sapienza asi cardiff cea chalmers iaps ifac ingv inpe neel  nnsu wits

   OLIMPO: a mm and submm telescope on a stratospheric balloon

Home Page
Science Instrument Collaboration Papers Photos

Scientific Program:

A different way of using the Cosmic Microwave Background


Compared to the sensitivity of current surveys, the CMB is a bright background light. OLIMPO uses this background light to detect aggregated matter at all stages of structure formation in the Universe. 

Most of the matter in the Universe is transparent for CMB photons, which have important interactions only with molecules (resonant scattering & absorption) and ionized gas (Thomson scattering & Inverse Compton scattering).

· Molecules form very late and mainly in galaxies, so occupy a very small solid angle; moreover their interaction with the CMB is masked by dominant galactic emission.

· Ionized gas, instead, is present at all scales, including the largest structures: clusters, superclusters, filaments. So we expect to be able to map the ionized matter in the universe observing the fine-scale anisotropy of the CMB induced by the ionized gas.

The preferred sites where ionized gas can be measured are clusters of galaxies. The potential well of the cluster is so deep that gas falling into the well heats up to millions of K : kT ~ 10 keV. For this reason galaxy clusters are powerful X-ray emitters.

The same gas scatters CMB photons (inverse Compton, also known as the Sunyaev-Zeldovich effect, SZ). Crossing a rich cluster, a photon of the CMB has about 1% likelihood of being scattered, and in case of scattering the energy gain from the hot electron is about 1%. So we expect a 0.01% effect in the temperature change of the CMB in the direction of clusters of galaxies. This is a large anisotropy, compared to the intrinsic anisotropy of the CMB, which is about one order of magnitude smaller.

The brightness change of the CMB has a characteristic spectral signature (see figure), due to the fact that the number of photons is conserved while the energy is shifted high. This has been studied in detail in Sunyaev R., Zeldovich Y.B., 1972, Comm. Astrophys. Space Phys., 4, 173 ;  Birkinshaw M., 1999, Physics Reports, 310, 97-195 ;  Rephaeli Y., 1995, Ann.Rev.A.A., 33, 541 ; J. E. Carlstrom, et al., 2002, Ann.Rev.A.A., 40, 643.

SZ spectra 

olimpo observing the universe

The OLIMPO telescope, spectrometer, and cryogenic detection system (left) and a simulation of the sky maps obtained during the observation of one rich cluster (right). Image Credit : Giuseppe D’Alessandro.

Studying the SZ with OLIMPO


The spectral coverage of OLIMPO has been optimized to match the relevant features of the SZ spectrum. The 145 GHz band will monitor the CMB brightness decrease, the 210 GHz band will be a null monitor, the 345 and 480 GHz channels will monitor the brightness increase and possibile dust contamination.

Of these bands, only the lowest one can be efficiently observed with ground-based telescopes.  OLIMPO complements, at high frequency and with similar angular resolution, the SZ surveys carried out by 10m-class ground-based SZ telescopes, like SPT and ACT.

OLIMPO features a low-resolution spectrometer (see paper): the sky brightness is measured in many (about 30) independent bands. This opens new possibilities. For example the spectral data within the so-called null band around 220 GHz can be used to measure the slope of the spectrum and the exact zero-crossing frequency, which depends on the absolute temperature of the CMB. Moreover, the measurements at high frequency are sensitive to the temperature of the ionized gas, via the relativistic corrections. A simulation of the typical spectral measurements quality is reported here.

We plan to observe a sample of 40 clusters, spectrally and photometrically, for each flight of the instrument. For each cluster, a set of 30 images at different frequencies will be obtained (see figure above), i.e. a data-cube with excellent information content, allowing for efficient separation of different components, and removing degeneracy in the estimate of cluster parameters (more info here ) .