Interstellar dust is omnipresent in our galaxy, but can falsify the observation of distant sky objects. Our view of them falls through a dirty window. It is all the more important to know these dust effects as precisely as possible. Now astronomers have created a detailed 3D card for the first time, which documented the dust distribution and features in the Milky Way and the Magellan clouds as detailed as ever. This enables the light absorption to be determined more precisely than before. This was made possible by 130 million stellar light spectra of the European GAIA mission and a good one million spectra of llamost patterns in China. With the help of an artificial intelligence, the astronomers were able to create their 3D card of the dustic in the dust. Her analyzes also produced new, sometimes surprising results for the properties of the interstellar dust itself.
When astronomers observe distant sky objects and measure their red shift, they must take into account a possible source of error. Because how reddish a star appears depends not only on its distance, direction of movement and cosmic expansion. Interstellar dust clouds that cross the star light also influence the spectrum. This dust makes distant cosmic objects appear reddish and dampens their brightness – however, both characteristics are important for the distance determination. For precise observations, astronomers therefore need to know how much dust is between them and their distant observation goals.
Gaia, llamost and an artificial intelligence
A new 3D card now helps to better assess the interface effects of interstellar dust. Xiangyu Zhang and Gregory Green from the Max Planck Institute for Astronomy in Heidelberg first evaluated the light spectra of around 130 million stars in the Milky Way and the Magellan clouds, which were measured by the European Space Telescope GAIA in the past ten years as part of its sky patterns. It is known from previous analyzes and models that the extent of the dust effect can be determined in such spectra based on a specific, wavelength-dependent absorption. Because cosmic dust particles absorb more with shorter wavelengths and less strongly with longer wavelengths. This wavelength dependency results in a specific “extinction curve”, the shape of which provides information about the amount and composition of the dust. So far, however, the information taken from the Gaia data about this dust effect in the local cosmos has been inaccurate because the Gaia spectra are low.
Zhang and Green have now solved this problem with the help of another data source and artificial intelligence. As they stated, around one percent of the stars analyzed by GAIA was also recorded by the lamost pattern of the Chinese national observatories. For these stars, this provided higher resolution and therefore finer light spectra, on which the characteristics of the stars and the dust -related extinction curve are easy to read. Therefore, the astronomers used this data to train a neuronal network. The task of the AI system was to recognize laws, with the help of which the exact dust-extracting curve can also be determined for the coarser Gaia spectra. “We therefore trained the model with 2.4 million star-type-specific parameters from the lamost data set,” the researchers explain. On this basis, the AI system learned to create realistic model spectra for different stars and dust properties. As a result, the astronomers were able to use the AI to reconstruct the extinction curve for the remaining 130 million Gaia spectra.
Extinction card and an unexpected finding
The result is a map that shows the distribution of the interstellar dust and its effects on astronomical observations for the Milky Way and the Magellan Clouds as detailed. “Earlier cards from the dust extinction assumed a constant value for R (V), which in our estimation caused a systematic uncertainty of around ten percent,” write Zhang and Green. “Our 3D card indicates the R (V) value for the foreground dust as a function of removal along any visual line in the Milky Way.” This now provides far more precise values. However, the new results also showed surprising. So far, astronomers have assumed that the extinction curve in heavenly areas flattened with higher dust density. The absorption effect of the dust is therefore less dependent on the wavelength in denser clouds. The reason for this is the increased proportion of slightly larger dust grains in dust clouds. This changes the absorption properties of those regions.
But the mapping now published shows something else: According to the new data, the extinction curve in areas with an increased but not maximum dust -density is steeper. The dust in these areas therefore absorbs star light increasingly in shorter wavelengths than for longer. Such zones can be found in the Milky Way, among other things, in the starry areas of the Scorpius and Ophuchus region as well as near other star weighing. According to Zhang and Green’s assumptions, the unexpectedly strong wave length dependency in these areas could arise from different dust sizes, but perhaps also by certain molecules in the cold molecular clouds of these areas.
Source: Xiangyu Zhang and Gregory M. Green (Max Planck Institute for Astronomy, Heidelberg), Science, Doi: 10.1126/science.ado9787