the diffuse emission we achieve a good fit if we assume that about 30-70% of the emission leaking the HII+PDR regions is able to leave the galaxy without passing through a diffuse ISM.
The final aim of the research project has been to get a better understanding of the role of dust in star-forming places: how it is distributed within the regions and how it is affected by the radiation field and the stellar winds coming from the stellar population that ionises the gas. We have made significant progress in this field: (i) we know now that the relative location of the stars, gas and dust affect the heating mechanism of the dust withing the star-forming regions, and therefore caution has to be taken when applying the dust emission as a SFR tracer. (ii) We have quantified the effect of leakage of ionising photons in the SFR estimations, showing that they can be affected up to 25%. Moreover, modelling the whole SED for a dwarf galaxy shows that a significant fraction of the stellar radiation is leaving the galaxy without heating the dust within it. Both results are crucial to quantify star formation in galaxies and therefore have implications in understanding the way massive star formation affects galaxy evolution.
The project has been carried out to a large extent in the nearby spiral galaxy, M33, which hosts a significant number of HII regions covering a wide range of luminosities and sizes and for which the required observations are available. It has been complemented with a study of two low-metallicity dwarf galaxies: NGC 1569 and NGC 4214 to cover a range in metallicities and environments. The project was divided into two primary goals. I briefly summarise them here and show the main results achieved during the project.
1.- Properties of the interstellar dust in a sample of HII regions with different morphologies and environments.
The new IR bands proposed as SFR calibrators are based on the strength of statistical correlations between the emissions in these bands and the Halpha emission, the classical SFR indicator. Considerable analysis has been carried out from a statistical point of view, but little has been done to compare the spatial distribution of the proposed SFR tracers and the location where the stars actually form within HII emitting knots (e.g. Relano & Kennicutt 2009). It is then necessary to study the spatial relation between the dust, gas and stars within the HII regions.
We have extracted the SED of a set of HII Regions in M33 covering a wide range of Halpha luminosities, sizes and different morphologies (compact, shells and mixed). We have looked for trends in the shape of the SED of HII regions and the morphology of the objects. We have found that the shells tend to emit less at the 24μm band and that the IR peak of the SED, which is related to the temperature of the dust within the HII region, is located at longer wavelengths for shells than for compact and mixed regions. Moreover, the logarithmic 100 μm/70 μm ratio for filled and mixed regions remains constant over one order of magnitude in Halpha and FUV surface brightness, while the shells and clear shells exhibit a wider range of values of almost two orders of magnitude. This could mean that for filled and mixed objects, the dust is so close to the stars within the regions that it is very efficiently heated and reaches a very well defined, narrow range of temperature, independently of the radiation field intensity. For shells and clear shells, other parameters may affect the dust-heating mechanism. It could probably be the location of the dust relative to the stars or the evolutionary state of the stellar population within the region that may lead to a dispersion in the correlation at low-intensity radiation fields.
We have modelled the individual SED of the HII regions in M33 with DUSTEM (Compiegne et al. 2011), which is an updated version of the classical dust model from Desert et al. (1990). The model includes three different type of grains: (i) large (radius 0.4 to 1 nm) polycyclic aromatic hydrocarbons (PAHs), (ii) carbonaceous nanoparticles (radius 1 to 10 nm) called very small grains (VSGs), and (iii) big grains (BGs, 10 to 100 nm). In the diffuse interstellar medium, small grains (PAHs and VSGs) undergo temperature spikes triggered by the absorption of stellar photons and cool by emission in the near and mid-IR range. Conversely, BGs, which have a longer cooling time and a shorter timescale between absorption of two photons because of their size, stay at constant temperature and emit like grey bodies. With DUSTEM we have obtained the mass fraction (relative to the hydrogen mass) of the different type of grains for each individual HII region in our sample. We found that the regions with shell morphology have a lower mass fraction of VSGs compared to mixed and compact regions. For these regions, shocks might be very important and break the BGs into VSGs very efficiently. A pixel-by-pixel SED fitting for the most luminous HII regions in M33 shows that the mass fraction of VSGs is higher in places close to the location of the stellar clusters, where the effect of shocks produced by the interaction of the stellar winds and the interstellar medium (ISM) can be very important.
2.- Effect of the leakage of ionising radiation in SFR measurements.
In the calibration used to convert the Halpha luminosity to the SFR it is generally assumed that every Lyman continuum photon emitted by the stars results in the ionisation of a hydrogen atom. However, if there is leakage of Lyman continuum photons from the star forming region, the observed Halpha flux would be lower than expected and the true SFR based on Halpha emission would be underestimated. Quantifying the fraction of ionising photons leaking from HII regions will be crucial in understanding better: a) the different results for the SFR measurements based on UV and Halpha indicators, and b) the heating of the dust associated with the diffuse ISM. Whether the bulk of the diffuse dust emission comes from dust heated by evolved stellar populations (and therefore not associated to current SFR) or by radiation escaped from star forming regions is the subject of a long debate (Lonsdale et al. 1987, Sauvage &Thuan 1992, Calzetti et al. 2010 and others) and represents one of the most important uncertainties in quantifying the star formation rate for whole galaxies. Therefore, disentangling the different heating sources of the dust is of particular importance to quantify the star formation both in nearby and distant galaxies.
We have found observational evidence that leakage of ionising photons from star-forming regions can affect the quantification of the star formation rate (SFR) in galaxies: leakage could decrease the SFR(Halpha)/SFR(FUV) ratio by up to 25 per cent. The evidence is based on the observation that the SFR(Halpha)/SFR(FUV) ratio is lower for objects showing a shell Halpha structure than for regions exhibiting a much more compact morphology. Although leakage cannot entirely explain the observed trend of SFR(Halpha)/SFR(FUV) ratios for systems with low SFR, we show that the mechanism can lead to different SFR estimates when using Halpha and FUV luminosities. Therefore, further study is needed to constrain the contribution of leakage to the low SFR(Halpha)/SFR(FUV) ratios observed in dwarf galaxies and its impact on the Halpha flux as an SFR indicator in such objects.
We have modelled the SED of the dwarf galaxy NGC 4214, separating the dust emission components: HII regions (plus their associated Photodissociation regions (PDRs) on pc scales) and the diffuse dust (on kpc scales). We analyse the full UV to FIR/submm SED of the galaxy using a radiation transfer model which self-consistently treats the dust emission from diffuse and star-forming complexes components, considering the illumination of diffuse dust both by the distributed stellar populations, and by escaping light from the HII regions. While maintaining consistency with the framework of this model we additionally use a model that provides a detailed description of the dust emission from the HII regions and their surrounding PDRs. We achieve a satisfactory fit for the emission from HII+PDR regions on pc scales, with the exception of the emission at 8 μm, which is underpredicted by the model. For the diffuse emission we achieve a good fit if we assume that about 30-70% of the emission leaking the HII+PDR regions is able to leave the galaxy without passing through a diffuse ISM.
The final aim of the research project has been to get a better understanding of the role of dust in star-forming places: how it is distributed within the regions and how it is affected by the radiation field and the stellar winds coming from the stellar population that ionises the gas. We have made significant progress in this field: (i) we know now that the relative location of the stars, gas and dust affect the heating mechanism of the dust withing the star-forming regions, and therefore caution has to be taken when applying the dust emission as a SFR tracer. (ii) We have quantified the effect of leakage of ionising photons in the SFR estimations, showing that they can be affected up to 25%. Moreover, modelling the whole SED for a dwarf galaxy shows that a significant fraction of the stellar radiation is leaving the galaxy without heating the dust within it. Both results are crucial to quantify star formation in galaxies and therefore have implications in understanding the way massive star formation affects galaxy evolution.