The ranking by impact (or citations per paper, a measure that lets smaller institutions compete equitably against the larger) puts the University of Toronto top, with Yale punching far above its weight class at #2. The real surprise in this table is the U.K. Royal Observatories at #5. This, the only institution outside North America to make the Top Ten in impact, was stripped of its largest research unit as a cost-cutting measure in late 1998. In the European area, the University of Durham (#12) squeaks ahead of Cambridge (#15), and then we descend to #28 for the next European entry, the National Research Council, Italy.

   Looking at the most-cited papers in the three cohorts, the most striking feature is the dominance of papers with a cosmological twist. Cosmology captures all six positions in '96, and four out of six in '98. In the '97 group five of the six address the problem of gamma-ray bursters, which are now known to be at cosmological distances. Two themes clearly emerge: the properties of galaxies at great distances and the cause of gamma-ray bursts. Progress in both areas is being driven by instrumental excellence.

   The Hubble Space Telescope (HST) is behind two '96 big hitters (paper #3 and #4), with its brilliant Hubble Deep Field (HDF) picture. This image of a small area of the sky, released in December, 1995, was made by adding 342 separate observations to create an effective exposure time of 10 days. This window on the distant early universe whisks astronomers to within a billion years of the onset of galaxy formation. HDF galaxies are small, with vigorous star formation, and they grow by mergers and cannibalism. Also in '96, results from the 1989 Cosmic Background Explorer continue to excite, with its maps that hint at structure formation only 10⁶ years after the Big Bang.

   On the ground, the mighty Keck Observatories—two 10-m telescopes belonging to Caltech and the University of California—has made a big impact. Paper #1 for 1998 is on a supernova explosion that took place before the formation of the Solar System. Supernovas of this kind are crucial landmarks in establishing the geometry of the universe. These observations suggest that the expansion will not reverse but is getting faster.

   Gamma-ray bursts hog the top rankings for '97. What these papers show is that in a split second detonation, gamma-ray bursts release more energy than any explosion since the Big Bang. Discovered in 1965, they have only recently been found to be very remote objects; the BeppoSAX mission (#2 in '97) participated in the crucial observations locating the bursts at cosmological distances.

   Among the journals that published these high-impact reports, Astrophysical Journal accounts for 1,483 papers that were cited 10 or more times—nearly half the papers in the survey. Astronomy & Astrophysics is next (521 papers), followed by Monthly Notices of the Royal Astronomical Society (362). By the measure of impact (of those journals that published at least 30 high-impact papers), Nature is tops with a cites-per-paper average of 29, followed by Solar Physics (27 cites per paper) and Science (23).

   The list of high-ranking papers could be gloomy reading for planetary astronomers, plain-vanilla astrophysicists, and austere theorists. They should take courage. Although these top-cited papers are driven by observational technology favoring "edge of the universe" problems, the next Science Watch survey, covering publications in 1999-2001, could look very different. Planetary science has made a quantum leap (despite the failure of the two recent Mars missions), and X-ray astronomy will be re-invented by the latest X-ray missions.