Äîêóìåíò âçÿò èç êýøà ïîèñêîâîé ìàøèíû. Àäðåñ îðèãèíàëüíîãî äîêóìåíòà : http://www.naic.edu/~tghosh/a2234/references/srt_report_all_v2.ps Äàòà èçìåíåíèÿ: Tue Apr 24 22:04:39 2007 Äàòà èíäåêñèðîâàíèÿ: Mon Apr 11 15:28:33 2016 Êîäèðîâêà: Ïîèñêîâûå ñëîâà: intracluster gas

The Sardinia Radio Telescope (SRT)
Science and Technical Requirements
Report of the SRT Working Group
IRA 371/05
The SRT Working Group consists of:
J. Brand 1 , P. Caselli 2 , M. Felli 2 (chair), K.-H. Mack 1 , S. Poppi 3 , A. Possenti 4 , I. Prandoni 1 , A. Tarchi 5
1 INAF - Istituto di Radioastronomia, Bologna, 2 INAF - Osservatorio Astrosico di Arcetri, Firenze
3 INAF - Istituto di Radioastronomia, Sez. di Medicina, 4 INAF - Osservatorio Astronomico di Cagliari,
5 INAF - Istituto di Radioastronomia, Sez. di Cagliari
28 January 2005
1

Contents
1 Introduction 4
2 Summary 5
3 General Requirements 6
3.1 Hardware Requirements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6
3.2 Software Requirements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6
3.3 Monitoring campaigns . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7
3.4 VLBI Requirements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8
3.5 The SRT Reference Parameters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8
4 Single-dish observations 10
4.1 Solar system . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10
4.1.1 Line observations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10
4.1.1.1 Radio spectroscopic observation of comets . . . . . . . . . . . . . . . . . . . 10
4.2 Galactic astronomy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11
4.2.1 Line observations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11
4.2.1.1 Galactic Masers with the SRT . . . . . . . . . . . . . . . . . . . . . . . . . . 11
4.2.1.2 Unbiased Surveys of Cloud Cores in NH 3 (1,1) and (2,2) and other molecules 13
4.2.1.3 Detection of DCO + (1{0) and N 2 D + (1{0) to trace the kinematics of high-
density molecular cores . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15
4.2.1.4 Complex Molecules in the Interstellar Medium . . . . . . . . . . . . . . . . . 16
4.2.1.5 The Evolution of Low-Mass Protostars and their Bipolar Out ows . . . . . . 18
4.2.2 Continuum observations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20
4.2.2.1 Active Binaries . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20
4.2.2.2 Measurements of Polarized Diuse Emission with the SRT . . . . . . . . . . 22
4.2.3 Pulsar observations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23
4.3 Extragalactic astronomy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 26
4.3.1 Line observations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 26
4.3.1.1 Extragalactic HI Blind Searches with the SRT . . . . . . . . . . . . . . . . . 26
4.3.1.2 H 2 O Megamasers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27
4.3.1.3 A Redshift Machine for the SRT . . . . . . . . . . . . . . . . . . . . . . . . . 28
4.3.2 Continuum observations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29
4.3.2.1 High-Frequency Extra-Galactic Sky Surveys with the SRT . . . . . . . . . . 29
4.3.2.2 Search for Sources with High Rotation Measures . . . . . . . . . . . . . . . . 32
4.3.2.3 High-Frequency Follow-up to Surveys . . . . . . . . . . . . . . . . . . . . . . 33
4.3.2.4 High-Frequency Mapping of Extended Sources . . . . . . . . . . . . . . . . . 34
4.3.2.5 Multi-frequency Monitoring of long- and short-term Blazar Variability . . . . 34
4.3.2.6 SRT and the Sunyaev-Zel'dovich eect . . . . . . . . . . . . . . . . . . . . . 37
2

5 VLBI 39
5.1 Galactic astronomy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 39
5.1.1 Line observations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 39
5.1.1.1 An Italian VLBI array for the 6.7 GHz CH 3 OH masers . . . . . . . . . . . . 39
5.1.2 Continuum observations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 40
5.1.2.1 Radio Emission from Galactic X-ray Binaries . . . . . . . . . . . . . . . . . . 40
5.2 Extragalactic astronomy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 42
5.2.1 Line observations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 42
5.2.1.1 Mapping of HI absorption regions in extragalactic sources . . . . . . . . . . . 42
5.2.2 Continuum observations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 43
5.2.2.1 Millimeter VLBI Observations . . . . . . . . . . . . . . . . . . . . . . . . . . 43
5.2.2.2 Wide-Field VLBI Imaging and Surveys . . . . . . . . . . . . . . . . . . . . . 44
5.2.2.3 A Study of Faint Extragalactic Radio Sources . . . . . . . . . . . . . . . . . 44
6 Geodesy with the SRT 46
7 Planetary Radar Astronomy 49
8 Space Science with the SRT 54
9 SETI 56
A Appendix 58
A.1 The SRT Receiving System Plan . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 58
A.2 A spectroscopic backend for the SRT . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 65
A.2.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 65
A.2.2 Spectroscopic backend characteristics . . . . . . . . . . . . . . . . . . . . . . . . . . . . 65
A.2.3 Implementation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 66
A.3 Pulsar Facilities . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 68
A.3.1 Overview of state-of-art de-dispersion techniques . . . . . . . . . . . . . . . . . . . . . 68
A.3.1.1 Incoherent de-dispersion using analogue lter-banks . . . . . . . . . . . . . . 68
A.3.1.2 Incoherent de-dispersion using digital lter-banks . . . . . . . . . . . . . . . 68
A.3.1.3 Coherent de-dispersion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 68
A.3.2 A pulsar backend for the SRT . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 69
A.4 A continuum backend for total intensity and polarization for the SRT . . . . . . . . . . . . . 69
3

Chapter 1
Introduction
In May 2004, the director of the Istituto di Radioastronomia (IRA), Prof. G. Tofani, on behalf of the Board
of SRT, asked eight researchers - active in a variety of elds of astronomical research - to form a working
group (WG) which was given the task to outline scientic programs that could be carried out with the
Sardinia Radio Telescope (SRT), and to describe the technical requirements necessary to lead these projects
to a successful outcome. The WG consists of Jan Brand, Paola Caselli, Marcello Felli (chairman), Karl-Heinz
Mack, Sergio Poppi, Andrea Possenti, Isabella Prandoni and Andrea Tarchi.
The installation of a WG was deemed necessary by the Board of the SRT, three years after the SRT
Symposium held in Cagliari on November 2001 (SRT: the impact of large antennas on Radio Astronomy and
Space Science, eds. D'Amico, Fusi Pecci, Porceddu & Tofani, published by SIF), in order to provide further
input from the potential user community in the early phases of the SRT's development, and to present
projects that maximize the scientic output of the SRT and which renders it competitive in an international
context. This input (as well as the results of the aforementioned Symposium) will allow the Board to
take decisions about the rst phase focal plane instrumentation and to anticipate future developments and
requirements.
This report includes contributions from outside groups, the particular areas of interest of which were not
adequately represented within the WG (e.g. planetary astronomy, VLBI, Geodesy and Radio Science). The
emphasis is on scientic themes and on the general features of the equipment necessary for successfully inves-
tigating them. In particular, sensitivity and observation time requirements are outlined, whereas technical
details of the antenna performance and focal plane instruments are beyond the scope of the present report
and are described only in rather general terms. Direct interaction between the scientic community and
those who are responsible for the realization of the hardware and software will be coordinated by the Board
of the SRT at a later stage.
Right from the start, it was felt by the WG that the report should not describe all possible scientic uses of
the SRT, but only those that re ect real interests in the Italian scientic community. Behind each scientic
project presented here, there is a group of researchers ready to actively work with the SRT once it has
become operational.
We have omitted the names of the people who have prepared the various sections to be consistent with the
philosophy that wants this report to be the result of the participation of the entire Italian scientic commu-
nity. Nevertheless, the WG wishes to express its gratitude to the following persons outside the WG for their
contributions and useful comments during the writing of this report: Roberto Ambrosini, Uwe Bach, Marco
Bondi, Maria Teresa Capria, Ettore Carretti, Elena Cenacchi, Claudio Codella, Gianni Comoretto, Gian-
franco De Zotti, Mario Di Martino, Luigina Feretti, Lars Fuhrmann, Carlo Giovanardi, Gabriele Giovannini,
Federica Govoni, Matteo Murgia, Simone Migliari, Luca Moscadelli, Stelio Montebugnoli, Enzo Natale, Luca
Olmi, Alessandro Orfei, Francesco Palagi, Paola Parma, Pierguido Sarti, Gian Paolo Tozzi, Corrado Trigilio,
Grazia Umana, Tiziana Venturi, and Mario Vigotti.
In order to obtain input for this report its rst draft was sent in November 2004 to more than 240
potentially interested scientist of the Italian astronomical community. The accompanying letter can be
found in http://www.ira.cnr.it/mack/SRT/, an electronic copy of this report can be downloaded from
http://www.ira.cnr.it/mack/SRT/srt report all.ps .
January 2005
4

Chapter 2
Summary
The report is organized as follows:
 A section where we provide general requirements, with particular attention to the software issues,
useful for an optimum use of the SRT.
 A large section is devoted to the single-dish use of the SRT, divided into classical astronomical sections:
solar system, galactic and extra-galactic, each one divided in turn into continuum and line.
 Four separate sections describe the projects and requirements for VLBI, Geodesy, Radar Astronomy
and Space Science..
***
New competitive opportunities for the SRT can be opened only following these general criteria:
 for single-dish applications:
{ use of the SRT at the highest possible frequency,
{ implementation of multi-beam arrays,
{ use of bandwidths as large as possible, for the continuum and for spectral line work, so that more
lines can be observed simultaneously, either over a large bandwidth at low resolution or at higher
resolution by centering several smaller bandwidths around dierent frequencies,
{ spectroscopy (galactic and extragalactic) in non-standard bands,
{ lower-frequency projects (as for instance pulsars, polarimetry) with ad hoc hardware.
 for VLBI
{ non-standard frequencies
{ frequency agility
{ Mark V
 for Geodesy: SX receiver systems, Mark V, co-located GPS
 for Radar: radar transmitter
 for Space Science: Doppler tracker
5

Chapter 3
General Requirements
This chapter deals with some hardware and software requirements necessary for a proper and eôcient use of
the SRT. Furthermore we discuss a series of observational programs that need to be carried out both before
and during regular SRT operations, in order to guarantee the collection of `clean' and well-calibrated data.
We would like to stress that software requirements are often underestimated in the development of a
new instrument. Our experience indicates that good software packages for operating the radio telescope and
managing the data ow is essential and of the same importance as a good performance of the antenna and
the focal plane instruments.
3.1 Hardware Requirements
For the operation of the multi-beam receivers we strongly recommend the employment of a built-in mechani-
cal derotator. During the observations the telescope's eld-of-view rotates as it tracks objects across the sky.
While this does not pose a problem in case of single beam (feed) observations, it aects multi-beam `grid
mapping' of extended sources because the sky position seen by each of the o-axis beams changes system-
atically during the observation. A derotator would compensate for this eect, thus permitting simultaneous
(long-integration) observations of as many positions as the number of beams in the receiving system. This
is of particular importance in spectral line studies (especially extragalactic ones) where long integrations are
required on each position in a map in order to obtain spectra with high enough signal-to-noise ratio.
For continuum mapping it often suôces when the derotation of the eld of view w.r.t. to the multi-beam
array can be recovered via software producing an (almost) full spatial coverage of a region of interests.
3.2 Software Requirements
All SRT operations are carried out by special software. Below we give a list of what we think is needed to
ensure proper functioning of the telescope.
1. A telescope drive programme that allows the SRT to point, focus, and to perform skydips (`antenna
tippings'). This software package should also be capable of executing the various observing modes
requested by the scientic projects outlined in this report, such as total power, position -, beam - , and
frequency switching, cross scans, raster scans, coverage of a matrix of points in any type of coordinates
and On-The-Fly maps.
A software package that is able to handle both continuum and line observations within the same
environment would be desirable.
2. Software that constitutes the interface between the observer and the telescope drive programme, i.e.
some template in which to set observational parameters, mapping mode, integration times, resolutions,
bandwidths etc.
3. Data acquisition software that reads out the backends and puts the data into a le, adds a header, and
writes the result to disk. Output les have to be written in a standard format (e.g. FITS, CLASS)
6

and the header should contain all the keywords needed by the various data reduction packages (e.g.
AIPS, Miriad, IDL, GILDAS).
4. On-line data monitoring and -reduction software (for example the acquired data could be written into
a second le, that can be accessed during the observations from a separate workstation). This would
allow the observer to monitor the data in real-time, and hence, to check the general performance of
the system. For this purpose it would also be useful to have several monitors on which the output of
various receiver/backend combinations are shown, so that one can immediately see if for instance the
backends are positioned correctly around the spectral line.
The on-line reduction software should preferably be of a type that allows one to write additional
modules that can be integrated into the main software package (e.g. IDL, GILDAS). Ideally several
reduction software packages could be available at the SRT; in this case software is needed to transform
the (say) FITS-les written by the data acquisition software package to the format required by the
specic data reduction software packages.
5. Remote observing capabilities: although the complexity of the SRT renders completely remote ob-
servations (almost) impossible and the presence of a team of trained operators at the telescope site
mandatory, the possibility for the observer to check remotely (i.e. from his/her home desk) the ow of
data for standard observations (e.g. long-integration spectroscopic observations), should be pursued.
6. Creation of a telescope archive. In this archive one would store information on all observed targets
(name, coordinates, observing modes, frequencies) to avoid unnecessary repeat observations and thus
to optimize the eôciency of the use of the SRT.
A backup archive of all observational data should be kept at the telescope. With time (taking into
account a certain proprietary period), certain parts of the data archive should be made publicly acces-
sible.
3.3 Monitoring campaigns
Before regular SRT operations can start, test observing campaigns are needed, to assure that useful sci-
ence data can be obtained. Once operations have started, continued monitoring of several parameters and
phenomena is required. Below we give an overview of what is needed.
1. The telescope must be able to accurately point to celestial sources. The pointing accuracy required
depends on the observing frequency. We therefore need to have a list of suitable pointing sources
at all operational frequencies. `Suitable' means continuum (or maser) sources, strong enough to be
detected in a short-integration cross-scan; the sources have to be point-like for the SRT at the relevant
frequency, and their distribution on the sky should be as uniform as possible (i.e. at every time of the
day at every azimuth and elevation there should be a pointing source nearby).
Likewise there is the need for continuum and line calibration sources, with known ux density at the
observing frequencies available at the SRT, and these should therefore be non-variable. These sources
are to be used to calibrate the bandpass and polarization (continuum observations). For spectral line
observations strong line sources are needed to verify the correct tuning of the receivers and to check
the centering of the backends.
Before beginning SRT operations, an extensive `pointing' and `calibration' campaign should be carried
out, to construct a pointing model and to establish the reference ux density values for the SRT at all
operational frequencies. During normal operations pointing and calibration sources should be observed
regularly to update the pointing model and to check the reference ux values.
As a start we could utilize the pointing and calibration source catalogues that are in use at the Eelsberg
100-m and IRAM 30-m telescopes.
2. The SRT is located in a `Radio-quiet zone': within a radius of 3 km around the telescope, no structures
can be erected that might adversely in uence the observations, without prior consent of the IRA.
7

Unfortunately there remains the distinct possibility that observations may suer interference from
transmitters that are well outside this radius (e.g. satellites!).
Periodic monitoring of radio interference at certain frequencies at which this is known to be a potential
problem is therefore needed. Software could then be created to either eliminate interference from the
scientic observations by inserting into the receiver a signal similar to the interference signal but with
opposite phase, or that shuts down the receiver temporarily.
Given the increasing problems of `in-house-made interference', caused by the electronic equipment of
modern radio observatories, adequate shielding, e.g. in the form of a dedicated `Faraday room' which
contains the strongest radio emitters, is essential.
3. In order to explore the possibility to use the SRT at higher frequencies (up to 100 GHz), it is very
important to start as soon as possible radiometer measurements at the SRT site of the sky transparency
at various frequencies of astronomical interest.
3.4 VLBI Requirements
With its large collecting area and the suite of planned receiver frequencies the SRT will be a powerful addition
to the various existing VLBI networks.
 For the participation in the European VLBI Network (EVN) frequency agility is very important. In
a possible priority list of the receivers to be built, the most requested EVN frequencies should be
considered rst. On the other hand, the EVN performance at frequencies  8:4 GHz and < 1:4 GHz is
still limited. The addition of the SRT capable to observe at these non-standard frequencies will yield a
signicant increase of sensitivity (factor 1.5 at 1.4 GHz, factor up to 4 at 86 GHz). Also the upcoming
launch of the Japanese VSOP2 should be taken into account when prioritising the receiver list.
 The SRT should be inserted in the Global mm-Array for high-frequency (> 40 GHz) VLBI.
 It should be attempted to establish a special relationship between the SRT and the VLBA similar to
the one already existing between the Eelsberg 100-m telescope and the VLBA.
 To facilitate any `national VLBI' observations (using the three IRA telescopes) an agreement with the
correlator facilities either at JIVE or the MPIfR should be formalised. Alternatively, the construction
of a small national correlator could be considered.
3.5 The SRT Reference Parameters
Reference values for the relevant telescope parameters are listed in Table 3.1. These values, unless stated
otherwise, have been used to compute the time estimates in the scientic projects collected in the present
report. The maximum instantaneous bandwidth (IF max ) of each receiver, reported in the third column of
Table 3.1 is at most 2 GHz; obviously, the actually usable maximum value depends on the frequency-range of
each receiver. In addition, where the frequency-range of the receiver allows this, one can select IF-values of
80, 400, and 800 MHz. Note that each receiver has 2 polarisation channels, for which the IF can be selected
individually. Depending on whether the backend allows this, spectral line observers could simultaneously
observe lines at dierent frequencies, while continuum observers may have a bandwidth of 2IF at their
disposition. Note that SEFD (Col. 8) stands for System Equivalent Flux Density.
8

Table 3.1: Reference values for relevant parameters of SRT receivers 
Receiver RF Band IFmax Rx Noise T Sys Noise T Ant e Ant Gain SEFD HPBW
number > (GHz) (MHz) (K) (K) (%) (K/Jy) (Jy) arcmin
1P 0.31-0.35 40 30 52 59 0.68 76 59.1
2P # 0.58-0.62 40
3P # 0.7-1.3 600
4P 1.3-1.8 500 5 20 59 0.69 29 12.6
5P # 2.2-2.36 160
6P # 2.36-3.22 860
7P # 3.22-4.3 1100
8P # 8.18-8.98 800
1BW 4.3-5.8 1500 15 20 58 0.67 30 3.9
2BW 5.7-7.7 2000 15 21 58 0.67 31 2.9
1G 7.5-10.4 2000 10 16 61 0.70 23 2.2
2G 10.3-14.4 2000 14 29 60 0.70 41 1.6
3G 14.4-19.8 2000 18 48 57 0.66 72 1.1
4G y 18-26.5 2000 21 81 56 0.65 124 0.88
5G 26-36 2000 14 34 54 0.63 54 0.63
6G 35-48 2000 40 60 52 0.61 98 0.47
7G 70-90 2000 91 171 40 0.46 370 0.24
8G 90-115 2000 106 186 35 0.40 460 0.19
 From E. Cenacchi, Laurea thesis, Univ. Bologna. Values refer to an elevation of 45 ô .
> Letters added to the receiver numbers indicate at which focus they will be mounted, and have
the following meaning: `P'='Primary'; `G'='Gregorian'; `BW'='Beam Waveguide'.
# Parameters not yet assessed.
y This is the only receiver already under construction and is a 7-horn system.
9

Chapter 4
Single-dish observations
4.1 Solar system
4.1.1 Line observations
4.1.1.1 Radio spectroscopic observation of comets
Scientic Background
The knowledge of the composition of cometary ices and of their physical properties is one of the main
issues of comet science. It is thought that cometary ices keep trace of the molecular composition of the solar
nebula in the giant planet region where they formed, as well as of the temperature environment in which
they condensed. In the last years, more than 22 molecules, radicals, ions and isotopologues were detected
from radio spectroscopic observations of comets; most of these new species were found due to the coming
of two extraordinary long period comets, C/1996 B2 (Hykutake) and C/1995 O1 (Hale-Bopp). From the
analysis of the new data available it seems that the ices found in C/1995 Hale-Bopp have many similarities
with the ices present in the star forming regions (Bockelee-Morvand & Crovisier 2002), thus suggesting that
the outer parts of the solar system inherited to a large extent the composition of the protosolar cloud, but
more data are needed.
The analysis of spectra from radio observations of comets is a mature eld; kinetic temperature and gas
out ow velocity, derived from the velocity at half-maximum of the lines, are used to estimate the production
rates of the sublimating gases. This is commonly done, for example, for OH (used in turn to derive water
production rate) and for NH 3 (Bird et al. 1997; Gerard et al. 1998; Colom et al. 1999). It must be stressed
that it is not enough to measure the production rate in one or few points close to the perihelion of the orbit,
but it would be highly desirable to obtain production rates spread along the orbit as much as possible. This
requires a reasonably easy access to a suitable radiotelescope like the SRT for short time intervals at selected
heliocentric distances.
Of the many subjects that could be investigated in a comet with a radiotelescope such as SRT, two are listed
hereafter as an example. It is clear that the opportunity to use SRT will give birth to many new ideas in
the cometary community.
 Search for new molecules. There are certainly unidentied lines in the radio domain and new
molecules to be detected at an abundance level with respect to water of 10 3 10 5 (Crovisier et al.
2004): we could cite, for example, water dimer (transition 2 0 E 1 0 E + at 24.284 GHz), deuterated
water (transition 1 10 1 11 at 80.578 GHz) and cyanodiacetylene (HC 5 N , transition 34 - 33 at 90.526
GHz; Bockelee-Morvan & Crovisier 2002). It is also thought that most of the organic molecules
identied in the interstellar medium could be found also on comets; new complex molecules could even
be searched for. The possible links between the formation of organic molecules on interstellar grains,
their incorporation in comets and their possible spreading in the solar system are some of the most
interesting subjects in cometary science.
10

 The ratio HCN/HNC. This ratio, regarded as an indicator for the origin of the ices, has been
measured for few comets. It seems to be very similar to that measured in warm quiescent molecular
clouds (Bockelee & Crovisier 2002): this is suggesting that cometary nuclei can be composed by
relatively unprocessed interstellar ices. Anyway, the strong variations of this ratio along the orbit,
measured in C/1995 Hale-Bopp, cast some doubts on the true signicance of this ratio. More data are
needed.
Although it is true that the Italian community working on such kind of observations is much smaller than,
for example, the one working on stellar formation, this community is nevertheless well-inserted in the inter-
national community. The availability of SRT time would be a wonderful occasion to strengthen and enlarge
our community and to establish cooperations with other communities, sharing knowledge and expertise.
Technical requirements
Cometary lines are narrow and rotational temperatures are low, so there is no problem of spectral confu-
sion. As a general requirement for the observations detailed above, in particular the search for low abundance
molecules, high sensitivity and good spectral resolution ( 0:1 km/s) are needed, along with a comet close
and bright enough.
The search for new molecules requires a continuous frequency coverage by the SRT receivers in the high-
frequency range above 10 GHz (excluding the atmospheric O 2 line region around 50 GHz). The HCN/HNC
ratio is measured at 90 GHz. A spectrometer with standard bandwidths of 10 MHz would be suôcient
for both projects. While the sensitivities for the search for new molecules cannot be predicted, typical
HCN/HNC ratio measurements require sensitivities of about 50 mK. A potential problem could be that, as
in other spectral domains, good molecular databases are necessary but are not always available.
Anyway, most important in order to carry out comet observations is that a kind of \target of opportunity"
policy should exist: usually bright comets arrive unexpectedly and approach Earth at high speed. Moreover,
even usually \quiet" comets can be subjected to unpredictable outbursts, during which their magnitude
increases by several times. It should be possible for our community to make observations on short notice.
References
Bird M.K., Huchtmeier W.K., Gensheimer P., Wilson T.L., Janardhan P., Lemme C. 1997, A&A 325,
L5
Bockelee-Morvan D., Crovisier J. 2002, EA&P 59, 83
Colom P., Gerard E., Crovisier J., Bockelee-Morvan D., Biver N., Rauer H. 1999, EA&P 78, 37
Crovisier J., Bockelee-Morvan D., Colom P., Despois D., Lis D.C. 2004, A&A 418, 1141
Gerard E., Crovisier J., Colom P., Biver N., Bockelee-Morvan D., Rauer H. 1998, Planet. Space Sci.
46, 569
4.2 Galactic astronomy
4.2.1 Line observations
4.2.1.1 Galactic Masers with the SRT
Scientic Background for water masers
Maser emission from the 6 16 5 23 rotational transition of water at 22.2 GHz is a common feature in both
circumstellar shells and in star-forming regions (hereafter SFRs). In both types of sources the maser emission
is highly variable. For the masers associated with Young Stellar Objects this variation is mostly erratic,
while those associated with late-type stars sometimes vary in phase with the luminosity of the central star,
and at other times may show highly irregular behaviour, including spectacular burst events.
11

During the last 15 years a unique monitoring program has been carried out at Medicina, in which water
masers in ca. 20 late-type stars and ca. 55 SFRs have been observed 3 4 times a year. These programs
have provided useful insights in the workings of the maser mechanism in both environments, and allowed
the recognition of correlations between parameters of the maser emission and the source properties. While
these observations have proven to be extremely useful for variation studies, it has also become clear that the
sensitivity, the spatial- and spectral resolution, and the velocity coverage oered by the Medicina telescope
and instrumentation does not favour the study of maser line proles and maser-burst events.
The SRT has a smaller beam size ( 53 00 versus 1. 0 9 for Medicina), which helps in separating dierent maser
sites, which is especially useful in SFRs. The SRT also has a superior quality of the surface of the dish, and a
`cleaner' beam, which results in a higher sensitivity and more eôcient observations. The SRT will therefore
allow the detection of maser emission of lower luminosity objects, thus enabling to more accurately set the
threshold luminosity at which the maser emission can be triggered.
The mere presence of the SRT will permit more frequent observations and therefore provides a denser time-
coverage, which will result in a higher probability of capturing maser bursts.
Scientic Background for other masers
A further advantage of the SRT for galactic maser research is the possibility to observe maser emission
of dierent molecules at various frequencies. For instance OH (1612 1712 MHz) and (eventually) SiO
(43 GHz). This is especially useful for circumstellar masers, as dierent parts of the expanding envelope are
sampled at these frequencies, thus allowing the determination of, e.g., the mass loss history. For SFRs there
is the possibility to observe masers from various inversion transitions of 15 NH 3 (22 24 GHz), CH 3 OH (12,
20, 23, 36 38, 44 GHz and higher), which would be impossible to do with Medicina/Noto.
The SRT as part of a VLBI network for water maser observations
A particularly exciting prospect is that the presence of the SRT will allow one to observe masers with an
all-Italian VLBI network (SRT-Medicina-Noto). Such a combination of telescopes has a maximum baseline
of 894 km, resulting in a synthesized beam size of ca. 2.5 mas. Thus, \in-house" pilot observations of
maser sites in SFRs can be performed, which serve as preparation for proposals to observe these sites with
the VLA or VLBI/EVN. The coverage of the u-v plane provided by the 3 Italian telescopes is shown in
Fig. 4.1.
Technical Requirements
At Medicina, with a system temperature T sys , one reaches an rms of
S Jy in a spectrum with a velocity-
resolution of
v km s 1 , in t ON minutes of ON-source integration time (with an equal amount of time spent
on an OFF-position, in position-switching mode), given by the expression:

S(Jy) = 1:32( 0:178
KtJ )( Tsys
350 K )( 1 kms 1

v ) 1
2 ( 1 min
t ON
) 1
2 .
Here KtJ is the Kelvin-to-Jansky conversion factor, which for Medicina is 0.178 at 22 GHz. Thus, the rms
reachable in a 5 minute ON-source integration, with a spectral resolution of 0.1 km s 1 , and a typical T sys
of 350 K, is 1.9 Jy per channel.
For the SRT we have:

S(Jy) = 8:36  10 2 ( 0:65
KtJ )( Tsys
81 K )( 1 kms 1

v ) 1
2 ( 1 min
t ON
) 1
2 .
Following Table 3.1 we assume T sys  81 K and KtJ  0:65, and therefore we expect to reach an rms of
 0:12 Jy per channel of 0.1 km s 1 in a 5 minute ON-source integration, an order of magnitude improvement
over Medicina, enormously facilitating the detection of fainter maser components. See also Table 4.1 for
dierent spectral resolutions.
For water maser observations a range of spectral resolutions is needed, depending on the specic type of
study one wants to undertake: from 0.25 km s 1 with a coverage of  < 500 km s 1 for detection of weak
outlying features (in SFRs; at large velocity-oset from the velocity of the cloud in which the maser is em-
bedded), to 0.01 km s 1 for detailed studies of line proles. An autocorrelator similar in its setup to VESPA
12

Table 4.1: R.m.s. noise in mJy reachable with the SRT y
t ON
v
0.01 0.10 0.25
(minutes) km s 1
5 374 118 75
10 264 83 53
30 152 48 30
60 108 34 22
y Assuming T sys = 81 K (see Table 3.1)
at the IRAM 30-m would be a good solution, as it allows for a large range of instantaneous bandwidths and
spectral resolutions to be used simultaneously.
Regarding the time requirement, the goal of the water maser patrol can be achieved only if a minimum of
5-6 observing runs per year are granted, each one of 2-3 days.
A very high spectral resolution will be useful for Zeeman measurements of H 2 O masers, where typical mag-
netic eld strengths ( 50mG) lead to line splitting of the order of 50 Hz. A backend such as Montebugnoli's
MSPEC0 would be very useful for this work (64,000 channels in an 8 MHz band, thus a resolution of 125 Hz
at 22 GHz). The possibility to do circular polarization observations is required for this.
Figure 4.1: Simulation of u-v-plane coverage for a VLBI water
maser experiment with the Medicina, Noto, and SRT telescopes.
The simulation is for a 12-hour observation of a source at ô = 30 ô ;
track labels are for the SRT-Noto (1), SRT-Med (2), and Med-
Noto (3) baselines, respectively. (Figure prepared by D. Dallacasa,
Dip.to di Astronomia, Univ. Bologna and IRA, Bologna)
Most recent publications of the water maser group
Brand J., Cesaroni R., Comoretto G., Felli M., Palagi F., Palla F., Valdettaro R. 2003, A&A 407, 573
Valdettaro R., Palla F., Brand J., Cesaroni R., Comoretto G., Felli M., Palagi F. 2002, A&A 383, 244
4.2.1.2 Unbiased Surveys of Cloud Cores in NH 3 (1,1) and (2,2) and other molecules
Scientic Background
Stars form in gaseous condensations (the so-called dense cores) within molecular clouds. The column
density (NH2  10 22 cm 2 ) and gas density (nH2  10 4 cm 3 ) toward these regions are so high that they
13

are opaque both in the visible and to the radiation from CO(1{0) transition, which is the common tracer of
diuse material in molecular clouds (nH2  10 2 {10 3 cm 3 ).
The spectral lines of the inversion transitions of NH 3 at 24 GHz are good tracers of the material in dense
cores of molecular clouds, particularly of those which form low-mass stars (e.g. Ho & Townes 1983; Benson
& Myers 1989). The transition (J,K) = (1,1) has 18 hyperne components separated into ve spectral
groups, thus allowing the determination of the optical depth. Moreover, the frequency separation of the
inversion transition (J,K) = (2,2) is less than 1 GHz from the (J,K) = (1,1) lines, so that both transitions
can be observed with the same telescope and receiver conguration, allowing one to make a more accurate
determination of the kinetic and excitation temperatures.
So far, most of the cores observed in the NH 3 (1,1) and (2,2) transitions were selected from the photo-
graphic plates of the Palomar Sky Survey (Myers & Benson 1983) or through the association with embedded
YSOs detected by the Infrared Astronomical Satellite (IRAS; e.g. Harju et al. 1993; Ladd et al. 1995).
Only few unbiased searches of cores have been done using the high density tracers CS (Lada et al. 1991;
Tatematsu et al. 1993) and in H 13 CO + (Onishi et al. 2002), but from these species it is diôcult to derive
physical parameters of the detected dense cores mainly because of the line optical depth and the fact that
both tracers are likely to deplete within the condensations, when the volume densities become larger than
a few 10 4 cm 3 (e.g. Tafalla et al. 2004; Lee et al. 2003). On the other hand, NH 3 is known to trace
densities as large as 10 6 cm 3 (e.g. Tafalla et al. 2002), so that its observations will give us information
on the whole core material. Moreover, the simultaneous observation of the two inversion transitions gives us
the unique opportunity of determining the gas temperature in dierent regions of the same molecular cloud,
extremely important to test the current calculation of the dust and gas temperature (e.g. Galli et al. 2002).
Finally, a detailed study of the ammonia line proles will shed light on the kinematics of the dense gas and
the molecular cloud complex as a whole, very useful to test the validity of current turbulent cloud models
(e.g. Ballesteros-Paredes et al. 2003).
Although they do not have the same potential as interstellar thermometers as does ammonia, also cyano-
polyynes (HC 2n+1 N, n=1{4) are useful tracers for cloud cores. Several molecules and transitions are easily
observable, as demonstrated by recent detections of HC 3 N(5{4) (45 GHz; Noto), HC 5 N(9{8) and HC 7 N(21{
20) (24 GHz; Medicina; Codella, priv. comm.), and can thus be used for unbiased surveys of cloud cores.
Technical Requirements
The SRT antenna, equipped with a multi-element receiver array, a high spectral resolution autocorrelator
and the possibility of using the frequency switching technique, will be the best instrument to carry out the
unbiased surveys described above. Considering that typical line widths of high density tracers in molecular
cloud cores are around 0.3-0.4 km s 1 , we will need velocity resolutions of at least 0.1 km s 1 (or 8 kHz at
24 GHz). For the unbiased survey, we can however consider a 20 kHz spectral resolution which will allow us to
detect the dense cores. These objects will then be successively mapped with a ner grid and higher spectral
resolution. Assuming a system temperature of 81 K at 24 GHz, a 20-kHz resolution and a total bandwidth
of  20 MHz, we will need about 5 minutes per position to reach a rms of 0.05 K (76 mJy) in antenna
temperature units (or about 0.1 K in brightness temperature units), needed to detect well all the ammonia
cores already known in regions such as Taurus. Considering a Half Power Beam Width (HPBW) of 50 00 and
the multi-beam array, we will need about 5 minutes to beam sample an area of about 20000 arcsec 2 , or 
50 hours to cover 1 square degree. Therefore, large portions of molecular cloud complexes such as Taurus,
Ophiuchus, Perseus, and Orion can be studied in a few days of telescope time. Future SRT ammonia surveys
will thus be compared with recent surveys performed in the millimeter and submillimeter continuum dust
emission (made with SCUBA, the Submillimetre Common-User Bolometer Array at JCMT, the James Clerk
Maxwell Telescope; MAMBO, the Max-Planck Millimeter Bolometer array at the 30m-IRAM antenna; the
Spitzer satellite; and in the future by Herschel PACS { the Photodetector Array Camera & Spectrometer {
and SPIRE { the Spectral and Photometric Imaging Receiver). We should point out that surveys in the Far
Infra Red and mm continuum (i) are limited in sensitivity to the densest and more centrally concentrated
structures (e.g. Johnstone et al. 2004), (ii) have diôculties in detecting extended emission because of the
chopping technique (e.g. Bacmann et al. 2000) and, of course, (iii) do not furnish any information on the
gas temperature and kinematics.
References
14

Ballesteros-Paredes J., Klessen R.S., Vazquez-Semadeni E. 2003, ApJ 592, 188
Bacmann A., Andre P., Puget J.-L., Abergel A., Bontemps S., Ward-Thompson D. 2000, A&A 361,
555
Benson P.J., Myers P.C. 1989, ApJS 71, 89
Galli D., Walmsley M., Goncalves J. 2002, A&A 394, 275
Harju J., Walmsley, C. M., Wouterloot J.G.A. 1993, A&AS 98, 51
Ho P.T.P., Townes C.H. 1983, ARA&A 21, 239
Johnstone D., Di Francesco J., Kirk H. 2004, ApJ 611, L45
Lada E.A., Bally J., Stark A.A. 1991, ApJ 368, 432
Ladd E.F., Fuller G.A., Padman R., Myers P.C., Adams F.C. 1995, ApJ 439, 771
Lee J.-E., Evans N.J. II, Shirley Y.L., Tatematsu K. 2003, ApJ 583, 789
Myers P.C., Benson P.J. 1983, ApJ 266, 309 Onishi T., Mizuno A., Kawamura A., Tachihara K., Fukui
Y. 2002, ApJ 575, 950
Tafalla M., Myers P.C., Caselli P., Walmsley C.M. 2004, A&A 416, 191
Tafalla M., Myers P.C., Caselli,P., Walmsley C.M., Comito C. 2002, ApJ 569, 815
Tatematsu K., Umemoto T., Kameya O., et al. 1993, ApJ 404, 643
4.2.1.3 Detection of DCO + (1{0) and N 2 D + (1{0) to trace the kinematics of high-density
molecular cores
Scientic Background
In the past few years it has been realized that abundant species such as CO and CS disappear from the
densest regions of molecular cloud cores, where protostars will soon form or are just formed (e.g. Caselli
et al. 1999; Tafalla et al. 2002; Belloche & Andre 2004). On the other hand, nitrogen-bearing species
(in particular NH 3 and N 2 H + ) survive in the gas phase until densities of about 10 6 cm 3 because of the
volatility of the parent species N 2 (e.g. Bergin et al. 2001; Caselli et al. 2002a; Tafalla et al. 2004). For this
reason, N 2 H + has been extensively used in recent work to trace the kinematics of the innermost regions of
dense cores and to search for the initial conditions of the star formation process (e.g. Caselli et al. 2002b;
Crapsi et al. 2004).
Even more interesting are the deuterated forms of species such as N 2 H + and HCO + (N 2 D + and DCO + ,
respectively), because their abundances are highly enhanced, relatively to the main isotopomers, in the dense
regions where CO is frozen onto dust grains (e.g. Dalgarno & Lepp 1984; Roberts & Millar 2000). Thus, they
better sample the core nucleus, giving us information on the chemistry (in particular the electron fraction;
e.g. Caselli et al. 1998) and kinematics (contraction motions) of core centers (e.g. Caselli et al. 2002b). We
point out that this applies not just to low-mass star forming regions, but also to high-mass pre-stellar cores,
recently detected by the Midcourse Space Experiment (MSX; e.g. Carey et al. 2000), and several projects
are currently under way to extend the observations of deuterated species in more massive and quiescent cores
(Caselli et al., in prep.).
Previous studies, carried out at the IRAM-30m antenna with bandwidths up to 40 MHz, are based on
observations of the N 2 D + (2{1) and DCO + (2{1) lines, given that so far it was not possible to extend the
3-mm receiver to below 80 GHz (where the J=1{0 transitions of the above two species are found: 72 and
77 GHz for DCO + and N 2 H + , respectively) and also because of angular resolution problems (the half power
beam width (HPBW) of the 30-m antenna at 80 GHz is about 30 00 , comparable to the size of a nearby (D
 200 pc) dense core nucleus). Unfortunately, the DCO + (2{1) line is optically thick and the N 2 D + (2{1)
line has 18 hyperne components heavily blended, from which it is diôcult to extract detailed information
on internal motions. On the other hand, the N 2 D + (1{0) line has seven hyperne components and three
15

well separated groups of hypernes, one of which consists of one single component. DCO + (1{0) also has
hyperne structure detectable with high spectral resolution (  < 5 kHz; Caselli & Dore, subm.). This is
crucial to study the line prole, to determine the infall velocity, and to unveil the radial velocity prole across
the core. Observations of the J=1{0 rotational transition lines will be important to trace the deuteration
process across the core and, together with the available J=2{1 lines, the gas density.
Technical Requirements
The frequency range between 70 and 80 GHz is typically not covered by millimeter antennas, such as
IRAM (although they are now exploring the possibility to extend the 3-mm receiver to 70 GHz). The only
antenna currently able to cover this range with reasonable angular resolution (but not with good sensitivity
and spectral resolution) is the 45-m Nobeyama antenna. The SRT-HPBW between 70 and 80 GHz will be 
15 00 , close to the angular resolution of the IRAM 30-m antenna at 2mm, where studies of the DCO + (2{1) and
N 2 D + (2{1) lines have been performed. Therefore, future SRT observations will be extremely useful to carry
out the proposed observations and to make an important step forward in our understanding of the kinematics
of star forming regions. For these studies, high sensitivities ( 20 mK, 31 mJy) and high spectral resolutions
( 3{10 kHz) will be required to resolve well the line proles and put stringent constraints on available
models of the radiative transfer in contracting clouds. Moreover, we strongly recommend the possibility of
using the frequency switching technique in particular for low-mass star forming regions, where the lines are
narrow and relatively simple, and the wobbler switching technique for more massive regions. A multi-beam
receiver will be ideal to make quick maps and easily extend the kinematic study from core center to edge
(necessary for a detailed comparison with current theories of core and star formation). To map a minimum
number of selected cores we estimate several weeks of telescope time, in good weather conditions.
References
Belloche A., Andre P. 2004, A&A 419, L35
Bergin E.A., Ciardi D.R., Lada C.J., Alves J., Lada E.A. 2001, ApJ 557, 209
Carey S.J., Feldman P.A., Redman R.O., Egan M.P., MacLeod J.M., Price S.D. 2000, ApJ 543, L157
Caselli P., Walmsley C.M., Tafalla M., Dore L., Myers P.C. 1999, ApJ 523, L165
Caselli P., Walmsley C.M., Terzieva R., Herbst E. 1998, ApJ 499, 234
Caselli P., Walmsley C.M., Zucconi A., Tafalla M., Dore L., Myers P.C. 2002a, ApJ 565, 344
Caselli P., Walmsley C.M., Zucconi A., Tafalla M., Dore L., Myers P.C. 2002b ApJ 565, 331
Crapsi A., Caselli P., Walmsley C.M., Tafalla M., Lee C.W., Bourke T.L., Myers P.C. 2004, A&A 420,
957
Dalgarno A., Lepp S. 1984, ApJ 287, L47
Roberts H., Millar T.J. 2000, A&A 361, 388
Tafalla M., Myers, P.C., Caselli P., Walmsley C.M. 2004, A&A 416, 191
Tafalla M., Myers P.C., Caselli P., Walmsley C.M., Comito C. 2002, ApJ 569, 815
4.2.1.4 Complex Molecules in the Interstellar Medium
Scientic Background
There are over 130 molecular species identied to date in interstellar space and a large fraction of them are
organic in nature. Among those with biochemical signicance are H 2 CO (formaldehyde), CH 4 (methane),
HCOOH (formic acid), H 2 C 2 O (ketene), CH 3 OH (methanol), NH 2 CHO (formamide), CH 3 CHO (acetalde-
hyde), CH 2 CHOH (vinyl alcohol), CH 3 COOH (acetic acid), HCOOCH 3 (methyl formate), CH 2 OHCHO
(glycol aldehyde), C 2 H 5 OH (ethanol), (CH 3 ) 2 CO (acetone), and C 6 H 6 (benzene). The simplest members of
several homologous series which are important in terrestrial biochemistry { aldehydes, acids, ketones, and
16

sugars { are also known interstellar molecules. Thus, there is potentially a strong link between interstellar
organics and prebiotic synthesis (Ehrenfreund & Charnley 2000). However, there is still much to be done
before we can answer crucial questions such as: (i) how much of the organic material in primitive Solar
System bodies, such as comets and asteroids, is pristine interstellar material?, (ii) to what extent does it
re ect chemical processing within the primordial nebula?
The search for new complex molecules and the abundance of organic species in dierent astrophysical
environments is a challenge for astrochemistry and awaits sensitive antennas and high spectral resolution
receivers able to span the whole range of frequencies from 1 GHz to the millimeter range, within which
many transitions of the above mentioned asymmetric rotors are present. This implies that the SRT will be
particularly suitable for this search and can be used: (i) to look for complex molecules in cold clouds, (ii)
to search for new species, whose spectra have been recently measured in the laboratory (e.g. Chen et al.
1998; McCarthy et al. 2004), and (iii) to try to identify some of the carriers of the Diuse Interstellar Bands
(DIBs) in the diuse medium.
In the rst case, we will learn how important dust grains are in the production of organic species (the
current view is that organic species are mostly formed on the surface of dust grains until a heating event
{ such as the formation of a protostar { will release the icy mantles in the gas phase; e.g. Caselli et al.
1993; Nomura & Millar 2004). The second project will denitely put stringent limits on the current chemical
models and shed light on the crucial questions mentioned above. For the last project, we point out that
the identity and nature of the DIBs have remained undetermined for over 70 yr (Herbig 1995). There are
suggestions that complex molecules may play an important role (e.g. Tulej et al. 1998; Rue et al. 1999), al-
though no denitive conclusions have been reached. Therefore, it will be a very attractive program to carry
out with the SRT. We point out that this ambitious project requires a continuous information exchange
with chemical research groups, and this is possible thanks to the active collaboration between the Arcetri
group and distinguished researchers and professors at the European Laboratory for Non-linear Spectroscopy
(LENS), associated with the University of Florence, at the Ciamincian Institute of Chemistry (University
of Bologna), at the Department of Physics and Astronomy of The Ohio-State University (Columbus, USA),
and at the Geo- and Radio-Astronomy Division and the Division of Engineering and Applied Sciences of the
Harvard-Smithsonian Center for Astrophysics (Cambridge, USA).
Technical Requirements
High sensitivity (10mK, 15 mJy), large bandwidths (up to 500 MHz) and spectral resolution ( 0.1 km/s)
are required to carry out these projects. Larger spectral coverage is required for the search of new species
and the identication of DIBs (see the recent discovery of complex molecules (propenal and propanal) by the
GBT in the 16 - 26 GHz band. A possibility, at least in the frequency range between 8.8 and 50 GHz, could
be to follow the prescription of the 45-m mm-wave radio telescope of the Nobeyama Radio Observatory, as
described in Table 1 of Kaifu et al. (2004), but a signicantly better backend is required to resolve well
molecular line proles in quiescent regions.
Time requirements: at this stage we can only estimate a minimum of one month of telescope time, in
good weather conditions, to give signicance to the project. But a search will be proposed any time new
frequencies will be measured in the laboratory.
References
Caselli P., Hasegawa T.I., Herbst E. 1993, ApJ 408, 548
Chen W., McCarthy M.C., Travers M.J., Gottlieb E.W., Munrow M.R., Novick S.E., Gottlieb C.A.,
Thaddeus P. 1998, ApJ 492, 849
Ehrenfreund P., Charnley S.B. 2000, ARA&A 38, 427
Herbig G.H. 1995, ARA&A 33, 19
Kaifu N., Ohishi M., Kawaguchi K. et al. 2004, PASJ 56, 69
McCarthy M.C., Thorwirth S., Gottlieb C.A., Thaddeus P. 2004, JChPh 121, 632
17

Nomura H., Millar T.J. 2004, A&A 414, 409
Rue D.P., Bettens R.P.A., Terzieva R., Herbst E. 1999, ApJ 523, 678
Tulej M., Kirkwood D.A., Pachkov M., Maier J.P. 1998, ApJ 506, L69
4.2.1.5 The Evolution of Low-Mass Protostars and their Bipolar Out ows
Scientic Background
Studying the early evolution of stars is diôcult because protostars form in regions which are deeply hidden
in molecular clouds and are surrounded by thick envelopes. As the circumstellar material is dispersed around
a young stellar object (YSO) by the action of its out ows, the spectral energy distribution (SED) of the YSO
evolves systematically. This process allowed the classication of YSOs in Classes 0, I, II, and III (e.g. Andre
et al. 1993). However, it is clear that this classication is too schematic, and it results, for instance, that
under the label of Class 0 sources (e.g. sub-millimeter protostars) we nd a rather heterogeneous collection
of YSOs. Indeed the main obstacle when studying the evolution of YSOs is the identication of a reliable
age indicator.
Out ows could have the key to rene the classication of YSOs, especially at low luminosities where
the ow geometry can be relatively simple. Mm-wave observations, mainly with the IRAM instruments,
have put in evidence the properties of out ows from Class 0 sources, which are highly collimated (see e.g.
Bachiller 1996). Out ows from Class I sources are much less collimated and have a much lower mechanical
power eôciency (Bontemps et al. 1996). The most recent observations show that there is also a kind of
time sequence for low-mass out ows based on well known objects and on the chemical changes that are
expected to be induced in the surrounding molecular out ows. In fact, the propagation of out ows lead to
shock waves which heat the surrounding medium and trigger chemical reactions, including also dust grains
that do not operate in more quiescient environments. Consequently, the abundance of some species (such as
SiO, CH 3 OH, H 2 O and several S-bearing molecules) can be strongly enhanced (e.g. Bachiller 1996, Bachiller
et al. 2001, Codella et al. 2005). Also a direct injection into the gas phase from the dust grains through
sputtering can play a major role (e.g. Caselli et al. 1997). Bachiller & Tafalla (1999) proposed a preliminary
scheme based on a limited number of objects. The empirical time sequence can be summarised as follows
(see also Santiago Garcia et al. 2005): (i) Jet-like out ows driven by Class 0 YSOs and associated with
extremely high velocities (EHV) molecular bullets that appear as secondary components in the spectra, (ii)
Chemically active out ows which are driven by a Class-0 object, show no bullets, but are associated with
very strong chemical anomalies with large abundance enhancements and prominent wings in several shocked
gas tracers, and (iii) Class I out ows with no chemical anomalies and associated with evacuated cavities
and HH objects. In other words, the chemically rich stage should be able to rene the low-mass protostar
classication. In particular, the abundances of several molecules at the high velocities of the out ows can
be used to further specify the characteristics of the Class 0 YSOs.
Unfortunately, the above classication is based on only a few objects, and further unbiased observations
are necessary to make it more rm and detailed. Thus, the chemical anomalies have to be characterised
in a sample of out ows from sources of similar low luminosity ( 10 L ), covering all of this empirical
sequence. The challenge is to nd molecular species (i) whose production is exclusively associated with
the high-temperature shocked gas, and (ii) whose abundance is clearly time-dependent to discern denitely
the Chemical Active Stage from the 1st and 3rd evolutionary phases. The answer can be found thanks to
standard shock tracers such as SiO, CH 3 OH, and HDO as well as to products of the S-bearing chemistry such
as SO, HCS + , OCS, H 2 CS, and SO 2 . Actually, Sulphur chemistry can be seriously aected by grain surface
reactions: shock waves can inject H 2 S into the gas phase with a consequent fast production of SO and SO 2
(e.g. Pineau des For^ets et al. 1993; Charnley 1997). Later on, some S-bearing species such as H 2 CS, OCS,
and HCS + are expected to denitely increase their abundances as a consequence of the injection of sulphur
into the gas phase.
Finally, it is worth to stress the importance of studying the line proles in detail. In the case of molec-
ular out ows, dierent excitation conditions as well as dierent gas compositions at dierent velocities are
expected (Codella et al. 1999, 2003, 2005). This makes essential surveys aimed at studying the chemical
and physical characteristics of the material owing from YSOs as a function of velocity.
18

Technical Requirements
The strategy of this project is to observe a sample of out ows in the lines of shock tracers as well as a
tracer of the ambient medium, like CS, to well dene the rest velocity. The selected molecules emit almost
all at frequencies larger than 20 GHz. When possible, two or three lines of each species will be needed to
discern between excitation and chemistry. The SRT HPBWs will be about 30 00 {48 00 in the 20-50 GHz range
and about 12 00 at frequencies above 70 GHz. These values allow one to map dierent positions located along
the out ow: for instance the size of the typical chemically rich out ow, L1157, is about 5 arcmin, whereas
the out ow associated with the typical Class 0 source, L1448, is 4 arcmin.
The nal list of the lines to observe will be selected in the next years and will be based on the results
of forthcoming projects. In any case, examples of suitable lines can be found in the 20{50 GHz range,
observable with 4G, 5G, and 6G receivers: CS(1{0), SiO(1{0), CH 3 OH(1K {0K ), HCS + (1{0), OCS(2{1), and
H 2 CS(1 01 {0 00 ). On the other hand, by using the 7G and 8G receivers in the 70-110 GHz range it is possible
to observe higher excitation lines: CS(2{1), SiO(2{1), CH 3 OH(2K{1K ), HCS + (2{1), OCS(6{5), H 2 CS(3 13 {
2 12 ), SO(3 2 {2 02 ), HDO(1 10 {1 11 ), and SO 2 (3 13 {2 02 ). In order to: (i) carefully investigate the line prole at
high velocities, and (ii) detect EHV bullets, velocity resolutions of at least 0.2 km s 1 (i.e.  13 kHz) and
bandwidths larger than 100 { 200 km s 1 (6.7 to 13.4 MHz) are needed.
At this stage, only very preliminary time estimates can be made by using the values reported in Table
3.1. By assuming a system temperature of 70 K at 40 GHz, an rms of 0.02 (0.04) K should be reached in
about 30 (5) minutes on-source. At 90 GHz, assuming a system temperature of 180 K one should have an
rms of 0.03 (0.08) K again in 30 (5) minutes on-source. The derived rms values are low enough to detect
the proposed lines with a good S/N ratio, since we expect brightness temperatures larger than 1 K for the
CS, SO, and CH 3 OH lines, and about 0.1{0.5 K for the emission due to the other selected tracers (Codella
& Bachiller 1999, Bachiller et al. 2001, Codella et al. 2003, 2005).
References
Andre P., Ward-Thompson D., Barsony M. 1993, ApJ 406, 122
Bachiller R., 1996, ARA&A 34, 111
Bachiller R., Tafalla M., 1999, in The Physics of Star Formation and Early Stellar Evolution, NATO
Advanced Science Institute, Ed C.J. Lada, Kluwer, Dordrecht
Bachiller R., Perez-Gutierrez M., Kumar M.S.N, Tafalla M., 2001, A&A 372, 899
Bontemps S., Andre P., Terebey S., Cabrit S., 1996, A&A 311, 858
Caselli P., Hartquist T.W., Havnes O., 1997, A&A 322, 296
Charnley S.B., 1997, ApJ 481, 396
Codella C., et al., 2005, in preparation
Codella C., Bachiller R., 1999, A&A 350, 659
Codella C., Bachiller R., Reipurth B., 1999, A&A 585, 598
Codella C., Bachiller R., Benedettini M., Caselli P., 2003, MNRAS 341 707
Codella C., Scappini F., Bachiller R., Benedettini M., 2002, MNRAS 331 893
Pineau des For^ets G., Roue E., Schilke P., Flower D.R., 1993, MNRAS 262, 915
Santiago Garcia J., et al., 2005, in preparation
19

4.2.2 Continuum observations
4.2.2.1 Active Binaries
Scientic Background
RS CVn type systems are close binaries that share a number of common features. In particular, they
display all the manifestations of solar activity (spots, chromospheric active regions, X-ray and radio coronal
emission, ares) but to a greater degree because of the presence of strong magnetic elds generated by eôcient
dynamo action. Those systems have been found to be conspicuous stellar radio sources. The interest to
study them comes from the opportunity to investigate stellar coronae at radio wavelengths and relate to the
radio emission with other activity diagnostics, in solar-like magnetic structures.
The statistical properties of a large sample of RS CVn have led to a self-consistent picture for the radio
emission mechanism interpreted in terms of gyrosynchrotron radiation by electrons of a few MeV or less,
spiraling in elds of 10 to 1000 Gauss. This conclusion is based mainly on the high variability of the radio
ux and the observed moderate circular polarization (which excludes a synchrotron regime) in combination
with the derived brightness temperature, TB  10 91 K. In particular the radio ux density usually shows
two dierent regimes: active periods, characterized by a continuous strong aring which can last for several
days, and quiescent periods, during which the ux density goes down to a few mJy.
Although these stars have been studied very extensively in the radio since their discovery, only during the
past few years radio multi-frequency observations were carried out. Due to the erratic nature of the radio
ares most of the radio observations were carried out during quiescent periods (Jones et al. 1994; White
& Franciosini, 1995). Only very recently Osten et al. (2004) reported on 4-epoch multi-wavelength (radio,
UV, X-ray and EUV) observations of HR1099, including an active period (up to 200 mJy at 6 cm). Still,
the complete evolution of the radio are, from its onset to its decline, is not well documented. In particular,
information on the trend at higher frequencies, where most of the energy releases as well as radiative and
collisional losses are evident, is missing.
VLBI observations with their high-resolution capability are, at present, the only technique capable of
probing the topology of coronal magnetic elds conning the hot coronae of these systems. Recently, VLBA
has provided maps of the radio coronae of RS CVns with unprecedented detail, obtained in both active and
quiescent periods (Beasley & Gudel, 2000).
The presence of magnetic activity on both K and G stars in the RS CVns and their proximity led to
theories involving the existence of a system of loops, connecting the two components, due to the interactions
of magnetic structures of both stars (Uchida, 1986). However it is still not possible to discriminate between
the possibility that the radio emission during ares originates from a magnetic structure surrounding the
binary system and probably formed by the interaction of the magnetic eld of the individual stars, or that
it is originating from large loops connected to only one of the stars.
Since 1991 single-dish 6-cm monitoring of active binary systems, carried out using the 32-m telescope
at Noto, has been conducted. Even if the principal aim of this monitoring program was to activate ad-hoc
VLBI observations once an active period has been detected, the collection of a quite large database has
allowed us to study the variability of radio emission from active binaries on short as well as long timescales.
In the following we will summarize results achieved so far with the single-dish monitoring at Noto to
evaluate the impact of the SRT on the study of the radio emission from active binaries. If we compare the
actual performance of the Noto telescope at 5 GHz, with those expected for the SRT, an improvement of
at least a factor of 5 in the theoretical sensitivity is foreseen. Therefore, it should be possible not only to
extend the sample of sources to be monitored, which are now limited to the brightest systems, but also to
follow, with high time-resolution, the development of the most energetic events. Furthermore, the frequency
agility will allow us to gather the spectral information, both in quiescence and in aring state, necessary to
fully understand the physics of coronal plasma.
One of the main scientic results of Noto active binaries monitoring was to assess the existence of extended
periods of activity, during which ares occur one after the other and the radio ux never reaches its quiescent
value.
These kinds of aring events have been observed in HR 1099 (Trigilio et al. 1993, Umana et al. 1995)
and UX Ari (Trigilio et al. 1998). The active periods can last from a few days up to several months, and
are characterized by a sequence of multiple 4-5 hour ares of variable intensity. The existence of very long
quiescent periods is a piece of observational evidence against recent models of the quiescent radio emission
20

from active binaries (Chiuderi et al. 1993), which foresee very high aring rate. Systematic radio observations
will also shed light on open questions such as morphology of radio emitting regions and possible correlation
with other diagnostics of magnetic activity.
UX Ari has been observed, with a good temporal coverage, in dierent epochs during strong active
periods. During a very active period, with a continuous series of radio ares of ux densities up to 350mJy,
the radio emission consists of two dierent components: a long-term modulation with an amplitude of
200  300 mJy and time-scale of few days and a succession of aring events. The amplitude of the radio
ares and their time of occurrence do not depend on the orbital phase, suggesting that the region where
they originate is always visible from Earth. Such a behaviour has been modelled in terms of a polar spot
as origin of the radio ares plus an active region, located in the corona of the K star and systematically
occulted by the K star itself, as the source of the modulated component (Trigilio et al. 1998).
While a rapid succession of intense radio ares is always visible during active periods, the modulated
emission is not always present. It was not observed during another epoch of UX Ari observations in 1995,
but was present again in 2000 with a minimum at similar phase as in 1993. This seems to indicate that
coronal magnetic structures, large enough to determine a rotational modulation, are not always present in
the corona of UX Ari; however, during strong aring periods, they appear to form always in the same part
of the K star hemisphere, where they are regularly occulted by the star itself at orbital phase 0.4. This does
not support the hypothesis that the emitting regions are related to interbinary magnetospheres.
Radio observations of active binaries oer the opportunity to study the coronal plasma and to relate the
radio emission with other activity diagnostics, in solar-like magnetic structures.
The decay phase of a giant are was followed, for the rst time, in H and in the radio (6 cm) in UX Ari
(Catalano et al. 2003). Since H is formed in the chromosphere this result indicates that the are involves
the whole stellar atmosphere.
With this kind of observations the non-thermal energy emitted by the radio corona, signature of high-energy
non-thermal electrons, and the chromospheric thermal H-losses can be computed, providing important
constraints on energy release models operating during ares.
Technical Requirements
 Development of specic acquisition software:
{ Scanning technique: The study of highly variable events requires high sensitivity. The rms
of the ux density with the Noto telescope at 5 GHz (with a T sys  35 K at zenith), obtained
observing with the on-o technique, is about 20{30mJy. Preliminary results of a prototype of
an acquisition software, based on the raster scan technique, at Noto indicate an rms even better
of 10 mJy in less than 30 min of integration time. This promises a rms of about 2 mJy with the
SRT, which makes the SRT a powerful telescope for this kind of study.
{ Frequency agility: The acquisition software packages should give the possibility to analyze the
data on-line in order to recognize the onset of a are event. In this case, the construction of
dynamic spectra is very important. This requires the possibility to switch quickly receivers in
order to obtain microwave spectra in the range from 1.6 to 45 GHz in less than 30 min.
 Multi-Beam receivers: The sensitivity would be clearly improved in the case of multi-beam receivers,
that eliminate the uncertainty due to the uctuations of the atmospheric brightness. This is particularly
important at high frequencies ( > 15GHz).
References
Beasley A.J., Gudel M. 2000, ApJ 529, 961
Catalano S., Umana G., Cafra B., Frasca A., Trigilio C., Marilli E. 2003, in The Future of Cool-Star
Astrophysics: 12th Cambridge Workshop on Cool Stars, Stellar Systems, and the Sun eds. A. Brown,
G.M. Harper, and T.R. Ayres, (University of Colorado), 2003, p. 981-985.
Chiuderi Drago F., Franciosini E. 1993, ApJ 410, 301.
21

Franciosini E., Chiuderi Drago F. 1995, A&A 297, 535
Jones, K.L., Stewart R.T., Nelson G.J., Duncan A.R. 1994, MNRAS 269, 1145
Osten R.A., Brown A., Ayres T.R., Drake S.A., Franciosini E., Pallavicini R., Tagliaferri G., Stewart
R.T., Skinner S.L., Linsky J.L. 2004, ApJSS 153, 317
Trigilio C., Umana G., Migenes V. 1993, MNRAS 260, 903
Trigilio C., Leto P., Umana G. 1998, A&A 330, 1060
Uchida Y. 1986, Astro. Spa. Rev. 118, 127
Umana G., Trigilio C., Tumino M., Catalano S., Rodono M. 1995, A&A 298, 143
White S.M., Franciosini E. 1995, ApJ 444, 342
4.2.2.2 Measurements of Polarized Diuse Emission with the SRT
Scientic Background
The polarized diuse emission of our Galaxy provides us with information of the large-scale structure
of the galactic magnetic eld. Moreover, it allows the investigation of structures typical of the polarized
emission like Faraday screens and regions with rapid magnetic eld changes, which generate structures only
in polarized emission.
The study of the diuse polarized emission is of great importance also for Cosmic Microwave Background
Polarization (CMBP) investigations. The galactic synchrotron radiation is expected to be the leading con-
taminant for CMBP experiments and its study at radio frequencies, where it is dominant, allows a clearer
view of its contribution.
The diuse polarized emission has been explored mainly around the galactic plane and at frequencies lower
than 2.7 GHz. The high galactic latitudes are less explored: to-date the investigation has been carried out
only up to 1.4 GHz, at which a complete mapping of the sky is almost completed by Wolleben et al. (2004)
and Testori et al. (2004). These data, not yet released, will allow the exploration of the large-scale structure
of the polarized emission down to the degree scale with a sensitivity of 15 mK.
However, at this low frequency the Faraday eects are still signicant, and surveys at higher frequencies
are desirable to investigate the galactic synchrotron emission in a condition of negligible Faraday eects.
Bernardi et al. (2003) show how this situation should be realized starting from about 5 GHz.
A large survey at 4.8 GHz, starting from the galactic plane, can thus signicantly enhance our knowledge and
provide signicant input to the study of both the galactic magnetic eld structure and the contamination
of the CMBP by synchrotron emission. The survey will consist of about 10 ô  10 ô maps, measuring the
relative values of the Stokes parameters within the observed patch (mean values in each patch are lost, as is
typical of this type of surveys). In combination with 1.4- and 2.4-GHz results it would allow the evaluation
of Faraday rotation eects and other important features like the frequency behavior of the polarized galactic
synchrotron emission.
The high sensitivities of the SRT receivers make the telescope suitable for this project.
Technical Requirements
The technical requirements can be summarized as follows:
 Polarimetric backend. A correlation polarimeter is needed for the realization of this project. In
particular, it has to perform the correlation between the Right- and Left-Handed circular polarizations
to allow the simultaneous measurement of both the Stokes Q and U parameters.
 Sensitivity. Assuming the mean spectral index  = 2:8 of the diuse synchrotron emission in the
1.4-10 GHz range (Platania et al. 1998), a survey at 4.8 GHz requires sensitivity of  4:8
px = 0:5 mK
(0.74 mJy), well achievable in 1 second by a receiver having T sys = 20 K and 1500 MHz bandwidth. A
wider band would, in any case, be preferable. Sensitivity requirement is thus  4:8
0 = 0:29 mKs 1=2 (0.43
mJys 1=2 );
22

 Instrumental polarization. On-axis instrumental polarization should be less than 1%; O-axis
instrumental polarization should be less than 1%, achievable with a cross-polarization pattern of the
optics better than 35 dB.;
 Scanning capability. With a single-dish antenna the maps are made by scanning the sky in or-
thogonal directions. This requires antenna software and hardware allowing precise pointing during the
scans (10{20 arcsec). The polarimeter would also acquire the data at an appropriate rate: considering
a sampling better than two pixels per beam, a FWHM = 3.9 0 beam size at 5 GHz, and an antenna
speed of 240 00 /s, the proper sampling rate frequency is SMP >40 Hz (as for the device installed at
Medicina). High antenna speed allows a fast coverage of the sky to be mapped.
 Data Reduction Software. Maps of both diuse emission and discrete sources require calibration
and map-making software. The one currently under development for the Medicina telescope will satisfy
the requirements of a SRT-like telescope.
 Time necessary. To map the Northern Sky, about 530 hours at 1.5 GHz bandwidth) of eective
integration time are necessary. As typical of such large surveys the project can be carried out in 2{3
years.
References
Bernardi G., Carretti E., Cortiglioni S., Sault R.J., Kesteven M.J., Poppi S. 2003, ApJ 594, L5
Platania P., Bensadoun M., Bersanelli M., de Amici G., Kogut A., Levin S., Maino D., Smoot G.F.
1998, ApJ 505, 473
Testori J.C., Reich P., Reich W. 2004, A Large-Scale Radio Polarization Survey of the Southern Sky
at 21cm. The Magnetized Interstellar Medium, p. 57{62.
Wolleben M., Landecker T.L., Reich W., Wielebinski R. 2004. The DRAO 26-m Large Scale Polariza-
tion Survey at 1.41 GHz. The Magnetized Interstellar Medium, p. 51{56.
4.2.3 Pulsar observations
Scientic Background
After nearly 40 years since the original discovery the pulsars rapidly rotating highly magnetized neutron
stars keep on having many exciting scientic applications, in elds ranging from ultra-dense matter physics
to relativistic gravity, cosmology and stellar evolution. A striking example has been the conrmation of the
existence of gravitational radiation, as predicted by Einstein's general theory of relativity. In the last 10
years, the Italian Pulsar group has carried out a series of successful pulsar experiments using the Parkes
64-m dish in Australia. In 1996 a large-scale survey of the southern hemisphere at 430 MHz (Manchester
et al. 1996, D'Amico et al. 1998) discovered more than 100 new pulsars, including 20 millisecond pulsars.
More recently, using a new generation 1.4-GHz multi-beam receiver, the Italian group has been involved in
an unprecedented boom of radio pulsar counting, doubling the number of known objects in the galactic eld
(D'Amico et al. 2001a; Manchester et al. 2001; Kramer et al. 2003). A deep search of the Globular Cluster
(GC) system has found 12 millisecond pulsars in 6 GCs for which no associated pulsars were previously
known, contributing a 25% to the number of clusters containing known pulsars (Possenti et al. 2003). A
high latitude survey for millisecond pulsars has found many new interesting objects, including the rst ever
known double-pulsar (Burgay et al. 2003; Lyne et al. 2004). In fact, increasing the pulsar counting allows
the discovery of many objects which are intrinsically rare in the population, but very interesting for their
physical applications. The Italian group has identied several young energetic pulsars, relativistic binary
systems, binary pulsars with a massive star companion and millisecond pulsars in tiny orbits with a body
of planetary mass. Future deep searches with similar equipment would eventually open the possibility of
detecting a pulsar orbiting a black hole.
In this scientic scenario, the Italian pulsar group proposes to use the SRT with two initial aims.
A) Search for and modelling of millisecond pulsars in the Galaxy and in the GC system. In fact millisecond
pulsars can be considered as test masses for probing gravitational eects and most of them are also extremely
23

stable clocks, allowing for accurate measurements of their rotational parameters, position and apparent
motion in the sky. Discovering more millisecond pulsars will allow one to address many interesting (astro-)
physical issues, ranging from the neutron star Equation of State (Cook et al. 1994) to the binary evolution
(with emphasis on the eclipsing millisecond pulsars, Nice et al. 2000; D'Amico et al. 2001b), to the population
statistic of this kind of sources (Lyne et al. 1998). When detected in GCs, millisecond pulsars prove to be
valuable tools for studying the GC-potential well (D'Amico et al. 2002), the dynamical interaction in the
GC-core (Colpi, Possenti & Gualandris 2002), the neutron star retention (Rappaport et al. 2001) and the
gas content in a GC (Freire et al. 2001).
B) Understanding gravity and gravitational waves. On the one hand, this can be done with the search and
follow-up of highly relativistic binary systems, similar to (or even more extreme than) the double-pulsar; on
the other hand with the timing of millisecond pulsars on the long term. In particular a timing array of many
millisecond pulsars can be used for detecting gravitational waves (Hellings & Downs 1983).
Technical Requirements
Given the aforementioned scientic framework and accounting for the pulsar projects ongoing at other
major radio-telescopes, the Italian pulsar group believes that a development plan for a competitive pulsar
research activity with the SRT on a short-medium term should be based on systems operating at the three
frequencies described in the following:
System 1: 325-MHz (low frequency) observations: The Radio Frequency Interference (RFI) environment
at the SRT site indicates that the 325-MHz band is relatively free of interferences (much more than the
408-MHz band) and that a relatively large clean bandwidth is available. Depending on the dynamic range
and robustness of the receiver system, a bandwidth up to 80 or even 100 mHz could be exploited. A 325-MHz
low noise cooled receiver system would be ideal in order to undertake large scale surveys at high galactic
latitude. Such a survey would probe the population of millisecond pulsars with unprecedented sensitivity in
 100 days of observation. The system should be equipped with a 1024,2048 or even 409632-kHz lter-bank
for each polarization (the amount of channel depending on the actual RFI situation). The same receiver,
equipped with a coherent de-dispersing system (like those developed at CalTech and Swinburne), can also
provide excellent performance in high precision timing observations.
System 2: 1.3{1.8 GHz (intermediate frequency) observations: As a short-term plan for a 21-cm pulsar
system, we propose to build a high resolution (210240.50 MHz) de-dispersing system to be used with the
1.3{1.8 GHz receiver already planned. Such a system can be very well suited for a deep search of the GC
system (requiring some tens of days of observations) and other selected targets (for instance SNRs), and
can be used for regular timing observations of non-millisecond pulsars. The outstanding results obtained at
Parkes at 21 cm, strongly suggest that this is a prime frequency for pulsar search. However, the key-feature
of the success of the 21-cm Parkes experiments was the availability of a 13-element multi-beam receiver. The
slightly shorter focal ratio of the SRT (f/0.34) compared to that of Parkes (f/0.4) constraints a multi-beam
receiver to probably no more than 7 beams: for covering a given sky area, the typical integration time would
then probably be reduced by a factor of 2 compared to that adopted in the Parkes survey (35 min). Thus,
in order to keep the same sensitivity, a bandwidth twice as large as the Parkes' one (288 MHz) should be
considered. The 1300{1800 MHz frequency interval in Sardinia is nominally relatively interference-free, but
the lower end of the band is very close to the frequency of a civil aviation radar system. This is the typical
situation that needs to be checked with a systematic RFI campaign, rather than a single shot, as the eective
impact of these spurious eects might strongly depend on azimuth, and on the time of the day. Furthermore,
we are now convinced that a state-of-the art 21-cm pulsar system should be equipped with a much higher
frequency resolution than adopted at Parkes (3 MHz), which strongly limited the discovery of millisecond
pulsars. With a double band per beam and a much larger number of frequency channels in the de-dispersing
system, such a system would require the development of a major backend. In summary a careful analysis of
the eective RFI environment and a comparative evaluation of the scientic interest of other groups should
be considered here before taking a decision about a 21-cm multi-beam receiver.
System 1+2: The case for a dual frequency receiver: The state-of-the-art for observations in the context
of the pulsar timing is represented by dual band receivers: infact they provide the large frequency baseline
(  > 60% of the upper frequency) and simultaneity necessary to measure with high accuracy the dispersion
delay of most millisecond pulsars, and estimate with ultra-high precision (rms 0.1{1.0 sec) any secular or
transient eect aecting their pulse arrival times at innite frequency. These receivers are also unrivalled
24

instruments for studying the class of the so-called eclipsing pulsars (Nice et al. 2000), the emission of giant
pulses (Cairns 2004) and the insterstellar medium along the line-of-sight to a radiopulsar (Bhat et al. 2002).
Recent technology allows to minimize the loss of eôciency (few percent at worst) with respect to single band
receivers operating at the same frequencies and hence dual band systems also optimise observation times and
telescope scheduling: a given sensitivity in both the bands can be attained in a much shorter observation
time than that required if observing at the two frequencies separately.
In particular the availability of a dual frequency receiver for performing pulsar timing observations has
become of paramount importance in recent years, within the framework of the international programme
called Pulsar Timing Array, for which there is a strong pressure for SRT to be involved in. Other big
radiotelescopes collaborating with this project have been already equipped (or are going to be equipped)
with dual band systems: at Parkes has been recently installed a 10cm/50cm receiver whose performances
are signicantly better than those of the two receivers previously available at those frequencies (and compa-
rable with the best single band receivers at 10 cm or 50cm installed elsewhere). At GBT the construction
of a dual 340MHz/820MHz feed capable of performing simultaneous observations at both bands has been
highly recommended, being extremely useful scientically (Report from the NRAO-GBT Pulsar Workshop
of November 2004) and largely preferable over other kinds of dual feed systems (Report from the NRAO-
GBT Pulsar Workshop of November 2004). In view of this, a natural conguration for SRT would be that of
a 90cm/21cm receiver, which would allow unique investigations by itself but might in turn overlap with the
Parkes (and future GBT) systems in term of frequency baseline, signicantly improving the overall precision
of the international Pulsar Timing Array. Of course this dual band receiver would replace the two systems
(System 1 & System 2) mentioned before.
System 3: 3 GHz (high frequency) observations: Relatively high frequency ( 3 GHz) is also very useful
to undertake a program (typically requiring 6 hrs every 2 weeks, for a total of 6 days a year) of extremely high
precision timing observations like those carried out in the case of millisecond pulsars, or in the measurements
of relativistic eects in double neutron star systems. This frequency choice requires a rather large bandwidth
(  > 1 GHz), in order to compensate the lack of ux due to the steep spectra of pulsars. According to the RFI
environment, a relatively clean bandwidth is available at the SRT site in the frequency interval 3{4 GHz.
This is a bit higher than the typical centre frequency ( 3 GHz) adopted for these applications, but still
well-suited. The pulsar backend necessary for these applications should be a digital lter bank (10241
MHz) similar to that developed at Parkes. The decision of the focal location of such a receiver is a matter of
compromise with other scientic interests. For pulsar applications, side-lobe suppression and spill-over are
not of paramount importance, while the key parameter is the eective antenna gain; so concerning pulsar
research, this receiver with mono-feed could be better accommodated in the primary focus. On the other
hand, should the SRT Board decide for the construction of a BW focus receiver operating at a similar
frequency, the Italian pulsar group could cope with that.
References
Bhat N.D.R., Gupta Y. 2002, ApJ 567, 342
Burgay M., D'Amico N., Possenti A. et al. 2003, Nature 426, 531
Cairns I.H. 2004, ApJ 610, 948
Colpi M., Possenti A., Gualandris A. 2002, 229, 409
Cook G.B., Shapiro S.L., Teukolsky S.A. 1994, ApJ 424, 823
D'Amico N., Stappers B.W., Bailes M. et al. 1998, MNRAS 297, 28
D'Amico N., Kaspi V.M., Manchester R.N. et al. 2001a, AIP Conference 587, 555
D'Amico N., Possenti A., Manchester R.N. et al. 2001b, ApJ 561, L89
D'Amico N., Possenti A., Fici L. et al. 2002, ApJ 570, L89
Freire P.C., Kramer M., Lyne A.G. et al. 2001, ApJ 557, L105
25

Hellings R.W., Downs G.S. 1983, ApJ 265, L39
Kramer M., Bell J.F., Manchester R.N. et al. 2003, MNRAS 342, 1299
Lyne A.G., Manchester R.N., Lorimer D.R. et al. 1998, MNRAS 295, 743
Lyne A.G., Burgay M., Kramer M. et al. 2004, Science 303, 1153
Manchester R.N., Lyne A.G., D'Amico N. et al. 1996, MNRAS 279, 1235
Manchester R.N., Lyne A.G., Camilo F. et al. 2001, MNRAS 328, 17
Nice D.J., Arzoumanian Z., Thorsett, S.E. 2000, ASP Conf. Ser. 202, 67
Possenti A., D'Amico N., Manchester R.N. et al. 2003, ApJ 599, 475
Rappaport S., Pfahl E., Rasio F.A., Podsiadlowski P. 2001, ASP Conf. Ser. 229, 409
4.3 Extragalactic astronomy
4.3.1 Line observations
4.3.1.1 Extragalactic HI Blind Searches with the SRT
Scientic Background
These years have seen the completion of the whole-sky HI (21 cm line) blind surveys performed at 70-m
class antennas equipped with state of the art multi-beam receivers. This remarkable eort was aimed, mainly,
at determining the shape of the HI mass function. In fact, one of the principal discrepancies between Cold
Dark Matter (CDM) structure formation theory and the real world is the large dierence between the large
predicted abundance of satellite halos around giant galaxies and the scantly observed satellite population.
In other words, there is no agreement on the logarithmic slope of the faint end of the galaxy mass function.
Since this is hard to measure directly, eorts have concentrated mainly on the optical luminosity function,
and on the HI mass function. In this way limits can be put at least on halos with sizeable contents of stars
and/or gas.
Another diôculty of the `missing satellite problem' is to determine the possible changes of the mass
function with the environment. Recent studies on the optical luminosity function show denite variations
of the faint end in diering environments: steeper faint ends (more dwarfs) are found in dense clusters. On
the contrary, the results on the HI mass function indicate that the population below 10 8 M  is depressed in
clusters respect to the eld environment. Obviously, we are still far from a comprehensive understanding of
the building process of the galaxy population.
Anyway, even the most extended, in terms of sampled volume, of the available HI surveys (HIPASS,
Parkes + HIJASS, Jodrell Bank) sample a lower mass limit just below M(HI) = 10 8 M  and no galaxies
have been reliably detected with M(HI) < 10 7 M  , independently of distance. A vast improvement in
the statistics below 10 8 M  is necessary to resolve the controversy over the faint end slope of the HI mass
function.
Presently, even for a dedicated antenna, it would be too demanding, in terms of telescope time, to embark
on whole-sky surveys deeper than those just completed. The projects in this area are therefore focusing on
deep blind surveys in selected areas. There is therefore a good chance for instruments of the 70-m class to
produce interesting science by scanning with long integration times restricted areas with a focused rationale.
Technical Requirements
The instrumentation required is quite standard nowadays: a spectral backend capable of a coverage of
 15000km s 1 , that is 70MHz, and a resolution of  20 km s 1 , that is 1000 channels per polarization.
With modern frontends (T sys  20 K), the noise expected at a 65-m antenna would be about  3 mJy
(channel to channel) with 5000 s of integration time. In practice this translates to a 2- detection limit of
 2:5
10 7 M  of HI in the Virgo cluster (17 Mpc), that is a large distance for the problem at hand.
26

This would be a remarkable result if it could be pursued in an area of some square degrees. No time would
be lost in o-scans since these would be acquired in adjacent elds as part of the project. The observations
should of course be coordinated with the other teams working on the subject AND, as in all blind surveys, it
could greatly prot of the availability of a multi-beam 1.4-GHz receiver like those presently used at Parkes,
Jodrell Bank and Arecibo, just to cite a few.
References
For more info on the subject and related references see for example the extragalactic white-paper at
http://alfa.naic.edu/.
4.3.1.2 H 2 O Megamasers
Scientic Background
This project aims at using the SRT to detect 22-GHz H 2 O masers within the nuclei of galaxies. In some
nuclei, the water masers populate accretion disks around the super-massive black holes (SMBHs), lying as
close as 0.1 pc to the nuclear `monster'. Once discovered, Very Long Baseline Interferometry (VLBI) can be
used to map the angular distribution of maser features throughout individual disks, and hence, to obtain a
direct imaging of disk structures typically obscured or not resolved by optical and infrared observations (e. g.
Miyoshi et al. 1995; Greenhill et al. 1995). So far, no other astronomical technique can provide images of
accretion disks in such a proximity to the SMBHs. Depending on the properties of the galactic nucleus (e.g.,
orientation to the line of sight and black hole mass), from such images important physical quantities can be
derived with extreme accuracy, for example: accretion rate, degrees of disk warping and orientation, and the
cosmic distance. The latter is particularly important because the technique used to obtain the distance is
entirely geometric and thereby independent of the luminosity calibrations that aect almost all other distance
measurement techniques. A `geometric distance' is available for the galaxy NGC 4258 (see Herrnstein et al.
1999), and it represents the most accurate distance measurement so far. Hopefully, geometric distances to
NGC 4258 and several more galaxies will ultimately provide anchors in the local universe, against which
calibration of the extragalactic distance scale can be checked.
Technical Requirements
For the project previously outlined we can prot by the 22 GHz multi-beam receiver option, planned for
the SRT, performing internal beam switching observations without losing time on the o position. We recom-
mend a spectrometer that (at least) supports simultaneously two 200 MHz bands in two polarizations with
8192 spectral channels each. Alternatively, the SRT could be provided with a digital correlation spectrometer
(e. g. that the CfA SAO4K; http://cfa-www.harvard.edu/ lincoln/swis/sao1k/html) with bandwidths of up
to 400 MHz and 4096 spectral channels. This corresponds to about 1.3 km/s at the rest frequency of the
H 2 O line (22.2 GHz), enough for detection purposes.
Such unusually wide bandwidth capabilities are necessary because known H 2 O masers in active galactic
nuclei (AGN) display typically sub-Jy spectral lines distributed over a frequency interval of about 500 to
2000 km s 1 , where the interval is dictated by the rotation speed of the disk at radii where the maser action
is excited. Because it is impossible to predict such interval, the widest possible instantaneous bandwidth is
critical for eôcient surveys of large numbers of galaxies.
The isotropic luminosities L iso of water megamasers range from 20 L  to 6000 L  (the `gigamaser' in
TXFS2226-184; Koekemoer et al. 1995). The maser features widths are also very dierent, normally between
0.5 to tens of km s 1 . For pure detection purposes, we can take a prototypical line with linewidth of dv =
30 km s 1 at a redshift z of 0.03 (the most distant maser detected, in the radio galaxy 3C 403, is at z =
0.06; Tarchi et al. 2003). The isotropic luminosity of a maser line is given by the equation:
L iso
[L  ] = 0:023

D 2
[Mpc]

Z S
[Jy]

dv
[km s 1 ]
Hence, 20 and 6000 L  features at z = 0.03 (D = 120 Mpc, assuming H 0 = 75 km s 1 Mpc 1 ) will have
a peak ux density of 2 and 600 mJy, respectively. Therefore, in a survey the possibility to detect als the
weakest megamasers (L iso = 20 L  ) imply a sensitivity per channel not worse than 0.7 mJy (3). With the
80 K system temperature and an antenna eôciency of 56 % at 22 GHz (Table 3.1), the SRT should be able
27

to obtain a sensitivity of 0.7 mJy in 60 and 25 hours, for a channel width of 1.3 and 5 km s 1 , respectively.
Such huge integration times and the large number of AGN that need to be searched for maser emission force
us to conclude that the outlined SRT survey will necessarily focus on high-luminosity megamasers (with
isotropic luminosities > 50 L  ), with integration times of (only) about 5 to 10 hours per target (dependin
on the channel width required).
Cosmological water megamasers: Standard 22-GHz receivers have nominal frequency bandwidths be-
tween 18 and 26 GHz. However, if the object that emits the water maser line(s) has a redshift z greater
than 0.25 the maser line falls outside the receiver bandwidth, at frequencies < 18 GHz. Hence, to extend
our survey to cosmologically-relevant water megamasers it would be necessary to extend the low-frequency
coverage o the 22-GHz receiver and/or to build an ad hoc receiver capable to work between 16 and 20 GHz.
So far, such a receiver is not yet available in any of the other large antennas (e.g. GBT and Eelsberg).
The SRT would then be able to perform the rst high-redshift water megamasers survey ever tried. Of
course, the integration times computed for z = 0.25 would rise by huge factors, limiting the possible survey
detections only to very luminous objects (L iso > 500 100L  ). However, as recently speculated, the number
of such objects might be higher in the `earlier' Universe (e. g. Townsend et al. 2001), because of enhanced
AGN activity and merging phenomena. The result of such a survey might be used as a pathnder for similar
studie that will be performed in the future by the Square Kilometer Array (SKA).
References
Greenhill L.J., Jiang D.R., Moran J.M., Reid M.J. et al. 1995, ApJ 440, 619
Koekemoer A.M., Henkel C., Greenhill L.J., Dey A. et al. 1995, Nature 378, 697
Miyoshi M., Moran J., Herrnstein J., Greenhill L. et al. 1995, Nature 373, 127
Tarchi A., Henkel C., Chiaberge M., Menten K.M. 2003, A&A 407, L33
Townsend R.H.D., Ivison R.J., Smail I., Blain A.W. et al. 2001, MNRAS 328, L17
4.3.1.3 A Redshift Machine for the SRT
Science Background
(adapted from the science case for the Large Millimeter Telescope [LMT] redshift machine,
http://www.lmtgtm.org/overview/sci/node4.html)
Current surveys are unable to associate unambiguously sub-mm sources to their optical/IR/radio coun-
terparts. Thus, the redshift distribution, luminosities and star-formation history of submm-selected galaxies
at redshifts > 1 is still unclear.
Using a stable, sensitive, very broad-band mm-wave receiver with an instantaneous bandwidth of  35 GHz
(the Redshift Machine) it will be possible to determine CO spectroscopic redshifts of mm-/submm sources
without the need of optical/IR spectroscopy, a task that may even be impossible to carry out at very high
redshifts: at z > 6 the redshifted Ly passes into the near-IR band and the intervening neutral hydrogen
clouds along the line of sight completely absorb the UV light of galaxies. Near-IR (NIR) spectra, obtained
with the new generation NIR multi-object spectrographs on 10-m class telescopes, will help somewhat to
elucidate the nature of star-forming galaxies at z > 6, but only for those systems that are not heavily dust-
extinguished or even quiescent. For these sub-mm/mm sources, molecular or atomic emission lines in the
mm bands (1 and 3 mm) may be the only means to obtain reliable redshifts. The distances between two
adjacent J-transitions of the CO molecule shrink to
 < 35 GHz (the approximate width of the mm-wave
atmospheric windows) at z > 2. This means that a 35-GHz bandwidth spectrometer is virtually guaranteed
of covering (if not detecting) at least one CO line for any galaxy with z > 2. Some redshifts will allow more
than one line to fall in the bandwidth. If the same object is observed by two (or more) of such spectrometers
covering adjacent wavelength bands, there is always the chance to detect at least two lines and determine
the object's redshift, even if (or in particular when) optical spectroscopy is not possible.
Given the initial redshift from the Redshift Machine it will then be possible to accurately tune more `con-
ventional' narrower-bandwidth heterodyne systems, that cover higher or lower frequency passbands, to the
28

appropriate redshifted frequency of additional CO transitions. The higher velocity resolution of the more
conventional receivers, and additional line ratios, will provide important kinematic and chemical information.
Apart from the above specic considerations, a broad-band facility would also oer the possibility to search
for new lines in extragalactic sources.
As an example of what one can expect, we refer to the various line surveys carried out toward Orion-KL, in
various bands between 70 and 857 GHz. Thousands of molecular emission lines were found in these surveys {
many will be too weak to detect in external galaxies, but lines of e.g. CO, CS, SO, CH 3 OH, H 2 CO, SO 2 may
be detectable. More precise detectability estimates for the SRT have to await a careful analysis of expected
line strengths, required redshift, receiver sensitivity, and available backends and bandwidth.
Technical Requirements
Observations at frequencies  > 100 GHz with the SRT are not possible. Hence, we can think of a lower-
frequency Redshift Machine composed of several spectrometers covering the 1 cm and the 3 mm bands. It
is conventionally assumed that the largest continuum band technically feasible can be at most as large as
 40% of its central frequency (still a challenging task). The SRT could then be equipped with a 35 GHz
wide-band spectrometer (central frequency 86 GHz) in the 3 mm band and two narrower-band spectrometers
continuously covering the 1-cm band (for instance two 10 GHz wide-band spectrometers centred at 43 and
32 GHz respectively).
Another issue is the sensitivity-feasibility of such high-z line measurements. Olmi (2003) has demonstrated,
using a simple model for the high-z galaxy ISM (Olmi & Mauskopf 1998) and correcting for cosmological
eects, beam dilution, telescope/receiver performance and atmospheric absorption estimated at the SRT
site, that the strongest CO transitions can be detected (5) in the 3-mm band in a time of  < 1 hr at
z  2:5 5. However, in the 1-cm band detections may take much longer, from a few to tens of hours. For
a detailed discussion of this issue we refer to Olmi (2001).
References
Olmi L. 2003, in Conf. Proc. 81, SRT: The Impact of Large Antennas on Radioastronomy and Space
Science, D'Amico, Fusi Pecci, Porceddu, Tofani (eds.), p.83
Olmi L., Mauskopf P. 1998, APS Conf. Ser. 146, 371
4.3.2 Continuum observations
4.3.2.1 High-Frequency Extra-Galactic Sky Surveys with the SRT
Scientic Background
The recent development of multi-beam receivers will enable single-dish telescopes, like the SRT, to survey
large areas of sky at high radio frequencies ( 15 GHz). High-frequency extragalactic sky surveys are
expected to have a major impact on astrophysics. Some of the key topics that can be addressed by such
surveys are listed below.
 They provide adequate samples to get a really unbiased view of rare, but very interesting, classes of
sources with at spectrum up to high frequencies, that, at low frequencies, are swamped by more
numerous populations which fade away as the frequency increases. One example are at-spectrum
radio quasars (FSRQ), for which much more extended samples are required to properly cover their
parameter space. A second example are BL Lac objects: it is still debated whether radio selected and
X-ray selected BL Lacs are the extremes of a single class of objects whose properties are determined
by the beam orientation or have intrinsically dierent properties.
 They open a window on new classes of sources, such as those with strong synchrotron or free-free
self-absorption corresponding to both very early phases of nuclear radio-activity (extreme GHz Peaked
Spectrum - GPS - sources or high-frequency peakers - HFP) and late phases of the evolution of Active
Galactic Nuclei (AGNs), characterized by low accretion/radiative eôciency (ADAF/ADIOS sources),
as well as to early phases of the evolution of radio afterglows of gamma-ray bursts (GRB).
29

 They produce surveys of the Sunyaev-Zeldovich eect in distant clusters of galaxies and perhaps on
galactic scales, extremely important both to understand the formation of large scale structure and the
heating of the intergalactic medium.
 They play a vital role in the interpretation of temperature and polarization maps of the Cosmic
Microwave Background (CMB), by allowing us to characterize and remove the contamination by as-
trophysical foregrounds.
A census of the expected contribution from dierent source populations in high frequency extragalactic
samples is listed in the Table below.
Table 4.2: Expected number density (per sq. degr.) for various source populations at 20 GHz. Based on the
model described in De Zotti et al. 2004.
N(> 1 mJy) N(> 10 mJy)
FSRQ 2.7 0.57
BL Lac 2.5 0.2
Steep-spectrum 26 2
ADAF 0.12 0.0042
GRB Afterglows 0.0045 0.0034
In the following we focus on a particular topic: the study of the very early phase of nuclear radio-activity.
The nature of the mechanism that causes the triggering of extragalactic radio sources is an open question. To
answer it a large sample of young sources is needed. Current models of radio source growth that consider the
eects of self absorption on the source synchrotron emission indicate that very young radio sources should
have radio spectra which peak at frequencies > 1 GHz, with younger sources peaking at higher frequencies
than older sources (O'Dea 1998; Snellen et al. 2000).
This is due to a well-known relation between linear size of a radio source and turnover frequency: the younger
the source, the smaller the size, the thicker the optical depth, the stronger the self-absorption eects, the
higher the turnover frequency. Typically, sources with linear sizes < 1 kpc have turnover frequencies > 1
GHz.
This class of extragalactic radio sources (assumed to be hundreds of years old) is called Gigahertz Peaked
Spectrum (GPS) population. Extreme GPS sources with turnover frequencies > 10 GHz (called High
Frequency Peakers, HFP) would have linear sizes < 10 pc and should be the youngest sources (tens of years
old). GPS sources are rare ( 20% of sources are GPS in 2.7-GHz samples and  3% are HFP in 5-GHz
samples). Today we know a total of a hundred of such sources. To obtain a better census and understanding
of the initial stages of the radio source physical evolution, high frequency extragalactic sky surveys are of
crucial importance.
Any blind radio survey is biased toward sources with spectra peaking close to the selection frequency, hence
blind surveys at frequencies > 10 GHz provide a means of eectively generating a sample rich in HFPs and
thereby testing the models of source growth in very young sources. On the other hand, high-frequency sky
surveys have started to become feasible only very recently. If we focus on surveys at frequencies 10 100
GHz (which are the relevant ones for the study of very young radio sources), the most interesting one is the
so-called 9th Cambridge survey (9C) carried out at 15 GHz with the Ryle Telescope (Waldram et al. 2003).
This survey has covered 520 deg 2 to 25 mJy, 190 deg 2 to 10 mJy and 12 deg 2 to 5 mJy (the deepest survey
currently available at  > 10 GHz). A follow-up on a sub-sample at other radio frequencies is currently
under way to select at- and/or inverted-spectrum sources (eg. GPS/HFP candidates). Preliminary results
indicate that the fraction of HFP candidates decreases from 33% at uxes > 100 mJy to 20% at uxes > 25
mJy, to 10% at uxes > 5 mJy (Waldram & Pooley 2004).
Another survey which could give a contribution to this topic in the future is the planned all-sky ATCA 18
GHz survey which will be undertaken in the southern hemisphere to a few hundred mJy (a pilot survey
covers about 1200 deg 2 to 100 mJy, Ricci et al. 2004).
High frequency sky surveys could also give a very important contribution in investigating the so-called faint
radio population (mJy and sub-mJy uxes at 1.4 GHz), a mixture of dierent kinds of sources (mainly low
luminosity AGNs and star-forming galaxies), which is still quite elusive. The possibility of reaching in the
30

near future (sub-)mJy ux detection limits at frequencies > 10 GHz will allow the study of the spectral
properties of such sources and the test of dierent accreting models proposed for low-luminosity AGN (eg.
advection dominated accretion ows - ADAF - vs core-jet models).
Technical Requirements
The SRT could give an important contribution to the topics discussed above. We can prot of the 22 GHz
multi-beam receiver (7 elements, 2 polarizations) which is already under development for the SRT to carry
out extragalactic surveys. The multi-beam option is crucial to eôciently conduct extensive surveys with the
SRT due to the very small primary beam of the 64-m antenna ( 54 00 FWHM at 22 GHz, see Tab. 3.1). In
absence of a derotator mounted at the telescope, a derotation software should be made available at the SRT
to correctly align all the beams (see Chapter 3).
Assuming an instantaneous bandwidth of 2 GHz for continuum observations, 2 polarizations and the antenna
sensitivity parameters provided so far for the SRT at 22 GHz (see Sect 3.5 and Tab. 3.1), it is possible to
estimate that a 1 s exposure would provide an rms noise of  2 mJy/beam, whereas a 1 min exposure would
give an rms noise of  0:25 mJy/beam. This translates to an observing campaign of, say, 20 days to map
either (A) an area of 500 deg 2 to 2 mJy/beam (rms), or (B) an area of 10 deg 2 to 0.25 mJy/beam (rms),
driving the telescope in raster scanning mode (as necessary for blind surveys).
Case (A) would be more suitable to pick up rare objects like the GPS/HFPs. From the existing counts at 15
GHz, we estimate that we would detect about 1300 extragalactic radio sources at 5 ( 10 mJy), about 20%
of them ( 260) being HFPs candidates (see Waldram & Pooley 2004). Such a survey alone could provide
nearly three times the number of GPS/HPF sources currently known. Case (B), on the other hand, would
be very suitable to study the faint radio population, providing  280 sources above  1:3 mJy (5).
To fully exploit the SRT capability, a survey should reach the SRT confusion limit, which is 50 70 Jy/beam
(rms) at 22 GHz (as estimated by extrapolating to lower uxes the 15 GHz number counts of Waldram et
al. 2003). A noise level of 60 Jy can be reached by the SRT with 18 min long exposures. This means an
observing campaign of 40 days to map an area of 0.5 deg 2 and detect 75 sources at S > 0:3 mJy (5). While
still feasible with the SRT, such a campaign is very long, and bad atmospheric conditions could make it even
longer. In this respect a multi-beam receiver with more than 7 elements would be very useful. Allowing
for a response loss of  25% for the external elements, we can accommodate a grid of 5  5 feeds in the
SRT 22-GHz focal plane. With a 25-element multi-beam system the observing time necessary for the survey
mentioned above would be reduced to 15 days.
In a future perspective, it would be also very interesting to equip the SRT with a 30 or 40-GHz multi-beam
receiver, suited for deep continuum observations (largest instantaneous bandwidth). Indeed, extragalactic
sky mapping is very scarce at these higher frequencies. The only existing surveys are the few very bright
surveys associated with recent CMB experiments: for instance the VSA at 34 GHz (S lim  100 mJy) and
the WMAP at 41 GHz (S lim > 1 Jy). Possible other surveys will be performed during the Planck mission.
From Tab. 3.1 we can see that the SRT sensitivity at 30 GHz is a factor of two better than at 40 GHz. In
addition, the beamsize decreases with increasing frequency, making sky surveys increasingly challenging.
Nonetheless, since a 30-GHz multi-beam receiver is already being developed for the Torun 32-m single-
dish, a 40-GHz multi-beam receiver could be considered preferable for the SRT. In this respect we notice
that the smaller beamsize at 40 GHz could be counterbalanced by a larger number of beams that can be
accommodated in the multi-feed array.
Another important issue to be taken into account before establishing the best frequency for a high-frequency
sky survey with the SRT is the atmospheric transparency at  30 GHz and its seasonal variations which
should be monitored at the SRT site.
A tentative estimate of the confusion limit of the SRT at 30/40 GHz can be obtained from the 15 GHz
number counts of Waldram et al. (2003), by assuming at-spectrum sources ( = 0). This yields rms limits
of 20 30 Jy at 32 GHz and of 10 15 Jy at 43 GHz. Given the sensitivity parameters tabulated in
Tab. 3.1, to achieve the confusion limit for statistically relevant samples (> 100 sources) in a reasonable time
(< 30 days), we need multi-feed receivers with at least 33 elements at 30 GHz and 10x10 elements at 40
GHz.
References
De Zotti G., Perrotta F., Granato G.L., Silva L., Ricci R., Baccigalupi C., Danese L., Toolatti L.
31

2003, in SRT: The Impact of Large Antennas Radioastronomy and Space Science, eds. D'Amico, Fusi
Pecci, Porceddu, Tofani (SIF), p. 57
De Zotti G., Ricci R., Mesa D., Silva L., Mazzotta P., Toolatti L., Gonzales-Nuevo J., 2004, A&A,
in press
O'Dea C.P. 1998, PASP 110, 493
Ricci R., Sadler E.M., Ekers R.D., Staveley-Smith L., Warwick E.W., Kesteven M.J., Subrahmanyan
R., Walker M.A., Jackson C.A., De Zotti G. 2004, MNRAS 354, 1020
Snellen I.A.G., Schilizzi R.T., Miley G.K., de Bruyn A.G., Bremer M.N., Rottgering H.J.A. 2000,
MNRAS 319, 445
Waldram E.M., Pooley G.G., Grainge K.J.B., Jones M.E., Saunders R.D.E., Scott P.F., Taylor A.C.,
2003, MNRAS 342, 915
Waldram E.M.& Pooley G.G. 2004, astro-ph/0407422
4.3.2.2 Search for Sources with High Rotation Measures
Scientic Background
The measurement of rotation measures (RMs) of extragalactic sources is a powerful instrument in several
respects. The RM is proportional to the product of the magnetic eld and the electron density along the
line of sight and thus bears considerable information on physical conditions in the source and along the
line-of-sight towards us. The measured RM is composed of a galactic (GRM) and an extragalactic (RRM)
contribution. The separation of these two parts can be done by the observation of neighbouring sources.
Since the galactic contribution is expected to vary on much larger scales than the extragalactic part, the
mean RM of neighbouring sources at similar galactic latitude is a reliable measure of the galactic part (Oren
& Wolfe 1995). Observing large samples with high source densities is therefore essential to guarantee a
separation of GRM and RRM. The extragalactic component RRM can provide information on any gaseous
material along the line-of-sight to the source (intervening galaxies, intergalactic clouds or protogalaxies,
intracluster gas of material associated with the radio source itself).
The unambiguous determination of RMs in such samples will push forward our knowledge in several aspects:
 characteristics of the medium in groups and clusters. A RM study will disclose the conditions of the
medium in galaxy clusters and groups, in particular at higher redshifts. Carilli et al. (1997) presented
evidence that most sources with detected RM have the Faraday screen local to the source. The simplest
interpretation would be that these sources are situated in dense, cluster-type environments, as has been
found at lower redshifts. E.g. Pentericci et al. (1997) and Athreya et al. (1998) reported observations
of a radio galaxy with a high RM at a redshift of 2.2, which is evolving to become a cD galaxy.
 At high redshifts the correlation between high RM and the existence of damped Ly clouds can be
studied. It has been claimed that large RM screens are neither galactic nor associated with the source,
but cosmologically intervening material (Kronberg & Perry 1982; Wolfe et al. 1992; Oren & Wolfe
1995). There is some indication that high-redshift sources with high RMs are also good candidates to
nd damped Ly systems. The main problem of these studies is the poor statistics.
 With suôciently closely sampled RM measurements it is also possible to `map' the GRM distribution
over large parts of the sky. These data are of great interest for galactic halo research.
Technical Requirements
As eventually at least three measurements at dierent well-spaced frequencies are required for an unam-
biguous determination of the RM, all broad-band continuum receivers should be equipped with polarimeters
for Stokes Q and U.
As an example the following time estimate is given for the planned 5-GHz receiver system (1BW in
Tab. 3.1). For the calculation a single horn is assumed, although a double-horn system would warrant a
32

more eôcient cancellation of atmospheric eects at frequencies  5 GHz (more important for total-intensity
measurements). Assuming an aperture eôciency of 58% for 5 GHz, a system temperature of about 20 K and
a receiver bandwidth of 400 MHz a total power ux density limit of 1.1 mJy/beam can be reached within
1 second of integration time. Given that the confusion limit at this frequency is estimated to be around
0.7 mJy/beam and 0.24 mJy/beam in total power and polarisation, respectively, this would be reached in
about 2.5 s pure on-source observing time. In cross-scan mode the total telescope time would then sum up
to about 90 s, if four subscans for better statistics and redundancy are made. This means that a sample of,
say, 1000 sources with ux densities > 40 mJy and assumed fractional polarisation of 3% can be observed
down to the confusion noise limit in polarisation in about 35 hours on a 5 level in each polarisation channel.
This observing mode requires driving software for cross-scans (as will be required also for standard pointing
checks).
References
Athreya R.M., Kapahi V.K., McCarthy P.J., van Breugel W. 1998, A&A 329, 809
Carilli C.L., Rottgering H.J.A., van Ojik R., Miley G.K., van Breugel W.J.M. 1997, ApJS 109, 1
Kronberg P.P., Perry J.J. 1982, ApJ 263, 518
Pentericci L., Rottgering H.J.A., Miley G.K., Carilli C.L., McCarthy P. 1997, A&A 326, 580
Oren A.L., Wolfe A.M. 1995, ApJ 445, 624
Wolfe A.M., Lanzetta K.M., Oren A.L. 1992, ApJ 388, 17
4.3.2.3 High-Frequency Follow-up to Surveys
Science Background
It is well-known that radio sources undergo intermittent activity. This is e.g. re ected in numerous
high-resolution images of radio galaxies which exhibit dierent components obviously produced at dierent
epochs. The nal exhaustion of engines in radio galaxies is followed by rapid synchrotron cooling of the
hotspots and lobes, which is the reason why only very few relic radio sources are observed at the present
epoch. In order to explore the evolution of radio sources it is hence mandatory to investigate the spectral
evolution of a large and complete sample of radio sources. One of the best suited samples to explore the
spectral characteristics of sources over a wide frequency range is the B3VLA survey (Vigotti et al. 1989).
It consists of a subset of 1049 sources selected from the B3 survey (Ficarra et al. 1985) in ve ux bins
by narrowing the survey strip in declination and restricting it to high galactic latitudes so that the com-
pleteness is very carefully controlled. This survey, some 30 times deeper in ux density than the 3C has
measured ux densities between 74 MHz and 10 GHz (Vigotti et al. 1999). It is now essential to extend the
ux measurements to radio frequencies beyond 10 GHz as it is here where either possible ageing (and thus
convex curvatures of the spectra) or the in uence of new synchrotron components (and thus high-frequency
attening) are visible rst.
Taking the B3VLA survey as example we have extrapolated the known broad-band continuum spectra to
probable future continuum frequencies at the SRT at 20, 32, 43, and 86 GHz. In order to give an idea which
noise levels ought to be reached we have estimated the median ux densities at these four frequencies for
the 1049-source B3VLA survey: 20 GHz: 13.4 mJy; 32 GHz: 8.0 mJy, 43 GHz: 5.7 mJy and 86 GHz: 2.5 mJy.
Technical Requirements
This observing programme requires broad-band (multi-beam) receivers at typical continuum frequencies
greater than 10 GHz (e.g. 20 GHz, 32 GHz, 43 GHz, 86 GHz, preferably with polarimeters). Assuming the
values reported in Tab. 3.1 at 20 and 32 GHz show that it is well possible to observe the majority of the
sample in relatively little time at three lower frequencies, while measurements at 86 GHz would be taken
for a suitable subsample only. The time-consuming factor will be the deep mapping of extended sources,
which will be increasingly necessary with decreasing beam sizes at higher frequencies. The employment of
multi-beam arrays should be of great help to accelerate the latter observations. Both cross-scanning and
33

raster-scanning modes would be required.
References
Ficarra A., Grue G., Tomassetti G. 1985, A&AS 59, 255
Vigotti M., Grue G., Perley R., Clark B.G., Bridle A.H. 1989, AJ 98, 419
Vigotti M., Gregorini L., Klein U., Mack K.-H. 1999, A&AS 139, 359
4.3.2.4 High-Frequency Mapping of Extended Sources
Science Background
The class of giant radio galaxies (GRGs) is formed by radio galaxies with linear sizes exceeding 1 Mpc.
Some 130 mostly nearby (z < 0:2) GRGs are known today, most of which were discovered serendipitously
(http://www.astro.uni-bonn.de/mjamrozy/grglist1.html). Their angular sizes range from a few arcminutes
to more than a degree. The aims of studying GRGs are twofold: Firstly, since their radio properties are
extreme, they are an ideal laboratory for studying the physics of radio sources. Due to their large angular
sizes, even single-dish observations with their relatively low resolutions reveal a wealth of detail in these
sources. Secondly, these are the only sources that can probe their environment and the intergalactic medium
(IGM) over such a large scale. Studying these sources can therefore constrain physical parameters of the
IGM. While most of these sources have been thoroughly studied at frequencies up to 10 GHz (e.g. Mack et
al. 1997, Schoenmakers et al. 2000), their characteristics at shorter wavelengths have not been observed yet.
Mapping of GRGs with the SRT should therefore be done at frequencies > 10 GHz. The use of single-dish
instruments is mandatory at such high frequencies due to increased missing spacing eects of interferometers.
Technical Requirements
This project depends on the availability of multi-beam receivers (for beam-switching to remove weather
eects) or better even horn arrays, large bandwidths (e.g. 2 GHz at 32 GHz) and polarimeters. The SRT at
20 GHz would oer a beam size of 56 00 , ideal for mapping the largest GRGs. From extrapolation of intensities
found in the 10 GHz maps obtained with the 100-m Eelsberg telescope an rms noise level of less than 0.3
mJy/beam seems mandatory for a signicant progress of high-frequency studies of these sources. Assuming
a system temperature of 81 K and a bandwidth of 2 GHz, we expect to reach a noise level of 1.7 mJy in 1
s. 0.3 mJy/beam are therefore reached in a pure observing time of about 32 s. Further assuming a raster-
scanning mode with one horn only, one would therefore require about 12 coverages. Typical durations for
one coverage for the largest sources would be about 6.5 hours (assuming a pixel size of 15 00 , scan speed
of 20 0 /min and total map sizes of 60 0  30 0 ), thus summing up to almost 78 hrs per source. Obviously the
employment of multi-beam receivers can greatly reduce this time.
References
Mack K.-H., Klein U., O'Dea C.P., Willis A.G. 1997, A&AS 123, 423
Schoenmakers A.P., Mack K.-H., de Bruyn A.G., Rottgering H.J.A., Klein U., van der Laan H. 2000,
A&AS 146, 293
4.3.2.5 Multi-frequency Monitoring of long- and short-term Blazar Variability
Scientic Background
One of the outstanding characteristics of blazars is the frequently observed strong variability (often by a
factor of two or more) across the whole electromagnetic spectrum on time scales of months to years. The
mechanisms of this variability are still not well understood. Possibilities discussed so far include shocks in
jets (e.g. Marscher & Gear 1985, Aller et al. 1985, Marscher 1986) and changes in the direction of forward
beaming e.g. due to helical trajectories of plasma elements in the jets (Camenzind & Krockenberger 1992)
or a precessing binary black-hole system introducing ux-density outbursts (lighthouse eect, Begelman et
34

al. 1980; Sillanpaa et al. 1988). Thus, variability studies furnish important clues into size, structure, physics
and dynamics of the radiating region in these sources. The wide variety of models calls for new and ongoing
long-term as well as multi-frequency ux-density campaigns capable of providing the necessary observational
constraints.
About 20 years ago, a new type of variability in extragalactic radio sources could be established: variations
on time scales of hours up to two days were found as a frequent phenomenon in compact blazar cores (Witzel
et al. 1986, Heeschen et al. 1987). There is still a debate about the physical origin of such rapid, IntraDay
Variability (IDV) in total ux density as well as polarization. IDV has been found to be present in about
30% of these compact objects (Quirrenbach et al. 1992) and reveal extremely tiny dimensions of the emitting
region ( light-hours) with brightness temperatures of up to TB ' 10 21 K, far in excess of the allowed inverse
Compton limit of T max
B ' 10 12 K. The cause of the variations seen in these sources is currently controversial
with claims being made for either: 1) a source-intrinsic or 2) extrinsic origin due to scattering in the
interstellar medium (ISM) similar to the phenomenon of pulsar scintillation (e.g. Wagner & Witzel 1995 and
ref. therein). While an intrinsic interpretation requires excessive Doppler boosting or shock-in-jet models
with uncomfortable special source geometries, interstellar scintillation (ISS) must contribute to the cm-radio
IDV due to the involved small source structures with sizes smaller or comparable to the scattering size set by
the ISM. In order to disentangle both eects, intensive new multi-frequency and polarization observations
are essential. In particular, the discovery of a seasonal dependence of the variability time scale throughout
the year has attracted a lot of attention. A scintillation model, which takes the Earth's motion with respect
to the scattering medium into account, can fully explain the eect and clearly identies ISS as origin of the
rapid variations (e.g. Dennett-Thorpe & de Bruyn 2000, 2003). At present, an annual modulation has been
detected only in the two most extreme IDV sources (time scales  0.5 hours). The detection of such seasonal
cycles in a larger number of IDVs will help to disentangle between an intrinsic or extrinsic origin of the IDV
phenomenon. In particular, the discovery of seasonal cycles in the \classical" type-II IDV sources (type II:
IDV on time scales  0.5{2 days; type I: > 2 days; Heeschen et al. 1987) is important to investigate the role
of ISS also in this class of objects (Fuhrmann 2004, Fuhrmann et al. 2002). As a by-product, the modeling of
seasonal cycles constrains important properties of the scattering medium and thus is an important new tool
to study the ISM. Furthermore, the size of the scintillating compact source component can be constrained
to angular scales presently not accessible with any radio interferometer (a few tens of as).
Technical Requirements
The multi-frequency and polarization capabilities of the SRT will provide a great new tool to investigate the
above mentioned topics. Radio outbursts in the long-term light curves of blazars usually show time delays
(days to weeks) between higher and lower frequencies due to a combination of optical-depth eects and
plasma motion. Dedicated multi-frequency, ux-density monitoring campaigns with the SRT over months
and years (e.g. simultaneous with observations in the IR, optical and at higher energies as well as VLBI)
will allow one e.g. to search for time delays, correlated variability over a wide range of frequencies, follow
the spectral shape evolution and therewith constrain jet models. Furthermore, an italian long-term radio
monitoring of blazars would be highly desirable in view of the next launch of the italian -ray-satellite
AGILE as well as the Large Area Telescope (LAT) -ray-detector on the GLAST satellite, which both
will likely detect many of these blazars in the GeV energy band. Among important blazar identication
and coordinated multi-wavelength monitoring campaigns of newly detected sources, e.g. the combination
of a SRT radio monitoring with optical observations (e.g. within the WEBT collaboration 1 ) will allow to
detect aring states which might trigger further ToO observations by these satellites. However, a broad
and good sampled frequency coverage of the SRT (between 1{100GHz, as planned and shown in Table 3.1)
will be important to obtain simultaneous broad-band radio spectra, in particular towards higher frequencies
( 10 GHz). In order to obtain a good time coverage, the sources should be monitored frequently over the
year (e.g. bi-monthly) during observing runs lasting for about 24 hours (depending on the number of sources
and frequency coverage).
As far as IDV studies are concerned, typically longer observing runs are essential to detect the rapid
variability and to determine the variability time scales with high accuracy. The total as well as polarized
ux density of \classical" IDV sources showing time scales of about 1{2days should then be observed with
1 http://www.to.astro.it/blazars/webt
35

a very fast duty cycle of about one measurement per hour over a continuous period of  5 days. In order to
detect seasonal cycles in the variability time scale and model the variations with a scintillation model, this
procedure should be repeated frequently over the year (ideally monthly or more often).
The observations discussed here require precise total ux-density measurements and thus broad-band (as
given in Table 3.1), multi-beam receivers (preferably at all continuum frequencies). The nal measurement
accuracy will mainly be aected by the local weather conditions, receiver performance/stability, pointing
accuracy and focus stability/accuracy. The possibility of multi-beam (double-horn) observations are of
particular importance to reduce the atmospheric contribution, especially towards higher frequencies. The
frequent observations of a set of suitable secondary calibrators (nearby, non-IDV sources) with the same duty
cycle as the target sources will allow one to trace the receiver/system and weather conditions during a run.
This data can subsequently be used to correct for telescope gain as well as systematic time-dependent eects
in the IDV light curves. The typical peak-to-peak variations seen in IDV sources are of the order 5{20%.
In order to detect even the modest variability amplitudes on a 3 -level, the nal scatter in the secondary
calibrator light curves due to the afore mentioned eects should be of the order  1.5{2 %.
The sources are usually point-like and suôciently strong ( 0.5 Jy) at the considered frequencies. Taking
the SRT receiver characteristics as given earlier, all sources will be detectable with a suôcient signal-to-noise
ratio. Thus, cross-scans are the method of choice as ideal compromise between suôcient accuracy and a fast
enough duty cycle. As by-product, the pointing deviation of the telescope can be determined and corrected
simultaneously during the observations.
At frequencies  10 GHz, the frequent measurement of secondary calibrators as well as opacity correc-
tions are of particular importance to minimize the strong in uence of atmospherical eects. Since a Sky-Dip
procedure of frequent -measurements is not possible in a densely time-sampled IDV experiment, a ra-
diometer will be very usefull (as suggested in section 3.2). However, excellent weather conditions during
the observations are essential. In order to study polarization and polarization angle variations the receivers
should be equipped with correlation polarimeters providing the simultaneous information of Stokes Q and
U. The instrumental polarization should be not greater than 1 %. The extremely fast duty cycle requires the
possibility to switch quickly between receivers. Switching times of  one minute will allow fast duty cycles
for a suôcient number of sources (1 scan per source and frequency in about one hour).
References
Aller H.D., Aller M.F., Hughes P.A. 1985, ApJ 298, 296
Begelman M.C., Blandford R.D., Rees M.J. 1980, Nature 287, 307
Camenzind M., Krockenberger M. 1992, A&A 255, 59
Dennett-Thorpe J., de Bruyn A.G. 2003, A&A 404, 113
Dennett-Thorpe J., de Bruyn A.G. 2000, ApJ 529, L65
Fuhrmann L. 2004, Ph.D. Thesis, University of Bonn
Fuhrmann L. et al. 2002a, PASA 19, 64
Heeschen D.S. et al. 1987, AJ 94, 1493
Marscher A.P. 1996, in ASP Conf. Ser. 100, 45
Marscher A.P., Gear W.K. 1985, ApJ 298, 114
Quirrenbach A. et al. 1992, A&A 258, 279
Sillanpaa A., Haarala S., Valtonen M.J., Sundelius B., Byrd G.G. 1988, ApJ 325, 628
Wagner S.J., Witzel A. 1995, ARA&A 33, 163
Witzel A. 1986, Mitt. Astron. Ges. 65, 239
36

4.3.2.6 SRT and the Sunyaev-Zel'dovich eect
Scientic Background The Sunyaev-Zel'dovich (SZ) eect is the modication of the spectrum of the
Cosmic Microwave Background radiation (CMB) as it transverse a cluster of galaxies or any other reservoir
of hot plasma. Photons interact with free electrons in the ionized medium through the inverse Compton
process resulting in a distortion in the spectrum of the outgoing radiation. The observed fractional change
in antenna temperature is proportional, in the rst approximation (i.e. neglecting the peculiar velocity of
the cluster and the relativistic correction), to the Comptonization parameter y /
R
n e T e dl, a line-of-sight
integral of the electron density and temperature through the cluster.
Measurements of the eect yield directly the properties of the hot gas, and the total dynamical mass
of the cluster. An extremely valuable property of the SZ eect is its redshift independence that makes it
a useful cosmological probe (Zel'dovich & Sunyaev, 1969). As matter of fact the SZ can be used for the
determination of the Hubble constant without the sistematic error due to the source evolution as is the case
of other methods.
The SZ eect was measured for the rst time with single-dish radiometers. More recently, the best
images of the eect (Carlstrom et al. 2002) has been obtained for about 50 clusters, mostly with ground-
based interferometric arrays (BIMA and OVRO) operating at low microwave frequencies (30 GHz). The use
of single-dish radiometers in the study of the SZ eect is less used since early 1990s with the exception perhaps
of 45m dish of Nobeyama Observatory. The combined results of Nobeyama observations, for instance, at 21
and 43 GHz and JCMT observations at 350 GHz allowed for the rst time in 1999 an unambiguos detection
of the eect in the Wien region of the spectrum (Komatsu et al. 1999). The advent of the generation
of large single-dish telescopes operating up to millimeter wavelenghts, such as GBT and SRT, open new
opportunities.
Measured typical values of the Compton parameter y at 32 GHz (Mason et al. 2001) is of few units in
10 5 . It turns out that the antenna temperature decrement is of the order of 90 - 200 K at this frequency.
Slightly lower or higher values of y are expected respectively at 20 GHz and 40 GHz.
The best frequencies for SRT should be 20, 30 and 40 GHz. Unfortunately, the 90-GHz band might
not be useful due the large amount of the water vapour. At this frequency, the estimated atmospheric
transparency in the best conditions at SRT is of the order of 90%; however more stringent limitation will
come from atmospheric absorption uctuation which will induce an increase of the (1=f) n noise. This will
also prevent the use of high performance, large bandwidth, bolometer arrays. It is worth noting that the
frequency band centred around the water vapour line at 22 GHz can be easily eliminated at the SRT owing
to the large receiver bandwidth. In this case, the atmospheric opacity becomes about a factor of 2, in optical
depth, less than the opacity at 90 GHz.
Technical Requirements
The angular extents of clusters are of several arcmin (e.g. a linear size of 1 Mpc is 9.5 0 at z=0.1, 3.2 0 at
z=0.5 and 2.7 0 at z=1, with WMAP cosmology). Indeed, images of the SZ eect in clusters over a wide
redshift range show cluster sizes of about 2 0 - 5 0 (see e.g. Carlstrom et al. 2002). Thus, given the SRT beam
of 52 00 at 20 GHz, the multibeam will be crucial in mapping the nearby clusters, but it will be needed also
for the clusters at redshift  0.5.
The technical requirements can be summarized as follows:
 Receiver: Multibeam receivers for 20 GHz and 40 GHz band
 Backend: total power
 Sensitivity : The expected sensitivity of SRT at 20 and 40 GHz, assuming for the receiver noise
temperatures and for antenna eôciencies the values reported in Table 1 of this Report, and an IF
instantaneous bandwidth of 2 GHz, are respectively about 3:2mKs 0:5 and 2:5mKs 0:5 ; for a cluster
with y = 5  10 5 , the expected signal is of the order of 0.1 mK which imply an integration time of
about 1000 sec at 1 sigma. An IF bandwidth larger than 2 GHz would be, in any case, preferable.
The half power beamwidths at 20 and 40 GHz are 52 00 and 28 00 , respectively. To observe a region
5 0 5 0 with the 20 GHz receiver focal plane array (7 double polarization receiver) at 0.1 mK sensibility,
37

the observing time would be about 50 hours. The beam separation in the sky between two adiacent
receivers is about 82 00 , so an undersampling factor of about 10 must be considered.
 Scanning capability: Beam switching is required. Likely raster scan and/or on-the- y techniques
will be used for the map acquisition.
References
Zel'dovich Ya., Sunyaev R.A. 1969, Ap&SS 4, 301
Carlstrom J.E., Holder G.P., Reese E.D. 2002, ARA&A 40, 643
Komatsu E., Kitayama T., Suto Y. et al. 1999, ApJ 516, L1
Mason B.S., Myers S.T., Readhead A.C.S. 2001, ApJ 555, L11
38

Chapter 5
VLBI
5.1 Galactic astronomy
5.1.1 Line observations
5.1.1.1 An Italian VLBI array for the 6.7 GHz CH 3 OH masers
Scientic Background
The formation of high-mass stars is still a poorly known process. Main open problems are: 1) the
accretion mechanism, whether massive stars form primarily via an accretion disk, or by accretion onto lower
mass cores followed by coalescence to form a more massive object; 2) the determination of the evolutionary
sequence of high-mass star formation; 3) the cause(s) terminating mass accretion. Present millimeter (and
submillimeter) interferometers lack the necessary angular resolution to study the physical conditions and the
gas motions in the proximity of the high-mass protostars. However, such a region can be studied with very
high angular resolution ( 1 mas) by means of VLBI (Very Long Baseline Interferometry) observations at
centimeter wavelengths of maser sources. In fact, H 2 O and Class II CH 3 OH (methanol) maser transitions
are commonly observed toward regions of high-mass star formation, the strongest emission lines (up to
thousands of Jy) being those at 22.2 GHz for water and those at 6.7 and 12.2 GHz for methanol.
The astrophysical environment traced by the 6.7 GHz CH 3 OH masers is still to be more precisely de-
termined. Single-dish and interferometric surveys (Walsh et al. 1997; Szymczak et al. 2000; Codella and
Moscadelli 2000) seem to indicate that, like the water masers at 22.2 GHz, the methanol 6.7 GHz maser
emission may trace a pre-UC HII phase. VLBI observations have shown that the 6.7 GHz maser spots (the
individual maser emission centers) are often distributed along lines or arcs (of sizes from 100 mas to 1 arcsec)
with, occasionally, a monotonic velocity trend along these structures (Norris et al. 1998; Minier et al. 2000).
This strongly suggests that the methanol masers trace ordered motions, and some authors have proposed
that they originate from circumstellar accretion disks.
Presently, the only VLBI array capable to observe at 6.7 GHz is the European VLBI Network (EVN),
but observations of the CH 3 OH masers have been hampered by poor knowledge (only a few arcmins) of
their absolute positions. The SRT together with the other two Italian antennae at Medicina and Noto,
will constitute a small, yet sensitive, array, which will be able to determine both the absolute position and
the space-velocity structure of the strong 6.7 GHz masers. The achievable angular resolution at the maser
frequency of 6.7 GHz is 10 mas, implying that relative positions of compact maser features detected with
signal-to-noise ratio 10 are obtainable with accuracy 1 mas. The main advantage with respect to the
EVN is that, observing with only three baselines it allows to use very short ( 0.5 sec) integration times for
correlating the visibilities, making the instantaneous eld of view of the array suôciently wide to compensate
for the uncertainty in the target position.
In the Spring of 2003 and 2004, we have carried out two exploratory VLBI runs at 6.7 GHz using the
single-baseline Medicina-Noto. In order to derive the absolute maser positions we applied the phase reference
technique, referring the visibility phase of the maser target to a closeby reference continuum source, whose
position is known with milliarcsec accuracy. The results are encouraging: with only two antennae, absolute
positions accurate to within a few tenths of an arcsec are obtained for maser sources with peak ux densities
39

in the range of 10 { 100 Jy. The SRT will bring two major improvements to these measurements: 1)
working with three antennae will allow selfcalibration of the visibility phase, removing the atmospheric phase
uctuations; 2) the SRT 64-m dish will signicantly lower the sensitivity threshold for maser detection. We
are condent that the Italian 6.7 GHz array will be able to determine positions with an accuracy of tens of
mas and space-velocity structures for maser sources with uxes as weak as a few Jy.
Technical Requirements
One of the main justications of the SRT project is to build a big antenna in a strategical position (in
the middle of the Mediterranean Sea) to provide sensitive, long, North-South oriented baselines for the EVN
array. Conceivably, VLBI operability should be a primary goal for the SRT already at `rst light'. In the
following, we indicate the main requirements for the 6.7 GHz VLBI operation:
1. Hydrogen maser to set the time of the VLBI station
2. MkV (or MKIV) VLBI recorder
3. Superheterodyne, cooled receiver, working in double circular polarization at the CH 3 OH maser fre-
quency of 6669 MHz
The correlator is obviously a fundamental component for VLBI. Our 6.7 GHz Medicina-Noto experiments
have been so far correlated at the Max-Planck-Institut fur Radioastronomie (Bonn, Germany). We do not
doubt that the Bonn correlator will be available for correlating the `Italian' 6.7 GHz experiments also in
future years. However it is predictable that, in the immediate future, the cross-correlation of a small number
of baselines can also be performed using a cluster of fast PCs. Then, at a relatively modest cost, we might
be totally independent in the correlation process, too.
References
Codella C., Moscadelli L. 2000, A&A 362, 723
Minier V., Booth R.S., Conway J.E. 2000, A&A 362, 1093
Norris R.P., Byleveld S.E., Diamond, P.J., Ellingsen S.P., et al. 1998, ApJ 508, 275
Szymczak M., Hrynek G., Kus A.J. 2000, A&AS 143, 269
Walsh A., Burton M.G., Hyland A.R., Robinson G. 1997, MNRAS 291, 261
5.1.2 Continuum observations
5.1.2.1 Radio Emission from Galactic X-ray Binaries
Scientic Background
Relativistic jets, collimated bipolar out ows of matter travelling at a signicant fraction of the speed of
light, appears to be a common feature of accreting supermassive black holes in the centres of active galaxies,
in some cases completely dominating the power output of such systems (Ghisellini & Celotti 2001). However,
the mechanisms of jet formation, their acceleration and collimation, the way in which they couple to the
accretion process and also their matter and total energy content are yet to be well understood. In the past
decade it has become apparent that such jets are also commonly associated with stellar-mass black hole and
neutron star X-ray binary systems within our own galaxy: since the characteristic timescales of accretion
and jet formation should be proportional to the accretor size, and hence mass, then by observing X-ray
binaries on timescales of days to decades we are probing the equivalent of otherwise unobservable timescales
of tens of thousands to millions of years or more for active galactic nuclei.
The key observational aspect of X-ray binary (XRBs) jets lies in their synchrotron radio emission. Such
jets appear to come in two types: milliarcsec-scale 1 continuous jets with at radio-mm spectra, and arcsec-
scale optically thin jets resolved into discrete plasmons moving away from the binary core with highly
relativistic velocities (see Fender 2004 for a review).
1 corresponding to 1 AU at 1 kpc
40

In the past decade or so, thanks to coordinated multi-wavelength campaigns, considerable progress has
been made in our understanding of the link between the production of jets and the in ow of matter in
XRBs. Specically, by means of simultaneous radio/X-ray observations, a quantitative correlation has been
established between accretion and the production of steady jets in black hole XRB systems displaying hard X-
ray spectra, in terms of a tight non-linear scaling between the radio and the X-ray luminosities (Hannikainen
et al. 1998; Corbel et al. 2000; Gallo, Fender & Pooley 2003). Probably the most notable consequence of
this nding is that, in spite of their low radiative eôciency, these jets may carry away a dominant fraction
of the dissipated accretion power, requiring a substantial modication of the existing models (Gallo, Fender
& Pooley 2003). Our knowledge about jets in X-ray binaries is mostly based on observations of black hole
candidates; this is because, in general, neutron star X-ray binaries appear to be less radio loud than black
holes (Fender & Kuulkers 2001; Migliari et al. 2003). However, recent deep coordinated radio and X-ray
studies of neutron stars have shown an overall pattern of behaviour very similar to that of black holes
(Migliari et al. 2003, 2004; Muno et al. 2004). By comparing these two systems we can gather important
information on the driving mechanism for the production of jets and on the role played by the characteristics
typical of the compact object involved, as e.g. the spin, the presence of an event horizon in black holes or of
a solid surface and a magnetic eld in neutron stars.
A number of key questions in XRB jet physics need to be addressed, with a broader relevance for the
study of the time-variable jet/accretion coupling on all mass scales; among others:
What are the conditions for a collimated out ow to exist? Specically: Does the presence of a jet
require a geometrically thick accretion disc? Does the jet production mechanism switch o at very low
accretion rates?
The high-energy spectrum of compact objects accreting matter at low rates, from the active nuclei to
XRBs, are successfully modelled in terms of advection-dominated accretion ows (e.g. Narayan et al.
1997), i.e. geometrically thick two-temperature in ows in which most of the released accretion power,
instead of being radiated, is stored in the ions and advected towards the accretor. Advection-dominated
solutions are highly unstable to convection and large entropy gradients; as such, they are expected to
power strong winds, possibly collimated (Blandford & Begelman 1999). Gallo, Fender & Hynes 2004
have argued that relativistic out ows in stellar mass black holes exist down to Eddington luminosities 2
as low as 10 6 L Edd . However, the collimated nature of such out ows remains a matter of speculation,
and so does the mere existence of relativistic out ows below LX=L Edd = 10 6 . In fact, the minimum
Eddington ratio at which a steady jet has been resolved on milliarcsec scales is LX=L Edd = 10 2 (in the
10 solar mass black hole XRB Cyg X-1, with a ux density of 15 mJy; Stirling et al. 2001). Sensitive
VLBI observations are needed in order to address this issue.
What is the role of the magnetic eld/event horizon? Specically: Do high-magnetic eld neutron
star XRB host jets? Is advection across an event horizon needed in order to reproduce the observed
luminosity dierence between black hole and neutron star XRBs, or can the jets account for it?
A possible explanation for the fact that XRBs hosting neutron stars seem to be less radio loud than
black hole XRBs is that the presence of a magnetic eld anchored to the neutron star surface might
somehow inhibit the jet production. A systematic investigation of high-magnetic eld neutron star
XRBs is needed to address this issue (note that none of the XRB pulsars has ever been convincingly
detected as an incoherent synchrotron radio source; however, the existing upper limits are not con-
clusive, as they are actually compatible with radio detections of some low-magnetic eld neutron star
XRBs).
On the other hand, `quiescent' (LX=L Edd < 10 5 ) black hole X-ray binaries are signicantly less
luminous { in the X-ray band { than the neutron stars. This has been interpreted as the single
strongest evidence for the existence of an event horizon in accreting black holes, as the advected power
would be realised when it impacts the neutron star surface, while it would cross the horizon and be
lost forever in the case of a black hole (see Narayan et al. 1997). An alternative explanation, which
does not require advection across an event horizon, is the existence of `jet-dominated' states in XRBs
2 The Eddington X-ray luminosity is the maximum luminosity for a spherically accreting object, corresponding to 1:3  10 38
erg/sec per solar mass of the accreting object
41

(Fender, Gallo & Jonker 2003), i.e. states in which the dominant fraction of the accretion power is
carried away by the radiatively ineôcient jet. Crucial to this interpretation is the steepness of the
radio/X-ray correlation in neutron star XRBs (Migliari et al. 2003), for which conrmation is sought
after.
Technical Requirements
The SRT can help in solving the above questions by means of, respectively:
VLBI observations of low luminosity XRBs would have a major impact in the eld of XRB jet re-
search. In particular, achieving milliarcsec resolution for micro-Jy sources would give us the possibility,
for the rst time, to resolve the radio emission region of `quiescent' sources, representing the majority
of the XRB population.
Single-dish observations of neutron star XRBs to be possibly coordinated with X-ray observations.
The required rms noise level for this kind of study is of 0.01 mJy/beam, achievable by SRT at 22
GHz in 12 hours.
References
Blandford R., Begelman M. 1999, MNRAS 303, L1
Fender R. 2004, to appear in `Compact Stellar X-ray Sources' (astro-ph/0303339)
Fender R., Belloni T., Gallo E. 2004, MNRAS 355, 1105
Fender R., Gallo E., Jonker P. 2003, MNRAS 343, L99
Fender R.P., Kuulkers E. 2001, MNRAS 324, 923
Gallo E., Fender R., Hynes R. 2004, MNRAS, in press
Gallo E., Fender R., Pooley G. 2003, MNRAS 344, 60
Ghisellini G., Celotti A. 2001, MNRAS 327, 739
Migliari S. et al. 2003, MNRAS 342, L67
Migliari S. et al. 2004, MNRAS 351, 186
Muno M. et al. 2004, submitted. /astro-ph/0411313
Narayan R., Garcia M., McClintock J. 1997, ApJ 478, L79
Stirling A. et al. 2001, MNRAS 327, 1273
5.2 Extragalactic astronomy
5.2.1 Line observations
5.2.1.1 Mapping of HI absorption regions in extragalactic sources
Scientic Background
A critical ingredient of any orientation unication scheme of radio-loud Active Galactic Nuclei (AGN) is
an obscuring region of atomic or molecular gas surrounding the central engine. This circum-nuclear material,
which is believed to be in a disk or torus, eectively shields the inner few parsecs from view, if the axis of
the radio source lies at a large angle to the line of sight (e.g. Antonucci 1993). It is likely that this gas also
plays an important role as fuel for the central engine and in collimating the bipolar out ow in the radio jets.
It therefore forms a vital element of our understanding of AGN.
Although the composition of the obscuring material in the torus may be varied, it seems likely that
at some radii and scale heights, there will be signicant amounts of neutral atomic hydrogen gas (Neufeld
42

& Maloney 1995). Young radio-loud AGN, known as Compact Symmetric Objects (CSO) and Gigahertz-
Peaked Spectrum (GPS) sources, provide a unique opportunity to study this gas. In contrast to other types
of compact radio sources, they are not dominated by emission coming from Doppler-boosted components at
the base of a nearly face-on radio jet, but from isotropically radiating hot-spots or mini-lobes mostly seen
with the radio axis at a large angle to the line of sight. Since this makes it likely that a large fraction of
the radio-emitting plasma is located behind obscuring material, it makes them ideal probes to study HI in
absorption.
Indeed, observations show a large incidence of HI in absorption against young radio sources ( 50%:
Conway 1996, Vermeulen, in press, astro-ph/0012352, Mack et al. in prep.). Unfortunately, since most of
the redshifted HI absorption frequencies fall outside the standard frequency bands, follow-up observations at
high spatial resolution using VLBI-arrays are only rarely possible. This while VLBI observations are found
critical for our understanding of the location and distribution of the absorbing gas, and subsequently for the
interpretation as a possible torus. It is therefore important to increase the number of instruments which can
be tuned to detect redshifted HI.
Technical Requirements
At present, only 6 EVN telescopes are equipped with an UHF receiver, i.e. Eb, Wb, On, Tr, Jb1 and Ar.
For this reason, the addition of another large antenna, such as SRT, would signicantly contribute to the
performance of the EVN array. To accurately map the HI absorption, in particular to image dierent parts
of the line prole separately, and obtain kinematic information at a reasonable dynamic range, the requested
sensitivity is  1 mJy/beam/channel. Such values could be achieved by the EVN with the addition of SRT,
if we consider for instance a channel width of 15 kHz, and an integration time of two hours. At a more
general level, the inclusion of SRT in the UHF EVN array improves image (and channel) sensitivity by a
factor of  1.5. The instantaneous bandwidths required for typical extragalactic HI absorption features
is about 3 MHz at 750 MHz and 10 MHz at 1400 MHz. At least 3 channels should cover the absorption
features, thus at least 256 channels might be necessary. As the planned programme is not a blind survey
but based on previous detection measurements the redshift and the expected line widths will be well-known.
It is important, however, to point out that a wide band is necessary for this receiver in order to warrant
a continuous coverage in redshift. The UHF band is in the range 750 - 1400 MHz, however the frequency
coverage is dierent for the various antennas, and the above mentioned performance holds only in the
frequency range 1000 - 1200 MHz. Outside this range, the number of available telescopes decreases, and for
this reason the addition of a large telescope in the EVN is even more important.
References
Antonucci R. 1993, ARA&A 31, 473
Conway J.E. 1996, IAUS 175, 92
Neufeld D.A., Maloney P.R. 1995, ApJ 447, 17
Peck A.B., Taylor G.B., Conway J.E. 1999, ApJ 521, 103
Vermeulen R. 2002, in he Universe at Low Radio Frequencies, IAU Symp. 199, eds. A. Pramesh Rao,
G. Swarup & Gopal-Krishna, p. 91
5.2.2 Continuum observations
5.2.2.1 Millimeter VLBI Observations
Scientic Background
Nowadays, mm-VLBI observations are regularly performed at 86 GHz (3.5 mm), where images with an
angular resolution of up to about 50 microarcsec are obtained, but are still limited in sensitivity. For this
reason, only a few dozen compact objects with typical ux densities S(86 GHz)> 0.5-1 Jy can be reliably
imaged. This limitation can be overcome by (i) adding more collecting area, (ii) increasing the observing
bandwidth and data recording rate, (iii) correcting phase uctuations introduced by the atmosphere (water
vapour radiometry, dual VLBI). Another important limitation is the sparse and non-uniform uv-coverage
43

of mm-VLBI observations (to date limited to 4 European antennas), and in particular the lack of short
interferometer baselines. The implementation of the SRT in the European VLBI network will improve the
high-frequency sensitivity of the network by a factor of up to 4. In order to fully exploit the capabilities
of a 64m antenna, the 86 GHz receiver should allow to record data with large bandwidth ( 1 GHz), and
possibly in dual polarisation. In this context, with its large collecting area, high-frequency performance and
appropriate location within Europe, the SRT can make a signicant contribution to the scientic research
of compact galactic and extragalactic radio sources.
Technical Requirements
Receiver at 86 GHz, large bandwith ( 1 GHz), dual polarisation.
5.2.2.2 Wide-Field VLBI Imaging and Surveys
Scientic Background
The steadily increasing sensitivity of the VLBI arrays is opening up new areas of VLBI research such as
the study of the sub-mJy radio source population using deep wide-eld VLBI surveys. What is currently
routine for the VLA is becoming possible for the VLBI arrays, thanks to the improved performance of the
antennas (new receivers and improved surfaces) and of the correlators. VLBI observations of the sub-mJy
radio source population have suôcient resolution to distinguish between AGN and starburst activity in these
optically faint radio sources (e.g. the EVN observations of the Hubble Deep Field). New telescopes in the
near future (the SRT is one of them), and upgrades of the existing ones (in terms of better surface and/or
new receivers) are planned. In particular, new receivers must have low system temperatures (possibly less
than 30 K) and large bandwidth in order to reach the highest sensitivity despite the radio interferences
present mostly at the lower frequencies. The target is to perform continuum observations with r.m.s noise
levels of a few microJy, imaging a large fraction of the primary beam of the largest antenna in the array
(several arcminutes). Dozens of faint sub-mJy and microJy radio sources will thus be simultaneously imaged
with milliarcsecond resolution, full uv-coverage and micro-Jy sensitivity.
Technical Requirements
Receivers at 1.4 and 5 GHz, largest possible bandwidth, dual polarisation
5.2.2.3 A Study of Faint Extragalactic Radio Sources
Scientic Background
The radio sky at metre to centimetre wavelengths is dominated by extragalactic radio sources. At ux
densities ranging from tens of mJy to Jy or more most of them are active galactic nuclei (AGN). At lower
ux densities radio sources associated with starburst galaxies are the dominant population in the source
count statistics (Condon 2004). Most of the faint sources have radio luminosities L 1:4 < 10 23 W/Hz, steep
spectra and angular sizes comparable to, or smaller than, their parent galaxies.
A number of new radio surveys covering large parts of the sky down to low ux densities are now available,
in particular: the NRAO VLA Sky Survey (NVSS) and Faint Images of the Radio Sky (FIRST) were both
carried out at 1.4 GHz, but with dierent resolutions (Condon et al. 1998; Becker et al. 1995). A parallel
survey was carried out at 325 MHz with the Westerbork Radio Synthesis Telescope (WSRT), the WENSS,
which is similar in terms of sky coverage and resolution to the VLA survey NVSS (Rengelink et al. 1997). In
the optical band, the Sloan Digitised Sky Survey is providing a wealth of information, which is an invaluable
support to the radio surveys.
With these new surveys, covering large sky regions and reaching ux density limits much lower than
previously available, it is now possible to study the population of low luminosity radio galaxies, which were
very rare in the less sensitive surveys carried out before the 1990s. In particular, it is now possible to study
the nature of faint radio sources, associated with elliptical galaxies with radio power  10 23 W/Hz. This
value is an order of magnitude lower than the radio power typically found in the samples obtained from the
B2 catalogue, and may be suggestive of a dierent population of radio galaxies.
In the entire WENSS we expect to nd about 70 elliptical galaxies with radio power  10 23 W/Hz. This
number is large enough to perform a statistical analysis on the properties of elliptical galaxies with very low
radio power. The main questions to be addressed by such studies are the following:
44

1) Is the jet phenomenon still present in low power radio sources? If so, how do their inner jets compare
to those in radio galaxies of intermediate to high power?
2) Do the well known correlations found in intermediate and high power objects still hold at low radio
powers? An example of such correlations is for instance the ratio between the nuclear and the extended
radio power.
3) Which is the dominant radio emission mechanism in these objects? Is the radio emission still driven
by a central AGN, or is it dominated by starburst emission? Or both?
In order to tackle these issues it is necessary to look into the very inner region of these radio galaxies, and
this implies the need of the angular resolution provided by the Very Long Baseline Interferometry (VLBI)
technique.
Imaging at parsec-scale resolution will enable us to distinguish between AGN radio activity (if compact
cores and radio jets are detected) and starburst emission (which is usually characterised by extended and
diuse emission). Furthermore it will allow one to study the properties of both classes of radio sources.
The selected sample The WENSS is used as a starting point to construct a sample of radio sources
identied with bright elliptical galaxies, i.e. m r < 16:5 (de Ruiter et al., in preparation). We expect to nd
about 70 bright elliptical galaxies with radio power  10 23 W/Hz. Inspection of the FIRST survey (20 cm,
resolution of  5 arcsec) will allow us to select the most compact sources on the arcsecond scale, i.e. best
suited for follow-up VLBI studies. As an indication, the bulk of the faint sources selected from the WENSS,
have ux densities of a few tens of mJy on the FIRST survey, corresponding to surface brightness of the
order of a fraction of mJy/arcsec 2 (0.2 - 1 mJy/arcsec 2 ).
Technical requirements and feasibility
The EVN is currently the most sensitive VLBI array in the world, its best performances being reached at
1.4 and 5 GHz. This is possible especially thanks to the new MK5 recording system, which already allows
recording rates up to 1 Gbps (corresponding to up to 128 MHz bandwidth). Using the very sophisticated
calibration and imaging techniques developed at JIVE (Garrett et al. 2001), i.e. phase referencing and wide
eld imaging, it is possible to reach detection levels of the order of 40 Jy/beam in few hours of observations.
The beam of the EVN at 1.4 GHz is of the order of 5 - 15 mas, depending on the array. This means that
radio sources with surface brightness of 1 mJy/arcsec 2 are already detectable.
The addition of SRT will have the following major impacts on the EVN: 1) it will improve the uv{coverage
at short baselines, and this is crucial for the detection of extended nuclear features; 2) it will improve the
number of very large telescopes, re ecting on the whole array sensitivity. For instance, at the frequencies
requested to carry out this project, the presence of SRT in the EVN would improve the image sensitivity
(i.e. the thermal noise) by a factor of  2, going down to a value of  20 25 Jy/beam for a 256 Mbps
recording, 2-bit sampling and a total time on-source of 2.5 hours.
In order to carry out this project it is essential that the SRT is equipped with the most updated recording
technique for VLBI observations, and with 1.4-GHz and 5-GHz receivers.
References
Becker R.H., White R.L., Helfand D.J. 1995, ApJ 450, 559
Condon J.J., Cotton W.D., Greisen E.W. et al. 1998, AJ 115, 1693
Condon J.J. 2004, in Radio Astronomy at 70: from Karl Jansky to millijansky, eds. L.I. Gurvits, S.
Frey, S. Rowlings, EDP Sciences, in press
Garrett M.A. et al. 2001, A&A 366, 5
Rengelink R.B., de Bruyn A.G., Miley G.K. et al. 1997, A&AS 124, 259
45

Chapter 6
Geodesy with the SRT
Scientic Background
1) Science with geodetic VLBI
Geodetic VLBI is a useful and powerful tool for acquiring precise space geodetic observations and enhanc-
ing scientic knowledge of the Earth and its dynamics. This technique plays a unique role in determining
the International Celestial Reference Frame (ICRF) (Gontier et al. 2002; Fey et al. 2004) and, constraining
the scale of the frame, strongly contributes to the denition and realization of the International Terrestrial
Reference Frame (ITRF) (Altamimi et al. 2002). VLBI also contributes to Earth Orientation Parameters
(EOP) determination: it is the only space geodetic technique capable of accurately and simultaneously mea-
suring Universal Time (UT1) (Capitaine et al. 2003), polar motion (Ma 1978), precession and nutation
(Herring et al, 2002). These products are important for geophysical applications as well as for astronomy
and astrophysics. Geodetic VLBI can also be applied to investigate Earth's surface deformation, troposphere
water vapour content (e.g. Negusini and Tomasi 2004; Schuh et al. 2004) and ionosphere Total Electron
Content (TEC). A complete review of geophysical applications of VLBI technique can be found in Robertson
(1991).
2) Geodesy with co-located techniques
The impact of the SRT on geophysics and its performance in astronomy would be greatly enhanced by
realizing a co-location with other space geodetic techniques. The scientic perspectives would considerably
increase and the role of the observatory within the geodetic networks would be of greater importance. A
very easy, eôcient and rewarding way to co-locate a VLBI telescope is realized by means of the Global
Positioning System (GPS): it is based on cheap, easy-to-use and easy-to-manage devices with a wide range
of scientic, engineering and commercial applications (e.g. Homan-Wellenhof et al. 2001). A co-location
with the Galileo system and its dierent services must also be evaluated
(http://europa.eu.int/comm/dgs/energy transport/galileo/index en.htm).
a - International scenario
Simultaneous observations of dierent space geodetic techniques can independently study the same geo-
physical phenomena and estimate related parameters. Integrated Precipitable Water Vapour (IPWV) and
TEC can be determined by GPS data and can be used to model and improve astronomical VLBI observa-
tions (e.g. Ros et al. 2000; Erickson et al. 2001). Final EOP values are estimated combining VLBI with
other space geodetic observations. Integration of techniques and observations is the scientic strategy that
has recently been adopted by the International Association of Geodesy (IAG) for dening its future policy:
space geodesy as a tool and a service for other sciences by means of Integrated Global Geodetic Observing
System (Beutler 2004). Combination and integration of space geodetic observations are mandatory in order
to investigate technique-dependent biases and improve the accuracy of the products (Ray 2000). Within this
international frame, co-location of dierent space geodetic instruments are important and must be realized
and maintained (Altamimi 2004).
46

b - National scenario
The intense seismicity that interests Italy is an expression of the complex tectonic features that character-
ize the Mediterranean region (Mantovani et al. 2002). Earth's surface deformations are strictly related to the
geodynamical evolution of the area. Italian VLBI telescopes have proved to be very important in studying
crustal motion in the Mediterranean and in Europe (e.g. Campbell et al. 2002). The SRT will have a key role
in the evolution of Italian space geodesy: it can form a polyhedron of four national VLBI telescopes, which
would represent a unique facility in Europe. Scientic perspectives are similar to those of Japan or USA,
where analogous national interferometers can be found. A local VLBI-based crustal deformation monitoring
will be possible, simultaneously ensuring a global framing of the phenomena with very high accuracy (Key
Stone Project: http://ksp.nict.go.jp/). Furthermore, the activity of the national GPS reference network
(Biagi et al. 2001) will be preciously supported by the four Italian VLBI telescopes. A better framing into
ITRF can be ensured, a link to ICRF can be provided and long-term technique-dependent trends can be
monitored (Imakiiere et al. 2004). A complete integration between national VLBI and GPS networks will
be possible only if the SRT is co-located with a GPS system. This is already the case for Medicina, Noto
and Matera observatories.
3) Realization and maintenance of co-locations
Nevertheless, co-locations alone do not solve the problem of integration of space geodetic techniques.
Eective co-locations are ensured by provision of accurate local-ties (Rothacher 2000). Their importance
has been largely acknowledged by the International Earth rotation and Reference systems Service (IERS)
and the IAG. Lately, the attention of the geodetic community has been focused on computation strate-
gies of accurate local ties and Solution INdependent EXchange format (SINEX) les generation (Sarti and
Angermann 2004). A few rigorous methods have proved to meet all IERS requirements for space geodetic
observations combination and ITRF computation (Dawson et al. 2004, Sarti et al. 2004b). These methods
are also consistent to one another (Sarti et al. 2004a) and will allow the computation of a reliable and
accurate SRT-GPS connection.
4) Maintenance of a radioastronomical observatory
A local ground control network is an essential infrastructure that must be created at the SRT site. The
local network will be used to monitor the stability of the radio telescope structure, to externally monitor the
gravitational deformation of the dish, to locate and monitor its geometrical reference point and to monitor
the stability of the site. This latter point is particularly important to distinguish between local and global
signals in order to enhance data signicance (Sarti and Vittuari 2003). If a co-location is present, the local
ground control network is an essential tool for monitoring the eccentricity between the reference points of
the space geodetic techniques. These tasks are nowadays considered as attractive scientic issues but will
be considered as mandatory and challenging maintenance operations in the near future. It is therefore im-
portant to carefully design the shape and the geometry of the local network as well as its extension, and
carefully consider the position of the ground pillars and the technological solutions adopted for geodetic
marker materialization (Sarti and Angermann 2004).
Technical Requirements
Fulllment of scientic applications described in point 1 will be guaranteed with S and X-band receivers
and the Mark V recording system. A co-location with a GPS system (antenna and GPS-GLONASS receiver)
will be of enormous advantage for enhancing the SRT role at national and international levels (as described in
a previous section). The best implementation of the SRT co-location will be ensured by the results described
in 1.3, both for scientic and technical/commercial purposes. Temperature, humidity and pressure sensors
(as specied by International GPS Service standards) are also necessary in order to evaluate IPWV and
eventually use the SRT data for climate change and environmental studies. As described in a previous
section, the design and the realization of a local ground control network is essential. Scientic and technical
expertise for a proper evaluation of local requirements and for an optimized design of the ground network,
the pillars and the ground markers can be found within the structure of the new INAF.
47

References
Altamimi Z. 2004, Position paper on: ITRF and collocation sites. IERS Tech. Note 33, in press.
(http://www.iers.org/workshop 2003 matera/)
Altimimi Z., Sillard P., Boucher C. 2002, ITRF2000. J Geophys Res 107, 10, 2214.
Beutler G. 2004, J Geodesy 77, 560
Biagi L., de Lacy M.C., Sanso F., Vespe F., 2001, Proc. Conferenza Nazionale ASITA, 1, 215-220
Campbell J., Haas R., Nothnagel A. 2002, TMR Networks, Geod Inst, Univ Bonn - ISBN 92-894-0763-8
Capitaine N., Wallace P.T., McCarthy D.D. 2003, A&A 406, 1135
Dawson J., Johnston G., Digney P., Twilley B. 2004, IERS Tech. Note 33, in press.
(http://www.iers.org/workshop 2003 matera/)
Erickson W.C., Perley R.A., Flatters C., Kassim N.E. 2001, A&A 366, 1071
Fey A.L., Ma C., Arias E.F., Charlot P., Feissel-Vernier M., Gontier A.-M., Jacobs C.S., Li J., MacMil-
lan D.S. 2004, AJ 127, 3587
Gontier A.-M., Le Bail K., Feissel M., Eubanks T.M. 2001, A&A 375, 661
Herring T., Mathews P.M., Buett B.A. 2002, J Geophys Res 107, B4, 10, 1029/2001JB000165.
Homan-Wellenhof B., Lichtenegger H., Collins J. 2001, GPS: theory and practice. SpringerWien-
NewYork - ISBN 3-211-83534-2
Imakiiere T., Hatanaka Y., Kumaki Y., Yamagiwa A. 2004, GEONET: Nationwide GPS array of Japan.
GIS@development, March 2004.
Ma C. 1978, Nasa Tech. Memorandum 79582, Greenbelt M.D.
Mantovani E., Viti M., Albarello D., Babbucci D., Tamburelli C., Cenni N. 2002, Boll Soc Geol Italiana
121, 99
Negusini M., Tomasi P. 2004, in: International VLBI Service for Geodesy and Astrometry 2004 General
Meeting Proceedings, edited by Nancy R. Vandenberg and Karen D. Baver, NASA/CP-2004-212255,
p. 456
Ray J. 2000, in: Rummel R., Drewes H,. Bosch W., Hornik H. (eds) Towards an integrated global
geodetic observing system (IGGOS). IAG Symp 120. Springer, Berlin, Heidelberg New York, p. 19
Robertson D.S. 1991, Rev. Mod. Phys. 63, 899
Ros E., Marcaide J.M., Guirado J.C., Sardon E., Shapiro I.I. 2000, A&A 356, 357
Rothacher M. 2000, In: Rummel R., Drewes H., Bosch W., Hornik H. (eds) Towards an integrated
global geodetic observing system (IGGOS). IAG Symp 120. Springer, Berlin Heidelberg New York, 41
Sarti P., Vittuari L. 2003, Reports on Geodesy, Warsaw, 2(65), 73.
Sarti P., Angermann D. 2004, IERS Tech. Note 33, in press.
(http://www.iers.org/workshop 2003 matera/)
Sarti P., Dawson J., Johnston G., Sillard P., Vittuari L. 2004a, Abs.N.EGU04-A-06824, Presented at
EGU 04, Nice, France
Sarti P., Sillard P., Vittuari L. 2004b,- J Geodesy, available online, DOI 10.1007/s00190-004-0387-0
Schuh H., Snajdrova K., Boehm J., Willis P., Engelhardt G., Lanotte R., Tomasi P., Negusini M.,
MacMillan D., Vereshchagina I., Gubanov V., Haas R. 2004, in: IVS 2004 General Meeting Proc., eds.
N.R. Vandenberg and K.D. Baver, NASA/CP-2004-212255.
48

Chapter 7
Planetary Radar Astronomy
SRT as Radar for Asteroid and Space Debris Studies
Introduction Among the Solar System bodies, asteroids are the largest population. Most of them orbit
around the Sun in a region of space, called the \main belt", located between the orbits of Mars and Jupiter.
Two groups (the \Trojans") are located in the L4 and L5 Lagrangian points of Jupiter, while others (be-
longing to the Aten, Apollo and Amor groups, collectively known as NEAs, Near-Earth Asteroids) follow
trajectories that cross the orbits of the terrestrial planets, therefore representing a potential hazard for the
Earth. In fact, in the main belt, catastrophic collisions among asteroids can occur, and the fragments which
result from these collisions can be \injected", due to complex dynamic processes, in th inner regions of the
Solar System. The hazard of a catastrophic impact with a large body is not so unlikely. According to
the Torino-scale (dened in 1999 during the international IMPACT congress, held in Turin and regarding
the potentially hazardous asteroids), it is estimated that an impact producing sever consequences to the
terrestrial ecosystem can occur every few hundred thousand years. The rst step toward the \mitigation" of
such a threat is space surveillance, in order to determine the orbits of most of the objects whose trajectories
are close to the Earth's orbit.
In the last years, the Planetology Group of the Turin Astronomical Observatory, in collaboration with the
IRA, started radar observations of NEAs, which is one of the most powerful techniques for the dynamical and
physical study of these bodies, oering huge advantages with respect to the traditional optical observations.
In fact, from the dynamical point of view it is possible to improve the accuracy of the orbits by orders of
magnitude, a key factor to evaluate the impact hazard: unlike the \classical" optical measurements, the
radar astrometric observations can make the dierence in predicting if an asteroid will hit our planet or
not. From the physical characterization point of view, it is possible to obtain some parameters, such as
size, shape, rotational state and supercial structure, which cannot be derived with other techniques or are
derivable only in a very rough way and with time consuming optical observations.
Asteroid Radar Astronomy A radar observation of an asteroid consists in the transmission of a signal
with xed parameters and in the subsequent registration of the signal echo. The great advantage with
respect to other \passive" techniques lies in the observe control of all the characteristics of the coherent
signal (especially the wave form, time/frequency modulation, and polarization) used to illuminate the target
(Ostro, 1993).
From the echo analysis it is possible to determine the orbital elements of an object very accurately,
allowing a very high precision of ephemeri calculation. In fact, measuring the signal propagation time with
an accuracy better than 10 6 s, the radial distance of the target can be estimated with an error of some tens
of meters. Moreover, the component of the asteroid velocity, VLOS , along the \line-of-sight" (connecting
the radar antenna with the target), produces a Doppler shift in the frequency of the echo signal, which,
measured with a accuracy of  0:01 Hz, allows one an estimate of VLOS with an error of the order of
1 mm/s. Furthermore, the spin of the object generates a Doppler frequency dispersion of the signal; the
measurement of the power distribution of the echo signal, as a function of the time delay and of the frequency,
allows one to obtain bi-dimensional images with spatial resolutions less than 100 meters if the transmitted
signal is strong enough.
49

Finally, by measuring the polarization properties of the echo signal, radar observations permit one to
infer some surface characteristics of the target such as roughness, albedo and abundance of metallic elements,
information which cannot be directly derived by using other astronomical techniques. In fact, the re ection
due to a single re ecting plane surface reverts the handedness, or helicity of a circularly polarized wave,
so single back-re ections from dielectric interfaces, whose sizes and radii of curvature greatly exceed the
wavelength, yield echoes almost entirely in the opposite circular (OC) polarization. On the contrary, same
circular (SC) echo power can arise from multiple scattering, from single backscattering from interfaces with
wavelength-scale radii of curvature (e.g., rocks), or from subsurface refraction eects. Therefore, the circular
polarization ratio, that is, the ratio between the radar cross-section (dened as 4 times the re ected power
per unit solid angle and per unit ux of incident power) measured in the two opposite polarization ways, SC
and OC , of the re ected signal (C = SC =OC ), is a useful gauge of the target near-surface wavelength-
scale complexity, or \roughness", from which important information about the nature of the surface regolith
as well as the radar albedo and also the presence of supercial metallic elements, can be inferred (Ostro
1993).
Consequently, radar observations are a unique, very powerful tool to study the macroscopic physical
properties (diameter, shape, rotation period and spatial orientation of the spin axis) and surface properties
of the target.
Space Debris Radar Observations Another fundamental application of the radar technique is the study
and monitoring of the space debris orbiting around the Earth, and the interaction of meteoroids with the
atmosphere. The articial material orbiting around our planet consists of actually operative structures only
to a minimum extent; most of them consist of out-of-use spacecrafts, rocket stages and fragments generated
by explosions and collisions of articial satellites.
Nowadays, optical and radar sensors are used to locate centimeter-sized particles (in Low-Earth Orbit,
or LEO) or decimeter-sized objects (in Geostationary Orbits, or GEO). The USA, and, to a lower extent,
Russia, have a space surveillance system, while Japan is starting such a survey. So far, Europe does not
have these facilities: only one radar (the TIRA of the FGAN Institute, in Germany) an one telescope have
been occasionally used by the European Space Agency (ESA) in order to detect and monitor space debris.
In particular, in Italy such campaigns have never been undertaken.
The problem of articial space debris is now analysed in a strict relation with the meteoroid environment.
ESA dened the Meteoroid and Space Debri Environment Reference Model (MASTER) to determine the
ux originating from the environment of particles following orbits close to those of space shuttles, in the
LEO, MEO and GEO regions. The analysis of material coming from space and \in situ" experiments
contributed to extending the knowledge on natural particles with millimetre and micron sizes, which are the
most abundant. As for spacecrafts in LEO-type orbits, impact velocities range from 5 to 15 km/s with a
mean value of about 10 km/s for space debris, and from 12 to 72 km/s with a mean value of 17-20 km/s for
meteoroids. Studies on impact residues, carried out on the LDEF, EURECA and HST satellites in order to
discriminate between impacts by natural meteoroids and by articial space debris have not been conclusive,
because of the complexity of the targets.
In GEO-type orbits, the ux of natural particles prevails over the ux of articial particles, whereas it is
assumed that collisions in LEO-type orbits are mostly due to articial objects. Particles coming from mete-
oroids are dangerous for orbiting structures because their impact energy exceeds 2 kJ, a value corresponding
to 3 to 10 g for some meteoroids and to 10 to 30 g for particles belonging to a meteor stream.
Therefore, in order to monitor the natural and articial debris population and to understand their evolu-
tion both on the short and long run, it is absolutely necessary to complete the models through observations
carried out from the Earth and from space.
The information provided by a radar system can be exploited to validate current models of debris envi-
ronment; they can also improve the precision with which we know the orbital parameters of those catalogued
debris for which a close encounter with an operative satellite, or a manned space shuttle, is predicted. Fi-
nally, they can verify the integrity of big wrecks and update, if possible, the catalogues of big debris being
currently tracked.
SRT as Radar for Asteroid and Space Debris Studies Our interest in radar observations has been
fostered by the future realization of the Sardinian Radio Telescope (SRT), that can be used both as a
50

receiving antenna of a bistatic system and as an independent (monostatic) system if it would be supplied
with a power transmitter. Of course, this second conguration, which provides the best combination of the
necessary technical parameters and full independence of the observations, is the most desirable.
In December 2001 the rst intercontinental experiment (Italy - Ukraine - USA) for the radar detection of
an asteroid took place, involving the Osservatorio Astronomico di Torino, the Istituto di Radioastronomia
(IRA) in Bologna and NASA-JPL, under the local supervision of M. Di Martino (Turin) and S. Montebugnoli
(Bologna). From the Goldstone (Mojave desert, California) and Evpatoria (Crimea, Ukraine) antennas,
monochromatic radio signals were transmitted towards asteroid 1998 WT24, less than 2 million km from
the Earth at the time. Echoes from the asteroid were detected by the 32-m VLBI antenna in Medicina, and
analyzed by means of a high-resolution, high-eôciency spectral analyzer. After approximately 12 s from rst
detection, the echo could be sharply resolved on the screen of the receiving station back-end, thus successfully
achieving the intended goals of the experiment (Di Martino et al. 2004). This experiment could represent
the rst step towards an integrated intercontinental network for the monitoring of potentially dangerous
NEA's, a network in which SRT could play a key role.
Radar Astronomy with SRT The scientic value of a radar experiment depends mainly on echo strength
and receiver sensitivity. The received power PR inversely scales with the fourth power of the target distance
R, and can be computed by means of the radar equation:
PR = P T G T GR  2 
(4) 3 R 4
(7.1)
where  is the transmitted wavelength, P T the transmitte power, G T the gain of the transmitting antenna,
GR the gain of the receiving antenna, and  the radar cross section of the target. The antenna gain is given
by:
G = 4
 2 A e (7.2)
where A e is the eective area of the antenna, obtained by multiplying the geometric area of the antenna by
its eôciency.
The power of the received signal can be very small compared to the syste noise power in the receiving
system, which is given by:
Pn = k T s
 (7.3)
where T s is the noise temperature in the receiver, k is Boltzmann's constant and
 is the bandwidth of the
receiver.
The echo signal can be detected when the received power PR is signicantly above the threshold given
by the RMS of the noise power. A quantitative measure of the extent to which the signal is stronger than
the noise, that is the quality of a radar measurement, is expressed by the SNR. It can be shown that the
SNR value expected for a radar observation is given by the following equation:
SNR = P T G T GR  2:5 ^
 D 1:5 P 0:5
p

t
8:96
10 3 k T s R 4 (7.4)
where ^  is the target radar albedo, dened as the ratio between its radar and geometric cross sections, D is
the target diameter, and
t is the integration time.
SRT as Receiver of a Bistatic Radar We consider two dierent possibilities for the use of SRT as a
planetary radar: SRT as a receiver in a bistatic radar system, and SRT as a monostatic radar.
In the rst conguration, SRT would act as the receiving station. The specications for the receiver
presented in Grue & Ambrosini (1998) provide a highly versatile instrument, covering a frequency range
spanning from 300 MHz up to 100 GHz. It is thus certainly possible to receive a signal transmitted, for
example, by the DSS14 antenna at Goldstone, operating in the X-band ( = 8560 MHz,  = 3.5 cm). In
this case, the values for the terms in the radar equation are P T  500 kW, G T  75.6 dB,  = 3.5 cm.
The gain for SRT in the X-band is approximately 73 dB, T s is around 30 K, and the diameter is 64 m, for
51

an eective area of 1945 m 2 . It is also necessary that the target is in a visibility window common to both
antennas, meaning that the target must have a North declination greater than 45 degrees.
SRT as Monostatic Radar For a completely independent radar system, it is necessary to provide the
radiotelescope with a power transmitter. If a comparison is made between the sensitivities of SRT and DSS14
in the X-band, it can be seen that SRT is approximately four times less sensitive. A better opportunity is
oered by the use of a transmitter operating in the Ka band ( = 34 GHz ,  = 0.8 cm P T  1 kW): writing
explicitly the eective dependence from the wavelength in the radar equation, one obtains:
PR = P T  A 2
e
4  2 R 4
(7.5)
the relative power of the echo signal produced by a transmitte operating in the Ka band is thus given by:
 2
X
 2
Ka
= 3:5 2
0:9 2  15 (7.6)
that is 15 times larger than the one produced by a transmitter operating in the X-band, for the same radiated
power. In terms of SNR, the net gain is reduced, because, as can be seen from equation (7.4), it scales
as  1:5 . Furthermore, it is necessary to take into account the noise temperature of the receiver, which
increases with frequency.
The following table illustrates the performance of SRT in the two bands:
Band G (dB) T (K) G=T (dB)
X 73 30 58
Ka 85 50 68
To obtain the net gain in terms of SNR for a Ka-band transmitter over an X-band transmitter, for the
same integration time, it is suôcient to substitute the gures listed above in equation (7.4), obtaining that
SNR 2
Ka =SNR 2
X  5.
SRT and Space Debris The seemingly unstoppable increase in the number of space debris in low orbits,
especially below 2000 km of altitude, poses increasingly larger threats to all space activities in that region.
Any prediction of the evolution of their population, especially in the long term, and any protection and
mitigatio measure requires adequate knowledge of their present distribution, also following peculiar events
such as an explosion or a loss of material.
Low-orbiting space debris with a size larger than 10 cm (now around 10,000 in number) are routinely
monitored mainly by the US-based USSPACECOM surveillance systems; their orbits are known only in terms
of the so-called \two lines elements" (http://www.amsat.org/amsat/keps/formats.html) provided by NASA.
To this class belong also space vehicles out of control, especially those who are re-entering the atmosphere:
because predictions of place and time of their re-entry are aected by large uncertainties, it is important to
have frequent updates of their orbit.
Our knowledge of the smaller debris, in particular those of millimetrer size, is in general indirect and
statistical: On the other hand, they are capable of producing signicant damage to space systems, and they
constitute a large fraction of man-made debris, larger than natural space debris. (The number of debris is
inversely proportional to their size.)
The main technique for monitoring them is radar detection, requiring large-size instruments: for this
task, the SRT would play an extremely signicant role, also in the context of international space policy,
especially that of ESA. All major space agencies have in fact initiated ambitious programs for monitoring
and prediction of, and protection from, space debris. Italy is a member of the Inter-Agency Space Debris
Coordination Committee (IADC); among the four main areas in which the space debris problem is studied
(Measurements, Environment and Database, Protection, Mitigation), the Italian contribution is particularly
weak (practically absent) in the rst one: the only radar observations of space objects performed in Italy
are those of natural meteorites. Europe in general is not well equipped in this respect; the main instrument
available is the military radar TIRA belonging to the German FGAN institute, which, under ESA control,
has occasionally provide important observations. Thus, there is plenty of opportunities for SRT in this eld
of research.
52

Conclusions In the light of the above, it is proposed that the SRT antenna be used for planetary radar
observations, both in bistatic and in monostatic mode. In the second case, which is deemed to be more
eôcient both for achieving scientic results and for monitoring circumterrestrial space, it is proposed to
provide SRT with a power transmitter and associated auxiliary structures, such as two parallel receiving
channels for simultaneous reception of same-sense and opposite-sense circular polarizations. In a study
performed at JPL (Interoôce Memorandum 335.1-95-038), the technological feasibility of a transmitter
operating in the Ka band with a 1 MW power has been demonstrated, using either a single transmitter or
a pair of 500-kW transmitters. The total cost of the device was estimated at 2,000,000 US$, corresponding
to about 1.500.000 Euro (exchange rate of 25-11-2004).
This kind of operation requires high stability and accuracy both in time and frequency measurements,
and, in the case of a bistatic system, in the synchronization between the transmitting and receiving stations.
Furthermore, the availability of such a complete instrument of investigation for both planetary and space
debris studies would allow the creation of an entirely national radar network together with the two twin
32-m antennas of Noto and Medicina.
References
Di Martino M., Montebugnoli S., Cevolani G., Ostro S., Zaitsev A., Righini S., Saba L., Poppi S.,
Delbo M., Orlati A., Maccaferri G., Bortolotti C., Gavrik A., Gavrik Y. 2004, Planet. Space Sci. 52,
325
Grue G., Ambrosini R. 1998, Planet. Space Sci. 46, 1393
Ostro S.J. 1993, Rev. Modern Phys. 65, 1235
53

Chapter 8
Space Science with the SRT
Since the start of the SRT project, `Space Science' applications have been included among its primary
research objectives. The reasons are twofold. Firstly, the Istituto di Radioastronomia has successfully
participated in many observational campaigns of Doppler Tracking of interplanetary spacecrafts with the
Medicina radio telescope: rst with the Ulysses mission and, more recently using the upgraded parabola at
Noto, with Cassini, even up to Ka-band (32 GHz). The primary observables in this type of experiments are
the phase and the amplitude and polarization of the radio signals making the radiolinks that connect the
spacecraft to Earth, both in the up and down-stream interplanetary paths. The extremely high frequencies
of these downlinks, now at X-band (8.4 GHz) and Ka-band (32 GHz), and the utilization of an atomic
frequency standard at the ground stations, allow the measurement of residual Doppler frequencies at the
level of fractions of microHertz over integration times of a thousand seconds. Such a tremendous resolution
opens observational windows suitable to make accurate checks of General Relativity like for example the
search for an experimental evidence of gravitational waves and the measurement of the Gamma constant
with uncertainties two orders of magnitude less than reached before, as recently obtained by the Italian
Radio Science team. The name `Radio Science' is used to identify all classes of scientic applications of the
Doppler tracking technique that now includes such diverse topics as:
 Planetary atmospheric proles and ionospheric compositions (one way, downlink only),
 Structure of planetary rings (one way, downlink only),
 Planetary surface characteristics (in the bistatic radar conguration),
 Planetary gravitational elds, shapes, masses, and ephemerids (two way, full up+down),
 Solar corona and solar wind (two way, full up+down links).
The second reason for doing `Space Science' with the SRT is related to the interest shown by the Italian
Space Agency in using such an antenna for sending uplink commands to spacecraft, if properly equipped
with special `transmitting' hardware. To this end, the SRT was designed for use as a full-tracking ground
station with simultaneous receiving and transmitting capability at four bands: X up (7.3 GHz), X down
(8.4 GHz), Ka up (34 GHz) and Ka down (32 GHz). The necessary equipment (mirros and electronics)
would be installed in the Beam Waveguide cabin, in the tertiary focus of the SRT. This second application
of course would have put extremely tight requirements on the operational reliability, compatibility, security
of the SRT and the allowance for special transmitting licenses.
At this moment, only the receiving conguration is planned for Space Science with the SRT. The inter-
ested scientists, apart from the Radio Science team, are then restricted to those willing to collect downlink
data from a spacecraft at a higher data rate (due to the extremely large collecting area of the SRT that gives
a higher S/N), and its availability at transit times dierent from those at other ground stations, normally
used for the tracking.
In the years to come the SRT, if equipped with dedicated data acquisition systems suitable for Space
Science applications will be able to play a strategic role, in special circumstances due to its much higher
sensitivity than the typical 34-m tracking antennas (a factor of between 4 and 8 is expected, both for the
54

larger collecting area and the active surface concept). The double frequency, X + Ka, primary focus receiver,
originally designed for and successfully used in Noto, will be easily installed on the SRT which has the same
F/D ratio for the main re ector. Its use will open new opportunities in Space Science experiments not only
with the last part of the Cassini mission, now supposed to end by 2009, but also with the new NASA - ESA
missions like Bepi Colombo and others to come.
55

Chapter 9
SETI
Scientic Background
Are We Alone in the Universe? This is a fundamental and crucial question, which has haunted mankind
since we realized that the points of light in the night sky are other suns. Today we have the technology to seek
a denitive answer. This is the aim of the Search For Extra Terrestrial Intelligence (SETI), an international
eort to look for articial radio signals coming from the outer space. A hypothetical result could represent
for mankind the biggest discovery of the millennium. This is a very hard challenge for modern physics,
technology and philosophy/theology. The Drake formula attempts to give us an estimate of the number of
communicative civilisations.
N = Rf p n e f l f i f c L
N = The number of civilisations trying to contact us
R = The average rate of star formation
f p = The fraction of stars that are suitable for planet formation
n e = Number of Earth-like planets
f l = Fraction of Earth-like planets where life develops
f i = Fraction of Earth-like planets where life has intelligence
f c = Fraction of intelligent species who want to communicate
L = Lifetime of a civilisation
Where Should we look?
 Frequency: Radio from 1 GHz to 10 GHz for the low background noise level and particularly at the
HI and OH emission frequencies (the so-called `waterhole')
 Polarization: Circular, linear, modulated
 Direction: Targeted search, All-Sky survey, Serendipitous mode (piggy-back observation), Radio Leak-
age (unintentional signals)/RF leakage (Kashunen-Loewe Transformation is requested)
 Narrow band signals { preferably pure continuous wave (CW) { (intentional signals) (Fourier-Transformation
is requested)
What signal should we look for?
Intentional radio message (CW):
 A signal with Doppler shift
 A signal with Doppler shift would indicate non-terrestrial. CW signal to get attention
 A CW beacon would stand out from Galactic noise
56

 Can be seen with low bandwidth receiver
Unintentional radio message:
 A radio signal with a probably unknown modulation.
This will spread the signal over a wide bandwidth and make the signal hard to detect
Technical requirements
We need to look at as much bandwidth as possible and at the same time we need to divide in very
narrow channels to increase the sensitivity. This means t operate with an extremely high resolution spectrum
analyzer (at least 0.7 Hz). The backend that could be used is the already existing Serendip IV high-resolution
spectrometer at present operating at the Medicina 33-m VLBI dish (15 MHz at 24 million channels). With
a given Eective Isotropic Radiated Power (EIRP) from an extraterrestrian transmitter, the beacon can be
detected at the distance of
D =
r
EIRP
A eff
4kTB
where EIRP is beacon power
k is Boltzmann's constant
B is frequency resolution
T is antenna temperature
A eff is the eective aperture of the receiving antenna
Having a larger antenna means the detection of the same signal coming from larger distances. In other
words, the distance covered (at a given power) by an extraterrestrial transmitter is dependent also on the
square root of the collecting area. For this reason the SRT is more suitable for this search than the present
Italian antennas. The starting point for this program is to install the existing high-resolution Serendip IV
spectrometer in the SRT. The main characteristics of this spectrometer are:
 Input bandwidth 16 MHz
 Number of channels 24.000.000
 Integration time 1.7 sec.
 Realtime processing.
57

Appendix A
Appendix
A.1 The SRT Receiving System Plan
In the framework of the planning of the SRT construction, the Board of the SRT has established several
working groups with the aim to provide the antenna with specic tools able to obtain a state-of-the-art
radioastronomical instrument. One of them is the receiver group, to which the design of the receivers and
of the complete receiving system architecture was assigned.
The key points of the receiving system can be summarized as follows:
 Design receivers that can be used for single dish as well as for VLBI, which avoids having more receivers
than the available space will permit at all three foci. It will also avoid the need of a manual mounting
and dismounting and oer an automatic and remote `frequency agility'. It should be noted that this
accomplishment will allow a receiver change in less than 4 minutes, in the worst case!
 Oer a continuous coverage of the RF (Radio Frequency) band from 300 MHz to 100 GHz. There are
exceptions to this in the lower part of the frequency spectrum due to man-made interference. All the
receivers will be dual channel.
 Design and construct large bandwidth feed systems (up to about 33% of the nominal frequency),
keeping good performance in terms of return loss, cross polarization purity, insertion loss and sidelobes.
 Oer a receiver noise temperature as low as possible by cooling most part of the feed system together
with the Low Noise Amplier (LNA)
 Oer dierent and selectable instantaneous bandwidths (IF), up to 2 GHz, leaving anyway the pos-
sibility to add electronics for particular and `extreme' receivers aimed at exploiting the full RF band
(for example RF polarimetric and/or total power receivers). The IF band will be positioned from 100
MHz to 2100 MHz spanning all the RF bandwidth by tuning a proper value of the local oscillator.
Other IF lters available are 80/400/800 MHz wide.
 Organize a receiving architecture and interconnections such that all the bands and bandwidths will be
automatically selectable and will provide the signal for backends mounted both on the antenna and in
the control room (500 m away, by using ber optic links).
58

Table
A.1:
Instrumentation
Requirements
for
Single-Dish
Observations
Derived
from
the
WG
Report.
Requirements
of
Scientic
WG
Project
Scientic
Keywords
Reference
Backend
Monofeed/ Multifeed
RXs
(GHz)
Instantaneous Bandwidth (MHz)
Sensitivity
Resolution (km/s)n
kHz
Feed
system
Antenna mode
Observation
Mode
1
Comets
Spectrometer
Monofeed
>
10
10
50
mK
0:1n
-
-
2
Galactic,
Masers
Spectrometer
Monofeed
1.612,1.712 12 20,2224 3638 43

200
0.12
Jy
0.01
to
0.25
n
0.05
-
-
-
3
Galactic,
NH3
Spectrometer
Multifeed
24
20
50
mK,
70
mJy
<0.1n8
-
-
-
4
Galactic,
DCO+,
N2D
+
Spectrometer
Multifeed
7080
?
20
mK,
30
mJy
n3
-
wobbling
Frequency
switching
5
Galactic,
ISM
Molecules
Spectrometer
Monofeed
1100
500
10
mK,
15
mJy
0.1n
-
-
-
6
Galactic,
Active
binaries
Total
Power
Monofeed,
Dualfeed
1.6

45
largest
possible
1
mJy
-
-
raster
scan
Frequency
agility
7
Galactic,
CMBP
Polarimeter
Monofeed
4.8
5001500
0.5
mK

p
s
-

35
dB
cross
cross-scan 4
0
/s
Mapping
software
8
Galactic,
Pulsar
Spectrometer
Monofeed
&
Multifeed
0.325,1.4,3
32,600,1000
8.5,0.9,?
mJy
p
s
n32,500,1000
-
-
-
9
Galactic,
Protostars
Spectrometer
Monofeed
better
Multifeed
20,
40,
70,
90
15
20
mK
0.2
n
10
-
-
Mapping
Software
10
Galactic,
X-Ray
Binaries
?
?
22
2000
0.01
mJy
-
-
-
-
11
Extragal.
HI
blind
Spectrometer
Multifeed
<
1.42
70
3
mJy
20n100
-
-
-
12
Extragal.
H2O
Mmaser
Spectrometer
Multifeed
22.2,1620
>
200
0.7
mJy
1.3n100
-
-
-
13
Extragal.
CO
Spectrometer
Monofeed
2737 3848 69104
10000 10000 35000
0.1
mJy
0.3
mJy
1.1
mJy
20n2000 20n3000 20n8000
-
-
-
14
Extragal.
sky
survey
Total
power
Multifeed Multifeed Multifeed
22 30 40
2000 2000 2000
0.050mJy 0.030mJy 0.015mJy
-
-
raster-scan
-
15
Extragal.
RM
sources
Polarimeter
Monofeed
5
500
0.24
mJy
-
-
cross-scan
Mapping
software
16
Extragal.
source
survey
Total
power
&
Polarimeter
Monofeed better Multifeed
20 32 43 86
2000 2000 2000 2000
13.4
mJy
8
mJy
5.7
mJy
2.5
mJy
-
-
cross-scan raster-scan
Mapping
software
17
Extragal.,
GRG
mapping
Total
power
&
Polarimeter
Monofeed Dualfeed Multifeed
20 32
2000 2000
0.3mJy 0.3mJy
-
-
raster-scan 20
0
/min
Mapping
software
18
Extragal.,
Blazar
monit.
Total
power
&
Polarimeter
Dual
Feed

10
2000

5mJy
-
<
35
dBcross
-
Frequency
agility
1
min.
19
Extragal.,
Sunyaev-Zel'd
Total
Power
Multifeed
20,40
>
2000
0.1
mK
-
-
beam
switch. wobbling raster
scan
-
20
Space
Science
tone
extractor
Dual
Frequency
8.4
/
32
-
-
-
-
21
SETI
Spectrometer
Monofeed
110
15

20
-
n
0.0007
-
-
-
22
Planetary
Radar
Spectrometer
Monofeed
8.5,
34
?
?
?
-
-
-
59

Table
A.2:
Instrumentation
Requirements
for
VLBI
Observations
Derived
from
the
WG
Report.
Requirements
of
Scientic
WG
Project
Scientic
Keywords
Reference
Backend
Monofeed/ Multifeed
RXs
(GHz)
Instantaneous Recording Rate
Sensitivity
Resolution (km/s)n
kHz
Feed
system
Antenna mode
Observation
Mode
1
Redshifted
hydrogen
MK5
Monofeed
0.8

1.3
-
-
-
-
Frequency
Agility
2
Wide-Field
VLBI
Imaging
and
Surveys
MK5
Monofeed
1.4,5
up
to
1Gbps
-
-
-
-
Frequency
Agility
3
Methanol
line
MK5
Monofeed
6.7
up
to
1Gbps
-
-
-
-
Frequency
Agility
4
AGN
MK5
Monofeed
8.4
up
to
1Gbps
-
-
-
-
Frequency
Agility
5
AGN,
X-ray
Binaries
MK5
Monofeed
22
up
to
1Gbps
-
-
-
-
Frequency
Agility
6
AGN
MK5
Monofeed
43
up
to
1Gbps
-
-
-
-
Frequency
Agility
7
Millimiter
VLBI
MK5
Monofeed
86
up
to
1Gbps
-
-
-
-
Frequency
Agility
8
High
Sensitivity
Array
MK5
Monofeed
0.33,0.61,1.4,5 8.4,15,22,43
256
Mbps
-
-
-
-
Frequency
Agility
9
Geodesy
MK5
Coax
Dualfrequency
2.3
/8.4
256Mbps
-
-
-
-
Frequency
Agility
60

Table
A.3:
Frontend
Congurations
Planned
by
the
Receiver
Group
Planned
by
Receiver
Group
Project
Planned RXs
(GHz)
Instantaneous Bandwidth (MHz)
Sensitivity (continuum
2ch)
Resolution feasibility (Km/s)nKHz
Feed
system feasibility
Antenna mode
Observation Mode
Remarks
1
1.31.8 10.314.4 1826 3548
80/500
80/400/800/2000 80/400/800/2000 80/400/800/2000
2.3
mJy

p
s
0.7
mJy

p
s
1.7
mJy

p
s
1.6
mJy

p
s
Yes
-
-
-
2
1826
80/400/800/2000
1.7
mJy

p
s
Yes
-
-
-
3
7090
80/400/800/2000
5.1
mJy

p
s
Yes
-
Yes
To
be
implemented
wobbling
time
=
0.7
s,
o=5*HPBW
4
see
planned
RF
band
list
80/400/800/2000
see
this
column
Yes
-
-
-
5
see
planned
RF
band
list
80/400/800/2000
see
this
column
-
-
Yes
Yes
6
4.35.8
80/400/800/1500
0.54
mJy

p
s
-
Yes
Yes
Previous
experience
7
0.310.35,1.31.8,3.224.3
40,500,1100
8.5/0.9/?
mJy

p
s
Yes
-
-
-
8
18
26
26
36 3548 7090
80/400/800/2000 80/400/800/2000 80/400/800/2000 80/400/800/2000
1.7
mJy

p
s
0.85mJy

p
s
1.6
mJy

p
s
5.1mJy

p
s
Yes
-
-
-
9
1826
80/400/800/2000
1.7
mJy

p
s
Yes
-
-
-
10
1.31.8
80
2.3
mJy

p
s
Yes
-
-
-
11
1826,14.419.8
80/400/800/2000
1.7/1.2
mJy

p
s
Yes
-
-
-
12
2636 3548
7090
+
90115
2000 2000 2000
0.85
mJy

p
s
1.6
mJy

p
s
5.1+
6.4
mJy

p
s
Yes
-
-
-
Widest
BW
at
30
and
40
GHz
possible,in
principle,
by
additions
to
planned
receiver.
69104
very
problematic.
13
1826 2636 3548
2000 2000 2000
1.7
mJy

p
s
0.85
mJy

p
s
1.6
mJy

p
s
-
-
Yes
-
Widest
BW
at
30
and
40
GHz
possible,
in
principle,
by
additions
to
planned
receiver.
Requested
sensitivities
can
be
achieved
with
planned
RX.
14
4.35.8
400/1500
1.1/0.54
mJy

p
s
-
-
Yes
Previous
experience
15
1826 2636 3548 7090
2000 2000 2000 2000
1.7
mJy

p
s
0.85
mJy

p
s
1.6
mJy

p
s
5.1
mJy

p
s
-
-
Yes
Previous
experience
BW=10
GHz
needs
additions
to
planned
receiver.
Requested
sensitivities
can
be
achieved
with
planned
RX.
16
1826 2636
2000 2000
1.7
mJy

p
s
0.85
mJy

p
s
-
-
Yes
Previous
experience
17
see
planned
RF
bands
list
2000
see
this
column
-
Yes
-
Yes
61

Table
A.3:
Frontend
Congurations
Planned
by
the
Receiver
Group
(cont'd)
Planned
by
Receiver
Group
Project
Planned RXs
(GHz)
Instantaneous Bandwidth (MHz)
Sensitivity (continuum
2ch)
Resolution feasibility (Km/s)nKHz
Feed
system feasibility
Antenna mode
Observation Mode
Remarks
18
18

26
35

48
2000
1.7
mJy

p
s
1.6
mJy

p
s
-
-
Yes
-
Widest
BW
possible,
in
principle,
by
additions
to
planned
receiver.
19
No
planning
20
1.31.8,2.363.22,3.224.3 4.35.8,5.77.7,7.510.4
80/400/800/2000
see
this
column
Yes
-
-
-
21
7.510.4,2636
80/400/800/2000
see
this
column
-
-
-
22
1826
80
1.7
mJy

p
s
Yes
-
-
-
62

Table A.4: Starting from the planned receivers bands this table identies which projects use it together with
which backend and type of receiver
Planned RXs Bands vs Requested Projects
Planned RF
Bands (GHz) Requesting Projects Requested Backend Requested RX
0.310.35 Pulsar
VLBI
Spectrometer
MK5 Monofeed
0.580.62 VLBI MK5 Monofeed
0.71.3 VLBI MK5 Monofeed
1.31.8
Maser
ISM Mol.
Active Binaries
Pulsar
Extragal. HI
VLBI
SETI
Spectrometer
Total Power
MK5
Mono/Dual
Multifeed
3.224.3
Pulsar
Active Binaries
SETI
Spectrometer
Total Power Mono/dualfeed
4.35.8
ISM Mol.
Active Binaries
CMBP
Extragal. RM
VLBI
SETI
Spectrometer
Polarimeter
Total Power
MK5
Monofeed
5.77.7
VLBI
Active Binaries
SETI
Spectrometer
Total Power
MK5
Monofeed
2.3/8.4 VLBI
Planetary Radar
Spectrometer
MK5 Monofeed
10.314.4
Maser
ISM Mol.
Active Binaries
Blazar Monitoring
Comets
Spectrometer
Total Power
Polarimeter
Mono/Dualfeed
14.419.8
ISM Mol.
Megamaser H 2 O
Active Binaries
VLBI
Blazar Monitoring
Comets
Spectrometer
Total Power
MK5
Polarimeter
Mono/Dual/Multifeed
63

Table A.4: Starting from the planned receivers bands this table identies which projects use it together with
which backend and type of receiver (cont'd)
Planned RXs Bands vs Requested Projects
Planned RF
Bands (GHz) Requesting Projects Requested Backend Requested RX
1826
Maser
Gal. NH 3
ISM Mol.
Active binaries
Megamaser H 2 O
Extragal. sky survey
Extragal. source survey
Extragal. GRG
VLBI
X-RAY Binaries
Protostar
Blazar Monitoring
Sunyaev - Zel'dovic
Comets
Spectrometer
Total Power
Polarimeter
MK5
Mono/Dual
Multifeed
2636
ISM Mol.
Active Binaries
Extragal. CO
Extragal. sky survey
Extragal. source survey
Extragal. GRG
Blazar Monitoring
Planetary Radar
Spectrometer
Total Power
Polarimeter
Mono/Dual/Multifeed
3548
Maser
ISM Mol.
Active Binaries
Extragal. CO
Extragal. survey
Extragal. source survey
VLBI
Protostar
Blazar Monit.
Sunyaev-Zel'dovich
Comets
Spectrometer
Total Power
Polarimeter
MK5
Mono/Dual/Multifeed
7090
DCO + , N 2 D +
ISM Mol.
Extragal. CO
Extragal. source survey
VLBI
Protostar
Blazar Monitoring
Comets
Spectrometer
Total Power
Polarimeter
MK5
Mono/Dual/Multifeed
90115
ISM Mol.
Extragal. CO
Blazar Monitoring
Comets
Spectrometer
Total Power
Polarimeter
Mono/Dual/ Multifeed
64

Mixer
Mixer
Receiver 1 Pol.+OMT
Mixer
Mixer
Pol.+OMT
Receiver N
IF 1 Left
IF 1 Right
IF N Left
IF N Right
Local Oscillator 1
Local Oscillator 2
Figure A.1: Conceptual schematics for a multi-beam front-end receiver. Each receiver has two separate
polarizations. The two polarizations can be tuned independently within the receiver band.
A.2 A spectroscopic backend for the SRT
A.2.1 Introduction
The SRT will have a frequency coverage from 300 MHz to 86 GHz (up to 115 GHz if the active surface will
allow it). The receivers will have a single beam and dual polarization, except for the 18-26.5 GHz band that
will have 7 beams with double polarization. Nevertheless, for eôciency considerations, it is advisable that
in the future all bands will be equipped with multi-beam receivers. All receivers should be able to operate
in three modes: `VLBI', continuum and spectroscopic.
The planned congurations are:
 VLBI mode: The receiver signal will be sent using a ber-optic connection to the control room. There
will be a total of 6 bers: 2 for each of the 3 foci: Primary, Gregorian and Beam Waveguide (BW)
 spectroscopic and continuum modes: The backends will be located in the vertex room, to avoid long
radio frequency interconnections. Only the integrated data, with a much lower data rate, will be sent
to the control building using a digital optical ber.
Preliminary studies of the receiver systems consider an intermediate frequency bandwidth of 2 GHz.
Separate local oscillator could be used for the two polarizations. Moreover, the central pixel of all the multi-
beam receivers will be used for VLBI operations. A conceptual schematic for a multi-beam receiver is shown
in Fig. A.1.
A.2.2 Spectroscopic backend characteristics
The main feature of the spectroscopic backend must be its exibility, both for single beam and for multi-beam
receiver observations. The operating modes can be summarized as follows:
1. Single beam
 Dual polarization
 Single polarization:
{ simultaneous observations of two bands arbitrarily positioned in the full receiver bandwidth
65

{ observation of the full 2 GHz bandwidth
2. Multi-beam
 Dual polarization
 Single polarization: simultaneous observations of two bands, arbitrarily positioned in the receiver
instantaneous bandwidth
The minimum velocity resolution should be around 0.03 km/s in any mode.
A possible implementation for the spectrometer consists of 32 channels of 512 spectral points each, that
can be combined together as 16 independent channels of 1024 points each (mode A: multi-beam), or as a
single channel of up to 16000 points (mode B: single beam). These are described in greater detail in the
following section, and in Figs. A.2, and A.3, respectively.
In Table A.5 the required frequency resolution ô, in kHz, is reported as a function of the observing
frequency and velocity resolution ôv (km/s). For mode A (1024 channels) the corresponding bandwidths
can be obtained reading the same values as expressed in MHz. We assume that the sampling frequency
will be around 160 MHz, and therefore the maximum instantaneous bandwidth will be 80 MHz. The
combinations in italics in Table A.5 denote bandwidths larger than this value, that cannot be obtained with
this implementation.
Observing Velocity resolution ôv (km/s)
frequency 0.03 0.1 0.3 1 3
1 GHz 0.1 0.33 1 3.3 10
5 GHz 0.5 1.66 5 16 50
10 GHz 1 3.3 10 33 100
20 GHz 2 6.7 20 67 200
45 GHz 4.5 14 45 140 450
90 GHz 9 30 90 300 900
Table A.5: Frequency resolution, in kHz, as a function of the receiver frequency and of the desired velocity
resolution ôv, in km/s. Values in italics denote a bandwidth too large for the assumed 160-MHz sampler.
It should be noted that using a 160-MHz correlator we can cover for dual polarization only a fraction
of the total 2-GHz instantaneous receiver bandwidth in mode B. It is possible to observe the full 2-GHz
instantaneous bandwidth only in single polarization and single-beam mode. To be able to cover a 2-GHz
bandwidth with 16 correlators per polarization, a sampling frequency of 250 MHz is required. This would
allow one to use a 250-MHz sampler for mode A, thus permitting greater velocity coverage for the higher fre-
quency receivers. However currently available Field Programmable Gate Arrays do not guarantee operations
at frequencies above 160 MHz, and thus this lower operating frequency has been assumed.
The possibility of observing two independent frequency bands arbitrarily placed within the full receiver
bandwidth requires the use of two independent local oscillators, as shown in Fig. A.1.
A.2.3 Implementation
A possible structure of the SRT correlator would include 32 sections with 512 spectral points each. Dedicated
software and a Crossbar Switch will allow to congure the correlator as follows:
 16 inputs with 80 MHz bandwidth each
 2 inputs with 1 GHz bandwidth each
 1 input of 2 GHz bandwidth
In multi-beam operation (Mode A, Fig. A.2), the IF signals from the multi-beam receiver, after down-
conversion and anti-aliasing ltering are sampled at 160 MHz and sent to a digital SSB (Single Side Band)
66

sampler
160 MHz
sampler
160 MHz
sampler
160 MHz
2 GHz Digital
receiver
Switch Correlator
0
2 GHz Digital
receiver
2 GHz Digital
receiver
filter
filter
filter
1
2
3
31
30
IF 1 Left
IF 1 Right
IF 16 Right
160-240 MHz
160-240 MHz
160-240 MHz
Figure A.2: Multi-beam mode (mode A). Signals from the multi-beam receiver are converted, ltered, and
sampled at 160 MHz. The frequency region extracted (up to 80 MHz) is the same for all beams and
polarization channels. The bandwidth can be further narrowed by a digital receiver. The 32 correlator
channels can be grouped to analyze fewer of the 32 receiver outputs with increased resolution.
0
31
30
1
2
3
0
31
30
1
2
3
Switch Correlator
2 GHz
filterbank
digital
tunable
ALMA
ALMA 2GHz
sampler
IF D Left
Figure A.3: Single-beam mode (mode B). Signals from the receiver is sampled at 2 GHz, and processed by
a tunable digital lter bank derived from the ALMA design. Up to 32 independently tunable channels can
be extracted and sent to the correlator
lter for the required bandwidth selection (from 0.1 to 80 MHz). The second local oscillator is the same for
all receivers.
These signal are nally sent to the inputs of the correlator. For dual polarization modes, the 32 correla-
tor sections are independently used to analyze up to 16 beams in two polarizations. For single polarization,
signals are sent to the even inputs of the correlator, in which two adjacent sections are connected in se-
ries to obtain the required 1024 spectral points. The electronic components are at low cost and available
commercially; the realization should be awless.
For large bandwidth, single-beam operation (Mode B, Fig. A.3), the approach developed for ALMA is
envisaged. After the sampling, the signal is `divided' in 32, 62.5 MHz wide, sub bands using sophisticated
digital lters followed by re-sampling. Each sub-band can be independently tuned across the 2 GHz instan-
taneous bandwidth, both to synthesize a larger band or to analyze regions of interest. These 32 signals are
sent directly to the 32 correlator sections. Sections can be cascaded to increase spectral resolution. This
conguration will require components partially not commercially available. Also the realization could be
quite complex and expensive.
For operation with much larger instantaneous bandwidth, say 10 GHz, like those required by 90 GHz
receivers, the acousto-optic technology must be considered as demonstrated by Acousto-Optic Spectrometers
on board of SWAS (Submillimeter Wave Astronomy Satellite) and Herschel.
67

A.3 Pulsar Facilities
A.3.1 Overview of state-of-art de-dispersion techniques
A.3.1.1 Incoherent de-dispersion using analogue lter-banks
Incoherent de-dispersion techniques are usually adopted in pulsar search observations. This technique implies
the use of a proper lter-bank system, whose resolution and number of channels depend on the central
frequency adopted and the maximum dispersion measure searched. For instance, in the Parkes multi-beam
surveys, a lter-bank system with 96 3 MHz-wide channels covering a bandwidth of 288 MHz for each
polarization of each beam was used, making a total of 2496 channels. This system was developed by the IRA
in collaboration with Jodrell Bank and the ATNF (Sydney). In this case, detected signals from individual
frequency channels are added in polarization pairs, high-pass ltered with a cuto at approximately 0.2 Hz,
integrated and 1-bit digitized every 250 or 125 ms, and recorded on magnetic tape (DLT) for subsequent
analysis.
It was proven that the 3-MHz choice was successful in terms of the total number of pulsars discovered
(more than six hundred), but it also became clear that a relative lack of millisecond pulsars discovered in
the Parkes multi-beam survey was due to some extent to the relatively poor frequency resolution adopted.
A parallel experiment carried out at Parkes, namely a deep search of the Globular Cluster system, thanks to
the small number of pointings required, used the single central beam only of the same multi-beam receiver
equipped with a much higher frequency resolution de-dispersing system (2x512x0.5 MHz). The experiment
was extremely successful in the discovery of millisecond pulsars.
A.3.1.2 Incoherent de-dispersion using digital lter-banks
An analogue lter-bank does not provide the stability required to achieve the desired accuracy in high
precision timing observations. If observing at relatively high frequency, say around 3GHz, which is an ideal
frequency for achieving high timing precision (because the signal is relatively free from multi-path scattering
eects), dispersion can be easily removed using a digital lter-bank. The disadvantage of a digital lter-
bank with respect to the classical analogue spectrometers used for pulsar searches, is that it produces an
intrinsically much higher data rate, but this is not an issue for timing observations which target a limited
sample of sources. A good state-of-art example of a digital lter-bank is that under construction at Parkes
and planned as the standard backend for timing observations at the 10-cm band of the dual band 50/10cm
receiver. The lter-bank spans a total bandwidth of 1 GHz (two polarizations) and has a frequency resolution
of 1 MHz. Because the dispersion smearing scales as  3 , this is almost equivalent to a frequency resolution
of 125 kHz at 21cm, which is good enough to provide high timing precision.
A.3.1.3 Coherent de-dispersion
Precise timing observations are also carried out at relatively low frequency (300-1800 MHz) at which coherent
de-dispersion is much more eôcient. Although interstellar scattering is much more severe in this case, low
frequency observations are necessary and complementary to high frequency observations because they provide
the large frequency baseline necessary to measure with high accuracy the dispersion delay, and estimate with
high precision the pulse arrival times at innite frequency. While in incoherent de-dispersors, like in those
based on analogue lter-bank, the signals of the individual frequency channels are sampled at post-detection
level, in a coherent de-dispersor the baseband radio signal is sampled. The advantage of this technique is
that de-dispersion can be removed totally by Fourier-transforming the radio signal and applying a proper
convolution lter. The disadvantage is that it requires rather high data rates and powerful real-time data
processing power. A state-of-the-art coherent de-dispersor, called CPSR-2, has been developed as a joint
project between Swinburne University and CalTech. It allows on-line coherent de-dispersion of a 128 MHz
band, and it is based on modern DMA cards developed at Caltech and a cluster of 196 fast CPUs available
on-line.
68

A.3.2 A pulsar backend for the SRT
In general, dierent sets of lter-banks and coherent de-dispersion systems must exist in a versatile equipment
for performing pulsar observations: namely analogue lters for survey observations, digital lters for precise
timing observations at high frequency and machines for coherently de-dispersing the signal for low and
intermediate frequency timing observations.
In particular, following what is described in Sect. 4.2.3 (Pulsar Observations), a state-of-art pulsar system
at the SRT should include:
1. A set of analogic lter-banks or digital lter-banks based on FPGA polyphase lters with 2048 channels
(1024 per polarization), each of them 32-kHz wide, for a total instantaneous bandwidth of  32; 64 or
128 MHz. It will be used for pulsar search at low frequencies (300-400 MHz).
2. An analogic lter-bank with 2048 channels (1024 per polarization), each of them 500 kHz wide, for a
total instantaneous bandwidth of  500 MHz. It will be mostly devoted to pulsar search at 1.4 GHz
and higher frequencies.
3. If a multi-beam system will be available at 1.4 GHz, each beam should be equipped with a lter-bank
like the one described at point 2. This will allow to undertake very large-scale searches for pulsar at
1.4 GHz.
4. A digital lter-bank with a total instantaneous bandwidth of 1 GHz and a frequency resolution of 1
MHz to be exploited in timing observations at high frequency (> 3 GHz).
5. A coherent de-dispersion system, capable of handling up to 512 MHz of instantaneous bandwidth, in
order to perform high precision timing observation both at low (300-400 MHz) and intermediate (1.4
GHz) frequencies. Although such a system does not exist yet, it can be easily assembled by the time
the SRT will be operational. It will require to combine 8 machines like the already available CPSR-2
machine developed at CalTech and Swinburne University.
A.4 A continuum backend for total intensity and polarization for
the SRT
The polarimeter available at the Medicina VLBI radiotelescope, can be considered as a prototype of the
continuum backend for the SRT. It consists of an analogue correlator which as inputs has the two circular
intermediate frequencies (IF left & IF right, obtained from the RF signal after a down-conversion with one
or more local oscillators). It gives 4 channels as outputs which are the two total power channels (left and
right) and the 2 Stokes parameters denoting linear polarization (Q & U). The outputs are simultaneously
sampled by 4 analogue-to-digital converters working at a maximum rate of 40 Hz. By means of this feature,
it has been possible to develop ad-hoc strategies for map-making by the SPOrt team (IASF-CNR Bologna
and IRA-CNR), because the output channels are sampled at a proper rate to sample fully the sky, during
the raster scanning. Recent measurements (Carretti E., Poppi S., 2004, Diuse Polarized Maps with the
Medicina Polarimeter: Map-Making Procedure and Observing Tests, internal report, in prep.) showed the
map-making capabilities in continuum for both total intensity and polarization.
The working scheme of a polarimeter is summarized in Fig. A.4, where the Q and U outputs measure the
linear polarization, the TP1 and the TP2 are the total power detection of left and right polarization. The
advantage of a polarimeter working with the two circular polarizations from the receiver is in terms of stability
and time eôciency, as both the Q and U Stokes parameters are directly obtained (Kraus, Radioastronomy,
1986). Moreover, the use of the IFs as inputs for the polarimeter satises the requests of versatility for the
SRT backends (i.e. backends able to work with as as many SRT receivers as possible); otherwise if RF inputs
(Radio Frequency signal without any down-conversion) should be used, ad-hoc backends for each receiver
would be necessary.
Furthermore, it is worth noting that in designing the receivers particular care must be taken to minimize
the instrumental polarization; for an analysis of all the possible sources of spurious polarization see Carretti
et al. 2001 (New Astron. 6, 173) and Cortiglioni et al. 2004 (New Astron. 9, 297).
69

Figure A.4: Polarimeter scheme
The polarimeter for the SRT will work with a large instantaneous bandwidth, at least 1 GHz, in order
to give the highest sensitivity. The outputs will be sampled at a rate of 100 Hz, suôcient to have enough
points per beam during a raster scan at maximum antenna speed. A possible choice for the correlation unit
is a digital correlator, which would be possible if the two input IF are sampled at the proper rate: for a
1-GHz-wide instantaneous bandwidth the sampler will work at a rate greater than 2 Gsamples/s, according
to the Nyquist theorem. The sampled signal can be digitally ltered and the IF can be divided into narrower
bands, allowing to exclude those aected by interference.
70