considerations on x-ray astronomy - memorie della...

Mem. S.A.It. Vol. 84, 472c© SAIt 2013 Memorie della

Considerations on X-ray astronomy

A start in X-ray astronomy

R. Giacconi

Johns Hopkins University, Department of Physics and Astronomy 3400 N. Charles StreetBaltimore, MD 21218 USA, e-mail: [email protected]

Abstract. Remarks on the history of x-ray astronomy. The successes of x-ray astronomywere not a fluke, but the result of the bounty of Nature, the aspirations of many people, afairly rigorous and methodical research effort, and the development of new technology andnew operational approaches. In the last 50 years we have made great steps in our physicalunderstanding of what the x-rays have shown us. We have made a progress of 10 billions insensitivity. We have developed the know-how and the methods to build even more powerfulobservatories that will carry us into the next 50 years of discovery and understanding.

Key words. X-ray astronomy

1. Introduction

The last fifty years have seen a rapid develop-ment of observational x-ray astronomy fromthe discovery of the first extra solar x-raysource Sco X-1, to the detection of celestialobjects 10 billion times fainter. This remark-able improvement has allowed detection andstudy of the x-ray emission from all knowncelestial objects and has opened to our viewthe high energy universe, from the formationof young stars to the study of the most dis-tant clusters of galaxies and AGNs. The storyof the early beginning of x-ray astronomy hasbeen very well described by Richard F. Hirsh inhis PhD Thesis (Hirsh 1979) and summarizedin his book “Glimpsing an Invisible Universe”(Hirsh 1985). A more personal view was givenin “The X-ray Universe” by Tucker & Giacconi(1985) and in “Secrets of the Hoary Deep” by

Send offprint requests to: R. Giacconi

Giacconi (2008). I thought that the most usefulcontribution I could give to this meeting, de-signed to celebrate fifty years of accomplish-ments and to look to fifty years in the future,would be to attempt to clarify some confusionabout the past and to express my hopes for thefuture of the discipline.

2. Opportunities in x-ray astronomy

Given that x-rays can not penetrate the Earth’satmosphere, one can set the start of interestin exploring the sky in x-rays to the develop-ment of the V-2 rockets during War World IIand more clearly to the launch of Sputnik onOctober 4,1957.

The Naval Research Laboratory group ledby Herbert Friedman made use in 1948 of V-2 rockets, captured by US troops in Germany,to investigate the radiation producing the ion-ization of the Earth’s atmosphere with the

Giacconi: Considerations 473

clear conclusion that it was due to solar coro-nal x-rays. The NRL group was the undis-puted leader in solar x-ray astronomy duringthe 1948-1959 period. Although the group at-tempted to observe emission from other stars,they were unsuccessful due to the very low fluxfrom extra solar sources and the design of theirexperiments.

The launch in 1957 of “The New RedMoon” by the Soviet Union had very signif-icant repercussions in the US due to whatwas considered to be not only a major as-sertion of Soviet technological prowess, butalso to have defense implications. In responseto this challenge President Eisenhower estab-lished the National Aeronautics and SpaceAdministration to integrate civilian efforts in1958. In the same year the National Academyof Sciences created the Space Science Boardwith the task of examining the scientific oppor-tunities presented by space investigations. TheSSB proceeded as usual by establishing severalcommittees in Physics and Astronomy.

As early as October, 1958, John ASimpson, Chair of the Committee of “Physicsof fields and particles in space” suggested in aninterim report mapping the sky in gamma andx-rays (Simpson 1958).

Lawrence H. Aller (Optical and RadioAstronomy Committee) pointed out in 1959that x-rays could reach the Earth from galac-tic distances and that many stars and nebu-lae would be detectable in wavelengths shorterthan 20 Angstrom, provided one developed thenecessary instruments (Aller 1959).

Leo Goldberg in the report of thesame committee (Oct. 24, 1959) emphasized,“Instrumentation for x-ray optics is in a veryrudimentary state and for image forming opticsnon existent”.

Bruno Rossi participated in these discus-sions and was appointed Chairman of theAd Hoc Committee of Space Projects onSept. 30, 1959.

2.1. Learning

In September 1959, I had joined AmericanScience & Engineering, a small private com-pany (27 people) created by its President,

Martin Annis, to carry out scientific re-search for the US Government in the fieldsof Defense, Education and Medicine. BrunoRossi was the Chairman of the Board.

The company hired me at the end of myFulbright Fellowship at Indiana and PrincetonUniversities to initiate space research activi-ties. I met Bruno Rossi for the first time in theFall of 1959, at a party in his house. He sug-gested that among other possibilities I couldconsider x-ray astronomy as a potential subjectof study. In typical Italian fashion he did notrelate to me the discussions, which had takenplace at SSB, presumably with the view thatthe young man could figure it out by himself.

Up to then I had been working in cos-mic ray and particle physics at the Universityof Milano with Giuseppe Occhialini, at theIndiana University with Robert W. Thompsonand at the University of Princeton in the cosmicray group of George Reynolds. I had thereforeto start anew in learning about x-ray physicsand x-ray astronomy. I studied the Comptonand Allison 1935 fundamental book “X-raysin theory and experiment” and brought my-self up to date by reading the relevant partsof Springer Verlags “Encyclopedia of Physics”edited by Flugge. I found a Russian publicationof 1958 “The Russian Literature of Satellites”,which in an article by S.L. Mandel’shtam andA.I.Efremov (Mandel’shtam & Efremov 1958)described in some detail H. Friedman’s solarwork in the period 1948-1958 and some the-oretical studies by Elwert and De Jager. Theyalso mentioned the upper limit for extra solarsources of 2 × 10−8 erg cm−2sec−1 for 6 keVphotons reported by NRL.

2.2. Thinking

The first obvious result from this learningwas the very low fluxes of X-rays to be ex-pected from non-solar sources. While solarcoronal emission below 20 Angstrom resultedin 106 cts cm−2 sec−1, the emission from Siriusat a distance of 8.6 light years would pro-duce at most only 0.25 cts cm−2 sec−1 evenif we adopted the extreme assumption that itemitted as much in x-rays as in visible light.Coming from a discipline (cosmic rays) that

474 Giacconi: Considerations

Fig. 1. A telescope for x-ray astronomy (Giacconi& Rossi 1960), shows a single reflection design us-ing nested parabolic mirrors at grazing incidence tofill the aperture.

could no longer contribute significantly to par-ticle physics research due to the low particlefluxes, I was quite concerned that x-ray astron-omy would be severely constrained by poorstatistics.

Two years of observations at the TestaGrigia Observatory (elevation 3500 meters)had been necessary to collect the 80 proton in-teractions on which to base my thesis researchon the Fermi fireball model. I had dreamedthen of a magic magnetic funnel that couldconcentrate cosmic rays on my cloud chamber.Could I realize such a collector for x-rays? Theanswer was yes! By using a parabolic graz-ing incidence mirror I could focus incoming x-rays from a large collecting area onto a smalldetector thereby increasing the sensitivity ofthe instrument by several orders of magnitude.Such a telescope did not exist in x-ray astron-omy, but I felt that there were no physics obsta-cles but only technical ones in its development.Thus I was certain that after developing such atelescope x-ray astronomy could be done.

As soon as I hit on this idea Martin Anniscalled Bruno Rossi, who expressed great in-terest, and immediately came up with an im-provement, namely the idea of nested mir-rors (Fig. 1). We submitted a paper, “A tele-scope for soft x-ray astronomy” in December1959 (Giacconi & Rossi 1960). In coopera-tion with George W. Clark and Bruno Rossiof MIT by January 15 1960 I wrote a techni-cal note “Brief Review of Experimental andTheoretical Progress in X-ray Astronomy”

(Giacconi, Clark & Rossi 1960). We consid-ered several mechanisms of x-ray productionincluding black body, optically thin thermalbremstrahlung, synchrotron emission and in-verse Compton scattering of relativistic elec-trons on optical or infrared photons. We con-sidered several potential sources as shown inTable 1, and we came to the conclusion thatfluxes from extra solar sources would indeedbe most likely of order of 10−6 the flux fromthe solar corona. Following Aller’s suggestionthat interstellar space was transparent (as far asthe galactic center) to x-rays of energy greaterthan 1 keV (Aller 1959) and that grazing in-cidence total reflection could efficiently focusphotons of less than 10 keV, we noted the veryimportant window between 1 and 10 keV forX-ray Astronomy (Fig. 2).

Fig. 2. Attenuation of radiation in space due to theinterstellar medium. Vertical lines indicate the shortwavelength cutoff for grazing incidence optics andthe long wavelength cutoff for seeing the galacticcenter.

I also examined critically the experimentscarried out by the NRL group to detect extrasolar sources. While their approach was sen-sible for ionospheric and solar x-ray studies,it seemed to us unsuitable to study the faintfluxes expected from other stars.

I thought there were three principal prob-lems with the NRL surveys: I) Exploring thesky for unknown sources with a narrow field ofview (3 degrees) would require a large numberof flights (∼ 100) to scan the entire sky, thus


Table 1. Possible sources of x-rays and their estimated fluxes, taken from Giacconi, Clark &Rossi (1960).

Source Maximum Mechanism EstimatedWavelength for Emission Flux

Sun < 20Å Coronal Emission ∼ 106 cm−2sec−1

Sun @ 8 Light Years < 20Å Coronal Emission 2.5 × 10−4 cm−2sec−1

Sirius if Lx ∼ Lopt < 20Å ? 0.25 cm−2sec−1

No convective zone

Flare Stars < 20Å Sunlike Flare? ?

Peculiar Stars < 20Å B ∼ 104 GaussLarge B ?Particle acceleration

Crab Nebular < 25Å SynchrotronEe ∼ 1013 eV in B = 10−4 Gauss ?Lifetimes?

Moon < 20Å Impact from solarwind electrons 0 − 1.6 × 103 cm−2sec−1

Φe = 0 − 1013 cm−2sec−1

Sco X-1 (1962) 2 − 8Å ? 28 ± 1.2 cm−2sec−1

providing a 1% chance to sweep through anyparticular source. II) No attention was given tothe background produced by high-energy cos-mic ray particles traversing the counter. In theGeiger counters used, the cross section sensi-tive to cosmic rays exceeded the area of thewindow sensitive to x-rays by a factor greaterthan 20 (and possibly as large as 200). Sincethe cosmic ray flux is about 1ct cm−2 arcsec−1

and the expected x-ray photon flux for extrasolar sources is only 1 cm−2 arcsec−1, this im-plies a signal to noise ratio of less than 1/20.III) Clearly the cosmic ray rate had to be muchreduced either by making the cross section ofthe counter sensitive to cosmic rays not muchlarger than the area of the x-ray window and/orby using an anticoincidence system to acceptx-rays and reject cosmic rays. This requiredthe ability to count single photons, rather thanmeasuring the average flux. The use of anti-

coincidence required the transmission of dig-ital information on an analog telemetry sig-nal. The net result was that the cosmic raybackground was reduced to a negligible frac-tion of the isotropic extragalactic x-ray back-ground which could therefore first measured inthe June 1962 flight.

2.3. Planning

Planning for a new space program unfortu-nately included not only scientific considera-tions but also the need to find financial supportfor the design and development of the neces-sary instruments and for the means of placingthem above the Earth’s atmosphere.

We were able to interest John Lindsaya solar astronomer at Goddard Space FlightCenter to support the “Design, Construction


and Testing of a prototype X-Ray Telescope”(NAS-5-660) starting in October 1960. Thisprogram ultimately culminated in the flight ofa 30 cm diameter x-ray telescope as part of theATM instruments flown on the first US SpaceLaboratory SKYLAB in 1973.

Our proposal A measurement of soft x-raysfrom a rocket platform above 150 km (AS&EP -26, 17 February 1960) was not accepted byNASA Headquarters, with the sarcastic ques-tion by Nancy Roman of why should NASAsupport a search for x-ray stars which werenot known to exist. Actually because of J. E.Kuperion of Goddard Space Flight Center ad-vocacy, NASA supported two groups: PhilipFisher at Lockheed from Nov. 1961 andMalcolm P. Savedoff at the University ofRochester from June 1959 and partly sup-ported the NRL effort.

We turned then to the Air Force CambridgeResearch Laboratory and we were able to in-terest John W. Salisbury (Chief of the Lunarand Planetary exploration Branch) in the pos-sibility of detecting fluorescent x-rays from theMoon, which after all was the only extra solarsource we could count on.

Our scientific planning consisted thereforein following two paths: I) the development ofx-ray telescopes to be used initially for so-lar astronomy and later, when sufficiently ad-vanced, for stellar astronomy. II) the devel-opment of instrumentation more sophisticatedthan hitherto used, in the hope that even with-out the new x-ray optics we could observe thebrightest celestial sources. We considered thelimit to the flux of as reported in 1959 by theNRL (2×10−8 erg cm−2 sec−1) to be only a ten-tative number, given the very high noise in theirmeasurements and the tiny portion of the skyexplored. In any case we concluded that an im-provement by a factor of 50-100 in sensitivity,which we deemed feasible, could lead to newresults.

2.4. Do

Beginning in October 1960 we began to de-velop the x-ray grazing incidence imagingtelescopes using the optical design describedby Wolter for microscopes (Wolter, H. 1952;

Giacconi & Rossi 1960) and using a varietyof techniques (Fig. 3), which resulted in thedevelopment of optical systems with greaterand greater collecting area and angular reso-lution. We tested these telescopes in a numberof rocket flights (Fig. 4) for solar research andin 1973 we were able to study from Skylab thex-ray emission by the solar corona during sev-eral solar rotations with an angular resolutionover the entire sun of better than 5 arcsec.

Fig. 3. Wolter-I x-ray telescope made as a Ni replicafrom a polished mandrel in 1963.

As to the search for extra solar sources,we developed Geiger counter detectors with10cm2 window area each. The detectors werehoused in a scintillator well, which providedthe anticoincidence signals for cosmic ray re-jection. Three such detector systems were in-stalled in each rocket flight and the field ofview was made as large as 120 degrees to in-crease our chance of success (Fig. 5). Afterrocket failures in 1960 and in 1961, the firstsuccessful flight in 1962 led to the discoveryof Sco X-1 (Giacconi et al. 1962).

The results are reproduced in (Fig. 6) andthey show the detection of Sco X-1 and of theisotropic extragalactic x-ray background mea-sured in two bands of the x-ray spectrum. Hadwe flown at a different time of the year wewould have detected the Crab Nebula emissionas we did in fact observe in the October 1962flight.


Fig. 4. Progress in solar x-ray observations from 1963 to 1973 and the Apollo Telescope Mount (ATM)telescope with better than 5 arcsec resolution.

I hope to have made clear in my discussionthat impetus for this research was not due toa single person intuition, but to a broad con-sensus of interests among physicists and as-tronomers. It also should be clear that the dis-covery of Sco X-1 was not a serendipitous dis-covery as sometime described, but the success-

ful result of a carefully planned and executedexperiment.

3. The Post Sco X-1 Era

In the preceding sections I have used as sub-titles “Learn-think-plan and do”, a brief re-


Fig. 5. Rocket payload from 1962 that discoveredthe non-solar x-ray source Sco X-1, and the cosmicx-ray background.

minder of our methodology. This is because Ibelieve that method was an important factor inour approach. I will return to this point when Iconsider the impact of x-ray astronomy on as-tronomy as a whole.

3.1. Plan

Scientifically the discovery of Sco X-1 openedup new horizons. The x-ray flux from this ob-ject was 1000 times the flux in the optical do-main, and its intrinsic x-ray luminosity was1000 times the luminosity of the Sun at allwavelengths. Thus we had found some newprocess of x-ray generation and some new typeof star. The fact that such new and unexpectedresults could be obtained with relatively sim-ple instrumentation opened the field to a large

Fig. 6. The results from the June 1962 rocket flightshowing the detection of Sco X-1 as well as theisotropic x-ray background.

number of astronomers all over the world tofurther investigate the high energy Universe.

My group at AS&E at this point found it-self in a rather difficult situation. The Air Forcewas no longer willing to support our research,which was pure astronomy. NASA on the otherhand had not yet started to support AS&E re-search on stellar emission. Herbert Gursky andI decided that NASA Headquarters would notsupport our research unless we proposed a wellthought out 5 year plan (1964 to 1969), whichincluded both rocket and satellite instrumentsand which would be appropriate to advance thefield. We were back to the “Plan” phase.

We submitted this new program onSeptember 25, 1963 when only three sourceswere known, Sco X-1, the Crab Nebula andCygnus (Giacconi & Gursky 1963).

The plan included a scanning satellitewhich became UHURU in 1970 and a 1.2 me-ter diameter x-ray telescope, which after a longobstacle course became CHANDRA, launchedin 1999 and still operating today, (Fig. 7).

The 1.2-meter mirror was proposed toachieve sufficient sensitivity to image discretesources or the granularity of the background onthe scale of 1 arc minute, (Fig. 8).

3.2. Execution (Do)

The path that was followed in the executionof the program is illustrated in Fig. 9, show-ing the steps, which had to be realized to reachCHANDRA. The only project in this figure


Fig. 7. Time-line for the 1963 AS&E proposal to NASA for an x-ray astronomy program.

Fig. 8. Sketch of the 1.2 meter x-ray telescopethat would eventually, many years later, become theChandra X-ray Observatory (nee AXAF).

that was not executed was the use of LAMARoptics on a large observatory. Notwithstandingthe many recommendations to NASA by sev-eral review committees it could never be turnedinto reality.

As I already mentioned this program ledto the use of a 30 cm diameter telescope onSKYLAB for an extended solar study fromMay 14, 1973 to November 16, 1973. The highangular resolution (< 5 arc sec) and the op-portunity to study several solar rotations ledto a fundamental change in our understandingof the physics of the corona as discussed byVaiana and Rosner (Vaiana & Rosner 1978).

In 1970 NASA approved a program for aLarge Orbiting X-ray Telescope for extra solarx-ray astronomy, which included a high reso-lution (Wolter type I) 1.2 meter diameter tele-scope and a 1 meter (Kirkpatrick-Baez) highthroughput mirror. In 1973 the program wascanceled due to overruns in other NASA pro-grams.


Fig. 9. The path to Chandra as envisioned in 1980. Two parallel paths were taken. On the left non-imagingmissions to increase the census of sources and better locate them, and on the right the development of x-raytelescopes.

As part of the High Energy AstronomyObservatory (HEAO) program, a smaller ver-sion of the x-ray observatory consisting of asingle 60 cm diameter telescope, consisting of4 nested pairs, was designed, built, and tested

in 5 years and flown in 1978. It became knownas EINSTEIN and it opened up x-ray astron-omy to the study of all celestial objects. Wewere able to study in their x-ray radiationauroras on planets, forming young stars, all


main sequence stars, normal galaxies and ac-tive galaxies, clusters of galaxies and the x-raybackground (Fig. 10).

To facilitate the use of these new data byall astronomers, several new practices were in-troduced: I) 30% of the time was reserved forobservations by guest astronomers. II) To per-mit analysis of the data by astronomers nottrained in x-ray astronomy, our group at theHarvard-Smithsonian Center for Astrophysicsprovided the community with calibrated dataalready transformed in physical quantities.This represented a radical increase in the re-sponsibility typically taken over by an ob-servatory. Rather than just granting guest ob-servers time, we committed ourselves to pro-vide them with data suitable for further analy-sis. This change was part of the overall ScienceSystem Engineering approach (Learn–Think–Plan–Do), which x-ray astronomers pioneeredand then introduced, in the scientific operationsof NASA Hubble Telescope, in the design, fab-rication and operations of the ESO Very LargeTelescope and NASA’s Chandra. This method-ology is now adopted in all of astronomy andI will describe its origin and content in a briefappendix.

3.3. Chandra

After many iterations (and versions), NASAfinally accepted the 1976 proposal byH. Tananbaum and me for a 1.2-meter tele-scope, which was flown in 1999, and isstill operating today. It is the most powerfulx-ray observatory ever placed in orbit, andits collecting area and exquisite angularresolution (0.5 arcsec) has led to a number ofimportant discoveries. To mention only oneas an example it permitted the solution of theproblem of the origin of the extragalactic x-raybackground first discovered in 1962.

The picture of the Chandra Deep FieldSouth (Fig. 11) was obtained with an expo-sure of about one million seconds. The to-tal sky area surveyed is only 14x14 arc min-utes and contains 346 x-ray sources, or about1.7/square arc minute. The x-ray flux from

these sources is about 10 billion times smallerthan that from the source Sco X-1 discoveredin 1962 (Fig. 12). Thus x-ray astronomy hasprogressed in 40 years as much as optical as-tronomy in 400.

4. The Future

Contrary to the point of view expressed by sev-eral astronomers, I am confident that we stillare in a discovery era. Given that the physicalnature of 97% of the matter in the Universe isstill unknown, I believe there is ample roomfor frontline research that will provide as manysurprises as have occurred in the last decades. Ithink that 2012 is similar to 1609 when astron-omy posed many of the unsolved questions,which physics had to solve.

We have learned in the last 50 years thathigh-energy phenomena are key to structureformation and evolution, and that they arethe norm not the exception in the Universe.Whenever we study explosions, high temper-ature plasmas or high-energy particles, x-rayobservations have shown to be essential to theirunderstanding and have provided a powerfuland unique tool for their study. It also is impor-tant to keep in mind that most of the baryonicmatter in the Universe is to be found in hightemperature plasmas.

In this article I have tried to explain thatthe successes of x-ray astronomy were not afluke, but the result of the bounty of Nature,the aspirations of many people, a fairly rig-orous and methodical research effort, and thedevelopment of new technology and new op-erational approaches. In the last 50 years wehave made great steps in our physical under-standing of what the x-rays have shown us, wehave made a progress of 10 billions in sensitiv-ity and we have developed the know-how andthe methods to build even more powerful ob-servatories.

I am therefore certain that in the next 50years x-ray astronomy will reach new heights.I hope the new generation of astronomers willhave the opportunity to Learn-Think-Plan andDo as we have had.


Fig. 10. On the left is a sketch of the Einstein X-ray Observatory (aka HEAO-B) which contained a 0.6 mdiameter X-ray telescope. On the right is an image of the nuclear region of the Andromeda Galaxy (M31)resolving the emission from individual binary systems, similar to those in the Milky Way.

Fig. 11. The Chandra Deep Field South 1 Msec im-age.

Acknowledgements. I am grateful to StephenMurray for his lively discussions regarding the his-tory of the field and the practice of science sys-tems engineering. Steve was very helpful in prepar-ing both this manuscript and the presentation atthe Milano conference. Thanks also to HarveyTananbaum for his careful reading and thoughtfulcomments.

Appendix A: The Method

I am well aware that most scientists, includingme, do their research with a total disregard ofthe views of philosophers of science. When Irefer to method, I mean something more mod-est and practical, namely what we have cometo call Science System Engineering, which hasbeen adopted over the years by my colleaguesand myself in x-ray Astronomy and then trans-ferred to optical Astronomy. Stephen Murray isteaching it as a course at JHU. As a mnemonicaid to myself I describe the method simply asLEARN–THINK–PLAN–DO (and TEACH).

I first learned it from a great particle physi-cist Robert W. Thompson, who established,with a magnetic field cloud chamber, the exis-tence of the theta zero meson by measuring itsmass to be 971+/- 10 electron masses as he re-ported at the Congress of Bagneres de Bigorrein 1953 (Thompson, Buskirk & Etter 1953).

I was shipped to work with him at theUniversity of Indiana in 1956, by my colleagueand mentor Beppo Occhialini as a FulbrightFellow. Thompson’s report, which I had stud-ied in the Proceeding of the Conference, hadimpressed me for its clarity and precision, andI became convinced that none of the other cos-mic ray groups could compete with his work.


Fig. 12. The logN-logS distribution of x-ray sources from the brightest Sco X-1 to the faintest seen in the1 Msec CDFS. The extended CDFS 4 Msec data extend the faintest source flux to ∼ 10−17 about 10 billiontimes fainter than Sco X-1.

Thompson had defined the physics prob-lem he wanted to solve, designed and con-structed the necessary instruments includingthe cloud chamber, the magnet and the stereo-scopic optics that allowed the required preci-sion for the measurement of momentum of theparticles from their tracks. He developed thedata reduction and analysis process, includingthe careful calibrations, the painstaking errordetermination and the methodical streamliningof the computation to a degree I had not yetencountered. In effect he had applied what wewould call today an end-to-end analysis to theentire work. I learned and used all of thesetechniques and I used his data for a search ofthe anti-lambda zero particle, unsuccessful be-cause of the scarce statistic. I find in retrospectthat his teaching has guided me throughout mywork.

The method consists simply in: I) learningwhat has been done in the field so far, including

theory, observational research and techniques.II) In thinking very carefully and criticallyabout previous experimental work in the field,and trying to come to some conclusions aboutthe potential for significant new discoveries.To analyze one’s strength and weakness in ac-tually executing the necessary work. To con-sider the difficulty and length of the technicaleffort required and commit to it. To carry outgedanken experiments to determine by simula-tion the required absolute and statistical accu-racy in the measurements. The influence of theexternal known conditions and backgroundsin making the analysis more complicated. III)Plan how to bring the experiment to conclusionas to time, resources and help required. IV) Do.Execute the plan using best available technol-ogy, skills and management to achieve resultsas quickly and as cheaply as one can. Noneof this can be done in a large program, weremany people are involved, unless they share in


the vision of what has to be done, and considerits overall goal as their own. Communicationup and down and sideways is essential. Honestcriticism of each others work is essential andshould be welcomed not rejected.

It is fair to say that my colleagues andI followed this approach during most of ourwork without feeling a need to write it downor document it at AS&E, CFA, STScI, andESO. At AS&E we simply assumed that ev-erybody worked that way. In 1973, at CFAwe found that most astronomers did not, andwe were simply carrying on with what wecalled the UHURU spirit. At STScI (SpaceTelescope Science Institute) Ethan Schreier,Rodger Doxsey, Don Hall and I had the chanceto start a whole institution from zero and, inter-preting the views of the Horning Committee,we decided that we had to satisfy the needsof the astronomy community, as we best un-derstood them. We were helped by the adventof computers in producing a rather sophisti-cated Science Operation System that was end-to-end from proposal support to in-flight op-erations, calibration, sequence of observations,on-line data reduction and finally random ac-cess archive. Much of the work was made pos-sible by automated systems. The preparation ofa new all sky catalog of stars brighter than 15thmagnitude prior to launch to permit pointing ofHubble efficiently is an example of how we ap-proached the science requirements.

At ESO we applied this same method to theentire design, fabrication, testing and operationof the Very Large Telescope as well as the end-to-end data handling system as I described inSecrets of the Hoary Deep (Giacconi 2008).

This changed view of the responsibilityby an observatory staff has permitted quickutilization of the data from major observato-ries and the ability to make repeated use ofarchival data by many users. Most of the re-sults have been good, except for a separationbetween builders and users, which I believe is

not healthy for the field. Particular attentionis needed to insure the training in “LEARN–THINK–PLAN–DO” of the new generation ofastronomy.

References

Aller, L. H., 1959, PASP, 71, 324.Giacconi, R., 2008, “Secrets of the Hoary

Deep”, Johns Hopkins University PressGiacconi, R., Clark, G. W. & Rossi, B. B.,

1960, American Science and Engineering,AS&E-TN-4

Giacconi, R. & Gursky, H., 1963, “AnExperimental Program of Extra-Solar X-rayAstronomy”, Proposal by AS&E to NASA,AS&E-449

Giacconi, R., Gursky, H., Paolini, F., & Rossi,B. B., 1962, Physical Review Letters, 9, 442

Giacconi, R. & Rossi, B. B.,1960, JGR, 65,773

Goldberg, L., 1959, “Summary Report fromthe Committee on Optical and RadioAstronomy”, National Academy of Sciences

Goldberg, L. & Dyer, E. P. Jr., 1961, “Galacticand Extragalactic Astronomy”, Science inSpace L. Brekner & H. Odishaw Editors

Hirsh, R. F., 1979, PhD Thesis University ofWisconsin

Hirsh, R. F., 1985, “Glimpsing An InvisibleUniverse”, Cambridge Universiry Press

Mandel’shtam, S. L. & Efremov, A. I., 1958,“Russian Literature of Satellites”, Academyof Sciences of USSR Vol II

Simpson, J. A. Space Science Board, 1958,“Interim Report of the Space ScienceCommittee of Fields and Particles in Space”,National Academy of Sciences

Thompson, R. W,̇, Buskirk, A. V. & Etter, L.R., 1953, Phys.Rev, 90, 1122

Tucker, W. & Giacconi, R., 1985, “The X-rayUniverse”, Cambridge University Press

Vaiana, G. S. & Rosner, R., 1978, ARA&A, 16,393

Wolter, H.,1952, Ann. Physik, 10, 94

considerations on x-ray astronomy - memorie della...

Documents