arXiv:astro-ph/0503234v1 10 Mar 2005

Accepted by the Astrophysical Journal (Supplement Series)

VLBA Imaging Polarimetry of Active Galactic Nuclei – An

Automated Approach

G. B. Taylor

, C. D. Fassnacht

, L. O. Sjouwerman

, S. T. Myers

, J. S. Ulvestad

, R. C.

Walker

, E. B. Fomalont

, T. J. Pearson

, A. C. S. Readhead

, N. Gehrels

, and P. F.

Michelson

ABSTRACT

We present full polarization Very Long Baseline Array (VLBA

) observations

at 5 GHz and 15 GHz of 24 compact active galactic nuclei (AGN).

These sources

were observed as part of a pilot project to demonstrate the fe

asibility of conduct-

ing a large VLBI survey to further our understanding of the ph

ysical properties

and temporal evolution of AGN jets. The sample is drawn from t

he Cosmic Lens

All-Sky Survey (CLASS) where it overlaps with the Sloan Digi

tal Sky Survey

at declinations north of 15

◦

. There are 2100 CLASS sources brighter than 50

mJy at 8.4 GHz, of which we have chosen 24 for this pilot study.

All 24 sources

were detected and imaged at 5 GHz with a typical dynamic range

of 500:1, and

21 of 24 sources were detected and imaged at 15 GHz. Linear pol

arization was

detected in 8 sources at both 5 and 15 GHz, allowing for the cre

ation of Faraday

rotation measure (RM) images. The core RMs for the sample wer

e found to have

an average absolute value of 390

100 rad m

−

. We also present the discovery

of a new Compact Symmetric Object, J08553+5751. All data wer

e processed

automatically using pipelines created or adapted for the su

rvey.

Kavli Institute of Particle Astrophysics and Cosmology, Me

nlo Park, CA 94025

National Radio Astronomy Observatory, P.O. Box O, Socorro,

NM 87801

Department of Physics, University of California, Davis, 1 S

hields Avenue, Davis, CA 95616

Owens Valley Radio Observatory, California Institute of Te

chnology, Pasadena, CA 91125

NASA Goddard Space Flight Center, Greenbelt, MD 20771

Department of Physics, Stanford University, 382 Via Pueblo

Mall, Stanford, CA 94305

– 2 –

Subject headings:

galaxies: active – surveys – catalogs – galaxies: jets – gala

xies:

nuclei – radio continuum: galaxies

1. Introduction

Among astronomical observations, Very Long Baseline Inter

ferometry (VLBI) imaging

is unique in providing details about parsec-scale structur

es in cosmologically distant objects.

The Pearson-Readhead (PR; Pearson & Readhead 1988) survey r

evealed a wide range in

morphologies and paved the way towards motion and variabili

ty studies. Subsequent VLBI

surveys such as the Caltech–Jodrell Bank Flat spectrum surv

ey (CJF; Taylor et al. 1996),

have imaged

∼

300 sources at 5 GHz, of which more than half (177) were observ

ed in full

polarization (Pollack et al. 2003). The VLBA Calibrator Sur

vey (VCS; Beasley et al. 2002)

included

∼

2000 sources, but at the expense of limited sensitivity and (

u, v

) coverage since the

driving goals were an astrometric grid of phase-referencin

g calibrators, and not high-quality

imaging.

A new survey of

∼

1000 sources with full polarization at 5 and 15 GHz would prov

ide an

unparalleled combination of size, depth, and polarization

information. The large sample of

high-quality data sets, especially when combined with surv

eys at other wavelengths, would

provide an excellent resource for studying the physics, env

ironments, and evolution of active

glactic nuclei (AGN). The key areas of study for a large surve

y are:

•

Gamma-ray AGNs:

This new Very Long Baseline Array (VLBA) survey, combined

with the survey data from the Gamma-Ray Large Area Space Tele

scope (GLAST) mission

(Gehrels & Michelson 1999), should revolutionize our under

standing of AGN jet physics.

After launch in 2007, GLAST will perform an all-sky survey 30

times more sensitive than

that of EGRET. The number of AGN expected to be detected is

∼

3000–4500, although

variability, especially in the gamma-ray properties, may s

trongly influence these estimates

(Vercellone et al. 2004). Correlations with milliarcsecon

d polarimetry are key to identifying

and understanding the gamma-ray sources. Extrapolating fr

om the correlation between 5

GHz radio luminosity and peak gamma-ray emission (Mattox et

al. 1997), we expect that

most of these objects will be compact, flat-spectrum sources

with 5 GHz flux densities over

30 mJy. EGRET results and theoretical studies indicate that

gamma-ray emission emanates

from jet outflows (e.g., Dermer & Schlickeiser 1994), where h

igh-energy non-thermal electrons

accelerated in shocks are thought to Compton-upscatter sof

t photons. Milliarcsecond radio

imaging can reveal the detailed morphological structure of

the core and outflow regions. With

a large number of sources, correlations between radio morph

ology and gamma-ray spectral

properties will be determined with high significance, thus p

roviding constraints on models

– 3 –

for baryonic acceleration and the microphysics of electron

-proton coupling. In addition, the

improved error box of GLAST (

∼

arcminutes) combined with the VLBA imaging may solve

the mystery of the unidentified high-latitude EGRET sources

•

Jet Magnetic Fields:

Recently, Pollack et al. (2003) have found a tendency for the

parsec-scale cores of quasars to have a magnetic field aligne

d in the direction of the jet, which

can constrain jet-launching/collimation models (e.g., Me

ier, Koide, & Uchida 2001). Similar

observations of BL Lacs have suggested that their magnetic fi

elds are oriented perpendicular

to the jets (e.g., Gabuzda et al. 2000). Unfortunately, the n

umber of BL Lacs and galaxies

with good polarization data are too small for a meaningful co

mparison between different

classes of objects. The 1000-source survey will address thi

s problem by providing

≥

100

objects in each class.

•

AGN Environments:

The survey data will be used to study propagation effects loca

to the jet, including free-free absorption in sources with c

ounterjets and Faraday rotation

measures (RMs) towards all sources with detected polarizat

ion. Typical quasars have RMs

which are surprisingly large (

1000 rad m

−

; Taylor 1998) and time-variable (Zavala &

Taylor 2001). Knowing the RM is also critical in order to make

sense of the jet properties

as otherwise the RM will tend to “smear-out” intrinsic relat

ionships.

•

AGN Evolution:

Compact Symmetric Objects (CSOs), by virtue of their small

sizes (

1 kpc), are thought to be examples of young (10

−

yr) radio sources that may

evolve to become eventually the well-known FRII-type radio

sources (Readhead et al. 1996).

Only small numbers of CSOs are known (

∼

40; Peck & Taylor 2000), limiting our ability to

study radio sources at early times in their evolutionary his

tory. A large VLBA survey would

approximately double the number of confirmed CSOs.

Here we present the results from a feasibility study for a pos

sible VLBI Imaging and

Polarimetry Survey (VIPS). While not a complete sample in it

self, this pilot study of 24

AGN was chosen to be a representative selection of strong and

weak sources and presents

a wealth of data on this modest sample. This includes one new C

ompact Symmetric Ob-

ject (J08553+5751) and the first Faraday rotation measure ob

servations of a sample of faint

sources. We have also developed and made available new pipel

ines for calibration and imag-

ing that allow rapid reduction of the VLBA data to finished pro

ducts. Throughout this

discussion, we assume H

=70 km s

−

Mpc

−

, Ω

= 0.27, and Ω

= 0.73.

– 4 –

2. Sample Definition

To facilitate multi-wavelength science, all targets lie wi

thin the

sr region of the North

Galactic Cap and the equatorial strips to be covered by the Sl

oan Digital Sky Survey (SDSS;

Abazajian et al. 2004, 2005). Thus multi-color imaging (UV t

o near-IR) will be available

for almost all targets, and spectroscopic information will

be present for the optically bright

targets (perhaps 33%). For optically faint sources, photom

etric redshifts may be estimated

from the

ugriz

imaging data (Weinstein et al. 2004). In this region there ar

e 2100 flat-

spectrum, compact (size

200 mas), and relatively bright (

GHz

50 mJy) sources above

◦

declination in Cosmic Lens All-Sky Survey (CLASS) (Myers et

al. 2003), a VLA survey

∼

12,000 flat-spectrum radio sources at 200 mas resolution. Fr

om this number we have

selected a sample of 1000 sources never before imaged at high

dynanmic range with VLBI.

The declination limit of 15

◦

is necessary to insure uniformly good radio imaging quality

. The

lower limit of 50 mJy has been chosen in order to have sufficient

SNR (

1.5 estimated at 15

GHz) on all compact sources with a spectral index

−

5 to permit self-calibration within

the coherence time at both 5 and 15 GHz, where the spectral ind

ex,

, is defined as

∝

Heavily resolved sources (sizes

20 mas) may still be undetectable, especially at 15 GHz

where 25 mJy is needed on scales

20 mas for self-calibration. We note that there is an

implicit avoidance of the galactic plane in our selection si

nce CLASS was selected to have

◦

. CLASS also had a spectral selection of

α >

−

5 from the parent Green Bank

Survey (GB6 – Gregory et al. 1996) at 4.85 GHz and the NRAO VLA S

ky Survey (NVSS

– Condon et al. 1998) at 1.4 GHz. A similar spectral selection

was employed for the CJF

survey to obtain a high success rate with VLBI imaging.

We note that there should be

∼

1000 GLAST sources found in the area covered by

VIPS. Since this is comparable to the number of VIPS sample me

mbers in the same region,

and compact flat-spectrum radio emission is believed to be a s

trong predictor of gamma-ray

emission (Mattox et al. 2001; Sowards-Emmerd et al. 2003), a

milliarcsecond radio imaging

survey of the VIPS sample may be important for the scientific i

nterpretation of the GLAST

all-sky survey (as discussed in

1).

Here we present observations of 24 sources (see Table 1) sele

cted within the first and

second data release areas of SDSS, and to be representative o

f the VLBA VIPS sample as a

whole. The integrated flux densities of the 8.4 GHz VLA parent

sample range from 52 mJy

for J08585+5552 to 850 mJy for J08546+5757 (JVAS 0850+581).

A total of 8, 7, 5 and 4

sources were selected from the ranges 50–100, 100–200, 200–

400, and

400 mJy.

– 5 –

3. Observations and Analysis

The observations were carried out at 5 and 15 GHz on 14 March 20

04, 15 March 2004,

28 June 2004, and 18 August 2004, using the VLBA of the NRAO

. Each observing session

lasted for 12 hours. The 5 and 15 GHz frequency bands were inte

rleaved in time, with 24

minutes scheduled on each target source at 5 GHz and 72 minute

s on source at 15 GHz.

Allowing for telescope slews and other overhead (6 minutes o

n average) the time needed

for each source was 1.7 hours. In order to improve (

) coverage and to allow for rotation

measure determinations, we spread the frequencies out over

the 500 MHz instantaneous

tuning range of the VLBA as follows: 5 GHz refers to 4607, 4677

, 4992 and 5097 MHz at

band center, while 15 GHz refers to 14904, 14970, 15267, and 1

5366 MHz at band center.

Each frequency was observed with 4 MHz bandwidth in each of ri

ght circular and left circular

polarization. The only significant loss of data were the loss

of 98 minutes due to a false fire

alarm at the Hancock, NH VLBA station on 18 August. Rain at som

e sites increased system

temperatures and reduced sensitivity, especially at 15 GHz

, in all four runs.

Amplitude calibration was derived using measurements of th

e system temperatures and

antenna gains at 4992 and 15366 MHz, which are close to the con

tinuum default frequencies

of 4999 and 15369 MHz. Fringe-fitting was performed with the A

IPS task FRING on the

strong calibrator 3C 279. Feed polarizations of the antenna

s were determined at 5 and 15

GHz using the AIPS task LPCAL and the unpolarized source OQ20

8. An initial phase self-

calibration was performed for each source using a point sour

ce model – no phase referencing

to nearby calibrators was used, as this would significantly r

educe the observing efficiency

and is unnecessary for such strong sources.

Absolute electric vector position angle (EVPA) calibratio

n was determined by using

the EVPA’s of 3C 279, J0854+2006, and J1310+3220 as listed in

the VLA Monitoring

Program

(Taylor & Myers 2000). Agreement in the EVPA correction dete

rmined from

these calibrators was generally better than 3 degrees for ob

servations within a few days. All

observations of EVPAs for the target sources have been corre

cted for Faraday rotation as

determined from the observations.

For each source, the 15 GHz data were tapered to produce an ima

ge at comparable

resolution to the full resolution 5 GHz image. For simplicit

y all sources were restored with

a fixed beam size of 3.2

1.6 mas in position angle 0

◦

at each frequency. The two images

The National Radio Astronomy Observatory is a facility of th

e National Science Foundation operated

under cooperative agreement by Associated Universities, I

nc.

http://www.vla.nrao.edu/astro/calib/polar/

– 6 –

were then combined to generate a spectral index map. It is imp

ortant to note that spectral

index maps made from two datasets with substantially differe

nt (

u, v

) coverages may suffer

from significant systematic errors, especially in regions o

f extended emission. Polarization

images were likewise made at matching resolutions in order t

o make rotation measure (RM)

images. All calibration and imaging was done semi-automati

cally using pipelines written

in AIPS (Greisen 2003) and Difmap (Shepherd 1997). These pip

elines can be found at the

VIPS web page

For 23 out of 24 sources the automatic imaging scripts were ab

le to produce an image

with the expected noise and a good fit (

sigma

) between the source model and the data.

The parameter

sigma

is the square root of the squared difference between the data a

the model divided by the individual variances implied by the

visibility weights. This is

close to the square root of the reduced

, except that a small change in the number of

degrees of freedom due to amplitude self-calibration has no

t been taken into account. This

simplification is generally benign and will only result in a

∼

10% reduction of

sigma

for

sources strong enough to amplitude self-calibrate (peak flu

x density brighter than 0.3 Jy at

5 GHz or 0.5 Jy at 15 GHz). The value of

sigma

indicates the agreement obtained between

the model and data in the self-calibration process with valu

es near unity indicating good

agreement. In one case a poor fit (

sigma >

4) was found to be due to bad data during 5

minutes from the Los Alamos, NM station. It is worth noting th

at in our inital automatic

imaging script we used a field size of 64

64 milliarcseconds, and this resulted in poor fits

for two sources with emission beyond the edge of the field. Jud

icial manual editing was

done to solve the first problem and imaging all sources with an

initial field size of 128

128

milliarcsec then provided good fits for all 24 sources. The fin

al images have rms noise values

within 30% of the predicted thermal noise.

The observing efficiency of 1.7 hours/source could be improve

d to 1.5 hours/source

without loss of sensitivity if the observing runs were exten

ded to 24 hours. This could be

done since the number of calibrator scans is the same for a 24 h

our run as it is for a 12 hour

run. The fraction of time spent on calibrators for each of the

12 hour observing runs in the

pilot project was 15%.

4. Results

Figure 1 displays total intensity images for all sources in t

he pilot project at 5 and 15

GHz. Contours are drawn starting at 0.8 mJy beam

−

. Further details can be found in

http://www.aoc.nrao.edu/

∼

gtaylor/VIPS/

– 7 –

Tables 2 and 3. All 24 sources were detected and imaged at 5 GHz

. At 15 GHz there was no

detection for J08553+5751 = JVAS 0851+580 (see

4.2.1), J08585+5552, or J15406+5803.

Since no observations were carried out with phase-referenc

ing, sources with a flux density

less than 25 mJy at 15 GHz were too weak to self-calibrate. The

lack of detections is due

to a combination of moderately steep spectrum sources and th

e lower sensitivity at 15 GHz

within the coherence time used for self-calibration of the p

hases. The sensitivity limit for

detection was a flux density of 10 mJy at 5 GHz and 25 mJy at 15 GHz

Linear polarization was detected in 8 of 24 sources at both 5 a

nd 15 GHz, and at 15 GHz

only in the source J16542+3950. In order to increase the SNR i

n the polarization images the

observations at 4607 and 4677 MHz were combined, as were thos

e at 4992 and 5007 MHz,

14904 and 14970 MHz, and 15267 and 15366 MHz. These 4 pairs wer

e subsequently combined

to generate images of the rotation measure by calculating th

e change in polarization angle

with wavelength squared on a pixel by pixel basis. The result

s are shown in Fig. 2, along with

the magnetic field polarization vectors corrected for Farad

ay rotation. Under the assumption

that the radiation is optically thin synchrotron emission i

n a homogeneous field, the vectors

shown indicate the projected source magentic field orientat

ion.

4.1. SDSS Magnitudes and Redshifts

We have searched for optical counterparts for the sources in

the SDSS Data Release 3

(Abazajian et al. 2005), with a counterpart defined as a catal

og member falling within 1

′′

the radio position. There are optical matches for 22 of the 24

VIPS sources. The detected

sources have

-band magnitudes between 17.5 and 22.1 (Table 1). The SDSS pi

peline pro-

duces a morphological classification of “STAR” for all but th

ree of the counterparts. These

sources are almost certainly quasars, and we have designate

d them as “Q” in Table 1. Fur-

thermore, SDSS redshifts are available for eight of the sour

ces, all of which have “STAR”

morphological classifications. The spectra for these sourc

es are all typical quasar spectra,

with broad emission lines of, e.g., CIV, CIII, MgII, or H

. The redshifts range between

= 0

5 and

= 2

4.2. Notes on Individual Sources

4.2.1.

J08553+5751

The morphology of this source revealed by our 5 GHz image (Fig

. 1) resembles that of

a Compact Symmetric Object (CSO). The identification of 0851

+580 with a galaxy, and the

– 8 –

lack of linear polarization, is also consistent with a CSO cl

assification, as is the lack of a

detection at 15 GHz.

4.2.2.

J14142+4554

This known CSO (JVAS 1412+461) (Peck & Taylor 2000) was inclu

ded in the sample

as a check on the ability of the survey to identify CSOs. In thi

s regard the clear detection

of two steep spectrum lobes is quite satisfactory. This sour

ce has been associated with a

galaxy of magnitude 19.9 and a redshift of 0.190 by Falco, Koc

hanek, & Mu ̃noz (1998). It

has a bent northern lobe and an edge-brightened southern lob

e. Gugliucci et al. (2005) find

no detectable motions of the lobes with an upper limit of 0.01

4 mas yr

−

, or 0.14 c. This

gives a lower limit on the age of the radio source of 2030 yr.

5. Discussion

5.1. Source Morphologies

Of the 24 sources observed we have manually classified them as

6 unresolved naked

cores, 10 short jets (less than 10 mas), 4 long jets (more than

10 mas), 2 very bent jets, and

2 CSOs (see Table 2). In Fig. 3 we plot the core fraction, define

d as the ratio of the peak

to the integrated intensity, versus the spectral index of th

e core component between 5 and

15 GHz. As expected, the naked core and short jet sources have

the flattest spectra (with

one exception of a long jet source). Such sources are likely o

riented at angles close to the

line-of-sight and are prime candidates for gamma-ray emiss

ion.

The two sources with highly bent jets have both steep spectra

and a low core fraction,

possibly indicating that they are viewed at moderately larg

e angles to the line-of-sight. In this

case their sharp bends are unlikely to be due to projection eff

ects, and must be intrinsically

large. For the one CSO detected at both 5 and 15 GHz, the core fr

action is overestimated

since the peak in the image corresponds to a hot spot, and not t

o the core.

5.2. Evolution of AGN

Compact Symmetric Objects (CSOs), by virtue of their small s

izes (

1 kpc), are thought

to be examples of young (10

−

yr) radio sources that may evolve to eventually become

the well-known FRII-type radio sources (Readhead et al. 199

6). Only small numbers of

– 9 –

CSOs are known (

∼

40; Peck & Taylor 2000), limiting our ability to study radio s

ources at

early times in their evolutionary history.

We present the discovery of a probable new CSO, J08553+5751.

Although this pilot

sample is far from complete, and with only 23 sources the stat

istics are unreliable, the

detection rate of 1/23 (not counting the CSO J14142+4554 sin

ce it was deliberately selected)

or 4% is interesting. Taylor & Peck (2003) found a detection r

ate of 2% for sources selected

from the southern part of the VCS. This is similar to that obta

ined for the northern part

of the VCS (Peck & Taylor 2000). As pointed out by Peck & Taylor

(2000), this detection

rate is much lower than that found in the PR sample of 7/65 (11%

), or 18/411 (4.4%) found

in the combined PR and CJ samples. We expect a slightly lower d

etection rate because

the VCS is comprised predominantly of flat spectrum sources.

Nonetheless, the significantly

lower detection rate found in the VCS is probably the result o

f the reduced sensitivity and

(

u, v

) coverage compared to the PR and CJ surveys. If the VIPS CSO de

tection rate is well

above 2% then that indicates that the drop previously seen is

not the result of looking at

the fainter end of the luminosity function. With a 4% detecti

on rate the VIPS survey would

add 40 CSOs to those known, roughly doubling the number of obj

ects. Assuming Poisson

statistics apply and we get 40 CSOs, then we should be able to d

etermine the CSO fraction

to an accuracy of

∼

0.6%.

5.3. Magnetic Fields in the AGN Environment

Typical quasars have rotation measures which are large (

1000 rad m

−

; Taylor 1998,

Zavala & Taylor 2003, 2004) and time-variable (Zavala & Tayl

or 2001). In the VIPS pilot

sample we find core RMs ranging from

−

45 rad m

−

in J16484+4104 to

−

931

20 rad

−

in J15457+5400 (Table 4). The average absolute value for the

8 sources shown in Fig. 2

is 390

100 rad m

−

. This value is somewhat less than the value of 640 rad m

−

found

by Zavala & Taylor (2004) for a sample of 40 strong AGN. Presum

ably the Faraday screen

orginates in close proximity to the AGN jet, possibly create

d by an interaction between the

jet and the ambient gas. Note that intrinsic RMs are larger by

a factor of (1+z)

. The

smaller RMs in the present sample could be the result of highe

r source redshifts, thinner

Faraday screens, or small number statistics.

Knowing the core RM value is also critical in order to make sen

se of the jet properties

as otherwise the RM will tend to “smear-out” intrinsic relat

ionships. With only 8 sources

we do not presume to look for any trends in the magnetic field or

ientation.

We note that the jets in 5 sources with extended polarization

detections all have low

– 10 –

RMs (in the range

50 rad m

−

), consistent with results from Zavala & Taylor (2004).

6. Conclusions

We demonstrate that reliable, high dynamic range, total int

ensity VLBA images can be

obtained at 5 GHz for 100% (24 of 24) sources with an integrate

d flux density of over 50

mJy at 8.4 GHz, using a survey mode with 1.7 hours total integr

ation time per source. For

21 of 24 sources (88%) high dynamic range images were also obt

ained at 15 GHz, allowing

for the production of spectral index images. We find a third of

the target sources can be

imaged in linear polarization as well, allowing us to determ

ine Faraday rotation measures.

A pilot project of 48 hours of observations yields a wealth of

information for 24 AGN. From

this sample we report on the discovery of a new CSO, JVAS J0855

3+5751.

Future observations of the complete VIPS sample of 1000 sour

ces will provide an un-

paralleled combination of size, depth, and polarization in

formation. The large sample of

high-quality data sets, especially when combined with surv

eys at other wavelengths, will

provide an excellent resource for studying the nature of Gam

ma-ray loud AGN, the jet

magnetic fields, the environs of the AGN, and AGN evolution.

This research has made use of the NASA/IPAC Extragalactic Da

tabase (NED) which

is operated by the Jet Propulsion Laboratory, Caltech, unde

r contract with NASA.

– 11 –

REFERENCES

Abazajian, K., et al. 2004, AJ, 128, 502

Abazajian, K., et al. 2005, AJ, submitted, astro-ph/041023

Beasley, A. J., Gordon, D., Peck, A. B., Petrov, L., MacMilla

n, D. S., Fomalont, E. B., &

Ma, C. 2002, ApJS, 141, 13

Condon, J. J., Cotton, W. D., Greisen, E. W., Yin, Q. F., Perle

y, R. A., Taylor, G. B., &

Broderick, J. J. 1998, AJ, 115, 1693

Dermer, C.D. & Schlickeiser, R. 1994, ApJS, 90 945

Falco, E. E., Kochanek, C. S., & Mu ̃noz, J. A. 1998, ApJ, 494, 4

Fanaroff, B. L. & Riley, J. M. 1974, MNRAS, 167, 31P

Gabuzda, D.C. et al. 2000, MNRAS, 319, 1109

Gehrels, N. & Michelson, P. 1999, Astroparticle Physics, 11

, 277

Gregory, P. C., Scott, W. K., Douglas, K., & Condon, J. J. 1996

, ApJS, 103, 427

Greisen, E. W. 2003, in Information Handling in Astronomy - H

istorical Vistas, ed. A. Heck,

Astrophysics and Space Science Library Vol. 285 (Dordrecht

: Kluwer), 109

Gugliucci, N. E., Taylor, G. B., Peck, A. B., & Giroletti, M. 2

005, ApJ, in press

Ma, C., et al. 1998, AJ, 116, 516

Mattox, J. R., Schachter, J., Molnar, L., Hartman, R. C., & Pa

tnaik, A. R. 1997, ApJ, 481,

Mattox, J. R., Hartman, R. C., & Reimer, O. 2001, ApJS, 135, 15

Meier, D.L., Koide, S., & Uchida, Y. 2001, Science, 291, 84

Myers, S. T., et al. 2003, MNRAS 341, 1

Pearson T. J., & Readhead A. C. S. 1988, ApJ 328, 114

Peck, A. B. & Taylor, G. B. 2000, ApJ, 534, 90

Pollack L.K., Taylor G.B., & Zavala R.T. 2003, ApJ 589, 733

– 12 –

Readhead, A. C. S., Taylor, G. B., Pearson, T. J., & Wilkinson

, P. N. 1996, ApJ, 460, 634

Shepherd, M. C. 1997, ASP Conf. Ser. 125: Astronomical Data A

nalysis Software and Sys-

tems VI, 6, 77

Sowards-Emmerd, D., Romani, R. W., & Michelson, P. F. 2003, A

pJ, 590, 109

Taylor, G. B., Vermeulen, R. C., Readhead, A. C. S., Pearson,

T. J., Henstock, D. R., &

Wilkinson, P. N. 1996, ApJS, 107, 37

Taylor, G. B. 1998, ApJ, 506, 637

Taylor, G. B. & Myers, S. T. 2000 VLBA Scientific Memo 26, Natio

nal Radio Astronomy

Observatory

Taylor, G. B. & Peck, A. B. 2003, ApJ, 597, 157

Vercellone, S., Soldi, S., Chen, A. W., & Tavani, M. 2004, MNR

AS, 353, 890

Weinstein, M. A., et al. 2004, ApJS, 155, 243

Zavala, R. T. & Taylor, G. B. 2001, ApJ, 550, L147

Zavala, R. T. & Taylor, G. B. 2003, ApJ, 589, 126

Zavala, R. T. & Taylor, G. B. 2004, ApJ, in press, astro-ph/04

05534

This preprint was prepared with the AAS L

X macros v5.2.

– 13 –

Fig. 1.— Total intensity contours of the VIPS pilot sources a

t 5 GHz (left), and 15 GHz

(middle), along with spectral index (right). The 15 GHz imag

e and spectral index image

cover the inner quarter of the 5 GHz field. Contours start at 0.

8 mJy beam

−

and increase by

factors of 2. Where redshifts are available, a 50 pc scale is i

ndicated. The synthesized beam

is shown in the bottom-left corner. Image parameters are giv

en in Table 2. The spectral

index image is overlaid with 15 GHz contours at a fixed resolut

ion.

– 14 –

No Detection at 15 GHz

Not Available

No Detection at 15 GHz

Not Available

Fig. 1.— Continued.

– 15 –

Fig. 1.— Continued.

– 16 –

Fig. 1.— Continued.

– 17 –

Fig. 1.— Continued.

– 18 –

Fig. 1.— Continued.

– 19 –

Fig. 2.— Polarization magnetic field (B) vectors at 5 GHz, alo

ng with images of the rotation

measure computed from polarization angles measured at four

frequencies. The polarization

angles have been corrected by the observed RM, and the vector

lengths are proportional to

the fractional linear polarization. Contours are at 5 GHz an

d drawn at the same levels as in

Figure 1.

– 20 –

Fig. 2.— Continued.

– 21 –

Fig. 3.— A scatter plot of the core fraction (peak intensity d

ivided by total intensity) versus

spectral index computed between 5 and 15 GHz for 21 sources. S

ources have been classified

as either naked cores, short jets (length

10 mas), long jets (length

10 mas), highly bent

jets, or Compact Symmetric Objects (see Table 2).

– 22 –

Table 1. VIPS Pilot Sample

Source

Alternate

8 GHz

Name

Dec

date

(1)

(2)

(3)

(4)

(5) (6) (7)

(8) (9)

J08474+5723 0843+575 08 47 28.0579

57 23 38.349

240 ... ...

J08490+5603

08 49 00.8546

56 03 50.122

Q 21.267 ...

J08499+5108

08 49 57.9836

51 08 28.997

325 Q 18.194 0.584 A

J08507+5159

08 50 42.2482

51 59 11.674

112 Q 18.961 1.892 A

J08546+5757 0850+581 08 54 41.9973

57 57 29.927

890 Q 17.739 1

.318 C

J08553+5751 0851+580 08 55 21.3558

57 51 44.091

160 G 19.843 .

J08575+5827

08 57 31.4511

58 27 22.538

Q 22.082 ...

J08585+5552

08 58 30.4786

55 52 41.302

G 18.473 ...

J09030+4651 0859+470 09 03 03.9901030 46 51 04.13753 963 Q 18

.989 1.47 B

J09470+5907

09 47 04.8631

59 07 41.467

... ...

...

J09493+6039

09 49 20.2283

60 39 22.871

325 Q 19.917 ...

J09496+5819

09 49 39.8149

58 19 12.933

112 Q 21.308 ...

J14142+4554 1412+461 14 14 14.8535061 45 54 48.65427

G 20.

158 ...

J15406+5803

15 40 37.5779

58 03 34.398

Q 18.549 1.25 A

J15450+5135 1543+517 15 45 02.8222

51 35 00.874

635 Q 17.552 1

.93 A

J15457+5400

15 45 43.8287

54 00 42.764

145 Q 19.827 ...

J15465+5146

15 46 33.6208

51 46 45.455

Q 19.943 ...

J15521+5552

15 52 10.8942

55 52 43.211

131 Q 20.688 ...

J16445+3916 1642+393 16 44 34.4775

39 16 04.915

635 Q 19.808 1

.583 D

J16469+4059 1645+410 16 46 56.8603

40 59 17.167

Q 19.252 ..

J16484+4104 1646+411 16 48 29.2622

41 04 05.558

210 Q 18.741 0

.852 C

J16525+4013

16 52 33.2136

40 13 58.339

100 Q 20.465 ...

J16529+3902 1651+391 16 52 58.5096

39 02 49.807

330 Q 21.341 .

J16542+3950

16 54 12.7223

39 50 05.681

145 Q 19.948 ...

Notes - (1) J2000 source name in the IAU format HHMMd+DDMM; (2

) Alternate name; (3) Right

ascension and (4) Declination in J2000 coordinates from CLA

SS (Myers et al. 2003), except for 1412+461

and 1543+517 taken from the International Coordinate Refer

ence Frame (ICRF; Ma et al. 1998); (5)

Flux density at 8.4 GHz from CLASS; (6) Optical host galaxy id

entification; (7) Optical magnitude in r

band; (8) spectroscopic redshift from SDSS; (9) Date of obse

rvation, A = 14 March 2004, B = 15 March

2004, C = 28 June 2004, and D = 18 August 2004.

– 23 –

Table 2. 5 GHz VIPS Image Parameters

Peak Flux

rms

Source

Beam

Total Flux

(mJy

Fit Morph- Core

Name

(mas)

(

◦

)

(mJy)

beam

−

) beam

−

)

sigma

ology

∗

Fraction

J08474+5723 2.99

1.53

−

0.8

219.5

110.0

0.21 0.998

0.50

J08490+5603 3.07

1.50

−

4.0

72.6

50.2

0.23 1.014

0.69

J08499+5108 3.03

1.51

−

3.5

191.8

167.5

0.19 1.009

0.87

J08507+5159 2.97

1.50

−

3.8

103.5

102.3

0.24 1.064

0.99

J08546+5757 2.98

1.54

−

0.6

875.3

645.8

0.23 0.979

0.74

J08553+5751 3.11

1.61

−

0.7

44.4

26.2

0.21 1.015

CSO

0.59

J08575+5827 2.99

1.53

−

1.1

83.3

66.7

0.19 1.010

0.80

J08585+5552 3.14

1.63

−

2.7

30.1

17.1

0.17 0.995

0.57

J09030+4651 3.11

1.54

1.1

746.6

503.4

0.20 0.952

0.67

J09470+5907 3.96

2.67

−

6.1

208.2

181.7

0.29 0.992

0.87

J09493+6039 3.97

2.65

−

8.8

58.5

57.4

0.23 0.991

0.98

J09496+5819 3.95

2.65

−

10.5

37.3

0.21 0.968

1.00

J14142+4554 3.36

1.64

−

8.9

171.3

43.6

0.23 1.061

CSO

0.25

J15406+5803 3.01

1.54

1.4

35.2

28.7

0.18 0.990

0.82

J15450+5135 3.11

1.56

3.1

318.7

211.4

0.27 1.102

0.66

J15457+5400 3.09

1.52

0.5

192.5

192.4

0.21 1.036

1.00

J15465+5146 3.23

1.59

4.9

78.2

69.2

0.23 1.026

0.88

J15521+5552 3.17

1.56

5.6

59.9

39.4

0.20 1.000

0.66

J16445+3916 3.47

1.92 21.6

64.2

46.7

0.23 0.983

0.73

J16469+4059 3.44

1.93 20.4

247.5

209.3

0.21 0.985

0.85

J16484+4104 3.20

1.55 11.5

307.7

291.5

0.25 1.086

0.95

J16525+4013 3.20

1.53

9.9

76.6

59.2

0.21 0.992

0.77

J16529+3902 3.26

1.54 10.7

345.0

334.5

0.19 0.912

0.97

J16542+3950 3.48

1.93 20.6

91.4

88.5

0.24 1.017

0.97

∗

Source morphology is either ’nc’: naked core; ’sj’: short-j

et; ’lj’: long-jet; ’bj’: bent jet; or CSO:

Compact Symmetric Object.

– 24 –

Table 3. 15 GHz VIPS Image Parameters

Peak Flux

rms

Spectral

Source

Beam

Total Flux

(mJy

Fit

Index

Core

Name

(mas)

(

◦

)

(mJy)

beam

−

) beam

−

)

sigma

Fraction

J08474+5723 0.98

0.52

−

2.2

66.9

24.6

0.19 0.972

−

1.32

0.37

J08490+5603 1.00

0.53

−

1.9

44.6

33.6

0.18 0.979

−

0.35

0.75

J08499+5108 0.98

0.51

−

2.8

198.2

159.0

0.19 0.981

−

0.05

0.80

J08507+5159 0.98

0.52

−

4.3

54.3

49.2

0.21 0.996

−

0.64

0.91

J08546+5757 0.97

0.53

−

1.0

495.5

266.9

0.21 1.014

−

0.78

0.54

J08553+5751

J08575+5827 1.06

0.52 1.7

52.5

28.4

0.16 0.972

−

0.75

0.54

J08585+5552

J09030+4651 1.10

0.52 2.4

748.6

616.1

0.18 0.922

0.18

0.82

J09470+5907 0.98

0.53 7.4

133.2

59.9

0.31 1.019

−

0.97

0.45

J09493+6039 0.94

0.52 9.7

53.9

40.9

0.22 0.975

−

0.30

0.76

J09496+5819 1.00

0.53 1.8

45.0

0.23 0.975

0.16

1.00

J14142+4554 1.21

0.60 0.9

34.6

21.6

0.24 0.950

−

0.62

J15406+5803

J15450+5135 1.04

0.54 1.4

116.2

56.9

0.23 1.024

−

1.15

0.49

J15457+5400 1.02

0.53 0.3

156.6

155.9

0.18 0.993

−

0.18

1.00

J15465+5146 1.18

0.59 16.0

33.1

33.0

0.25 0.995

−

0.65

1.00

J15521+5552 1.16

0.59 17.0

42.6

36.4

0.27 1.003

−

0.07

0.80

J16445+3916 1.05

0.52 8.5

65.0

61.4

0.25 0.999

0.24

0.94

J16469+4059 1.05

0.52 8.4

265.1

179.5

0.17 0.983

−

0.13

0.68

J16484+4104 1.04

0.51 12.3

452.1

426.9

0.23 0.994

0.33

0.94

J16525+4013 1.04

0.51 12.9

63.7

50.7

0.23 0.991

−

0.14

0.80

J16529+3902 1.06

0.51 10.8

332.4

323.7

0.22 0.980

−

0.03

0.97

J16542+3950 1.05

0.52 7.4

77.7

57.2

0.21 0.987

−

0.38

0.74