Léo RamosAstrophysicist João Evangelista Steiner thought he was happy when studying X-ray astronomy and black holes early in his career at the Institute of Astronomy, Geophysics and Atmospheric Sciences of the University of São Paulo (IAG/USP). During his post-doctoral position at Harvard University, and when he was hired by the Smithsonian Institute as an American government employee, his vision of how science was carried out at a competitive level changed dramatically. Returning to Brazil in 1982, Steiner became an active science organizer and supervisor and an obsessive fighter for improved infrastructure for astronomical research.
The list of his work promoting Brazilian astronomy is long. The modernization of the Pico dos Dias Observatory, the establishment of the National Astrophysics Laboratory (LNA) and decisive Brazilian participation in the Gemini and Soar observatory consortia, both in Chile, are the best known. Steiner was also secretary general of the Brazilian Society for the Progress of Science (SBPC), Secretary for Research Unit Coordination of the Ministry of Science, Technology and Innovation (MCTI), and he directed the Institute for Advanced Studies (IES/USP). Today he is a harsh critic of Brazil’s participation in the European Southern Observatory (ESO).
This work has not slowed his work as a researcher. The years devoted to management since 1982 coexisted with frequent astronomical observations, publication of scientific papers, supervision of precision instruments for observatories and a continuous interest in what happens in the universe, the “biggest and best laboratory there is,” he often says. Now, in 2013, Steiner is eager to begin a study on the center of the Milky Way using a new technology recently installed at Gemini. Married with three children, the astrophysicist born in São Martinho, in the state of Santa Catarina, agreed to this interview with FAPESP.
|University of São Paulo (undergraduate, master’s and PhD)|
|Harvard University (post-doctorate)|
|Institute of Astronomy, Geophysics and Atmospheric Sciences (IAG/USP)|
You come from a very small town in Santa Catarina, settled by Germans. It is true that you only learned Portuguese at age 10?
My great-grandparents were German. They immigrated in 1860, during the War of Paraguay and it appears to have been because of the conflict. Dom Pedro II, who had strong connections with Austria—his mother was Austrian—wanted to populate the coast of Santa Catarina for geopolitical reasons. My great-grandparents came from the Rhine Valley, in Germany, and settled in the Capivari Valley in Santa Catarina. My father’s family came from Koblenz, where the Moselle river meets the Rhine. My mother’s family, the Boeings, came from Bocholt. She was descended from two brothers who left Germany to avoid military service. William went to Seattle and founded Boeing, a company that later began manufacturing airplanes. Werner went to Santa Catarina. The bad part of this story is that I was born on the poor side of the family. In São Martinho I spoke German because that was all that was spoken. Until the Second World War, classes were in German. I learned Portuguese at 10, on my own, because at school there comes a time when you must. But I have never had many scientific associations with Germany. I have visited there more frequently, recently, because I have a son who is an opera singer and lives in Hamburg.
How did your education continue?
Our schools were run by priests and nuns. Those who studied, which was rare, attended one or the other. I went to the school run by priests and my sister to a school run by nuns. Then I came to São Paulo to take the University of São Paulo (USP) entrance exam.
Did you already plan to study astronomy? Were you one of those kids who built things?
No. But I did build a telescope, out of curiosity. I also built a radio, and tried to make a computer that never worked. But I was curious about the universe. My parents were farmers and I remember, when I was about 7, I was clearing the pasture with my mother and I asked her where the world ended. And she said that the end of the world was very far away, farther than Germany. Twenty years later, a relative visited us from Germany. It was a tough trip, when Santa Catarina didn’t even have paved roads. It was dust, curves and mountains. He arrived, came into the kitchen, sat on the first chair he saw and said, “Now I know where the end of the world is!” I felt vindicated. I was already 27. This gives you an idea of how relative things are.
Weren’t you crazy about telescopes?
I was curious about everything. Physics attracted me, because it answered the most fundamental scientific questions. I wanted to study physics in the best physics department in Brazil, and everyone said it was at USP. I came here and enrolled in 1970. At the beginning of my third year, I decided that the best physics laboratory was the universe. Many of the questions of scientific interest could be answered through astrophysics, because anything that involved large scales, heavy masses, large gravitational fields, temperatures, magnetic fields—all the extremes are found in the cosmos. The problem is figuring out how to transform this into a laboratory. This is why we must extract a lot of information. The two fundamental questions in contemporary physics are dark matter and dark energy. Suddenly, we discovered that we knew little about the universe, since it has these two entities that dominate its dynamics. Not that I predicted these things. Not at all. But I was also not incorrect in believing that the universe is a good laboratory.
Did your interest in black holes begin during your master’s studies?
Exactly. X-ray astronomy was just beginning back then. Cygnus X-1 was the first X-ray source discovered in the Cygnus constellation. When they measured the mass of Cygnus X-1, they saw that it was much larger than a neutron star, or a white dwarf, and thus could only be a black hole. This was in 1973. My undergraduate research project was on identifying X-ray sources. In 1974 I began working on my master’s and created a theoretical model for Cygnus X-1. I worked with Prof. José Antonio de Freitas Pacheco, who lives in France now. My master’s thesis was interesting because it was related to discoveries made shortly before that time. Cygnus X-1 was the first candidate to have a black hole. My master’s consisted of building a theoretical model of an accretion disc (a structure made up of diffuse material orbiting around a star or black hole) and calculating the spectrum of X-rays that it would have to emit. I showed that the two matched.
And your doctorate?
For my doctorate, I took this disc model and applied it to white dwarf stars, neutron stars and supermassive and stellar-mass black holes, which are the four situations in which accretion produces the energy released. This energy does not come from a normal star, like the Sun, whose origin is the nuclear fusion that transformed hydrogen into helium and then into other, heavier elements. And the mass differential is transformed into energy. These objects are extremely compact and have a very deep gravitational well, in their form and in their capacity to accelerate material in their gravitational field. Any gas captured begins to spiral inward and the gravitational energy is transformed into kinetic energy, according to the law of conservation of energy. The kinetic energy in the innermost orbits is much greater than in the outer orbits, because the speed is much higher. Consecutive orbits have different speeds and this creates friction, which converts the kinetic energy into thermal energy. The temperature is so high—we’re talking about 100 million degrees—that it emits photons which escape in the form of radiant energy before the matter enters the black hole or neutron star or white dwarf.
Is this a way to know if there is a black hole in the region observed?
At first it was very difficult to determine whether it was a black hole or a neutron star, for example, or even a white dwarf, because they all emit X-rays. They are in a binary system of stars, and thus we can measure the mass of the two components. The white dwarf has an upper limit, the Chandrasekhar limit, which is 1.4 solar masses. For the neutron star, the limit is 3.5 solar masses, which is called the Volkoff-Oppenheimer limit. If the solar mass is greater than 3.5, it is a black hole.
But was this known at the time?
It was known, but the mass measurement was difficult to make. In that period, the X-ray source pulsed in many of the X-ray binary stars discovered. They are the X-ray pulsars. X-rays are produced by the gyrations of the magnetic poles, which emit X-ray beams like a lighthouse. A black hole does not have a magnetic field. So none of these pulsing sources could be a black hole. They had to be neutron stars, as was shown for the vast majority of them. It was a slap in the face, because we thought we had a series of black holes and it would be easy to study them when, in fact, the vast majority were neutron stars. So much so that we only see 20 black holes in the literature now, 40 years later.
Did you also study this in your doctorate? Was it hard to find an advisor?
I was the third doctoral student in astrophysics in Brazil. Prof. Abrahão de Moraes was very well known here at USP, and sent students to France for their doctorate studies. In 1972, shortly after his death, one of his former students—Pacheco—finished his doctorate and returned to Brazil. He was my advisor. Later, others returned from abroad and the community grew.
When did you go abroad?
Right after my doctorate, in 1979. I was very interested in X-ray astronomy. The first satellite capable of detecting X-rays was launched on December 20, 1970, from the coast of Kenya. It was called Uhuru, which is the Kenyan word for freedom. Many binary stars were discovered by this satellite, which was American. At Harvard, I worked with the first X-ray telescope, called the Einstein Observatory. The scientific base was at Harvard, even though the telescope belonged to NASA. At that time, there were no space telescopes. Uhuru was a small and very primitive device for detecting X-ray photons. Einstein was a telescope and had great photographic abilities. It was launched in 1979.
And why did you return to Brazil? You certainly could have found a position in the United States.
I did find a position. I was hired at Harvard by the Smithsonian Institute as a US government employee. It’s kind of a funny story. When Einstein was launched, the images obtained were all out of focus. Something similar happened with the Hubble telescope, years later. The researchers were desperate because they had spent a fortune to make the first large space telescope. NASA put all its staff to work on finding the problem—and it failed. Harvard also tried, unsuccessfully. I was there and did my first scientific work as a Harvard professor on quasars. Then a professor there suggested that I study the telescope’s problem. I said I had never seen a satellite in my life and he said that was exactly why: those who were familiar with them were not able to solve the problem. Who knows, maybe I could? They placed all the computers I wanted and two programmers at my disposal. I began to work on it, day and night, with the right to call the programmers in at any time to do calculations for me. It took me two or three weeks, but I found the solution. I showed it to them and guaranteed that they could take the photographs again and they would be in focus. They had the raw data in files and decided to do a test. They took a very poorly focused image and used the program with the system of 14 equations that I developed and tried it out. It came out perfectly. Actually, it was something simple. They had two optical telescopes on the satellite, which were fixed on two stars. But this telescope moves around the Earth, and the planet’s magnetic field varies. They had created a shield to prevent influence from the magnetic field. What I did was show that this shielding was 50 times worse that what had been specified and that the magnetic interference was causing the focusing errors.
Is this what led to your being hired?
This is an American cultural thing: when you show you are competent, your future is guaranteed. They are very objective and organized. Anywhere else in the world, I would still be just a Brazilian. But there I was the guy who solved the problem. This changes how you are treated. I was already a professor at IAG/USP when I asked for a leave to take a two-year post-doctoral position there, with a grant from FAPESP. The grant ended, I asked for a leave without pay, and they hired me. I stayed there for a year and had to decide whether to stay or come back. I came back for two reasons: the first was because my family wanted to. At that point I was married and had two children. Then, my third was born. The second was that I had never considered not coming back. I was educated here, with public money, in public institutions. I had FAPESP grants at every level. A person who receives a public education, like I did, who becomes a scientist thanks to this, owes the society that supported him. For me, this is a basic belief. I wanted to stay there a bit longer because I knew that when I returned to Brazil, in 1982, I would face a difficult situation.
What was Brazilian astrophysics like at that time?
At Harvard, we had tons of computers. We could calculate anything. When I returned to the Astronomy Department, we had five HP 25 hand calculators. If someone needed one, they had to get in line. We also had the Electronic Computer Center (CCE), which was University-wide, but located at the Polytecnical School. I did my master’s and doctorate there, using a Borroughs 6900 for calculations. I would need to carry boxes full of punch cards and submit them at the desk. They would tell me how long it would take, and after two days, for example, I would return to pick up the print out. We thought it was great. That was before I went to the United States. There I realized that we needed something better, and upon my return I began to galvanize the community to demand change. It was difficult, because many people were against it. They thought it was fine, because they did not realize how behind we were, in general. In 1985, I went to the INPE (National Institute for Space Research), and established an Astrophysics Division. We began to purchase equipment; we bought computers for image processing. My first student, Ivo Busko, did a thesis that included processing astronomical images. When he finished, he went to work at the Space Telescope Science Institute, in the US, and there was a bit of confusion, because he was there and he had designed the software to enhance the images. When the equipment was launched, they also discovered that the images were poor and he was the only person who knew how to process the images. Ivo went there with a post-doctoral grant. When the problem arose, the first thing NASA did was hire him.
With technological advances, everything has become smaller, with better resolution. Did this apply to the engineering of telescopes?
Yes, but there is another key issue—infrared technology—the most difficult range of the electromagnetic spectrum for science. We all emit infrared radiation. So do telescopes. Imagine building a telescope for the visible spectrum and filling it with light bulbs. When observing the stars, the background is very bright. The smartest way to solve this in the infrared spectrum is to build a telescope, launch it into space, and cool it so that thermal emissions are zero, or negligible. But achieving this is leading-edge technology, and very difficult. Iras was the first infrared telescope, launched 30 years ago. The images were blurry, still rudimentary. All infrared space equipment has to be refrigerated with liquid helium. Iras was a big success because it was able to operate for nine months. And this was great, since refrigerating liquid helium in space is extremely difficult. The James Webb, the next space telescope, which will replace the Hubble, will be placed outside Earth’s orbit so it will not be affected. It will have a screen to protect it from the Sun. Since it will be protected from both the Earth and the Sun, its temperature will naturally be very low.
When will it be launched?
Perhaps in 2015. It will be very interesting for us, because we work with data cubes, which we obtain through an IFU (integral field unit spectroscopy) device. Everything our group does here now is in the form of data cubes because they are information-rich. We have developed a series of methods and are becoming a point of reference in this area.
Is this treatment of data cubes something developed at IAG?
Yes, it is ours, my students and I developed it. And various other groups are already using it. We trained some Brazilian groups, as it is very difficult. The material is published, and the software is available, but training is still necessary. If you have a galaxy, traditionally you put a slit on top of it and obtain the spectrum. The scientific information comes from the spectrum. It is different with IFU spectroscopy. We create a matrix of lenses, and place an optical fiber under each. We take all the optical fibers, align them in a slit, and produce a spectrum for each fiber. The computer can reconstruct everything. Then we have X, Y and λ [lambda], the wavelength. Thus we have a three-dimensional array, a cube. The Gemini has two IFUs. One in the optical spectrum and another in the infrared. Both are very good instruments. The Webb will have five.
To work with this method, does the data have to be captured in three dimensions?
Yes. The Americans are worried, because they are still having large problems in dealing with data cubes. The Europeans have more experience and, better yet, they are saying that even the Brazilians have more experience in this area. This is something that can be learned, but after a certain age, a person can find learning new tools difficult. I started this because I was forced to. I helped build the Gemini and Soar telescopes. I was on the Gemini board for 5 years and the Soar board for 12 years. One of Gemini’s advances is that these instruments will be integral field. I planned to be able to do this type of science. When I was given the opportunity to undertake a project using the Gemini and its instruments, I did not hesitate. Now I am going to study galaxies and clusters of galaxies, which can be examined very well with this equipment. I received the first data cube and started working. I started to see lots of problems with the data and I asked for help: who knew how to deal with this? No one knew, anywhere. So my only alternative was to solve the problems. I came up with some methods, but there was still a heck of a lot of programming work to do. I had two students who were great at this, and we developed everything.
So you could say that decoding the data cube information was your contribution?
Conceptually and intellectually, yes. This is all relatively new and these things take some time to be assimilated. We began in 2009.
You were a part of large telescope projects. What was that like?
When I returned to Brazil, we had a recently-inaugurated telescope in Itajubá, at the Pico dos Dias Observatory (OPD). I realized that the telescope was being used in a precarious manner, with photographic plates used for spectroscopy. I began to call for modern instruments in 1982. So much so that I brought the first charge-coupled device (CCD)—a sensor used for digital images—to Brazil. I asked for funding to import a CCD chip, and my request was approved. I received the resources to import the chip, but the Pentagon vetoed it, considering it to be “sensitive technology.” And it was not even being imported from the United States, but from England. I then asked for funding to import an astronomical camera with an embedded CCD. I arranged with the seller to not specify that there was a CCD inside. I sent one of our researchers to help assemble and hide the CCD, and it worked. It was the first chip of this kind to arrive in Brazil, in 1986. It was installed in the OPD, and from that point on, Brazilian astronomers started doing modern science, with CCDs, computers and image processing. Before that, everything was done with pantographic plates, which were high-tech in 1890. In the 1980s, this technology was no longer used in the United States, having been replaced by modern, digital processes. There was another important subtlety. This telescope was managed by the National Observatory, but there were many disputes about control, many conflicts. In Brazil, the tradition was for each group to have its own instrument, each leader had his own “church.” And this doesn’t work in astronomy. So I proposed the establishment of the National Astrophysics Laboratory (LNA) to the National Council for Scientific and Technological Development (CNPq). The CNPq understood the proposal and agreed. It was the first national laboratory in Brazil, in 1985, fifteen years before the second, the National Synchrotron Light Laboratory. And it was a huge struggle, because it involved a new culture and new mindsets. And interests, of course. When interests come into play, people do not always act rationally.
Was the objective for there to be greater sharing of astronomical equipment?
The objective was to have shared infrastructure nationwide. Today, no one talks about national infrastructure— now it is international because no country can fund large projects alone. In order to establish the LNA, we built the equipment, the CCDs, the cameras, and all of this helped to modernize Brazilian astronomy. We published a series of papers based on this telescope and on the technologies we introduced. And everyone in Brazil had full access, because use was free. There was internal competition, but only based on scientific criteria. That was the principle. We created LNA not to be clever, but to survive. By making it shared, we had more resources to invest. We only needed a single investment—albeit large, it was only once. This is the language that the government began to understand. The level of research increased because we were all compelled to compete and manage projects according to best international practices. We learned to do this and this raised the level of Brazilian astronomy. When we became partners in Gemini, it was a kind of recognition of our success with the LNA, although the gap was unbelievably large. That’s when I had the idea of doing something at the intermediate level, Soar.
You began to advocate for the construction of Soar in 1993. Twenty years later, was it worth having it built?
Without a doubt. But things happen slowly in this area. Every telescope project takes at least 12 years to become ready. Starting from the initial idea, through design, planning, numerous commissions and committees … You also need to obtain resources. It takes several years before construction can begin. Then, until the telescope sees the first light, another 12 years. Then it needs to go through a year of commissioning, fine-tuning in order to work well. In other words, you cannot just plug it in and use it. Then, when the telescope is working, there are still problems. It has its instruments and each one is a separate stage. They are expensive, sophisticated, you always want the latest technology and even a little more to be competitive. These instruments take time to build and Brazil had no expertise in this area. For Soar, however, we built three spectrographs, the SIFS, the Steles and the BTFI – the latter by Cláudia Mendes de Oliveira, here at IAG. The BTFI (Brazilian tunable filter imager) is a high-tech piece of equipment that allows analysis of both the chemical composition and the internal relative motions of galaxies. It is ready, and now we can already start doing science. SIFS is a fiber optic integral field instrument. In other words, it took twenty years after Soar was envisioned for us to be able to use it, you have to understand that. And Gemini was no different—it was ready five years before Soar. In order to compare Gemini and Soar, you have to compare Gemini today with Soar five years from now. Soar is still far from reaching cruising altitude. Even Gemini is not there yet. It has already helped us and will continue to help us obtain many results, but some complain that it could produce more, have a greater impact. Still, Gemini produces articles published in Nature and Science at least every two months.
When will Soar reach that level? It still has a lot of critics.
They make sense. We all want to do better. We are evolving, but up to now, the rate has been slower than we would have liked. And this is related, fundamentally, to instrumentation, and not to the telescopes, which are great. One of the instruments that will be very useful for Brazilian science is the high-resolution optical spectrograph, Steles, made for Soar. The British made a similar one, in terms of functionality, for Gemini, but it did not work and they have started to design another. It is a deficiency in Gemini that will be remedied in Soar first. It was supposed to have been sent in November, but, as always seems to happen, there was a problem with one of the parts. Steles has 1,500 mechanical parts. Solving this in Soar before fixing it in Gemini will be a great leap for Brazilian astronomy. In 2013, the Soar’s high-resolution optical spectrograph will be ready. They have not even begun to build Gemini’s. The problems that occurred are real, but they happen with all telescopes. The most important thing, in the case of Soar, is that now we will start using tools built in Brazil, and thus impact and productivity will truly grow. In Gemini, we have a new type of equipment called conjugate adaptive optics, which allows us to correct the distortions in the images produced. It creates a tomographic scan of the entire atmosphere using four lasers. Time periods for astronomers to use this instrument were first distributed in November. Gemini is the first telescope to have this. I have a proposal to study the center of the Milky Way, which was approved. Augusto Daminelli, here at IAG, also had a proposal approved. We will be the first two users. The technology is leading-edge and we have great expectations.
Why do you use queues now?
This began with Gemini and Soar. This is how it works: a researcher who needs a short time for observation tells the astronomers working at the observatory what data he needs to obtain. They perform the observation and send the information to the researcher, who does not need to be there physically. Other observatories, like ESO, do not use this method. In these cases, when the researcher obtains a night, he goes to the observatory and makes his observation. In the case of Gemini, since we didn’t have much time, most projects did not get even one night. To optimize, we decided that the Brazilian projects would be done in a queue. It was an intelligent decision, because we were able to produce three times more papers per hour of observation than other partners, such as the Americans. We have a high level of scientific competitiveness. With Gemini, we decided on the queue method and did not distribute time in the classic manner. With Soar, we did both—the researcher can choose the queue method or the classic method. It depends on the project. For example, I want to take a spectrum of some celestial object using Soar. In order to do this, I need two hours of observation. It is silly to go to Chile for two hours of observation. The queue mode solves this. If I have two nights, I prefer to work from here and not in Chile, because here I can call in the entire team and watch in the observation room we have here at IAG. In Chile we have a technician who performs the necessary operations. He opens the dome, closes the dome, and points the telescope.
Do people still have that romantic image of astrophysicists looking through a telescope, as was done in the past?
When I tell first-year students that telescopes no longer have a viewfinder, students are very disappointed. It’s a shock. Nowadays, astrophysicists work behind a computer connected to highly sensitive cameras. It makes no difference if the telescope is upstairs or on the other side of the hemisphere,. Of course, some people are wild about telescopes. Leaving the dome, seeing that beautiful sky in the Andes, full of stars, is enchanting. But to produce good science and train good scientists, the logic is a little different and must be optimized.
What do you think of Brazil’s participation in ESO, the European consortium of telescopes located in Chile?
In my opinion, what is happening today was totally predictable. Brazil agreed to take part in the consortium at a cost of €255 million, which is almost R$700 million over 10 years; then, we will pay about €25 million per year for maintenance, for the rest of its useful life. I wrote an indignant, six-page letter about this three years ago and sent it to Sérgio Rezende, who was the head of the MCTI and made the decision without prior discussion or evaluation. He responded verbally saying I was wrong. I predicted that we would spend a lot of time and energy, and miss windows of opportunity and, ultimately, find that we do not have that money to spend.
And what happened?
Exactly what I thought would happen. When we looked like we were rich, not too long ago, we did not have this money. Right now our GDP is growing slowly and the derivative is negative. The government will not want to spend this money on the ESO. It is not even in the budget. To be approved, President Rousseff will have to send it to Congress, where it will have to be approved by five commissions and two plenary assemblies. And everyone knows how much our politicians love astronomy… Moreover, it is not just a question of money. With the Soar and Gemini observatories, we contributed X% of the money and use them X% of the time. At ESO, this proportionality does not exist. We pay €255 million for the right to compete with them on unequal terms—with rare exceptions. They are smart and I think they are right, from their point of view. It’s up to us to decide what our best development strategy will be. There are excellent alternatives that would cost at least 10 times less.
And why did the Brazilian government agree to these conditions?
It is hard to understand. The year 2010 was an election year. President Lula and his ministers were convinced that they had put Brazil on the map. And the best proof of this was that we were invited to join the world’s largest observatory, ESO. This is what they said. My analysis is neither political nor ideological. I’m saying that in that particular year, that was their reasoning. I think that’s the best explanation. The problem is that the minister signed a commitment for the rest of the observatory’s life two days before leaving government. And without discussing it with the minister who succeeded him, Aloizio Mercadante. This agreement will subsidize European science and technology with Brazilian taxpayers’ money.
Physicists, especially those who are not astrophysicists, were critical of the large sums of money spent on astrophysics while investment in the new synchrotron most certainly would have had a higher return.
The new synchrotron will be used by scientists in various scientific areas and by a community 20 times larger. With the money promised to the ESO, we could build a new synchrotron ring every five years.
In addition to the scientific aspects of astronomy, you directed the IEA. What made you decide to take that on?
I had some activities related to science and technology policy, I was president of the Brazilian Society of Astronomy and secretary general of the SBPC. I led Brazil’s entry in the Gemini project and I was responsible for much of the construction of Soar. During Fernando Henrique Cardoso’s second term, I was one of the secretaries in the Ministry of Science and Technology, under Minister Ronaldo Sardenberg. In 2003, I returned to USP and a few months later the position opened up at the IEA. I was invited to be on the short list and, imprudently, I accepted. The four years I spent there were not bad, but I do not think I’ve been successful in terms of building a new IEA. The IEA suffers from organizational problems and, in my opinion, the institution should have a strong strategic role within USP. But it will never play this role if USP does not want it to. A close partnership with the Dean’s Office is essential. Regardless, we did some interesting things. The journal Advanced Studies was almost a secret until we were able to put it on SciELO, despite the protests of my colleagues. Today it is the third most accessed journal in Brazil, with over 3 million hits a year. The first and second are in the area of public health. We also held several cycles of studies and debates, but I think it needs to have a more strategic position within the university. Before all that I was director of space sciences at INPE. Most of my involvement with science policy and administration happened due to the need to be able to fight more broadly than as a researcher or a group for better opportunities to do research. The Synchrotron is the best example to illustrate this idea. It is open, public infrastructure that needs large investments, but only in one piece of equipment that must be modernized. When it becomes obsolete, we will need another. That’s what we are going through now. But this type of investment was not part of Brazilian scientific culture. And in astronomy this is very visible. It was this need that led me to other battlefields, shall we say. So much so that, today, I feel I have already fulfilled my obligations. I am very happy writing papers and teaching. I have already accomplished my missions. Now I want to use the telescopes.