Wow! signal

[Jennie McGrath]

The WOW! Signal
Credit: The Ohio State University Radio Observatory and the North American AstroPhysical Observatory (NAAPO).

[Jennie McGrath]

The Wow! signal was a strong narrowband radio signal detected by Dr. Jerry R. Ehman on August 15, 1977 while working on a SETI project at the Big Ear radio telescope of Ohio State University. The signal bore expected hallmarks of potential non-terrestrial and non-solar system origin. It lasted for 72 seconds, the full duration Big Ear observed it, but has not been detected again. It has been the focus of attention in the mainstream media when talking about SETI results.

Amazed at how nearly the signal matched the expected signature of an interstellar signal in the antenna used, Ehman circled the signal on the computer printout and wrote the comment "Wow!" on its side. This comment became the name of the signal.

[Jennie McGrath]

Technical details

The circled letter code 6EQUJ5 describes the intensity variation of the signal. A space denotes an intensity between 0 and 0.999.., the numbers 1-9 denote the correspondingly numbered intensities (from 1.000 to 9.999...), and intensities of 10.0 and above are denoted by a letter ('A' corresponds to intensities between 10.0 and 10.999..., 'B' to 11.0 to 11.999..., etc). The value 'U' (an intensity between 30.0 and 30.999...) was the highest ever detected by the telescope. The intensity in this case is the unitless signal-to-noise ratio, where noise was averaged for that band over the previous few minutes.

The bandwidth of the signal is less than 10 kHz (each column on the printout corresponds to a 10 kHz-wide channel; the signal is only present in one column). Two different values for its frequency have been given, 1420.356 MHz (J. D. Kraus) and 1420.456 MHz (J. R. Ehman), but both very close to the frequency of the hydrogen line at 1420.405 MHz. Two possible equatorial coordinates are given:

R.A. = 19h22m22s ± 5s
R.A. = 19h25m12s ± 5s
Both coordinates give dec. = -27°03´ ± 20´ (epoch B1950.0).

The Big Ear telescope was fixed and used the rotation of the Earth to scan the sky. At the speed of the earth's rotation, and given the width of the Big Ear's observation "window", the Big Ear could observe any given point for just 72 seconds. An extraterrestrial signal, therefore, would be expected to register for exactly 72 seconds, and the recorded intensity of that signal would show a gradual peaking for the first 36 seconds -- until the signal reached the center of Big Ear's observation "window" -- at which time it would show a gradual decrease.

Therefore, both the length of the Wow! signal, 72 seconds, and its shape would correspond to an extraterrestrial origin

[Jennie McGrath]

Searches for recurrence of the signal

The Big Ear telescope used two beams to search for signals; the Wow signal was detected in one of these beams but not the other. It should also have appeared a mere 3 minutes later (or earlier); it didn't. Ehman unsuccessfully looked for recurrences of the signal using Big Ear in the month after the detection.

In 1987 and 1989, Robert Gray searched for the event using the META array at Oak Ridge Observatory, but did not re-detect it.

Gray also searched for the signal using the Very Large Array, which is significantly more powerful than Big Ear, in 1995 and 1996.

Gray later searched for recurrences of the event in 1999 using the University of Tasmania's Hobart 26m radio telescope. Six 14-hour observations were made at positions in the vicinity, but did not detect anything similar to the Wow signal

[Jennie McGrath]

Speculations on the origin

It has been speculated that interstellar scintillation of a weaker continuous signal — similar, in effect, to atmospheric twinkling—could be a possible explanation, although this still would not exclude the possibility of the signal being artificial in its nature. However, even by using the significantly more sensitive Very Large Array, such a signal could not be detected, and the probability that a signal below the Very Large Array level could be detected by the Big Ear radio telescope due to interstellar scintillation is low. Other speculations include a rotating lighthouse-like source or a signal sweeping in frequency.

Ehman has stated his doubts that the signal is of intelligent extraterrestrial origin: "We should have seen it again when we looked for it 50 times. Something suggests it was an Earth-sourced signal that simply got reflected off a piece of space debris."

He later recanted his skepticism somewhat after further research scientifically relegated an Earth-bound signal to be astronomically unlikely, due to the requirements of a space-borne reflector being bound to certain unrealistic requirements to sufficiently explain the nature of the signal. Also, the 1420 MHz signal is problematic in itself in that it is "protected spectrum" or bandwidth in which terrestrial transmitters are forbidden to transmit. In his most recent writings, Ehman resists "drawing vast conclusions from half-vast data."

[Jennie McGrath]

Location of the signal

The location of the signal in celestial coordinates was, at (epoch J2000.0)

Right Ascension (On the positive horn): 19h25m31s ± 10s

Right Ascension (On the negative horn): 19h28m22s ± 10s

Declination (Is the same for both horns): -26d57m ± 20m

This region of the sky lies in the constellation Sagittarius, roughly 2.5 degrees south of the fifth-magnitude star Chi-1 Sagittarii

[Jennie McGrath]

Explanation of the Code "6EQUJ5"
On the Wow! Computer Printout
By Jerry Ehman

 
The photo of the computer printout of the Wow! source shows not only my handwritten comment ("Wow!") but also the circling of the 6 characters "6EQUJ5" lined up vertically in a column. What is the meaning of this code?

Each of the first 50 columns of the computer printout shows the successive values of intensity (or power) received from the Big Ear radio telescope in each channel (10 kHz wide) in successive 12-second intervals (10 seconds was used for actual sampling and another approximately 2 seconds was needed for computer processing). In order to conserve space on the printout, Bob Dixon and I decided to use a coding method that would result in only one alphanumeric (i.e., either alphabetic or numeric) character for each intensity. The computer was programmed to keep a continuously-updated account for each channel of a baseline value and an rms value (rms stands for "root mean square", which is equivalent to the statistical term "standard deviation"). The actual intensity (after the baseline value was subtracted out) was then divided by the rms value to obtain a scaled value (i.e., the number of standard deviations above the baseline). Since there was space for only one character to be displayed, we decided to take only the integer value of this scaled intensity for values in the range 0 to 9.999... . The truncated value of zero was printed as a blank (space). The truncated value of 1, 2, 3, 4, 5, 6, 7, 8, and 9 were printed directly. For scaled intensities of 10 to 35, inclusive, the capital letters of the alphabet were used. Thus a truncated value of 10 was printed as an "A", 11 as a "B", etc. If the scaled intensity ever got to 36.0 or above, the program would simply start over again at zero (e.g., a truncated value of 38 would be printed the same as that of 38-35=3, namely a "3").

Thus, the "6EQUJ5" code in channel 2 means successive intensities as follows:
6 --> the range 6.0 - 6.999...
E --> the range 14.0 - 14.999...
Q --> the range 26.0 - 26.999...
U --> the range 30.0 - 30.999...
J --> the range 19.0 - 19.999...
5 --> the range 5.0 - 5.999...

The value "U", meaning the range 30.0 - 30.999..., was the largest value ever seen. We do not believe that the intensity ever got above 31.0 and hence no rollover (subtraction of 35) ever occurred. It would have been easy to spot in a sequence of 6 or 7 numbers that should follow the antenna pattern of the telescope.

The six successive values in channel 2 fit the antenna pattern of Big Ear very well. I have also done a correlation analysis of the six data points with the mathematical functions: (1) gaussian = normal curve; and (2)(sin(x)/x)^2. The data fit each of those two functions very well with correlation coefficients of over 0.99 (i.e., almost a perfect fit). I also fit the data to each of the two actual antenna patterns (of the two horns) using the moderately strong radio source OY372. The correlation coefficients were again over 0.99. There was not enough difference between the two correlation coefficients to determine which horn the Wow! source was received in.

http://www.bigear.org/6equj5.htm

[Jennie McGrath]

A VLA Search for the Ohio State "Wow"

Robert H. Gray
Gray Data Consulting, 3071 Palmer Square, Chicago, IL 60647; rgray@graydata.net
and
Kevin B. Marvel
American Astronomical Society, 2000 Florida Avenue NW Suite 400, Washington DC 20009; marvel@aas.org


Received 2000 April 26; accepted 2000 August 7

ABSTRACT

In 1977 a search for extraterrestrial intelligence at the Ohio State University Radio Observatory recorded a strong, narrowband, and apparently intermittent emission near the 21 cm hydrogen line. The detection displayed the antenna pattern signature of a transiting celestial radio source but was not repeated in subsequent transit observations. The event has been advanced by some as a candidate interstellar signal and dismissed by others as probable interference; no independent attempt to replicate the event with a spectral resolution comparable to Ohio State's has been reported. We used the Very Large Array to search for a possible underlying sourceartificial or naturalwhich could account for the detection by occasionally brightening because of scintillation, intrinsic variability, or some other mechanism. With a sensitivity greater than 100 times the original observations, we found two continuum sources within the Ohio State coordinate error boxes, but they displayed no unusual spectral features, showed no sign of flux variability, yielded normal spectral indices based on additional observations at 6 cm, and were too weak to account for the Ohio State detection. No narrow-bandwidth point sources were detected over a band of 1.5 MHz to a flux limit of about 20 mJy at the nominal coordinates. We conclude that the "Wow" was not due to a continuous source usually below Ohio State's several jansky detection threshold but occasionally increasing in flux by a factor of less than 100. Our search does not significantly constrain the possibility of intermittent sources because we dwelled for only 522 minutes per field.

Subject headings: extraterrestrial intelligenceradio lines: general

1. INTRODUCTION

     On 1977 August 15 a search for extraterrestrial intelligence (SETI) at Ohio State University recorded a strong (30 ), narrowband (10 kHz), and apparently intermittent emission near the 21 cm H I line (Kraus 1979). The detection persisted for six integration periods totaling 72 s and displayed the characteristic antenna pattern signature of a transiting radio source, although in just one of two beams, suggesting that it may have been intermittent or variable. Its frequency was reported as 1420.356 ± 0.005 MHz (J. D. Kraus 1990, private communication) and in a subsequent analysis given as 100 kHz higher (Ehman 1998)in either case near the 1420.405 MHz frequency of the H I emission line corrected to the local standard of rest, a frequency that has been suggested for interstellar signaling (Cocconi & Morrison 1959; Drake & Sagan 1973). The narrow bandwidth, match to the antenna pattern, and high intensity were so suggestive of an interstellar radio signal that one of the project scientists wrote "Wow!" on the printed record, which has become a name for the event.

     The apparent coordinates of the emission were R.A. = 19h22m22s or 19h25m12s (both ±5s), decl. = -2703 ± 20 (1950) (J. D. Kraus 1990, private communication); a subsequent review suggested right ascensions 3s and 5s later, respectively, with an uncertainty of ±10s (Ehman 1998). Two right ascensions result from the vertically polarized dual-feed antenna system forming two beams; which beam emission was detected in was not recorded because celestial sources were expected to appear in both beamsfirst in one, then the other 2m50s later in right ascension. The fact that the Wow was detected in only one beam could mean that it was not fixed on the sky and thus was likely due to man-made interference. However, the intensity pattern detected in that one beam so closely matched the signature of a transiting celestial source that we consider an alternative explanation for the single-beam detection: that the intensity of a source fixed on the sky varied significantly between beams, perhaps because of interstellar scintillation.

     Ohio State made approximately 100 additional transit scans at the same declination over the next few years but did not detect the emission again (Dixon 1985). The subsequent nondetections could, of course, be interpreted as evidence that the original detection was due to man-made interference, but the evidence for a celestial source (discussed in a following section) is quite strong.

1.1. Evidence Against a Noise Origin

     The Wow cannot be ascribed to a random noise fluctuation because a single 30  noise peak would not be expected in thousands of years at the Ohio State sampling rate and because six consecutive peaks over 5  in the same channel and mimicking the antenna pattern cannot be due to noise. Ohio State's 50-channel filter-bank receiver produced 50 samples every 12 s, each sample the average of 10 1 s integrations in a 10 kHz channel, with 2 s for processing (Dixon 1985). The resulting 3.6 × 105 samples day-1 would be expected to yield a few peaks over 4  each hour and one as high as 5  each day in the presence of noise alone, assuming a Gaussian distribution. In contrast, some fast Fourier transform (FFT) spectrometers used in more recent SETI experiments produce 1010 samples day-1 having an exponential distribution and yield noise peaks as large as 24  during 1 day and 30  over 1 yr (Colomb et al. 1995).

1.2. Evidence for a Celestial Origin

     Radio telescopes operating in transit mode provide a strong test for celestial origin because the power received from a point source fixed in the celestial coordinate system varies as the beam of the antenna system sweeps across the source. The resulting intensity curve is characteristic of the antenna system's gain pattern, and the Wow intensities fit a Gaussian model of the Ohio State antenna pattern very well (r = 0.99), shown in Figure 1. Only signals entering through the skyward pointing main beams would be expected to display the antenna pattern, ruling out ground-based transmitters.

[Jennie McGrath]

Fig. 1 CITED IN TEXT  |  HI-RES IMAGE (124kb)  |   

--------------------------------------------------------------------------------
Gaussian curve fit to observed intensities of the Ohio State Wow emission. Error bars show the uncertainty in time (±6 s) and intensity (±0.5 ).

[Jennie McGrath]

The observed intensity curve also indicates the length of time taken for a point source to pass through the beam, which depends on antenna beamwidth, cosine of declination, and the sidereal rate. For the Ohio State antenna, the expected duration of a point-source transit is 30 s between half-power points at  = 0 and increases to 36 s at the  = -27 of the Wow. The 39 s width of the observed intensity curve is consistent with the expected duration for a point source fixed in celestial coordinates at the declination observed, considering the uncertainty in time.

     Airborne or orbital vehicles would not in general remain in the beam for a time comparable to a transiting celestial source. Low Earth-orbiting satellites in east-west orbits are ruled out because they would cross the beam (8 half-power beamwidth [HPBW] in right ascension) in only a few seconds, assuming a motion of 4° per minute. A satellite in a polar orbit could conceivably cross the beam on a north-south line (where the beam is elongated to 40) with a fortuitous combination of velocity, range, and inclination that might mimic the transit duration of a celestial source. While the possibility of an interference origin cannot be entirely ruled out, it does not appear so likely as to justify dismissing the Ohio State data.

1.3. Limitations of the Ohio State Telescope

     Ohio State's failure to detect the Wow repeatedly may have been due to limitations of the telescope. One limitation was relatively low sensitivity, about 2 × 10-22 W m-2 per channel referenced to a 3.5  detection threshold. Reobservations could not rule out an underlying source with a flux less than a few janskys, which might occasionally brighten to produce a detection.

     A second limitation was the difficulty of detecting a possibly intermittent or varying source with a fixed transit telescope. The apparent source locale could be viewed for only 144 s day-1 (two beams, 72 s full width), and an intermittent or variable source might have been below the detection threshold during the brief time its position was viewed in subsequent transits.

     A third limitation was the relatively narrow band (500 kHz) of frequency covered by Ohio State's spectrometer. The Wow was detected in channel 2, near the edge of the band, and it is conceivable that it drifted into or out of the band during the few minutes between the passage of the two beams. That would require a fortuitous frequency drift ratein the range ±100±150 Hz s-1but is an alternative explanation for the single-beam detection.

1.4. Possible Origins

     The Wow frequency was near the 21 cm H I emission line, but there are several reasons to think it was not due to hydrogen. First, the line is usually detected over 100 kHz, yet this emission was not detected in adjacent 10 kHz channels (Ohio State subtracted a running baseline from each channel to remove the effect of such spatially extended sources). Second, H I flux would not be expected to vary on a timescale of minutes to produce the single-beam detection. Intermittent astrophysical radio sources are known but usually vary on a timescale either much shorter than minutes (e.g., pulsars) or much longer and are usually broad band. Relatively narrow spectral features are known (e.g., masers), but not at 21 cm. One goal of our observations was to search for an underlying astrophysical source, normally below Ohio State's detection threshold, which might account for the Ohio State detection by interstellar scintillation (Cordes & Lazio 1991), intrinsic variability, or some other mechanism.

     The Ohio State emission had several characteristics that make it seem worth investigating as a candidate interstellar signal as well. First, the 10 kHz bandwidth is suggestive of radio communication, which often uses essentially monochromatic carriers. Second, radio transmissions can vary in amplitude or switch on and off entirely, which could account for the single-beam detection. Third, there is strong evidence for a celestial origin, and narrow-bandwidth emissions from celestial sources are exactly what many searches for extraterrestrial intelligence seek. A second goal of our observations was to search for an underlying artificial source, possibly enhanced by scintillation to produce the Wow detection.

     Interstellar scintillation is known to cause flux variations of a factor of 2 in pulsar observations and has been noted as a possible cause of one-time apparent detections of narrowband signals in SETI (Cordes, Lazio, & Sagan 1997). According to this work, a large scintillation gain (possibly combined with receiver noise fluctuations) could occasionally produce a brief detection of a source normally below the receiver system detection threshold. Since the probability of a large scintillation gain g is extremely low, exp , the source might not be detected again even with many observations using a telescope of the same sensitivity. Reobserving with greater sensitivity, however, increases the probability of detection dramatically because one need not wait for a statistical peak in scintillation gain. A 10-fold increase over Ohio State's sensitivity would allow a scintillation hypothesis to be tested decisively, because the probability of such a large scintillation gain is extremely low, less than 10-5.

     Scintillation is, of course, not the only mechanism that might account for an interstellar signal being detectable on one occasion and not on others. One might imagine, for example, a highly directional transmission such as radar propagating from the surface of a rotating planet, sweeping across observers periodicallyperhaps once each extraterrestrial "day" (Sullivan, Brown, & Wetherill 1978). However, a decisive test of this sort of hypothesis would be difficult without some idea of the period (in this case, the length of the day) and extended observations. A scintillation hypothesis can be tested decisively with the simple expedient of more sensitive observations, which is the strategy we employed.

1.5. Prior Searches

     In 1987 and 1989 the Harvard/Smithsonian-META system (Horowitz et al. 1986) was used to observe the nominal Ohio State positions for 4 hr periods in a search for extremely narrow bandwidth and possibly intermittent radio signals over several 400 kHz segments of the 21 cm band (Gray 1994). With 0.05 Hz resolution, the 8.4 × 106 channel META spectrometer could detect ultra-narrowband radio signals to a limit of 4 × 10-24 W m-2 (with a 19.5  threshold, which noise peaks exceed only a few times per hour) but would not have detected the Ohio State signal if it had been wider than 10 Hz or if it had not been Doppler corrected to a constant frequency. No useful bounds were placed on natural sources because the extremely narrow channels were insensitive to normal spectral line and continuum radio sources.

     No other searches for the Wow at large radio telescopes appear in a summary of SETI experiments (Tarter 1995).

2. OBSERVATIONS

     We searched for evidence of an underlying source of the Ohio State emission using the Very Large Array1 (VLA) in observing runs in 1995 and 1996. Our observations improve on Ohio State's sensitivity by a factor of greater than 100, offering the prospect of detecting weak underlying sources. Our observations improve on the META search by using channel widths comparable to Ohio State's receiver, eliminating spectrometer resolution as a variable, and by covering the large uncertainty in declination. Our observations improve on all previous work by covering a larger range of frequency and by obtaining much higher spatial resolution.

     The full VLA has not previously been used for observations of interest to the SETI community, so a few comments on terminology and suitability are appropriate; theory and methods are discussed elsewhere (e.g., Perley, Schwab, & Bridle 1994). The VLA is a 27-element interferometer that can produce high spatial resolution maps or images of the radio sky. The size of the image is determined by the beamwidth of a single 25 m antenna (30 at 21 cm), but the spatial resolution is determined by the baselines between antennas, which can range up to tens of kilometers. Our 21 cm observations using the so-called BnA array configuration, for example, covered 1800 × 1800 primary fields, which were mapped to 1024 × 1024 pixel images containing 105 synthesized beams, each with a beamwidth of 39. In spectral line mode, images at many adjacent frequency channels yield a data cube with the dimensions of right ascension, declination, and channel. Data cubes can also be treated as sets of spectra10242 spectra at discrete coordinates in this example, although not all independent.

     Synthesis imaging interferometers offer several advantages for SETI purposes. One is that interference maps to a point source only in the direction of the celestial polethe only point where the time lags at all antennas are all the same. Another advantage, for observations near emission lines, is that the VLA's high spatial resolution configurations are relatively insensitive to large-scale structures such as hydrogen gas. A final advantage is that high spatial resolution allows radio source positions to be matched with optical objects; radio detections coinciding with stellar positions would be potentially interesting, since stars are rarely detectable radio sources (Becker et al. 1996; Wendker 1995). One disadvantage of the VLA for some purposes is limited spectral resolution, but we wished to replicate the Ohio State receiver's 10 kHz resolution, and the VLA's 6.1 and 12.2 kHz spectral line modes were an adequate match. The need to move the array off-source for phase calibration at frequent intervals (<1 hr) would be a disadvantage in a search for intermittent sources, but only continuous sources were within the scope of our investigation.

     Observations in 1995 were carried out using the BnA configuration with a spatial resolution of 39. This configuration has the southeast and southwest arms of the array in the B configuration (with a maximum antenna separation of 11.4 km), while the north arm is in the largest A configuration (with baselines up to 36.4 km), and provides a circular synthesized beam for sources where  < -15. The observations covered a band of 0.78 MHz with 127 channels and a resolution of 6.1 kHz. Both left- and right-circular polarizations were observed, but only the Stokes I images are discussed throughout this paper. Two fields centered on the nominal Ohio State coordinates were observed for 20 minutes each, with 10 minute integrations made 15 north and south and 5 minute integrations 30 north and south, for a total of 10 partly overlapping fields. Observations north and south covered twice the Ohio State declination uncertainty, to include sources outside of the Ohio State declination HPBW.

     Observations in 1996 used the D configuration array, where the longest baseline is approximately 1 km, yielding a spatial resolution of 44. These observations measured the 6 cm flux of sources previously detected at 21 cm to obtain spectral indices and also included a more sensitive search for narrowband 21 cm sources over a wider range of frequency. The two nominal fields were observed at 21 cm for 43 minutes each, with 127 channels and a resolution of 12.2 kHz, covering a band 1.5 MHz widedoubling our prior sensitivity in one case and improving sensitivity by 50% in the other case. Additional 5 minute snapshots were made at 21 cm, covering 6.25 MHz with 31 channels at 195 kHz resolution. Sensitivity in the various fields is summarized in Table 1.

[Jennie McGrath]

Table 1 CITED IN TEXT  |  ASCII  |  TYPESET IMAGE  |    -------------------------------------------------------------------------------- Sensitivity of 21 cm Observations FIELD NAME 1995 Sep 25 (39 resolution) 1996 May 7 (44 resolution) Continuum (0.78 MHz) Line, Typicala (6.1 kHz) Continuum (1.5 MHz) Line, Typicala (12.2 kHz) Continuum (6.25 MHz) 1922-265... 2.1 17 ... ... 1.3 1922-267... 1.4 8 ... ... 1.2 1922-270... 1.0 5.5 (7.4 max.) 0.8 3.5 (6.1 max.) 1.5 1922-273... 1.7 8 ... ... 1.4 1922-275... 2.6 13 ... ... 1.6 1925-265... 2.0 14 ... ... 1.0 1925-267... 1.5 9 ... ... 0.9 1925-270... 1.0 5.3 (7.3 max.) 0.5 3.7 (5.4 max.) 1.3 1925-273... 1.9 9 ... ... 1.9 1925-275... 2.4 15 ... ... 1.4 Note.Table entries are rms in mJy beam-1.      a Typical for channels 1560 and 70110. Bandpass filtering attenuated channels near the edge of the band, and channels near band center had higher noise because of H I emission.

[Jennie McGrath]

 1 The VLA is a facility of the National Radio Astronomy Observatory, which is operated by Associated Universities, Inc., under a cooperative agreement with the National Science Foundation.

2.1. Continuum Sources

     We first identified apparent continuum sources, to check their spectra for peculiar features and to later exclude them from tests for narrow-bandwidth sources. The AIPS task IMAGR was used to create "channel 0" maps, an average of the flux over the 75% of channels least attenuated by bandpass filtering. Channel 0 maps provide the highest sensitivity to continuum sources but could detect a single-channel source such as the Wow (assuming a flux of 60 Jy) at a signal-to-noise ratio (SNR) of 100 if it were continuous and an SNR of 10 if present for only 2 minutes during our longest integration. The primary goal of this step, however, was to identify continuum sources, not to search for a presumably rare brightening of the Ohio State source.

     Sources were identified and cataloged from channel 0 maps using the AIPS routine SAD ("search and destroy"). This unfortunately named feature extraction task searches for islands of flux in an image, fits a Gaussian model of the synthesized beam, and records coordinates, flux, and other parameters for the candidate source in a catalogoptionally subtracting (hence, destroy) the fitted feature to leave a map of residuals. Features with peak flux at least 8 times the typical noise in 6.25 MHz maps (0.91.9 mJy, depending on the field) were accepted as real sources and are plotted in Figure 2. Table 2 lists the sources detected, all of which were unresolved. All of the sources were also detected in the subsequent NRAO VLA Sky Survey (Condon et al. 1998).

[Jennie McGrath]

Fig. 2 CITED IN TEXT  |  HI-RES IMAGE (199kb)  |       

--------------------------------------------------------------------------------
Continuum sources detected at 21 cm. Source coordinates are at circle centers, and the area of each circle indicates relative flux. The error boxes surrounding each of the two right ascensions given by Ohio State are shown by dashed lines.

[Jennie McGrath]

2.1.1. Continuum Sources in the Error Boxes

     Two apparent continuum sources were detected in the Wow coordinate error boxes, defined by the Kraus right ascensions and a ±6s uncertainty (the Ehman positions with a ±10s uncertainty are covered in the analysis of entire fields in a subsequent section). That is not surprising, because the error boxes covered approximately 10% of the area observed, and 23 continuum sources were detected over the entire area. The sources were at R.A. = 19h22m220 and 19h25m131, both quite close to the right ascensions given by Ohio State, with fluxes of 32 and 16 mJy, respectively. Ohio State's single-channel sensitivity of several janskys was, however, far too low to detect continuum sources with such fluxes. We examined them in some detail to verify that they were continuum sources.

     Spectra of the two sources, shown in Figure 3, display emission over all channels as expected of continuum sources. There is no evidence of prominent single-channel features that might have been enhanced by scintillation or other mechanisms to produce the Ohio State detection. To verify that these sources radiate over the wide frequency range expected of continuum emission mechanisms, we later obtained 6 cm fluxes for all sources detected in 1995. The spectral indices fall in the range expected (0.51.1) for normal continuum sources (Kraus 1986), typically extragalactic objects with synchrotron emission mechanisms.

[Jennie McGrath]

Fig. 3 CITED IN TEXT  |  HI-RES IMAGES: 3a (233kb) 3b (352kb)  |       

--------------------------------------------------------------------------------
Spectra of sources at (a) R.A. = 19h22m220, decl. = -2654284 and (b) R.A. = 19h25m131s, decl. = -2708221. The resolution is 12.2 kHz.


Fig. 3a




Fig. 3b