Cosmology at the Beginning of a New Millennium
By Ronald Ebert
(Published in Skeptic Magazine, volume 8, number 3, 2000)
One of the most basic questions people have asked from the earliest times is where did it all come from? Religions have been invented in part to answer this question, but today we use science to discover the truth about the world. In recent times, science has discovered a great deal about the origin of the universe but all the questions are not yet answered. Today discoveries in this field are rapid and they often change the way we view our theories.
The Big Bang wins over its competitors
In the mid-twentieth century there were several competing theories for the origin and evolution of the universe. The Big Bang was based on three key observations. The most basic was that the universe was expanding. Spectral light of distant galaxies showed a shift of key features toward the longer, red end of the spectrum when compared to the spectral light of elements generated in a laboratory here on the earth. The consensus of opinion was and still is that the movement of distant galaxies away from the earth causes this redshift. It is very similar to the Doppler shift heard as a train passes you and moves away as it blows its whistle. What's more, for galaxies that were apparently twice as far away based on their brightness, the redshift was twice as great. The discovery of this remarkable linear relationship, now known as Hubble's law, supplied evidence for the notion of an expanding universe. In such a universe, every galaxy, or group of gravitationally bound galaxies, moves away from every other. If we project such a universe back in time, the universe is significantly hotter and denser than it is now.
This realization lead to two other predictions. Early in its history the universe must have been hot enough for a burst of nuclear reactions to occur. The proportions of light elements and their isotopes that would have been formed in this nucleosynthesis era were calculated and later verified by observations.
It was also realized that the early universe would have been so hot that matter would have been ionized - free electrons and protons as well as bare atomic nuclei would have been present. Photons, particles of light, would not have been able to travel far before colliding into other particles. (In fact the photons and free charged particles formed a sort of fluid.)
The universe at this time is opaque, similar to the surface of the sun today. But as the universe expands over time, it cooled. It was calculated that about 300,000 years after the beginning, the temperature
dropped low enough for the particles to combine into atoms. This caused a drastic drop in free charged particles, allowing photons to travel reasonably long distances before they collided with charged particles. The universe became transparent, and the light that was released expanded and cooled with the expansion of the universe. Today this radiation should be present in the microwave region of the radio spectrum. We have detected this radiation and we now call it the cosmic microwave background radiation (CMB).
The redshifts of galaxies, the proportions of light elements, and the CMB are the three key pieces of evidence that validate the Big Bang. But the Big Bang has had its detractors. The two main competitors were the steady state theory (its proponents later modified it and called it the quasi-steady state theory) and plasma cosmology. In its original form, the steady state model proposed that the universe is not only isotropic (the same in all directions) and homogenous in space but in time as well. A steady state universe has no beginning or end. It is infinitely old, and as it expands a continuous creation of matter is required to maintain the average density of the universe that we see today. Its proponents claimed that nucleosynthesis took place in stars rather than an early, hot dense universe. But with this mechanism, the calculated proportion of helium is different. What's more, the steady state universe does not predict a background radiation, and distant parts of the universe should look the same as nearby parts.
Observations have verified the Big Bang model and falsified the steady state one. The proportion of helium matches the predictions of the Big Bang. The CMB has been detected. And when we observe the distant universe we note many differences compared with the nearby universe, such as the existence of quasars only in the distant universe.
In the quasi-steady state model, the universe expands only in a limited way and many distant redshifts are the result of processes other than an expanding universe. The redshifts of quasars in particular are not due to their recession velocity in an expanding universe, but rather it's a Doppler shift caused by their ejection from the cores of galaxies. But if this were true, we should see many blue-shifted quasars, as many would be ejected in our direction by chance. No blue-shifted quasars are seen.
The other major competitor, plasma cosmology, proposes a non-expanding universe with redshifts that are due to processes other than expansion, a universe of half matter and half antimatter, and plasma rather than gravity as the dominant shaper of matter in the universe. But as with the steady state models, its predictions have not been borne out by observations. Very little antimatter has been found, what there is is seen as parts of jets from the cores of some galaxies. And gravity, not plasma, is dominant on the largest scales.
Recent observations of distant supernovas have falsified the contention that redshifts are not caused by an expanding universe. The light curves of distant supernovas, where light intensity versus time are plotted, show a relativistic time dilation compared to light curves of nearby supernovas.1 This can only happen in an expanding universe.
Plasma cosmology and the steady state theories proposed alternative mechanisms for redshifts. These included the Wolf effect and tired light. In the Wolf effect, the phases of the sources are supposed to combine in a way that results in a redshift. But what one sees depends critically on the alignment of the observer with the sources. With a different alignment, one should see a blueshift instead. Chance should provide half redshifts and half blueshifts. But very few galactic blueshifts are seen.
Tired light simply proposes that light loses energy through some unknown mechanism as it travels through space, and this results in the observed redshift. But no time dilation would be seen through a tired light redshift, so this too is falsified.
A strange beginning
If the equations of general relativity are used to project the universe back in time as far as possible, all the universe - matter, energy, space and time, gets smaller and smaller until it has finally shrunk to an object of literally zero size. When we've gone back in time that far, not only has the whole of space contracted to a mathematical point - time, which, in Einstein's theory of relativity is intimately linked to space, has itself come to an end. We've reached the border of spacetime itself, a weird object that physicists call a singularity. According to relativity, here time stops, and it is no more sensible to ask what came before it than it is to ask what is north of the North Pole.
However, this is not the whole picture. Twentieth century physics has not only given us Einstein's theory of gravity, general relativity, but also a revolutionary new description of the microscopic world: quantum theory. In a situation like the earliest phases of the big bang, we must expect quantum theory to play as big a part as gravity theory -- what happens there should be governed by a theory of quantum gravity, a theory that should encompass quantum theory and general relativity.
Unfortunately, formulating quantum gravity has proven very difficult and the problem of quantum gravity is basically unsolved. So we have no reliable description of what really happened when our universe was curled up as tiny as
10-35 meters, as predicted by the big bang model. Possibly, we will meet, after a finite time, some quantum version of the classical big bang singularity, although nobody knows what such a "quantum singularity" would look like. Theorists who are working to develop quantum gravity suggest there may be alternative scenarios, such as a "no beginning universe". This would stretch into negative time as well as positive time, a little like the positive and negative numbers on a thermometer, eternally existing with no origin needed, but this is very speculative. In any case, whatever the true state of the origin of the universe (assuming an origin model is the correct one), the fact that it would in some manner be in the realm of quantum theory suggests that the origin would have no cause. There is indirect evidence of this from quantum theory itself.
Uncertainty is an integral feature of quantum theory, as embodied in the Heisenberg Uncertainty Principle (HUP). HUP says that given two complementary attributes, like position and momentum, attempting to measure one attribute beyond a certain level of precision automatically and unavoidably results in a large degree of uncertainty in the other attribute. It is not a matter of using the smallest things around like sub-atomic particles to measure other smallest things around, as some textbooks assert. The uncertainty manifests even when you do an indirect measurement where you don't actually interact with a particle.
Einstein for one couldn't stomach this uncertainty. He agreed that quantum theory made accurate predictions but contended that it was incomplete. He claimed that there were factors that don't show themselves in experiments that are responsible for the uncertainty. He called these hidden variables. Einstein and two collaborators, Podolsky and Rosen, devised an experiment that would force the hidden variables to manifest. It involved separating two correlated particles and then measuring their attributes. The technology to actually do this EPR experiment, as it was called, did not exist in Einstein's time. But in 1964 John Bell devised an equivalent experiment involving a mathematical inequality that could actually be run. Definitive Bell inequality experiments were conducted in the past few decades.2,3
The results are unequivocal. Hidden variables, the causes of quantum events, were found not to exist. Quantum events are governed by a fundamental and irreducible randomness.
Since then others have attempted to go beyond Einstein and Bell, but all such attempts have failed. Many crazy quantum interpretations have been devised to introduce cause into quantum theory, but they all suffer from the fatal flaw that they are non-falsifiable - there is no way to test them, even in principle, to see if they are false. And if we can't test for their falseness, we also don't know which are true. Not only can't we know which if any are true, but we can't distinguish any one from any other one. None of them can be accepted by science for these reasons. No one has yet come up with a testable alternative, and until that happens, if ever, we have to accept the results of the Bell inequality experiments. Quantum events do not have causes.
Accepting this empirical evidence, we must ask if the singularity was a quantum event. We have indirect evidence that it likely was. HUP predicts the existence of virtual particles. Energy and time are two other complementary attributes. If we look at any point in space in a brief enough period of time, its energy becomes uncertain. This uncertainty manifests as the creation of virtual particle pairs. They exist very briefly and vanish back into nothing.
Do virtual particles really exist? Experiments verify that they do. They are responsible for a phenomenon called the Lamb shift, where the energy levels of electron orbitals in atoms have a different value than what they would have without them. The experiments that verify the Lamb shift also verify HUP's calculations for the energy-time relationship as well.4,5 HUP predicts that all known particles will manifest as virtual particles, but the more massive they are, the longer you have to wait before they come into existence.
Singularities (or whatever it is that quantum gravity will give us) may well form from such a quantum process. While we don't know much about the properties of singularities, it may be that they come and go just like ordinary virtual particles. The one that started our universe would have exploded before vanishing back into nothingness. In any case, if singularities form by a quantum process it would mean that the one that started our universe had no cause.
The fate of the universe depends on its density
Given that we live in an expanding universe, what will be its fate? That depends on its mass density. Gravity between the galaxies should be slowing the rate of expansion (although as discussed below, we're not quite sure if that is the case). The universe may have too little mass to ever halt the expansion. Or it may have more than enough mass to halt it, in which case the expansion will come to a halt after some finite time and then the universe will contract into a Big Crunch. Or the universe may have just enough mass to halt the expansion, but no more. In that case, the universe will continue to expand and will only halt after an infinite amount of time.
These three possible mass densities also determine the overall shape of our universe's four-dimensional spacetime continuum. We can't visualize 4-D spacetime, let alone how it curves, but we can use two-dimensional analogies. Too little mass results in a hyperbolic shape like a horse saddle. The critical amount of mass results in a flat shape like a tabletop. And if there is enough mass to re-collapse the universe, that would result in a spherical shape like the surface of a globe.
Inflation steps in to fill the holes in the Big Bang
There are puzzling observations about the universe that the Big Bang could not account for. The two main ones are the uniformity and flatness of the universe. If we compare opposite parts of the sky they are the same, yet neither light nor anything else has had time since the Big Bang to travel from one region to another. But measurements of the CMB show they have the same temperature and characteristics. How could they equalize like that without any communication between them? This is known as the horizon problem.
While we don't yet know the shape of spacetime, we do know it is either at or close to flat. Why is the mass density so close to the critical density instead of a thousand or a million times too high or low? This is the flatness problem.
There are a few other problems as well that we won't get into. Suffice it to say the Big Bang needed to be refined. In the 1980s such a refined theory was developed and was called inflation theory because of the extreme expansion that was a central part of it. Since then many variations have been developed, but the basic theory goes like this -
10-35 of a second after the first moment the universe expands and cools to its critical temperature of
1027 degrees Celsius, similar to water cooling to its critical temperate of
0o Celsius. The water should turn to ice. But it is possible to cool pure undisturbed water below its freezing point without it turning to ice. The water is then supercooled, and the slightest disturbance will cause ice to rapidly form.
The same thing happens to the universe. It supercools below its critical temperate and is now in an unstable state called false vacuum, for the same reason you might call supercooled water false ice. The false vacuum is disturbed, most likely by a quantum fluctuation, and a pocket of true vacuum forms and expands very rapidly. By the time this inflationary era is over
(10-30 of a second), the pocket of true vacuum has expanded at least
1050 times. The universe that we see today is at this time about 10 centimeters - the size of a grapefruit. It is part of a far larger universe, one that is some
1027 times bigger. Since that time both have expanded greatly. Today we can't see that larger universe since light from it hasn't had time to reach us.
At the start of the inflationary era three of the four fundamental forces, the electromagnetic, the nuclear strong and the nuclear weak (the fourth is gravity) are united, but during inflation break apart in a process called symmetry breaking. (Gravity is thought to be unified with the other three earlier but it breaks away before inflation starts.) During this process an effective negative pressure develops which fuels inflation.
For inflation to work, an as yet undeveloped theory of physics known as the Grand Unified Theory (GUT) must be applied. Among other things, GUT determines when and how symmetry breaking and negative pressure operate. (GUT should also tell us why the universe is all matter instead of half matter and half anti-matter.) Inflation is a theory that is unconfirmed at this time, which relies on GUT, another unconfirmed theory. This is different from the formation of the elements, the CMB, and all other physical processes that go on in the universe. These are all understood based on well-confirmed theories of physics. Inflation, its variants, and competing refinements of the Big Bang rely on their predictions being confirmed by observations even more critically than other processes do.
The more successful predictions a theory has the stronger our confidence in the theory becomes. Inflation has three main predictions that we should be able to confirm or disconfirm to a high degree of confidence in the near future. First, the universe is flat to within at least one part in
1050. This is caused by the extreme expansion during the inflationary era. It is like taking a balloon with an unbreakable skin and blowing it up to the size of the earth (except that the expansion under inflation is far more extreme). If a small part of the balloon's surface is examined, it will seem perfectly flat.
This aspect of the theory solves the flatness and horizon problems. The first should be obvious; the second is solved because a very tiny self-similar piece of spacetime is blown up far larger than the observable universe.
The other two predictions have to do with the structure of the CMB. During the inflation era, random quantum fluctuations take place. They result in irregularities or anisotropies that later get carried over to the CMB because of gravitational interactions. These should be scale invariant - you don't get more of them in any period of time as the universe blows up than in any other period of time. These scale invariant fluctuations should result in certain characteristics - they should have what is known as a Gaussian distribution (this comes from their complete randomness) and they should peak at one degree of the sky, which is a length that is twice as big as the full moon. Other characteristics of the CMB could further refine inflation models, and in any case will tell us many basic parameters of the universe.
But is inflation true?
A number of observations have been done to check inflation's predictions. In the past decade, studies have been carried out to determine the mass density of the universe. The dynamics of galactic groups as well as the movements of stars within galaxies indicate there is a great deal of matter that does not revel itself by glowing in any part of the electromagnetic spectrum. We don't know what this unseen matter is, but there is at least ten times more of it than the visible matter. Studies of the dynamics of these galactic groups give us the amount of matter, both visible and dark, in these groups.6,7
Galaxy counts in the Hubble Deep Field show us that the geometry of spacetime seems to be negatively curved. When galaxies are counted at different volumes of spacetime at different radii from us, we expect a certain number in each volume for a flat spacetime, but we find more galaxies than we expect in distant spacetime.8
Looking at X-ray emitting gas that is embedded in galactic clusters, we do a calculation of how much mass is required to hold this gas in place given it's temperature. This calculation finds the mass needed to prevent the X-ray gas from dissipating, including the dark matter.9
We look at several gravitational lenses, where a distant quasar is lensed by an intervening galactic cluster. Based on the timing of changes in the quasar images we can figure out the parameters of the geometry and then do a calculation to find the total mass of the galactic cluster - again the nature of this precludes not counting some dark matter.10,11
All of these studies indicate that there is only 20-30 percent of the critical mass density. Each of these studies has observational uncertainties and possible systematic errors. No one of them can be pointed to as evidence beyond a reasonable doubt for a low mass density. But since they are completely different methods for determining the mass density and they all agree with each other, we can be confident in their results. What's more, no study has shown either a critical or more than critical mass density.
Is the universe undergoing an accelerated expansion?
One other study to determine the mass density had a surprising result. This was a study of type Ia supernovas.12,13 When a white dwarf star is in orbit with a normal giant companion star, the white dwarf can draw material off its companion. When a critical amount of accumulated mass is reached, the white dwarf collapses into a neutron star in a supernova explosion. Such explosions that lack hydrogen lines in their spectrum are known as type Ia's. It is found that they have a characteristic brightness versus decay time. If we know the decay time of one, we then know its absolute brightness, and by comparing that to the relative brightness we can determine its distance independently from redshifts caused by the Hubble expansion of the universe.
Type Ia's show yet again that the universe has a low critical mass density, but they show something else as well. Distant type Ia's are fainter than they ought to be under a Hubble expansion. No apparent observational or systematic errors can account for this. Taken at face value, it means that the universe is undergoing an accelerated expansion. The mutual gravitational attraction of galaxies should be slowing the expansion, but if the expansion is speeding up, something else is going on.
The observations that show the universe has less than the critical mass density were a disappointment to inflation theorists, but these supernovae observations have given them an out. Inflation theorists contend that the accelerated expansion is caused by a cosmological constant. This is a slight imbalance in the quantum vacuum fluctuations that occur at every point in space. The result is a repulsive energy that itself provides an equivalent extra mass. When the cosmological constant equivalent mass is added to the normal mass density of the universe, you get a flat spacetime.
There are several problems with this. Unlike the studies of the normal mass density, the accelerated expansion is so far dependent on a single kind of study, that of the type Ia supernovas. While it seems to be well done, there can be unknown systematic errors that would invalidate the result. Indeed, it has been found that the light versus decay curves for distant type Ia's may not match those of the nearby ones.14 If this finding is confirmed, the accelerated expansion may not be true.
Even if the accelerated expansion is true, it may be caused by something other than a cosmological constant. Theorists have been gleefully devising alternate mechanisms. The leading contender is something called quintessence. Quintessence can be several things, such as the decay of dark matter particles, the "field" that drives inflation during the inflationary era, the influence of extra dimensions, or the decay of leftover false vacuum artifacts from the early universe. Unlike the cosmological constant, quintessence would not provide a steady repulsive force over time and would not contribute to the mass density of the universe. If it is quintessence, the universe is not flat and inflation is false.
What does the CMB say about inflation?
Inflation's prediction of flatness is as yet not directly confirmed. But flatness and the scale invariance of the CMB are related. Calculations predict a peak of the anisotropy scale at one degree of sky, and finding such a peak is an indirect confirmation of flatness. The COBE satellite, which studied the CMB in the early 1990s, could only look at the anisotropies down to seven degrees, but three earthbound studies with finer resolution have tentatively found the one degree
peak15 Even better data has been obtained more recently from two balloon experiments, Boomerang and
Maxima.16 They too, have found the one degree peak, although the error bars in the analysis of the data are still too large to draw definitive conclusions. But all is not well.
Inflation theory predicts a series of secondary peaks in the anisotropy spectrum. They trail after the first peak at one degree at finer angular scales. In the early universe, the dense spots have greater gravitational attraction than surrounding areas and pull more matter in, but then photon pressure repels that matter. This causes oscillations that are acoustical in nature. The result is a series of anisotropy peaks that are of the same nature as harmonic notes on a musical instrument.
Boomerang and Maxima should have been sensitive enough to see the second peak. But if it is present, it is much smaller than theory predicts. No variant of inflation can account for this. In addition to these findings, studies of the COBE data show that the anisotropies may not be
Gaussian.17 It would seem that whatever is going on, it’s more complicated than inflation and it’s more than anyone bargained for.
Alternatives to inflation
A few alternatives to inflation have been proposed in recent years. Theorists have tried to substitute a far faster speed of light for the inflationary era. It must be emphasized that these theories are no comfort for Creationists who claim that the speed of light was far faster in the past and has been steadily decreasing up to the present time. With these theories, the anomalies are over in far less than a second, after which light travels at the speed we know today. A far higher speed of light provides many of the same results as inflation without the disadvantages.
An even more bizarre alternative is that the universe does not have a simple geometry but is instead a complex fifth dimensional hyperbolic finite superspace. In this scenario, the universe could have any one of large number of possible shapes. The universe would only look infinite because light could travel unlimited distances in it, but light would wrap around - it is like the old venerable video game Asteroids, where a spaceship going off the right side of the screen re-appears on the left. If the characteristic length of such a universe is less than the distance to the formation of the CMB, we may see multiple images of the same galaxies and characteristic repeatable patterns in the CMB's anisotropies. Such a universe solves the horizon problem because we are looking at repeatable patterns when we look at different parts of the sky, and the flatness problem is solved because such a structure requires a specific hyperbolic curvature for each particular shape. The amount of hyperbolic curvature wouldn't be far away from flatness and would be in general agreement with the mass density findings.
The answers will be coming soon
The fine scale characteristics of the anisotropies of the CMB will provide us with a wealth of information that will allow us to distinguish between all these models. It will also pin down many fundamental parameters of the universe. In the spring of 2001 the Microwave Anisotropy Probe satellite is due to be launched. MAP will be able to examine the anisotropies down to a three degree length scale with far greater accuracy and precision than any previous experiment. In 2007 the European Space Agency will be launching its Planck satellite, with even more advanced technology. Planck will examine the anisotropies down to a one degree scale. These satellites will provide us with the answers to the questions we have always asked ourselves - how did the universe come into being, how did it evolve, and what will be its fate?
Acknowledgements
I have greatly benefited from discussions with Gordon Van Dalen, Taner Edis, Markus Pössel, Vic Stenger, John Baez, Nathan Urban, Aaron Bergman, Paul D. Shocklee, Ted Rosencrantz, and Richard. S. Gancarczyk.
References
1. Leibundgut B; Schommer R; Phillips M; Riess A; And Others.
Time Dilation In The Light Curve Of The Distant Type Ia Supernova Sn 1995k.
Astrophysical Journal, 1996 Jul 20, V466 N1:L21-L24.
2. Aspect, A.; Dalibard, J.; Roger, G.
Experimental test of Bell's inequalities using time-varying analyzers.
Physical Review Letters, 20 Dec. 1982, vol.49, (no.25):1804-7
3. Tittel, W.; Brendel, J.; Zbinden, H.; Gisin, N.
Violation of Bell inequalities by photons more than 10 km apart.
Physical Review Letters, 26 Oct. 1998, vol.81, (no.17):3563-6.
These are two of several Bell experiments which show the inequality is violated.
4. Boshier Mg; Baird Peg; Foot Cj; Hinds Ea; And Others.
Laser Spectroscopy Of The 1s-2s Transition In Hydrogen And Deuterium - Determination Of The 1s Lamb Shift And The Rydberg Constant.
Physical Review A-General Physics, 1989 Dec 1, V40 N11:6169-6184
5. Berkeland Dj; Hinds Ea; Boshier Mg.
Precise Optical Measurement Of Lamb Shifts In Atomic Hydrogen.
Physical Review Letters, 1995 Sep 25, V75 N13:2470-2473
6. Coles P; Ellis G.
The Case For An Open Universe.
Nature, 1994 Aug 25, V370 N6491:609-615.
7. Blakeslee, Jp; Davis, M; Tonry, Jl; Dressler, A; And Others.
A First Comparison Of The Surface Brightness Fluctuation Survey Distances With The Galaxy Density Field: Implications For H-0 And Omega.
Astrophysical Journal, 1999 Dec 20, V527 N2 Pt2:L73-L76.
8. Roche N; Shanks T; Metcalfe N; Fong R.
The Clustering Of Blue And Red Galaxies At B Similar-To 25.5 Mag.
Monthly Notices Of The Royal Astronomical Society, 1996 May 15, V280 N2:397-426.
9. Donahue, M; Voit, Gm; Gioia, I; Luppin
9. Donahue, M; Voit, Gm; Gioia, I; Luppino, G; And Others.
A Very Hot High-Redshift Cluster Of Galaxies: More Trouble For Omega(0)=1.
Astrophysical Journal, 1998 Aug 1, V502 N2 Pt1:550-557.
10. Tyson Ja; Valdes F; Wenk Ra.
Detection Of Systematic Gravitational Lens Galaxy Image Alignments - Mapping Dark Matter In Galaxy Clusters. Astrophysical Journal, 1990 Jan 20, V349 N1:L1+.
11. Smail, I; Ellis, Rs; Dressler, A; Couch, Wj; And Others.
Comparison Of Direct And Indirect Mass Estimates For Distant Clusters Of Galaxies.
Astrophysical Journal, 1997 Apr 10, V479 N1 Pt1:70-81.
12. Perlmutter, S; Aldering, G; Goldhaber, G; Knop, Ra; And Others.
Measurements Of Omega And Lambda From 42 High-Redshift Supernovae.
Astrophysical Journal, 1999 Jun 1, V517 N2 Pt1:565-586.
13. Riess, Ag; Filippenko, Av; Challis, P; Clocchiatti, A; And Others.
Observational Evidence From Supernovae For An Accelerating Universe And A Cosmological Constant.
Astronomical Journal, 1998 Sep, V116 N3:1009-1038.
14. Glanz, James.
Has A Cosmic Standard Candle Flickered?(Type Ia Supernovae May Not Be Accurate Measure Of Distance)(Brief Article) Science V285, N5424 (July 2, 1999):21.
15. Sincell, Mark.Firming Up The Case For A Flat Cosmos.(Observations Seem To Confirm A Measurement 'Hump' In The Background Radiation Of The Universe)(Brief Article)
Science V285, N5435 (Sept 17, 1999):1831.
16. Schwarzschild, Bertram.
Balloon measurements of the cosmic microwave background strongly favor a flat cosmos.
Physics Today v53, n7 (July, 2000):17
17. Kaminkowski, Marc; Jaffe, Andrew H.
New Troubles For Inflation. (Cosmology)
Nature V395, N6703 (Oct 15, 1998):639
Addendum, 26 August 2002
There have been more discoveries related to cosmology since my article was written. Most important is the confirmation of the harmonic overtones in the anisotropy spectrum of the CMB. Further analysis and additional data from the Boomerang and Maxima experiments, as well as new data from DASI, a CMB telescope near the South Pole, have demonstrated the first harmonic peak as well as hints of the second. New data from the Cosmic Background Imager, a CMB telescope with high angular resolution situated in the Chilean Andes, have definitely confirmed the third spectral peak as well as dampening peaks at even higher harmonics that attenuate into a "tail," something also predicted by the theory.
This is the critical evidence that was lacking at the time I wrote my article. If an alleged fundamental peak at one degree is seen as predicted by the theory of acoustical CMB anisotropy peaks, but the harmonic overtone peaks also predicted are not seen, that is enough to falsify the theory of CMB anisotropy peaks if we are confident that we had enough resolution and sensitivity to see the harmonic overtones. We would then have to conclude that the theory is wrong, the peak at one degree is not acoustically generated but is there from some unknown process of which we have no theory. This would also mean that the universe is not flat and inflation is wrong. And that was my tentative conclusion in the article.
But now the verification of the harmonic overtones, and especially the dampening tail seen on the higher peaks, turns this completely around. The theory of acoustical CMB peaks is now strongly confirmed. This is not a "smoking gun" for inflation, but it is well consistent with it. The smoking gun will probably come with the results and analysis of the Planck satellite, which should have the sensitivity and resolution to see higher order polarizations in the CMB anisotropy spectrum caused by gravity waves during the inflationary era - something that inflation but no other theory predicts.
The confirmation of the theory of CMB acoustical peaks also now confirms the existence of dark energy - the name now given for the cause of the accelerated expansion of the universe. We can now believe beyond a reasonable doubt that the one degree fundamental peak does mean the universe is flat. And since as I explain in the article, other observations show only about a quarter of the amount of baryonic matter and dark matter necessary to come to critical density and make the universe flat, the rest is made up by dark energy. The type Ia supernovae studies demonstrating dark energy required confirmation from a completely independent observation, and now we have it.