Cosmic Background Explorer

The Differential Microwave Radiometer (DMR) investigation uses three differential radiometers to map the sky at 31.4, 53, and 90 GHz. The radiometers are distributed around the outer surface of the cryostat. Each radiometer employs a pair of horn antennas viewing at 30° from the spin axis of the spacecraft, measuring the differential temperature between points in the sky separated by 60°. At each frequency there are two channels for dual polarization measurements for improved sensitivity and for reliability. Each radiometer is a microwave receiver whose input is switched rapidly between the two horn antennas, obtaining the difference in brightness of two fields of view 7° in diameter located 60° apart and 30° from the axis of the spacecraft. High sensitivity is achieved by temperature stabilization (at 300 K for 31.4 GHz and at 140 K for 53 and 90 GHz), by spacecraft spin, and by the ability to integrate over the entire year. Sensitivity to large-scale anisotropies is about 3E-5 K. The instrument weighs 120 kg (260 lb), uses 114 watts, and has a data rate of 500 bit/s.[7]

The Diffuse Infrared Background Experiment (DIRBE) consists of a cryogenically cooled (to 2 K) multiband radiometer used to investigate diffuse infrared radiation from 1 to 300 micrometres. The instrument measures the absolute flux in 10 wavelength bands with a 1° field of view pointed 30° off the spin axis. Detectors (photoconductors) and filters for the 8 to 100 micrometre channels are the same as for the IRAS mission. Bolometers are used for the longest wavelength channel (120 to 300 micrometres). The DIRBE sensitivity will be better than 2E-12 W/(cm² sr) in channels 1 to 3. Channels 4 to 8 will reach 6E-13 while channels 9 and 10, with their less sensitive bolometers but larger etendue, will reach 4E-12. These limits are achievable with existing detectors cooled to near the cryostat temperature of 1.6 K. The telescope is a well baffled, off-axis, Gregorian flux collector with re-imaging. The instrument weighs approximately 34 kg (75 lb), uses 100 W and has a data rate of 1700 bit/s.[8]

The Far Infrared Absolute Spectrophotometer (FIRAS) is a cryogenically cooled polarizing Michelson interferometer used as a Fourier transform spectrometer. The instrument points along the spin axis and has a 7° field of view. This device measures the spectrum to a precision of 1/1000 of the peak flux at 1.7 mm (0.067 in) for each 7° field of view on the sky (over the range 0.1 to 10 mm (0.39 in)). The FIRAS uses a special flared trumpet horn flux collector having very low sidelobe levels and an external calibrator covering the entire beam; precise temperature regulation and calibration are required. The instrument has a differential input to compare the sky with an internal reference at 3 K. This feature provides immunity from systematic errors in the spectrometer, and contributes significantly to the ability to detect small deviations from a blackbody spectrum. The instrument weighs 60 kg (130 lb), uses 84 watts and has a data rate of 1200 bit/s.[9]

Data from COBE showed a perfect fit between the black body curve predicted by big bang theory and that observed in the microwave background.

Comparison of CMB results from COBE, WMAP and Planck - 21 March 2013.

During the 15-year-long period between the proposal and launch of COBE, there were two significant astronomical developments:

• First, in 1981, two teams of astronomers, one led by David Wilkinson of Princeton University and the other by Francesco Melchiorri of the University of Florence, simultaneously announced that they detected a quadrupole distribution of CMB using balloon-borne instruments. This finding would have been the detection of the black-body distribution of CMB that FIRAS on COBE was to measure. In particular, the Florence group claimed a detection of intermediate angular scale anisotropies at the level 100 microkelvins[14] in agreement with later measurements made by the BOOMERanG experiment. However, a number of other experiments attempted to duplicate their results and were unable to do so.[6]
• Second, in 1987 a Japanese-American team led by Andrew E. Lange and Paul Richards of University of California, Berkeley and Toshio Matsumoto of Nagoya University made an announcement that CMB was not that of a true black body.[15] In a sounding rocket experiment, they detected an excess brightness at 0.5 and 0.7 mm (0.028 in) wavelengths.

With these developments serving as a backdrop to COBE's mission, scientists eagerly awaited results from FIRAS. The results of FIRAS were startling in that they showed a perfect fit of the CMB and the theoretical curve for a black body at a temperature of 2.7 K, thus proving the Berkeley-Nagoya results erroneous.

FIRAS measurements were made by measuring the spectral difference between a 7° patch of the sky against an internal black body. The interferometer in FIRAS covered between 2- and 95-cm−1 in two bands separated at 20-cm−1. There are two scan lengths (short and long) and two scan speeds (fast and slow) for a total of four different scan modes. The data were collected over a ten-month period.[16]

Data obtained at each of the three DMR frequencies — 31.5, 53, and 90 GHz — following dipole subtraction.

The DMR was able to spend four years mapping the detectable anisotropy of cosmic background radiation as it was the only instrument not dependent on the dewar's supply of helium to keep it cooled. This operation was able to create full sky maps of the CMB by subtracting out galactic emissions and dipole at various frequencies. The cosmic microwave background fluctuations are extremely faint, only one part in 100,000 compared to the 2.73 K average temperature of the radiation field. The cosmic microwave background radiation is a remnant of the Big Bang and the fluctuations are the imprint of density contrast in the early universe. The density ripples are believed to have produced structure formation as observed in the universe today: clusters of galaxies and vast regions devoid of galaxies.[17]

DIRBE also detected 10 new far-IR emitting galaxies in the region not surveyed by IRAS as well as nine other candidates in the weak far-IR that may be spiral galaxies. Galaxies that were detected at the 140 and 240 μm were also able to provide information on very cold dust (VCD). At these wavelengths, the mass and temperature of VCD can be derived. When these data were joined with 60 and 100 μm data taken from IRAS, it was found that the far-infrared luminosity arises from cold (≈17–22 K) dust associated with diffuse H I region cirrus clouds, 15-30% from cold (≈19 K) dust associated with molecular gas, and less than 10% from warm (≈29 K) dust in the extended low-density H II regions.[18]

Model of the Galactic disk as seen edge-on from Earth's position.

On top of the findings DIRBE had on galaxies, it also made two other significant contributions to science.[18] The DIRBE instrument was able to conduct studies on interplanetary dust (IPD) and determine if its origin was from asteroid or cometary particles. The DIRBE data collected at 12, 25, 50 and 100 μm were able to conclude that grains of asteroidal origin populate the IPD bands and the smooth IPD cloud.[19]

The second contribution DIRBE made was a model of the Galactic disk as seen edge-on from our position. According to the model, if the Sun is 8.6 kpc from the Galactic Center, then it is 15.6% above the midplane of the disk, which has a radial and vertical scale lengths of 2.64 and 0.333 kpc, respectively, and is warped in a way consistent with the HI layer. There is also no indication of a thick disk.[20]

To create this model, the IPD had to be subtracted out of the DIRBE data. It was found that this cloud, which as seen from Earth is Zodiacal light, was not centered on the Sun, as previously thought, but on a place in space a few million kilometers away. This is due to the gravitation influence of Saturn and Jupiter.[6]

In addition to the science results detailed in the last section, there are numerous cosmological questions left unanswered by COBE's results. A direct measurement of the extragalactic background light (EBL) can also provide important constraints on the integrated cosmological history of star formation, metal and dust production, and the conversion of starlight into infrared emissions by dust.[21]

By looking at the results from DIRBE and FIRAS in the 140 to 5000 μm we can detect that the integrated EBL intensity is ≈16 nW/(m²·sr). This is consistent with the energy released during nucleosynthesis and constitutes about 20–50% of the total energy released in the formation of helium and metals throughout the history of the universe. Attributed only to nuclear sources, this intensity implies that more than 5–15% of the baryonic mass density implied by big bang nucleosynthesis analysis has been processed in stars to helium and heavier elements.[21]

There were also significant implications into star formation. COBE observations provide important constraints on the cosmic star formation rate, and help us calculate the EBL spectrum for various star formation histories. Observation made by COBE require that star formation rate at redshifts of z ≈ 1.5 to be larger than that inferred from UV-optical observations by a factor of 2. This excess stellar energy must be mainly generated by massive stars in yet - undetected dust enshrouded galaxies or extremely dusty star forming regions in observed galaxies.[21] The exact star formation history cannot unambiguously be resolved by COBE and further observations must be made in the future.

On 30 June 2001, NASA launched a follow-up mission to COBE led by DMR Deputy Principal Investigator Charles L. Bennett. The Wilkinson Microwave Anisotropy Probe has clarified and expanded upon COBE's accomplishments. Following WMAP, the European Space Agency's probe, Planck has continued to increase the resolution at which the background has been mapped.[22][23]