The much-vaunted “perfect fit” curve published by John Mather et al in 1991 allegedly shows the exact alignment of theory and observation in the Microwave Background Radiation. It is indeed a wonderfully precise match, the result of years of intense scrutiny. In private correspondence, my friend and helmsman Professor Paul Jackson shared the experience. ”I remember following with excitement the build-up to that figure, which happened over many years from 1967.  Wobbly non-satellite data plots first emerged with huge error bars. The eventual super-accurate FIRAS results of COBE had error bars so small that they had to be multiplied by 400 to be visible on the plot.”  

The much-vaunted "prediction-matches-observation" plot made famous by John Mather.  Both the prediction and the observation are controversial.

 Sometimes it’s pure happenstance that grants one an insight to workings of another man’s craft. Shortly before he died in 2009, my friend and advisor Tony Bray nonchalantly changed the whole complexion of my understanding when he said, “Here, take a look at this.” In his hand was a print of a paper by specialist professor of radiology, Pierre-Marie Robitaille. It’s called COBE: A Radiological Analysis. It turns out that this paper is a vital link in the chain of argument surrounding the way that the microwave radiation is measured, and equally importantly, how patterns are extracted from the fuzz.

It was from COBE measurements that Dr Mather drew his graph, and as in the case of cosmological redshift, it turns out an entire body of theory was verified by patently flawed data. COBE carried three instruments. One of them, the Far Infrared Absolute Spectrometer (FIRAS), was designed to do nought but measure temperature and compare the radiation spectrum with that of a precise blackbody.

The diagram shows the incredibly precise blackbody spectrum supposedly measured by FIRAS (plotted as intensity against frequency). However, that’s all it does—it traces a perfect blackbody curve—and as we shall soon see, that in itself does not say anything meaningful about Big Bang Theory or expansion. Dr Robitaille, an acknowledged and acclaimed authority in the field of radiation measurement, gives an intensive appraisal of the design failures of that particular instrument, leading to the shocking revelation that methods employed by the COBE team were so flawed as to be utterly misleading.

Please allow me to quote rather extensively from this paper (with sincere apologies for the technical complexity). This is where my drill hits the nerve. Firstly, Professor Robitaille deals with the inadequacies of the FIRAS instrument itself (I added the bold type emphasis):

“Data released from FIRAS has been met with nearly universal admiration. However, a thorough review of the literature reveals significant problems with this instrument. FIRAS was designed to function as a differential radiometer, wherein the sky signal could be nulled by the reference horn, Ical. The null point occurred at an Ical temperature of 2.759 K. This was 34 mK above the reported sky temperature, 2.725 +/- 0.001 K, a value where the null should ideally have formed. In addition, an 18 mK error existed between the thermometers in Ical, along with a drift in temperature of ~3 mK. A 5 mK error could be attributed to Xcal; while a 4 mK error was found in the frequency scale. A direct treatment of all these systematic errors would lead to a ~64 mK error bar in the microwave background temperature. The FIRAS team reported ~1 mK, despite the presence of such systematic errors. But a 1 mK error does not properly reflect the experimental state of this spectrophotometer. In the end, all errors were essentially transferred into the calibration files, giving the appearance of better performance than actually obtained […] Neglecting to fully evaluate FIRAS prior to the mission, the FIRAS team attempts to do so, on the ground, in highly limited fashion, with a duplicate Xcal, nearly 10 years after launch … Despite popular belief to the contrary, COBE has not proven that the microwave background originates from the universe and represents the remnants of creation.”

Dr Robitaille goes into considerable and intimate detail on aspects of the FIRAS instrument which render its measurements fraught with inaccuracies, but it would be sensible to keep this précis fairly brief. Next, he makes a revealing summary of the blackbody issue:

“One hundred and fifty years have now passed, since Kirchhoff first advanced the law upon which the validity of the microwave background temperature rests. His law of thermal emission stated that radiation, at equilibrium with the walls of an enclosure, was always black, or normal. This was true in a manner independent of the nature of the enclosure. Kirchhoff’s law was so powerful that it would become the foundation of contemporary astrophysics. By applying this formulation, the surface temperatures of all the stars could be evaluated, with the same ease as measuring the temperature of a brick-lined oven… However, since blackbody radiation only required enclosure and was independent of the nature of the walls, Planck did not link this process to a specific physical cause. For astrophysics, this meant that any object could produce a blackbody spectrum. All that was required was mathematics and the invocation of thermal equilibrium. Even the requirement for enclosure was soon discarded. Processes occurring far out of equilibrium, such as the radiation of a star, and the alleged expansion of the universe, were thought to be suitable candidates for the application of the laws of thermal emission. To aggravate the situation, Kirchhoff had erred in his claim of universality. In actuality, blackbody radiation was not universal. It was limited to an idealized case which, at the time, was best represented by graphite, soot, or carbon black. Nothing on Earth has been able to generate the elusive blackbody over the entire frequency range and for all temperatures.”

This casts further material doubt upon the validity of the blackbody prediction, but right now, we are naturally anxious to hear about the invisible error bars. Was the match between theory and observation really that marvellously precise?

“Despite the presence of systematic errors, the FIRAS team is able to essentially sidestep the recordings of their thermometers and overcome their inaccuracy. […] (They) present a dozen values for the microwave background temperature, using varying methods. This occurs over a span of 13 years. Each time, there is a striking recalculation of error bars. In the end, the final error on the microwave background temperature drops by nearly two orders of magnitude from 60 mK to 0.65 mK. Yet, as will be seen below … FIRAS was unable to yield proper nulls ... Despite the subsequent existence of systematic errors, the FIRAS team minimizes error bars. […] Relative to error bars, the result obtained, using an average of many methods was analogous to ignoring the existence of known temperature error in the reference calibrators Xcal and Ical. The existence of imperfect nulls was also dismissed, as were all interferograms obtained while the Earth was directly illuminating FIRAS… It is well established, not only in physics, but across the sciences, that systematic errors can be extremely difficult, even impossible, to detect. Consequently, one must not dismiss those systematic errors which are evident … This treatment would discount attempts to lower the error bar to 1 mK in the final FIRAS report.”

The stage is now set for the coup de grace, an exposé of the precarious vulnerability of the famous Mather plot when its credentials are examined. After listing 15 serious problems with the COBE instrumentation, Dr Robitaille brings it to focus in a single devastating sentence, which I have emphasised in bold:

“Given the systematic errors on Xcal, Ical, the frequency drift, and the null temperature, it is reasonable to ascertain that the FIRAS microwave background temperature has a significant error bar. As such, an error on the order of 64 mK represents a best case scenario, especially in light of the dismissal/lack of data at low frequency. The report of a microwave temperature of 2.725 +/- 0.001 K does not accurately reflect the extent of the problems with the FIRAS instrument. Furthermore, the absolute temperature of the microwave background will end up being higher than 2.725 K, when measured without the effect of diffraction, and when data below 2 cm^-1 is included. Contrary to popular belief, the FIRAS instrument did not record the most perfect blackbody spectrum in the history of science. In the end, the methods to process the anisotropy maps are likely to be ‘creating anisotropy’ where none previously existed”

This list could go on and on; the bottom line is that if we graciously allow that the standard model makes at least some sense—invisible “dark sense”, perhaps?—then we must concede that the CMB makes much more sense as the limiting temperature of space heated by ambient starlight and radiation from astrophysical structures, including even the Earth itself, than the signature of a hypothesised primordial explosion.