Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

A technique for the analysis of total power radio continuum data Backhouse, Christopher James 1987

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata

Download

Media
831-UBC_1987_A6_7 B32.pdf [ 3.77MB ]
Metadata
JSON: 831-1.0096860.json
JSON-LD: 831-1.0096860-ld.json
RDF/XML (Pretty): 831-1.0096860-rdf.xml
RDF/JSON: 831-1.0096860-rdf.json
Turtle: 831-1.0096860-turtle.txt
N-Triples: 831-1.0096860-rdf-ntriples.txt
Original Record: 831-1.0096860-source.json
Full Text
831-1.0096860-fulltext.txt
Citation
831-1.0096860.ris

Full Text

A TECHNIQUE FOR THE ANALYSIS OF TOTAL POWER RADIO CONTINUUM DATA by CHRISTOPHER JAMES BACKHOUSE B . S c , The U n i v e r s i t y o-f A l b e r t a , 1985 THESIS SUBMITTED IN PARTIAL FULFILMENT OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF SCIENCE i n THE DEPARTMENT OF PHYSICS We accept t h i s t h e s i s as conf o r m i n g to the r e q u i r e d s t a n d a r d THE UNIVERSITY OF BRITISH COLUMBIA September 1987 © C h r i s t o p h e r James Backhouse, 1987 In presenting this thesis in partial fulfilment of the requirements for an advanced degree at the University of British Columbia, I agree that the Library shall make it freely available for reference and study. I further agree that permission for extensive copying of this thesis for scholarly purposes may be granted by the head of my department or by his or her representatives. It is understood that copying or publication of this thesis for financial gain shall not be allowed without my written permission. Department of Ph^sla  The University of British Columbia 1956 Main Mall Vancouver, Canada V6T 1Y3 Date Septe^Ur- 3 1^8 7 DE-6(3/81) i i ABSTRACT In the - f a l l of 1986 the G a l a c t i c Radio P a t r o l began t o t a l power o b s e r v a t i o n s with the new seven f e e d system on the 91 metre r a d i o t e l e s c o p e at Green Bank, W. V i r g i n i a . The data were taken at a wavelength of 6 cm, with N y q u i s t sampling i n t e r v a l s , i n the r e g i o n of the g a l a c t i c p l a n e c o r r e s p o n d i n g t o the c o o r d i n a t e ranges of 1=25 t o 225 degrees, and |b| <. 5.5 degrees. E x i s t i n g s u r v e y s at 6 cm p r o v i d e a coverage of the g a l a c t i c p l a n e over the g a l a c t i c l o n g i t u d e range 190 to 60 (through 360) degrees. T h i s work i s l a r g e l y complementary i n t h a t i t s coverage i s over the g a l a c t i c l o n g i t u d e range of 25 to 225 degrees. A mapping t e c h n i q u e has been developed t o f u l l y e x p l o i t t h i s d a t a . T h i s t e c h n i q u e w i l l a l l o w the mapping of the g a l a c t i c r e g i o n with a s e n s i t i v i t y s e v e r a l times t h a t of p r e v i o u s l y a v a i l a b l e s u r v e y s and with r e l i a b l e s t r u c t u r a l i n f o r m a t i o n on a n g u l a r s c a l e s of <. 1 degree. The above t e c h n i q u e was a p p l i e d to a t e s t r e g i o n c e n t r e d upon the supernova remnant G109.1-1.0. The r e s u l t a n t maps were compared t o a p r e v i o u s l y p u b l i s h e d map of t h i s supernova remnant i n o r d e r to determine the s t r e n g t h s and r e l i a b i l i t y of the p r e s e n t r e d u c t i o n method. TABLE OF CONTENTS i i i A b s t r a c t . i i T a b l e of Contents i i i L i s t of F i g u r e s and I l l u s t r a t i o n s i v Acknowledgements. > v i i 1. I n t r o d u c t i o n . . 1 2. The T e l e s c o p e , i t s Feeds and R e c e i v e r s . . . 4 3. The C o n t r o l System of the T e l e s c o p e 6 4. The Design of the Ob s e r v i n g Program 9 5. The A c q u i s i t i o n of the Data... 14 6. C a l i b r a t i o n s • 16 6.1 C a l i b r a t i o n s of the Antenna 18 6.2 C a l i b r a t i o n s of the R e c e i v e r s 48 7. Noi se. • 53 8. A n a l y s i s 57 8.1 The O b s t a c l e s 57 8.2 An Overview of the A n a l y s i s 63 8.3 The C o n v o l u t i o n Program..... 66 8. 4 Base 1 i ne Removal 69 8.5 Some Simple Atmospheric Removal Methods 74 8.6 The H y b r i d Method 77 9. R e s u l t s 79 10. Summary and C o n c l u s i o n s 90 R e f e r e n c e s . 92 Appendix A: The Data Format 93 Appendix B: T r a n s l a t i o n of F l o a t i n g P o i n t R e p r e s e n t a t i o n s 98 Appendix C: The South Bound Mapping 99 L i s t of F i g u r e s and I l l u s t r a t i o n s i v 1) The o r i e n t a t i o n of the f e e d assembly as seen on the sky d u r i n g a n o r t h bound scan 5 2) A c t u a l b e h a v i o u r of the t e l e s c o p e . . . . . ...7 3) The l i n k i n g of a no r t h and a south bound scan 7 4) The i n t e r l e a v i n g of the t r a c k s . . . . . . . .....10 5) The coverage a c h i e v e d f o r no r t h bound scans 12 6) The coverage a c h i e v e d f o r south bound scans 12 7) A f l o w c h a r t of the d a t a a c q u i s i t i o n 15 8) P o i n t i n g e r r o r i n r i g h t a s c e n s i o n f o r feed 7 .26 9) North bound d e c l i n a t i o n p o i n t i n g e r r o r f o r feed 7 .26 10) South bound d e c l i n a t i o n p o i n t i n g e r r o r f o r feed 7 27 11) A sample beam map of feed 3 27 12) A sample beam map of f e e d 1.. ...28 13) The r e s i d u a l s a f t e r s u b t r a c t i n g the best f i t model beam from the beam map of feed 1........ .28 14) The no r t h bound antenna g a i n curve f o r f e e d 1 29 15) The no r t h bound antenna g a i n c u r v e f o r f e e d 2 29 16) The north bound antenna g a i n curve f o r feed 3 30 17) The n o r t h bound antenna g a i n c u r v e f o r feed 4 30 18) The no r t h bound antenna g a i n curve f o r f e e d 5 31 19) The n o r t h bound antenna g a i n c u r v e f o r f e e d 6 31 20) The n o r t h and south bound g a i n c u r v e s f o r feed 7 32 21) The combined antenna g a i n c u r v e f o r feed 7. .....32 22) The n o r t h bound N-S beam width f o r beam 1 33 23) The n o r t h bound N-S beam width f o r beam 2 33 V 24) The n o r t h bound N - S beam width -for beam 3 34 25) The n o r t h bound N - S beam width f o r beam 4..... 34 26) The n o r t h bound N - S beam width f o r beam 5 ...35 27) The n o r t h bound N - S beam width f o r beam 6.... 35 28) The n o r t h and south bound N - S beam widths f o r beam 7....36 29) The combined N - S beam width f o r beam 7 36 30) The n o r t h bound E-W beam width f o r beam 1..... .37 31) The n o r t h bound E-W beam width f o r beam 2 ...37 32) The n o r t h bound E-W beam width f o r beam 3.......... 38 33) The n o r t h bound E-W beam width f o r beam 4 .....38 34) The n o r t h bound E-W beam width f o r beam 5 39 35) The n o r t h bound E-W beam width f o r beam 6... 39 36) The n o r t h and south bound E-W beam widths f o r beam 7....40 37) The combined E-W beam width f o r beam 7...... 40 38) The a n g u l a r s e p a r a t i o n of beam 7 and beam 1 ....41 39) The a n g u l a r s e p a r a t i o n of beam 7 and beam 2 41 40) The a n g u l a r s e p a r a t i o n of beam 7 and beam 3 ....42 41) The a n g u l a r s e p a r a t i o n of beam 7 and beam 4.... 42 42) The a n g u l a r s e p a r a t i o n of beam 7 and beam 5 43 43) The a n g u l a r s e p a r a t i o n of beam 7 and beam 6 43 44) The r o t a t i o n a n gle e r r o r f o r beam 1 44 45) The r o t a t i o n angle e r r o r f o r beam 2 44 46) The r o t a t i o n angle e r r o r f o r beam 3 45 47) The r o t a t i o n a n gle e r r o r f o r beam 4 45 48) The r o t a t i o n a n gle e r r o r f o r beam 5.. 46 49) The r o t a t i o n a n gle e r r o r f o r beam 6..... 46 v i 50) Comparison of n o r t h and south bound g a i n s f o r -feed 3....47 51) Comparison of n o r t h and south bound g a i n s -for feed 6.... 47 52) Gain s t a b i l i t y d u r i n g scan 2739 51 53) Gain s t a b i l i t y d u r i n g scan 2741 51 54) Gain s t a b i l i t y d u r i n g scan 2980 52 55) Gain s t a b i l i t y d u r i n g scan 2982 52 56) Raw d a t a showing the c h a r a c t e r i s t i c s of the p o o r l y behaving c h a n n e l s (from scan 1202)......... 56 57) Raw d a t a showing a smooth v a r i a t i o n i n atmosphere and s t r o n g g a i n v a r i a t i o n i n channel 10 (from scan 2739)....61 58) A c o n t o u r p l o t showing the e f f e c t s of atmospheric e m i s s i o n and s t r o n g i n t e r f e r e n c e 61 59) Raw d a t a showing r a p i d l y v a r y i n g atmospheric e m i s s i o n seen e q u a l l y by a l l f e e d s . . . 62 60) A map of the t e s t r e g i o n produced by Braun (1981) 80 61) A l a b e l l e d contour p l o t produced by the h y b r i d method with a neighbourhood of 5 p o i n t s and which shows s t r u c t u r e removal 82 62) The above map r e p l o t t e d without l a b e l s on the contours..83 63) A l a b e l l e d contour p l o t produced by the h y b r i d method with a neighbourhood of 200 p o i n t s 84 64) The above map r e p l o t t e d without l a b e l s on the contours..85 65) A s e n s i t i v e , l a b e l l e d c o n t o u r p l o t produced by the h y b r i d method with a neighbourhood of 5 p o i n t s 87 66) A s e n s i t i v e , l a b e l l e d contour p l o t produced by the h y b r i d method with a neighbourhood of 200 p o i n t s 88 v i i Acknow 1edgements The h e l p and h o s p i t a l i t y of the s t a f f a t Green Bank a re g r e a t l y a p p r e c i a t e d . Much i s owed to my c o l l e a g u e Huang J i a n Xu f o r h i s t i r e l e s s work in the a n a l y s i s of the p o i n t i n g d a t a w i th the POPS sys tem w h i l s t at Green Bank. I would l i k e to thank D r . W.H. McCutcheon and D r . W . L . H . S h u t e r f o r t h e i r s u g g e s t i o n s and h e l p . My f e l l o w g r a d u a t e s t u d e n t Graeme Luke was of g r e a t s e r v i c e i n t r a n s f e r r i n g the d a t a f rom tape t o t a p e . Many o t h e r s , too numerous to l i s t , o f f e r e d a s s i s t a n c e or a d v i c e . I must of c o u r s e thank my s u p e r v i s o r D r . P . C . Gregory f o r h i s d i r e c t i o n , p a t i e n c e and a s s i s t a n c e t h r o u g h o u t t h i s work. 1. INTRODUCTION The G a l a c t i c Radio P a t r o l began making s u r v e y s of the g a l a c t i c p l a n e i n 1977. The main goal o-f thes e s u r v e y s (Gregory and T a y l o r , 1981, 1986; T a y l o r and Gregory, 1983) was the f i n d i n g o-f and r e s e a r c h upon v a r i a b l e r a d i o s o u r c e s . To t h i s end the s u r v e y s a r e s e n s i t i v e and e x t e n s i v e . As such, the c u r r e n t survey p r o v i d e s an o p p o r t u n i t y t o c o n s t r u c t a new a t l a s of g a l a c t i c r a d i o e m i s s i o n with an a r e a of more than 1000 square degrees. There a r e many s u r v e y s which i n c l u d e the g a l a c t i c p l a n e . The a l l sky s u r v e y s of Has lam et a l . at 408 MHz (1981), B e r k h u i j s e n at 820 MHz (1972), Condon and B r o d e r i c k at 1400 MHz (1985) and R e i c h at 1420 MHz (1982), are such s u r v e y s . At f r e q u e n c i e s c l o s e r t o t h a t used i n the p r e s e n t survey, the s u r v e y s of Haynes et a l . at 5 GHz (1978) and of A l t e n h o f f e t a l . at 4.875 GHz (1978) have covered the i n n e r p o r t i o n ( near 1=0 ) of our g a l a x y . Although t h e r e i s some o v e r l a p of r e g i o n s , t h i s work (1 = 25 to 225 degrees) i s complementary i n coverage to those of Haynes et a l . (1=190 through 360 to 40 degrees) and A l t e n h o f f e t a l . (1 = 357 through 360 t o 60 d e g r e e s ) . Moreover, t h i s work has produced maps with meaningful c o n t o u r s at i n t e r v a l s of 25 mJy/Beam Area as compared wi t h the 100 mJy/Beam a r e a and 260 mJy/Beam a r e a i n the s u r v e y s of A l t e n h o f f e t a l . and Haynes et a l . r e s p e c t i v e l y . In the p a s t , d i f f e r e n t i a l mapping has been employed to remove the e f f e c t s of atmospheric e m i s s i o n and g a i n v a r i a t i o n . Because the e m i t t i n g r e g i o n of the atmosphere tends t o be l a r g e and c l o s e , i t s e m i s s i o n i s r e c e i v e d e q u a l l y w e l l by the c l o s e l y spaced beams. By s u b t r a c t i n g the s i g n a l of a beam o f f the so u r c e 1 from the s i g n a l of a beam on the source, t h i s atmospheric e m i s s i o n i s removed. U n l e s s t h i s second beam was not t r a c k i n g any other s o u r c e , s t r u c t u r a l i n f o r m a t i o n has been l o s t by t h i s s u b t r a c t i o n . For t h i s reason, d i f f e r e n t i a l mapping tends t o suppress extended s t r u c t u r e . D e s p i t e t h i s , the power of such non l i n e a r t e c h n i q u e s as the maximum entropy method (MEM) and Clea n (Hogbom 1974), have allowed the mapping of low i n t e n s i t y extended s t r u c t u r e (Braun 1981). The m i s s i n g i n f o r m a t i o n must be r e p l a c e d by some form of e x t r a p o l a t i o n . The v a r i o u s methods of replacement, MEM, Emerson's method and Cl e a n , tend t o be ve r y computing i n t e n s i v e . Moreover, t h e i r e x t r a p o l a t i o n of the m i s s i n g d a t a may be erroneous as i n the case of MEM's peak s h a r p e n i n g e f f e c t s (Nityananda, Narayan 1982), C l e a n ' s r i d g e s (Steer e t a l . 1984) or MEM's Gibb o s c i l l a t i o n s (Cornwell 1983). The new seven f e e d assembly on the 91 metre t e l e s c o p e i n c o r p o r a t e d r e c e i v e r s of such g a i n s t a b i l i t y t h a t they were used i n a t o t a l power mode. Each of the f e e d s had two cha n n e l s ( r i g h t and l e f t hand c i r c u l a r p o l a r i s a t i o n ) . The output from each r e c e i v e r was r e c o r d e d d i r e c t l y and thus c o n t a i n e d the c o n f u s i n g e m i s s i o n s and g a i n v a r i a t i o n a r t i f a c t s n o r m a l l y removed i n the d i f f e r e n t i a l mapping method. 2 D e s p i t e t h e s e c o m p l i c a t i o n s ( i n h e r e n t i n the t o t a l power o b s e r v i n g method), the d a t a r e d u c t i o n t e c h n i q u e developed i n t h i s work p r o v i d e d maps comparable i n q u a l i t y t o those produced by the Maximum Ent r o p y Method from d i f f e r e n t i a l o b s e r v a t i o n s (Braun 1981). The t e c h n i q u e i s f a r l e s s computing i n t e n s i v e than the MEM. S i n c e the c u r r e n t t e c h n i q u e need not remove i n f o r m a t i o n from s t r u c t u r e s l e s s than 1 degree i n e x t e n t , i t need not r e p l a c e the i n f o r m a t i o n by some p r o c e s s of e x t r a p o l a t i o n . I t i s t h e r e f o r e more r e l i a b l e . 3 2. THE TELESCOPE. ITS FEEDS AND RECEIVERS The seven feeds of the 91 metre telescope were arranged in the f i l l e d hexagonal pattern shown in figure 1. A small displacement of a feed in the focal plane resulted in the angular displacement of the beam associated with that feed. The constant of proportionality between these two displacements is known as the beam deviation factor. For a f a i r l y typical value <at about 60 degrees of declination), these seven feeds then scanned the sky with beams separated from their nearest neighbour by about 8.5 arc minutes. Rotating the feed assembly until feeds 1, 7 and 4 described a line rotated 19.1 degrees from the track in the sky of the central feed provided for seven equally spaced tracks with a separation of approximately 1 half power beam width. (see figure 1) The beam separations and other telescope parameters are dealt with more precisely in the calibrations section. Each of the seven feeds s p l i t the signal into a right and a left hand circularly polarised signal. These signals were then amplified by a cooled GaAs Fet receiver which had a ldb bandwidth of 500 Mhz. The output of these 14 channels was integrated for 0.2 seconds per sample. Typically, the system noise temperatures were about 60 Kelvin. This gave rise to typical rms noise fluctuations of: (Tiuri 1966) = <60 K) /S B x T * 6 mK where B: 3 db bandwidth (Hz), 1* : integration time (sec) By combining the two polarisations this value was reduced by a further factor of S 2. The sample noise fluctuations and system 4 noise levels are analysed in more detail in the Noise section. More important than the rms noise were the spurious signals caused by gain variations. The gain o-f the receivers varied in time. The smaller the amplitude of the gain variation the smaller was the associated time scale. These variations are explored more thoroughly in section 6.2 . Figure 1. The orientation of the feed assembly as seen on the sky during a north bound scan. 5 3. THE CONTROL SYSTEM OF THE TELESCOPE The 91 metre telescope was capable o-f being driven in declination only. The telescope was swept along in right ascension by the rotation of the earth i t s e l f . Although this was a straightforward method of using a telescope, i t had the unfortunate side effect of placing time constraints upon a l l movements. The telescope was controlled by a Honeywell 316 < H316 ) computer. This computer attempted to ensure that the telescope was at the correct declination and moving at the correct drive rate at the correct time. The H316 used either the variable speed motor <0 to about 140 arc minutes/sec) or the fixed speed motor (10 degrees/minute) to drive the telescope in declination. Delays were associated with the use of either of these motors. A detailed knowledge of how the H316 made its decisions of how and when to move the telescope, was vita l for a smooth observing run. I n i t i a l l y , my telescope scheduling programs assumed that the H316 would, after a short delay, smoothly accelerate the telescope to meet the start of the next scan. The characteristics of this behaviour were parametrized so that they could be easily varied. Unfortunately, the H316 accelerated, not to the starting position of the next scan, but to the extrapolated position of the scan at that time (see figure 2). Moreover, if the difference in declination between the current position and this extrapolated position were greater than a certain value (usually 90 arc minutes) the H316 would use the fixed speed motor. With the delays associated with using the fixed rate motor (delays to set and 6 t e l e s c o p e b e i n g \ d r i v e n a-l 10 d e g r e e s / n l n d e l a y w h i l e brakes s e t a n h slew n o t o r engage end of n o r t h bound scan d c l n y uli l i e bra lies s e t and v a r i a b l e r a t e d r l v i i s engaged t e l e s c o p e a t t e n p t s to get back on t r a c k bg d r i v i n g at n a x i n i u n v a r i a b l e d r i v e r a t e e x t r a p o l a t i o n of next scan s t a r t of next south bound scan Cas connandtd]^ TIME , t o l e s c o i K i s . back on t r a c k Figure 2. The behaviour of the telescope. e x t r a p o l a t i o n of south bound scan c a l i b r a t i o n scan e x t r a p o l a t i o n of c a l i b r a t i o n scan l... a f t e r a d e l a y t he t e l e s c o p e noves t o the e x t r a p o l a t e d p o s i t i o n a t the n a x l n u n d r i v e r a t e , c a t c h i n g up b e f o r e the commanded s t a r t t i n e . * " \ " ' . end of north— bound scan. / e x t r a p o l a t e d / dec 1i n a t i o n / i s a l s o / d e s t i nat ion dec l i n a t i r o n '.•*.—telescope has caught up v t o the e x t r a p o l a t e d p o s i t i o n . . . . . - . . . « - cnnnanded s t a r t or s o u t h bound scan. T i n e Figure 3. The linking o-f a north bound and a south bound scan. 7 release brakes), i t often took longer to move with the fixed rate motor than with the far slower variable speed motor. Occasionally the H316 was given erroneous positional information. In lieu of the true declination the H316 sometimes received a value of 0 degrees. Since the difference in declination was usually greater than the c r i t i c a l distance mentioned above, the H316 typically stopped the variable speed motor <if moving), applied the brakes, disengaged the variable speed motor, and fi n a l l y engaged and started the fixed rate motor speed motor. Although the positional errors were rapidly corrected, the time consuming nature of the actions taken by the H316 ensured that the telescope was well off the planned track. This problem was partially remedied by not permitting the H316 the use of the fixed speed motor during the night-time observations. Although the positional errors s t i l l caused the telescope to deviate from the planned track, these deviations were far smaller. My telescope scheduling program was modified to ensure that when the telescope completed a command and was about to commence a driven scan, the telescope was already on the extrapolated track of the the next commanded action (see figure 3). The H316, with its frequent positional errors and occasional crashes, created many gaps in the mapping and calibration coverage. 8 4. THE DESIGN OF THE OBSERVING PROGRAM The i n i t i a l goals o-f the mapping were! 1) It should cover the galactic longitude range 1=25 to 1=225 degrees. 2) It should be centred upon the line de-fined by galactic latitude b=0 degrees. 3) That i t maintain (at least) Nyquist sampling intervals. 4) That, in ten nights o-f observing, i t cover as large an area as possible. 5) The declination drive rate must be kept constant (2 degrees/minute) The reasons -for this are described in the calibration section. As described in the telescope control section. the seven -feeds, when properly rotated, traced tracks separated by approximately one hal-f power beam width. Since Nyquist sampling required twice this density of sampling, for each set of seven tracks, a second set of seven tracks was interleaved (see figure 4 ) . The set of seven tracks is referred to as a scan. The set of linked scans made upon a given day is referred to as a sequence. It was d i f f i c u l t to interleave the tracks. At a given declination, a given angular separation between the interleaved tracks corresponded to a time delay between the start of the f i r s t scan and the start of the second interleaved scan. Unfortunately, as the declination varied so did the required time delay. Thus if the scans were properly spaced at a declination of 0 degrees, they were too closely spaced at higher declinations. If the interleaved scan needed to gain more time it Order of feeds Is as f o r North bound t c m s , Tracks are r o t a t e d s l i g h t l y c o u n t e r - c l o c k w i s e f r o n the v e r t i c a l . D E C L I N A T I O N E E Q • 5 T " S E Q . " * " " R I G H T A S C E N S I O N 3 2 4 E l i Q U I - N C E 3 FEQUnNCC e S E Q U E N C E t ' S E Q U E N C E S E Q U E N C E 9 * S E Q U E N C E 9 F i g u r e 4. T h e i n t e r l e a v i n g o f t h e t r a c k s . 10 covered a slightly shorter b range. If the interleaved scan needed to lose time i t extended over a slightly larger b range (see f i gures 5 and 6). The non-linear nature of this problem required the use of an iterative algorithm in order to find a solution. The algorithm, given only the starting galactic longitude, guessed the latitude coverage. It then calculated the declination ranges, telescope travel and pause times in order to ascertain whether the scans had the correct angular separation. If not, then the latitude range was adjusted. Whenever the latitude range made the transition from too large to too small or vice versa, the adjustment amount was multiplied by -0.5. This is known as a binary search technique. The power of this approach is that with 20 guesses, the b range was found to within one part in a million. When the algorithm had found a suitable b range, i t used the endpoint of the f i r s t pair of scans as the starting point for the next pair. In this manner evaluation continued until the last endpoint produced had a galactic longitude greater than 225 degrees. As described in the telescope control section, the algorithm assumed that the control computer would move to the starting point of the next scan. Instead the computer directed the telescope to the extrapolated declination of the next scan. This caused almost a l l scans to start off the track. The adopted solution was to write a smoothing program. This smoothing program adjusted the lengths of the scans in 11 order to ensure that the telescope was always upon the extrapolated line o-f the next scan. By doing this, the control computer did not waste time by moving the telescope in the wrong direction. Even so, by the time i t was able to get the telescope moving, i t was -far off the desired declination. It had to catch up tD this desired point before the data taking could commence. The only way the telescope could catch up was to increase its declination drive rate. Since the point upon the extrapolated scan was moving at 120 arc minutes/minute, and the maximum drive rate was only slightly higher than this, the telescope caught up very slowly. Since the highest reliable speed of the telescope was about 140 arc minutes/minutes, the catch up speed was approximately 17 per cent faster than the normal drive rate. If the control computer delayed one second, then i t was necessary to allow the telescope 6 seconds to catch up. To add to the confusion, the delay introduced by the control computer varied. For reasonable r e l i a b i l i t y the telescope was allowed 40 seconds to catch up. Since the scans are quite short at the extremes toward 1=25 degrees and 1=225 degrees, this catch up interval has changed the region of coverage somewhat. (The scans are no longer centred upon the line b=0 degrees.) Although the scans were not perfectly centred upon the line b=0 degrees, the most important requirement, that of Nyquist sampling, was satisfied for both North and South bound scans. 13 5. THE ACQUISITION OF THE DATA The receiver outputs were sampled at an interval of 0.1 seconds, pairs o-f these samples were added to -form the 0.2 second integrated samples. The 0.2 second samples were then written (in the record format described in Appendix A) to a 1600 BPI tape by the Masscomp computer at the telescope. These tapes were then taken to another Masscomp where these records were rewritten in a more compact format, 4 records per block. The tapes of blocked data were then duplicated and the duplicates sent to Vancouver. The calibration information was dealt with slightly differently. The integration time for calibration d r i f t scans was 1 second rather than the 0.2 seconds used for a l l driven scans. The calibration data were transferred to a Modcomp computer at Green Bank from which the pointing corrections were obtained using the POPS analysis system. (described more f u l l y in the calibration section) (see figure 7) The observing program produced data at a rate of approximately 40,000 bytes per minute in a session which lasted a month. The vast amount of data involved in this project, over half a gigabyte, required that the data be stored e f f i c i e n t l y . By storing the data in a binary rather than ASCII format, the data was compacted by a factor of about 3. In this way, one third as many tapes were required, and subsequent processing took one third as long. Once at UBC, the 1600 BPI tapes were copied to the TK50 tapes (standard upon nVaxes) at the TRIUMF computing f a c i l i t i e s . Once this was done the Patrol's ^Vax could read the data. 14 S R N V O P C G E N E R A T E S T E L E S C O F E O B S E R V I N G COMMANDS O R I G I N A L I te © B B P I T F i P E I G A L A C T I C R A D I O P A T R O L U V A X U N D E R U N I X cuvnx T K 5 B T A F E 1 U V A X A L S O UNDER VMS O P E R A T I N G S V S T E M < E T H E R N E T > * V A X 7 8 3 U N D E R V M S O P E R A T I N G S V S T E M F i g u r e 7. f l o w c h a r t o-f t h e d a t a a c q u i s i t i o n , 13 6. CalibratIons The beam geometry and beam shapes were determined by the shape of the telescope surface. Since the telescope flexed as i t moved in declination, an accurate determination Df the geometry and beam shapes required the calibration of the telescope over the declination range used in the survey. The parameters required to determine the beam geometry and beam shapes were the relative beam displacements, the half widths and peak gains of each beam. The flex of the telescope surface also affected the direction in which the beams pointed. These effects, known as painting errors, were also determined over the declination range used in the survey. For each beam, the antenna collected the emission from the sky. The power/Hz collected is thus: P/Hz = k T A = 'A where: B ( 8 , 0 ) A m f ( e - 8 o , 0 - 0 o ) dfi k - Boltzmann constant T« - Antenna temperature ( units of Kelvin ) A m - effective collecting area of antenna < units of ma ) f ( 8 , 0 ) - power pattern of the antenna (normalized to unity) B ( 8 , 0 ) - sky brightness (units of W m_5B Hz - 1 steradian - 1) This collected power is usually expressed as an antenna temperature. This antenna temperature is defined as: 2k. B ( 8 , 0 ) A m f ( 8 - 8 o , 0 - 0 o ) dn For a point source at 8i, 0*, with a flux density S (in Jy)i B ( 8 , 0 ) = S n < 8 - 8 i . ) S ( 0 - 0 X ) Thus, 16 TA = S _ A „ 2 k or, S = 2Jk_J_A (with units of Jy) A m For an extended source we define the beam solid angle as: fib = (with units of steradian/Beam area) *»TT and the average sky brightness seen by the beam as: r BAV - Ofc," B(8,0) +(8,0) dn (with units of Jansky/steradian) With these definitions we have: TA = 1 2kJ B(8,0) A„ f(8-8o,0-0o) dn = A m BAV fib 2k or, BAV fit. =_2JkJ_A_ (with units of Jy/Beam area) A m For an extended source, the scale factor of 2k/Am converts the collected power/Hz from an antenna temperature <K) to units of Jy/Beam area. For a point source, the same scale factor converts the collected power/Hz from units of Kelvin to units of Jy. In this work the sky brightness is expressed in units of Jy/Beam. The required scale factor 2k/Am was determined over the declination range from the analysis of point sources. Associated with each feed were two receivers. Since the gains of these receivers (Volts/Kelvin) varied in time, the calibration of these gains were required as frequently as possible. These calibrations are described more f u l l y in section 6.3. 17 6.1 Calibration of the Antenna Calibration scans were made through two sets of isolated, unresolved sources of known (to within ~5 per cent) flux density. Each set consisted of sources spaced at intervals in declination of approximately 5 degrees from a declination of -10 to 70 degrees. These scans provided the information required to adjust the telescope's pointing and provided the values for the telescope parameters. Fisher and Payne (1982) have shown that the focal point of the 91 m telescope moves as the telescope moves in declination. Because of this the feeds were not normally centred on the focal point. This in turn could give rise to a rotation angle dependence in the beam characteristics. Contributions to this dependence may also arise from asymmetry in the telescope dish. In order to eliminate this problem, for each declination the telescope parameters were determined at the same rotation angle as was used in the mapping stage. The past experiences of the Galactic Radio Patrol indicated that there was a velocity dependence of the apparent half power beam widths in the driven coordinate (Picha 1986). This broadening effect is thought to be due to oscillations in the telescope structure. Since the detailed behaviour of this effect was unknown, i t was decided to keep the drive rate constant. The above constraints required that for each declination the calibration sources be mapped in exactly the same manner as the galactic region. If the drive rate was to be kept constant, then the East - West half power beam widths had to be obtained by 18 a n a l y s i n g the maps made of each c a l i b r a t i o n s o u r c e . One s e t of c a l i b r a t i o n s o u r c e s was used to make n o r t h bound s c a n s wi th the a p p r o p r i a t e r o t a t i o n a n g l e s . The second s e t was used s i m i l a r l y f o r the s o u t h bound s c a n s . A l t h o u g h i t was o r i g i n a l l y i n t e n d e d t h a t each beam would be • f u l l y mapped, w e l l i n t o the o b s e r v i n g s e s s i o n a breakdown of the r e f r i g e r a t i o n sys tem r e q u i r e d t h a t the f e e d assembly be remounted. R e c a l i b r a t i o n of the t e l e s c o p e was then r e q u i r e d . With the number of days then r e m a i n i n g , t h e r e was i n s u f f i c i e n t t ime to f u l l y map a l l the beams. To compound t h i s p r o b l e m , f o r two days the s o u t h bound beam mapping was done w i t h i n c o r r e c t f e e d assembly r o t a t i o n s . Though t h i s p rob lem was d i s c o v e r e d w h i l e s t i l l at Green Bank, t h e r e was no t ime a v a i l a b l e t o redo the e r r o n e o u s s c a n s . For the s o u t h bound mapp ing , o n l y f o r f e e d s 3 , 6 and 7 were t h e r e s u f f i c i e n t d a t a f o r a n a l y s i s . S i n c e the d r i v e r a t e dependence of the t e l e s c o p e p a r a m e t e r s was s m a l l r e l a t i v e t o the p a r a m e t e r s t h e m s e l v e s , i t was p e r m i s s i b l e t o use d r i f t s c a n s to a s c e r t a i n the c o a r s e r i g h t a s c e n s i o n p o i n t i n g c o r r e c t i o n s . The f e e d assembly was g i v e n the same r o t a t i o n t h a t i t would have had i n a d r i v e n s c a n at the same dec 1i nat i o n . A s e r i e s of d r i f t s c a n s were made by the c e n t r a l f e e d through t h e c a l i b r a t i o n s o u r c e s . The p o i n t i n g c o r r e c t i o n s in r i g h t a s c e n s i o n were a s c e r t a i n e d by c o m p a r i n g the r i g h t a s c e n s i o n of the p o i n t where the peak f l u x was found w i t h the known r i g h t a s c e n s i o n of t h a t s o u r c e . The POPS a n a l y s i s system at Green Bank was used f o r t h i s r e d u c t i o n . F o r a l l subsequent c a l i b r a t i o n s a 1 9 correction o-f 0.8 seconds o-f time was added to each commanded right ascension, (see -figure 8) A series o-f driven scans were made by the central -feed through the calibration sources. After reduction of the scans (again using POPS) the pointing error in declination was found to depend upon the drive direction. For the north bound scans, no corrections were deemed necessary (see figure 9). The larger south bound pointing errors were reduced by subtracting the correction function shown in figure 10. For a l l subsequent calibration scans these corrections were applied to each commanded declination. Since the north bound beam mapping was far from complete, only feeds 1,2 and 3 were well described. There were relatively few data for feeds 4,5 and 6. Describing the beam of the central feed was especially d i f f i c u l t . At each calibration source only one half of the central beam was mapped. Complete maps of the central beam were made by combining two complementary halves taken from declinations at most 3 degrees apart. Since the beam parameters varied slowly, this introduced a negligible error. The display of these beam maps showed that the beam of each feed was tracked close to the centre of each calibration source (see f i gure 11). Braun (1981) described his beam shapes as asymmetric gaussians with plateaux and fi t t e d his beam model only to the data which were at least 10 per cent of the peak height. In my work, 20 least squares -fits o-f anisotropic gaussians o-f the form: A exp( (-4 ln 2) ( b _ z ( d 0 - d)= + c~ 2<a 0 - a) 2) ) A - amplitude of beam response (in units of K/Jy) d,a - declination, right ascension respectively b,c - half power beam widths for declination, right ascension respectively were found to describe the beams well (see figures 12, 13). The gaussians were fitted only to the data which were greater than 10 per cent of the peak value. The residuals from the subtraction of the best f i t from the beam map were at most 10 per cent of the peak value at the sidelobes and much less (several per cent) at the peak (figure 13 is typical). As mentioned earlier, problems with the telescope control computer were frequent and caused the loss of considerable data. The beam map approach to the calibration work was especially vulnerable to these problems. If any one of the three scans closest to the centre of a beam was missing, then the f i t tD the beam map was poor. When dealing with such quantities as relative beam placements, two complete beam maps were required. These quantities were thus more d i f f i c u l t to determine. The f i t s of these anisotropic gaussians to the north bound beam maps provided the gain (figures 14 to 20), North-South and East-West half power beam widths (figures 22-28 and 29-34 respectively), and relative positions on the sky (figures 38 to 49) of the seven beams. The behaviour of the central beam was f a i r l y well defined from the analysis of both north and south bound maps. In figure 21 the best f i t parabolas to the north and south bound gain data were very similar. (Each parameter of each of these parabolas is within 21 one standard deviation of the other.) Similar behavior was found for the beam widths, (see figures 37, 29) For feeds 3 and 6, several south bound beam maps were analysed. The results Df these f i t s were plotted atop the north bound results in figures 50, 51. From these figures i t can be seen that there was l i t t l e dependence of these beam parameters upon the direction in which the telescope was driven. The north bound mapping was done using the north bound calibration data. The south bound mapping used the south bound calibrations for the central feed and the north bound calibrations for the peripheral feeds (for which there was l i t t l e i f any valid south bound calibration data). For the determination of the relative beam placements, the following assumptions were made: 1) There were no deviations of the beam shape from the model that would give rise to an error in the fitted position. 2) The pointing errors did not change from day to day. 3) The pointing errors did not change in the ~8 arc minute (maximum) span of declination from the peripheral feed to the central feed. All these assumptions appeared to be valid. A complete verification would have required far more calibration data. Since the beam modelling algorithm determined the positions of the beam centres (relative to the centre of the central beam) very accurately (typically within 4 arc seconds), the uncertainties in the calculated beam placements (the data points in figures 38 to 49) were negligible. The relative beam placements 22 varied slowly across the declination range. With the number of data points available, the highest order polynomial that could have been f i t t e d to the data was a quadratic polynomial. The uncertainties in the -fits arose from the possible presence of higher order components in these variations. The calculated beam separations (from the central feed to the peripheral feeds) indicated that these separations were quite stable (8 to 9 arc minutes) throughout the declination range, (see figures 38 to 43). There were few data points for these relative beam calculations and the f i t s are thus quite uncertain. The gridding program assumed a constant beam separation of 8.5 arc minutes between a l l feeds (within 0.3 arc minutes of the best f i t separations at a declination of 59 degrees in figures 38 to 43). Uncertainty in the beam separations was a major source of positional error. This positional uncertainty was approximately 1 minute of arc. In figures 44 to 49 the rotation angle errors were plotted for the six peripheral beams. It should be stressed that these rotation errors could have arisen from feed assembly rotation errors, feed placement errors or dish asymmetry. The strong variation of these angle errors with declination indicates that the major component was due to dish asymmetry (The other components were not expected to vary with declination). These values of actual angle minus the commanded angle were almost a l l negative. This suggests that the feed assembly was rotated too far clockwise (as seen on the sky) by about one degree. These rotation angle errors gave rise to positional errors of about 0.3 23 arc minutes. These positional errors have been neglected since they were less important than those due to the beam separations. The pointing error of the central feed (figures 8,9 and 10) was about 0.5 arc minutes at the declination of the test region. Again, this error was less than that due to the uncertainty in the beam separations and so i t too was neglected. The survey of Condon and Broderick (1985) (with the same telescope but with a different feed assembly and at a wavelength of 21 cm) suffered from extended smooth side lobes of such strength that the effects of Cassiopeia A (Gill.7-2.1) extended to and partially obscured the remnant of G109.1-1.0 several degrees away. Condon and Broderick attributed this to scattering from the N-S feed supports. In this work such effects of Cassiopeia A were less than a degree in extent and so did not affect the mapping of G109.1-1.0. The methods used to remove the effects of gain variations and atmospheric emission removed a l l slowly varying components of the signals. These methods therefore removed the signals which arose from low level side lobes of large extent (located far from the beam centre). Close examination of the beam maps indicated that the major deviations from the assumed model (a gaussian) hear to the beam centre occurred in the East and West directions. These deviations are thought to be due to the coma lobe and appeared as single, weak secondary gaussian peaks. The N-S and E-W cross sections of the central and secondary peaks appeared to have the same half power widths. The ratio of intensity contributions of the central 24 peak to contributions from the secondary peaks in a region of slowly varying extended emission was thus the ratio of the peak heights. Since these low level features of the beams varied rapidly in declination as well as from feed to feed, the uncertainty in the received signal due to contributions from the secondary peaks was estimated by averaging the relative peak heights for a l l calibration sources. The average secondary peak height was 5 per cent of the central peak height. The secondary peaks were never higher than 10 per cent of the central beam height. The crudeness of this analysis was necessitated by the limited beam map data. The effect of these secondary peak contributions was to introduce a limitation upon the dynamic range of the map. Typically this dynamic range limit was 20:1. The intensities of the maps are accurate (where there has been no subtraction of structure) to within 10 per cent of the value given plus 15 mJy/Beam. The 10 per cent figure is from the uncertainties in the gain curve f i t s and uncertainties due to contributions from the coma lobe. The 15 mJy/Beam limit arises from the short term gain variations (see next section). 25 4.Q -3 xn c Z.B o <_i 01 B.B4 -2.B4 ++ + + + 4,B J i n 1 1 1 1 1 1 1 1 1 1 1 1 1111 1 1 1 1 m 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 n 1 1 1 1 1 I W T J | ZB -IB B IB ZB 3B 4B 50 SB 7B F i g u r e 8. P o i n t i n g e r r o r (measured - true p o s i t i o n ) ln r i g h t ascension -for the north ( + ) and south bound (*) c e n t r a l beam. 3B SB i 3B -I : B . B 1. o •3B i -SB 4 -9B -2B 111111 r t - r r p T r T ' T i T r i T T T T r r r i - n - j - r T T i 1 1 1 1 1 1 1 1 1 1 1 i r r q SB 78 -IB 00 10 20 30 40 50 F i g u r e 9. P o i n t i n g e r r o r ln d e c l i n a t i o n f o r the north bound c e n t r a l beam, (measured - true) 2 6 *._ * 1 1 • " " i " ' 11111111111111111111111111| 111 II 11111111111111-; 11111111111111111111 -20 -10 00 10 20 30 40 50 60 70 Declination (degrees) F i g u r e 10. P o i n t i n g e r r o r in d e c l i n a t i o n f o r the south bound c e n t r a l beam, (measured - true) r ow 8 c e n t r a l track F i g u r e 11. A map of the beam of feed 3. Note the m i s s i n g data ln row 3 - these data were not recorded due to c o n t r o l computer problems. Peak shown r e p r e s e n t s 4.36 Jy, 27 J — \ _ J v. f i g u r e 12. A map of the beam of f e e d 1. At the end of each scan a r e s q u a r e wave l i k e a n o m a l i e s - the r e m a i n d e r s of the rows a r e s e t t o z e r o . Peak r e p r e s e n t s 3.26 Jy. J ~ \ . c e n t r a l t r a c k J L J—\_ F i g u r e 13. The map of f i g u r e 12 a f t e r the s u b t r a c t i o n of the best f i t beam model. S c a l e i s t h a t of f i g u r e 12. 28 62 ca CO cr» CO CD CO CO co CO co CO to CO CO CO CO CO CD CO CD (saajSap) u o i i B U i p a a I i i i I i I I I ! I I i i I ' I ' ' I I I I I I I I I I I I I I I I I I I I 1 I 1 I OO'O OT'O OZ'O F tu ~ Or'D 5- 3 OS'O j- £ 09'0 w oro F- OS'O • x P 3 £ > J Q f U | B 6 e u u s i U B punaq m a a u a q i "bT 3 J i i 6 | J 0£ "b psa* J Q * a A u n a uje6 s u u s j u s punoq q j j o u an± • £j aun6|d CO CO CO cr» co CO en en co CO CO CO CO CO CO ca co co co co co CO i i i 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 OO'O r OT'O i oz'O t 0£'0 |- =~ Qr'O !- s o s ' o |- £ o 9 ' 0 w OZ'O F- OB'O 06'0 CO co co cr. co co co co co co co co co CO CO co CO CO i -» co CO C O CO CO CO (S33j6op) u o i ; e i n p a Q -j-t. i i i I i i i i i i i I I i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i - r OO'O =• OT'O j- oz'O 1 g.Ot'O \ ~0r'0 \ s o s - o j - J 09'0 j- W OZ'O |- 08'0 i - 06'Q i i 11 i i 11 i M i i 11 i 11 i i 11 i 11 i i i n 11 i i 11 i 11 i i 11 i 11 i i 11 i 11 i i 11 i 11 i i 11 i 11 i i i D e c l i n a t i o n ( d e g r e e s ) CO CO m CD CO CM CD M CD CD CD CO CD i n CD CD CD CD F i g u r e 18. The north bound antenna g a i n curve f o r feed 5. 0,50 -. 0.80 4 0,70 _ 0.60 J 0,50 ~ i 0.40-5 3 0.30 S -E 0.20 4 0.10 0.00 l i i i i i i i i i i i i I I I I I I | I I I | i l l I I I I I I | 1 I I I I I i I I | I I I I I I I I I | I I I I I I I I I | D e c l i n a t i o n ( d e g r e e s ) CD CD CD CO CO CO CO CD CO CD CD CD CO CO CD i n CO CO CO F i g u r e 19. The north bound antenna g a i n curve f o r feed 6. 31 0.30 0.80 0.70 _ 0.60 J 0.50 ~ 0.40 -S TO 0.30 1 3 0.20 0.10 0.00 I I I I I I I I I 1 I I I I I I I I I I I I T I I I 1 I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I j I I I I I I I I I I D e c l i n a t i o n (degrees) CO CO CD CO co CO CO CO CO CO CO CD CD CO m - - -CO CO CO CO m CO CO c x r o "el- i n F1gure 2 0 . The north (+, s o l i d l i n e ) and south (•, dashed bound antenna g a i n c u r v e s fo r f eed 7 . CO CO CO r ~ I I I I I I I I I I I ' I I I I I I I I I I I I I I I I I I I I I I I I I I I | I I I I I I I I I | I I I I I I I I I | | D e c l i n a t i o n (degrees) CO CO CO CO CO CO CO CO CO CO r-> CO CO CO CO CO i n CO CO CO \0 CO r-F l g u r e 2 1 . The combined (north (+) and south ( O ) bound) antenna g a i n c u r v e f o r feed 7 . 32 •Z paa* J a t i n p j M uieaq s-N punoq g>jou auj. T.Z ajx\S\j tn m co <T> CO CO CO CO CO CO CO CO CO CO CO CO CO (saaj6ap) uor ieutpaf l i I I i i -i i i i i i i I i i i i i i i i i I i i i i i i i i i i i i i i i i i i i i i i i i i i i i i l i i i i i i i i i - i OO'T r-oo'Z E- £ 0Q'£ j- I Q0's 00'9 ro : {A •I paaj. J Q J mp|M uieaq s-N punoq ICJJOU am -zz 3Jn6|j CD CO CO CO in CO CO CO CO CO C-4 CO CO CO CO CO co c o CO CO CO (S33jBop) u o t i B u i p a a • i i i i • • • • I ' ' ' ' ' ' '' ' I ' ' ' •' ' ' ' ' I i ' ' ' ' ' • ' i » • ' ' ' ' i ' ' ' ' ' ' i i i i i i i i i i i i 00 TE- ^ OQ'Z r 5 oo's ro Q0'9 6.00 in CJ 5.00 c ••-41 e u 4.00 u 4 3.00 £ ^ 2.00 -1,00 r i i M 1 1 1 1 1 1 1 1 1 1 1 1 1 II 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 j 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 D e c l i n a t i o n (degrees) CD CD CD CD CD CD CD CM CD CD CD 1-5 CD CD CD CO CO CD in CD CO CD CO F i g u r e 24. The north bound N-S beam width f o r feed 3. 6.00 -| .—* z zx 2.00 £ 4 3= 1.00 i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i | II i i i i i i i | i i i i i i i i i | i i i i i i i i r~| D e c l i n a t i o n (degrees) CD CO CD CM CD CO CD CD CD CO CO CD i n CD CD CD U3 CO F i g u r e 23. The north bound N-S beam width f o r feed 4. 34 CO CO CO •9 p a a ) JO+ m p i « luesq S-N P " ™ q u } j o u a q i "^Z a j n B j J cr» CD CO co e n co CO CD CO CO C-4 CD CO CO CO CO CO I I I • I I 1 I I I I I I I I ( S 0 D J 6 0 P ) U O T l B U I I D D a i I i i i i i • • • i i i i i i i i i i i i i i i i i i i i i i i I •S p a a * J Q ) i ) i p | M uieaq S-N p u n o q i ) i j n u a m •9Z o-» *>- C-J CO CO CD CO CO CD CO CO -- - — - - - - CO CO CO CO CD CO CO CO co CO CD CO CD CD CO CO (soajBop) uo t ;eu ipoQ 11111111111111111111111 00*9 0-to C~l c n c r . H— M t~l *> i n cr . C3 CD CO CD C3 CD CD CD CD CD C3 C3) C3 C3 CD CD CD CD CD c •J It M cr H t CT 9/ Q 3 n I 0 - 3 a CT c* — 3- 3 HP -t. a o ~j — 3 * O It r* a 3-SJ -CT o c 3 Z I Q.eo 10,BB 29.08 I d ro 3B.B8 E 49.BB -Q 5B.BB Half Width (arc minutes) l i l r i i i 11 r . i 11 11 l . I 11 . n i i i iT l l l 11 i l i 11 r t i t l t . i 6B.BB -7B.BB J / 1 Half Width (arc ninutes) H 1 IM I 1 11 1 I I II I 1 I 11 M I H I I I 11 1 I t II II I ll 1 I 1 II I I III CT - i 0 3-c t 3 a 3 • 2 s tn 3-CT -r» + 01 -3 HI I a a. — r* a * 3 a o i — <1 3 <t a. a u \ i o • c 1B.BB 68.BB -7B.BB J 5.00 4.00 DJ 3 -2.00 i.00 - --0.00 i i 11 i i 11 i 11 i i 11 i i i i i 11 i 11 i i i n~i i i i 11 i 11 | i 11 i 11 i i i [ i 11 i i 11 i 11 i i 11 i 11 i Declination (degrees) C D CD CO CD CD CD CD CO CD f-O CO CO CD i n CO CD CD CO CD CD I— F i g u r e 3 0 . T h e n o r t h b o u n d E-W beam w i d t h f o r f e e d 1. 5.00 -q 4.00 01 z> -3.00 2,00 ±E -1 .00 - J 0,00 "] 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 M 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 Declination (degrees) CO co CO CO CD CO CO C-~l CD CO CO l-» CO CO CD CO CO CD i n CD to CD CD CO F i g u r e 3 1 . T h e n o r t h b o u n d E-W beam w i d t h f o r f e e d 2 . 37 5.00 4.00 | 4 3.00 2.00 ^ i.oo = 4 TO 0.00 + 4-I I I I I I 11 I 11 I I I I I I I I I I I I I I I I II I 1 I I I I I I 11 I I 11 I 11 I I 11 I 11 I I I I I I I I I I D e c l i n a t i o n (degrees) CD CD CD CD CD CD CD CD CO CD CO CD CO CD CO CD CO -H CM l-O -tf CD i n CD <J3 CD CO CD r-F l g u r e 32. The north bound E-W beam width f o r feed 3. 5,00 4.00 | -E| 3,00 u 2.00 ±= 1 1.00 0,00 - " 1 1 1 1 1 1 1 1 1 1 i i ' ' i i i 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 D e c l i n a t i o n (degrees) CD CO CO CD CD CO CD CO CD CD CO CD CO —i CM r-> -<J- m F i g u r e 33. The north bound E-W beam width f o r feed 4. co CO CD CD CD CO 33 5.00 -q | I I I I I I I I I I I ' I I I I I I I | I I I I I I I I I | I I I I I I I I I | I I I I I I I I I [ I I I I I I I I I | I I I I I I I I I | D e c l i n a t i o n (degrees) CD CO co CO CO CO CO CD r - l CD CD CD tn co co co wo co CO co F i g u r e 34. The north bound E-W beam width f o r -teed 3. 5.00 4.00 OJ 3.00 u ro 2.00 £ 4 1.00= -3 0.00 _±-I I I I I I I I I I I I I I I I 11 I I I I I I I 1 I I I I 11 I I 11 I 11 I I 11 I 11 I I 11 I 11 I I 11 I I I • I • I D e c l i n a t i o n (degrees) CO CO CO CO CO CO CO CD CD CO CD CD r - l co —i cx fi T in F i g u r e 35. The north bound E-W beam width -for feed 6. CD CD CD VO CO CD CD r~ 39 5,00 =J -c e -1,00= 4 fO — - - • c r s i 11 i 11 i i 11 i 11 i i 11 i 11 i i 11 i 11 i i 11 i 11 [ i 11 i 11 i i i [ 1 1 1 i i 11 i 11 i i 11 i 11 i i | D e c l i n a t i o n (degrees) CO CO CO CD CO CO ^ CD CO CD CM CD CO CO CD CO CO CO CO U3 CD CO F i g u r e 3 6 . T h e n o r t h s o l i d l i n e ) and s o u t h ( • , d a s h e d l i n e ) b o u n d E-W beam w i d t h s f o r f e e d 7 . 3,00 u 4 nj I I I I I I I I I | I i i I I I I I i | I I I I I I I I I | I l I i I I I I I | I I I I I I I I I | I I I I I I I I I | I I I I I I I I I D e c l i n a t i o n (degrees) CO CD CO CO CO CO CD CD CM CD CD CO in CD o CD CD r-F i g u r e 3 7 . T h e c o m b i n e d ( n o r t h (+) a n d s o u t h ( • ) b o u n d ) E-W beam w i d t h f o r -feed 7 . 40 C 1 u -0 H 3-IP Ol 3 c C3 CD C3 CD CD CD S e p a r a t i o n Cart minutes) CD CD 8.80 18.08 r? 20.08 i * 11 i i i 11 I I I I I ' i I I ' ' i ' • • i ' ' • r o -{ t o 38.08 zZ 1 l» » 3 48.88 H 9 3 a cr o u 3 53. BO 60.00 H \ \ 78.BB CD CD CO CO CO CO e u a> O i l l C 01 T Ul ro •o pi 8.88 18.89 2B.G9 =3 33.38 £ S e p a r a t i o n ( a r c minutes) i . i 11 • • i 111 • i 111 11 11. / or n> P 3 43.83 H 0 » 3 58.88 \ \ 68.00 H \ 7B.00 J B . B 8 <X) c V 3 c 91 -I it •a i t * 3 3 Separation (arc ninutes) • i i i i i i • i i i i i • ' i i • i ' i ' i ( i i CO CO I B . 8 0 Sf n cu Z B . 8 8 =3 CL to 3 3 . B O 5-4 B . B 8 H 5 B . B B -3 6 8 . B B H 7 B . 0 B CO CO li) e o —I U3 c 111 CT It 3 CO CO CO CO Separation (arc ninutss) 8 . B B 1 0 . 8 0 r ? ZB.ee 3 8 . 0 0 " 4 8 . 8 B H 5 B . 8 B -H 1 1 1 1 1 ' 1 1 • i 11 111 t. 111 11 111 11 111 r / / 6 8 . 8 0 7 8 . 8 8 3 10.00 - i VI Al -«t-» -=3 -C ••-4 — e -u -u -ro --C -o -•«-« — -ro -L. ro -o. -OJ -7.00 i i i i i i i i i i i i i i i i i i i i i i i-i i i i i i i i i i i i i i i i | i i i i i i i i i | i i i i i i i i i | i i i i i i i i i | D e c l i n a t i o n (degrees) en m CD ca co CO CD C D C M CD CO CO CO CO CO CO CD CD CD <J3 CO CO F i g u r e 42. The angular s e p a r a t i o n of beam 7 and beam 5. 10.00 -, 9.00 | 4 ro <= H 8.00 -2 CJ 1X1 7.00 1 1 1 1 1 1 1 1 1 i 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 D e c l i n a t i o n (degrees) CD CD CO CD CD CD CO CD C D CD CO CO CO CD CD CD CD » H C M r -J - * F i g u r e 43. The angular s e p a r a t i o n of beam 7 and beam 6. CO CO CO M3 CD CO C D 43 1 1 C3 *-— — — _ m CD CD CD C3 CD CD CD i l l C * H 0 3" 1 it cr i It • 31 r . 3 Ci a 3 •» 3 lO It ID T 1 • 3 n B.BB 1B.8B 2B.BB o to -3 -ro 3B.8B S^ 48.8B -58.88 68.B8 H Bean fingle E r r o r (degrees) • 111111111111111111111111111111111111111 70. BB 1 CD CD 1 C3 f-3 rO — — — CD CD CD CD CD CD CD B.BB 18.88 rj -> 11 28.BB ro -7 r 38.! 4e.ee -5B.B8 i 6B.3B H Bean Angle E r r o r (degrees) i n 111 II 11' 111111 m M n 11M111 II 11 7B.B8 4.00 Z.00 0,00 in CJ OJ u •2.0 cn c; e m eg cn -4.00 i i i i i i i i i i i i i M i i i i i i i i i i i i i i i i i i i i i i i i | i i i i i .i i i i | i i i i i i i i i | i i i i i i i i i | D e c l i n a t i o n (degrees) CO CO CO CD CD co CO co CD r-J CO CO CO CO co ID CO CD CD VO CO r— F i g u r e 46. The r o t a t i o n angle e r r o r (measured - commanded angle) •for beam 3. 4.00 -a in OJ OJ 2,00 CT OJ TO O 0.00 t cn •2.0 « e OJ OJ ~4.00 - 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 M 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 i D e c l i n a t i o n (degrees) CD CO CO CD CO CO CD CO CO CO r-l CO i n CO co CD ID CD CD CD r ~ F i g u r e 47. The r o t a t i o n angle e r r o r (measured - commanded angle) f o r beam 4. 45 4.00 -n V) Ol 01 01 - a o.oo £ -q cu « cn -2.0 £ E ro eg c a - 4.0 0 | i i i i i i I i i I i i i i i i i i i i i i i i i i i i i i i i i | i i i i I i. I i q I I I i i i i I i | i i i i i i i I I | D e c l i n a t i o n (degrees) ca co co co c o c o r-3 co m co CD c o co I— F i g u r e 48. The r o t a t i o n angle e r r o r (measured - commanded angle) f o r beam 5. 4.00 CJ Ol cn CJ XJ 0.00 j : cn -2.0 « -3 -4,00 "j 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 D e c l i n a t i o n (degrees) CO CO CO CO CD CM CO CD CO CD CD CO CO i n co co CD CD F i g u r e 49. The r o t a t i o n angle e r r o r (measured - commanded angle) f o r beam 6. 46 0.90 4 0.80 0.70 _ 0,60 J 4 0.50 ~ -E 0.40 -5 - E 0.30 S 1 0,20 0,10 0.00 , i • i i i • < i i i i i i i i i i i i m - r r n i i i i i i r r | i i i i i i i i 11 i i i i i i i i i | i i i i i i i i i D e c l i n a t i o n (degrees) CO co CO CD CO CO CD CD CM CO CD CD r-j co co CD CO CD ID CO \0 CD F i g u r e 50. A comparison o-f north and south bound antenna g a i n s •for feed 3. 0.90 0,80 0.70 _ 0.60 J -I 0.50 w 0.40 .5 0.30 5 0.20 0,10 0.00 • I I I I I M I I I I I I I I I I I I | I I I I I I I I I | I I I I I I I I I | I I I I I I I I I | I I I 1 I I I I I | I I I I I I I I I | D e c l i n a t i o n (degrees) CD CD CO co CD CO CO CD CM CD CD CD CD CO CD CD CD i n CD CD F i g u r e 51. A comparison of north and south bound antenna g a i n s f o r feed 6 47 6.2 Calibration of the Receivers The gains of the 14 channel receivers varied randomly in time and therefore had to be calibrated as frequently as possible. The most d i f f i c u l t problems encountered in this work arose from receiver gain variations. A noise diode was fired in the detector assembly before each scan. The change in output signal of the receiver (Volts) from the noise tube f i r i n g was recorded. The data reduction program obtained the scaling factor (the reciprocal of the receiver gain) by dividing the known noise tube level (Kelvin) by the output level change (Volts). It should be noted that the assumption was made that the output level change was due entirely to the f i r i n g of the noise diode. Although day to day gain variations as high as 10 per cent were encountered, the scan to scan (elapsed times of about 5 minutes) variations were far lower. As an example, the ratio of the scaling factors from one pair of adjacent north bound scans (used in the test map) were: channel scale (scan 1202) scale (scan 1204) rat io noise tube (Kelvin/Volt) (Kelvin/Volt) (Kelvin) 0 12.2751 12.2386 . 9970 5. 1 1 13.0053 12.9949 .9992 5.0 2 14.0690 14.0551 . 9990 5. 1 3 13.9166 13.9011 . 9989 5.2 4 12.8007 12.7973 . 9997 5.2 5 12.8807 12.8914 1.0008 5.4 6 10.8610 10.8705 1.0008 5. 1 7 12.5228 12.4327 .9928 5.3 8 12.4184 12.4186 1.0000 4.6 9 13.4470 13.4537 1.0005 5.2 10 16.9660 16.9730 1.0004 6.2 11 11.3575 11.3418 . 9986 4.6 12 13.0318 13.0116 . 9984 5.6 13 11.8892 11.8954 1.0005 4.9 48 In the above, the gain variations ranged as high as 0.8% (channel 7). This may seem small. Consider however, that the output from each channel consisted o-f about 60 Kelvin due to system noise and spillover radiation (see section 7). The true signal -from an extended structure in the galactic region was typically 50 milliKelvin. A gain variation o-f 0.8% introduced an a r t i f i c i a l signal of 0.5 Kelvin. Clearly this gain variation was important. These gain variations were not smooth. The gains varied randomly between the calibration points. Fortunately, the larger scale gain variations occurred on a larger time scale. The electronic gain calibrations were sometimes incorrect. On some occasions a l l 14 channels were assigned gains that differed from the last values by much the same ratio. It is thought that these gain anomalies were due to interference seen by a l l feeds during the calibration. In the test map region one such occurence left a l l 14 channels with gains 2 per cent lower than their gains in the last scan. As mentioned above, any gain variations gave rise to spurious signals. The noise tube firings allowed the correction for the scan to scan variations in the gain. It was of great importance to find how the gain varied between calibrations of the receiver. Sources of circular polarisation (pulsars and variable stars) are rare and are unresolvable with the 91 m telescope. Thus the mean difference (mean of 30 samples) between the left and the right hand circular polarisation channel output should not vary in time (except for the noise induced variations 49 which in this case are less than 1 mK and insignificant). This mean did vary strongly in time. Since the mean of the result was determined by the gains of the two channels, by seeing how this mean varied, i t was ascertained how rapidly the gains themselves varied. In each of figures 52 to 55 this mean difference was computed for 14 records (corresponding to an elapsed time of 84 seconds, a typical duration for mapping). The 14 values were then plotted (0 to 13) after subtracting from a l l the mean of the f i r s t record. As a very rough measure of variation, the means varied by about 15 mK per record. Each record represented a change in declination of 12 arc minutes and an elapsed time of 6 seconds. For the mapping of point sources the data of one record represented approximately the minimum length base line (base lines are discussed in section 8.4) that could have been used. The gain variation uncertainty over such a base line was 15 mK while the system noise contributed only 3 mK per Nyquist sample. (There were about three 0.2 second samples per Nyquist sample and each 0.2 second sample had a rms noise of ~6 mK) Thus, even for scans through point sources, the gain variations were more significant than the rms noise. 50 F i g u r e 52. S i g n a l s produced by g a i n v a r i a t i o n s d u r i n g scan 2739. mil IMIMI IIIIM i i i i n r j i n T r n i i j T T T n r m p m T i T i ] n m i i i i | M M i i m ] r n T i T n q i i n i i i n ] i n i n r n | T i n T n T T ] Tine (records) CO CO CD co CO r-0 CO CD CO i n co CO CD CD CO CD CO tn CO CO CO CO CO CO CD —I CM CO CD CD r-l F i g u r e 53. S i g n a l s produced by g a i n v a r i a t i o n s d u r i n g scan 2741. 51 ro OJ 'MMnmTiTiniiTiiiiTinmin |iiTiriiii|irnrrrn |nniTiiT|niinrri|inmni]riminipiitnnifniTnm|nrnini| Tine (records) co ro co co co co co co m ' c o c o c o c o c o co co co co co co co co co ro co co co co . . . . CO — I CH OO F i g u r e 51. S i g n a l s produced by g a i n v a r i a t i o n s d u r i n g scan 2980. v-.-.V. Tnmn imi11 |rnrrrnrrii m m IJMIII m m m i n i ipnnTmirnntni-n-rninnymnii i i j i n m i l l ] Tine (records) co co cn co o co CO CO CO CO CO CO CO CO CO CO CO CO CO CO CO CO CO c o c o e o c o cn -H r-J co —< CM ro -«r F i g u r e 55. S i g n a l s produced by g a i n v a r i a t i o n s d u r i n g scan 2982 52 7. NOISE System Temperatures: It was noticed in the early mapping stages and at Green Bank that many o-f the channels varied greatly in performance. As a f i r s t step in measuring the performance of the channels the system temperatures were estimated. Since sources of circularly polarised emission are rare, the left hand polarisation channel output was subtracted from the right hand polarisation channel output for each feed. The system temperatures were calculated for each feed from the standard deviation of this difference. The assumption was made that both channels contributed equally to the noise. To reduce the effects of gain variations this was done for the data of one record at a time (6 seconds). The 140 records used were taken from throughout the observing session. Thus: T B Y S = / ( B T / 2 9 ) X ^ i l ( T I - T A W B I * ) t B : Bandwidth in hertz T : Integration time in seconds (0.2 seconds) T i : The result of subtracting the i t H LH sample from the ith RH sample for the given feed. T A V E : Average of T± over the 30 samples in each record System temperature estimates (subtraction of polarisations)  channels feed temperature variation of temperature (1 stdev) 0,1 1 106 K 25 2,3 2 2 1 9 7 7 4,5 3 70 9 6,7 4 78 28 8,9 5 78 7 10, 11 6 92 14 12, 13 7 85 28 Given such a wide variation of performance, i t was important to isolate the very poor channels. The system temperature can be calculated by performing s t a t i s t i c s upon the results from the 53 subtraction of consecutive samples in a scan devoid o-f sources. Since there were almost 7 samples in a half power beam width, only if there were a powerful compact source would the s t a t i s t i c s be affected greatly. The scans used below were found by inspection of the maps to be devoid of such sources. Thus: T e v e = •(BT/14) x -Aii 2 ( T a i * i -T„)») t B, T : as above T n : n t h sample of given channel output (0 to 29) i : index ranging from 0 to 14 System temperature estimates (Kelvin)  (subtraction of consecutive samples from quiet scans) channel scan 0 1 2 3 4 5 6 7 8 9 10 11 12 13 1200 87 70 72 134 73 63 58 82 66 71 68 67 65 73 1204 89 68 75 134 72 60 65 91 68 76 76 67 70 75 1906 75 64 64 403 62 58 60 87 62 65 78 73 67 69 1910 87 69 71 438 78 64 70 83 61 84 83 72 62 66 2582 88 69 63 157 63 67 61 82 64 71 74 68 62 250 2586 93 70 67 176 71 59 60 89 62 66 87 71 66 74 3214 89 68 64 157 68 59 59 89 61 65 82 93 61 68 3218 84 66 70 157 77 64 64 85 66 69 93 78 64 71 4289 86 65 67 309 67 60 66 85 63 68 80 83 68 67 4293 91 70 75 317 72 65 66 85 66 64 76 81 69 70 mean: 87 68 69 283 70 62 63 86 64 70 80 75 65 70 stdev: 5 2 4 117 5 3 4 3 2 6 7 8 3 3 (The above mean and standard deviations do not include the value of channel 13's system temperature during scan 2582.) 54 Some Observed Characteristics of the Poor Channels: O - Occasional spurious spikes. 3 - Consistently high noise, highly erratic 13 - Occasionally channel appears to have extremely high noise. (See the above Tsys table -for anomalous value) Sometimes this noise appeared to be modulated, (noticed at Green Bank) In figure 56 the extremely high noise of channel 3 is evident. Close inspection shows weak peaks in channel 0 that are not seen in channel 1. Channel 13 shows two anomalous samples which are about 0.2 Kelvin from their neighbours. The best estimates of the channel performances are believed to be the means of the T S Y S values as calculated by the consecutive sample method. 55 J 0.2 K S 10 ^ V - . r ^ _ - - w r . , ^ _ - ^ w J " v - ' ' ^ . - . ^ ' ~ - ' ^ - - ^ . ^ v . v y ^ ^ ^ ^ ^ g ..v»v-~^ ~ ^v> - , - ^ ^ v , w . ^ - ' ^ . - . . , - . - , : ^ - , ; i ^ v ^ w A ; - ^ . ^ . , , v , . ^ , v w _ 5 4 3 CHrtHHEl TIME HUMBER F i g u r e 56. Raw d a t a showing t h e c h a r a c t e r i s t i c s of t h e p o o r l y b e h a v i n g c h a n n e l s (from s c a n 1202) 5 6 8. ANALYSIS 8.1 The O b s t a c l e s : Only r e c e n t l y have r e c e i v e r s been s u f f i c i e n t l y s t a b l e to j u s t i f y t h e i r use i n an a b s o l u t e power mode. The t e l e s c o p e at Green Bank was unique with i t s c o n f i g u r a t i o n of 14 c h a n n e l s i n an a b s o l u t e power mode. The uniqueness of the equipment opened the way f o r t e s t i n g new methods of a n a l y s i s . These methods had t o r e c o v e r the s i g n a l from a background of s e v e r a l forms of c o n f u s i n g e m i s s i o n . These e m i s s i o n s were: Sp i1 l o v e r : S i n c e t h i s p l a n e t i s o p t i c a l l y t h i c k at a wavelength of 6 cm, i t r a d i a t e s as a b l a c k body at a temperature of about 300 K e l v i n . As the t e l e s c o p e was d r i v e n i n d e c l i n a t i o n , the amount of ground r a d i a t i o n seen by the f e e d s v a r i e d . Atmospheric E m i s s i o n : The v a r i a t i o n s i n atmospheric e m i s s i o n a r e due, f o r the most p a r t , t o e m i s s i o n by water vapour (Hayfors 1976). The z e n i t h b r i g h t n e s s temperature due t o t h i s e m i s s i o n i s t y p i c a l l y 5 t o 5.5 K e l v i n ( H a y f o r s 1976). The v a r i a t i o n i n t h i s f i g u r e i s due t o the f l u c t u a t i o n of the water vapour c o n t e n t of the atmosphere. Although the e m i s s i o n v a r i e s , the atmosphere remains o p t i c a l l y t h i n (with a t y p i c a l o p a c i t y of 0.03 dB) ( H a y f o r s 1976) and has a n e g l i g i b l e a t t e n u a t i o n e f f e c t . As can be seen i n f i g u r e s 57 and 59, the atmospheric c o n t r i b u t i o n tended t o v a r y s l o w l y from the bottom t o the top of the map. (Weather f r o n t s and c l o u d s subtended l a r g e a n g l e s as compared t o the t e l e s c o p e beams.) 57 Gain Variations; With a gain variation of A G , a channel output varied by an amount A G X Tmy,m. Because T« y. was so much larger (about 65 Kelvin) than the signal due to a typical galactic source (a fraction of a Kelvin), each small gain variation created a large spurious signal (see receiver calibration section). Other Interferences and Signals Occasionally a l l feeds simultaneously received a very powerful but very short signal. These left artifacts in the shape of the feed pattern that showed strongly on subsequent maps. The signals were so strong that they were easily recognised as spurious (see figure 58). The galactic background emission is the background of thermal and non-thermal emission seen predominantly in and about the galactic plane. The non-thermal component is due to synchrotron emission from cosmic ray electrons and supernova remnants (Westerhout). The thermal component is free-free emission from HII regions. The unresolved component of this background emission has been described as a ridge with a half width of 2 degrees (Altenhoff 1968). This large scale structure will be almost entirely removed by the base line and atmospheric emission removal. The Severity of These Obstacles: As described in the receiver calibration section, gain variations produced strong artifacts. In order to reduce the 58 effects o-f these variations i t was necessary to remove base lines -from the data (this is described in the base line removal section). By doing so, the linear components o-f the con-fusing emissions were also removed. In the test regions examined, the con-fusing emissions varied slowly across the -field o-f the map. As a rough estimate o-f the e-f-fects o-f spillover the findings of Condon and Broderick (1985) were used. With the removal of a base line, only the non-linear components of this variation were of significance. The maximum concavity found in the aforementioned spillover graphs was about 2 mK per degree squared. The maximum deviation of the spillover from this f i t t e d (at the end points) line was then: = (1/2)(2mK per square degree)(2.6 degrees) 2 = 6 mK This was far smaller than the gain variation artifacts. Moreover, this worst case error gradually manifested i t s e l f across the map. (The gain variation artifacts varied far more rapidly). The powerful interference seen in figure 58 ( f i l l e d hexagonal patterns) was removed by requiring that a l l variations be consistent with a half power beam width of about 3 arc minutes. The interferences seen in the figure involved signal changes of about 0.4 Kelvin for a duration of one sample (a half power beam width in declination consists of about 7 samples). These values were replaced by interpolation. An indication of the relative significance of these spurious emissions is now given. When mapping the region about G109 (whose angular extent was roughly half a degree), a typical contour 59 interval was about 25mK (or about 50 mJy). The hemispherical structure o-f 6109 was visible at about this sensitivity. Superimposed upon this 25 to 50 mK signal were typically: ~200 mK - Variation in the spillover and atmospheric emission. ""100 mK - Variations due to gain -fluctuations. ~400 mK - Very strong, brief signals appearing in a l l channels. With the removal of base lines, the gain fluctuation and spillover emission contributions are much reduced. This leaves typical superimposed spurious signals of: ~200 mK - Variation in the atmospheric emission. ~6 mK - Variation in spillover emission, (maximum) ""15 mK - Due to gain fluctuations (see base line section) ~400 mK - Very strong, brief signals appearing in a l l channels. The data, with base lines removed, but with these spurious emissions remaining, are plotted in figure 58. 60 61 1 ° .2 * 10 • U** * » r * (scan 27 p i g u r e seen e M 8.2 An Overview o-f the Analysis The data, in the original format (appendix A) were read in (by the 'fisherman' program) from either tape or disk. Any data pertaining to the region to be mapped were collected by this program. The Masscomp computer at Green Bank wrote the floating point data using the IEEE floating point format whereas the uVax uses the DEC floating point representation. These collected data were translated to the DEC floating point format (appendix B) and it s positional information was converted from epoch 1950 to epoch 1986 coordinates. Once translated, the receiver gain calibration information (see section 6.2) in the data records was used to scale the amplifier outputs to units of Kelvin from the original units of Volts. After these procedures were carried out the data for one map were written to a relatively small disk f i l e (several megabytes in size). The data in this small disk f i l e of scaled, translated and precessed data were then read by the gridding program into arrays of 160 columns and 480 rows, where each column consisted of 480 consecutive samples in a scan. Each row represented an elapsed time of 0.2 seconds. The portion of the signal array that was actually displayed in the maps was slightly smaller and covered a region of sky of 2.56 degrees of declination by 2.65 degrees of right ascension. The array rows and columns were approximately aligned with the right ascension and declination axes respectively. There were three such arrays, two recorded the positional information and the third recorded the received signal i t s e l f . The received signal was calculated by averaging the two 63 polarisations of each feed (except for feed 2 - the performance of channel 3 was so poor that i t was not used). A fourth array (160 x 480) of integers recorded which feed contributed which sample. As described in the next section, base lines were determined and subtracted from each column of the signal array. The four arrays were then written to disk in binary form (total f i l e sire was about 1 megabyte). The next stage of the analysis was the removal of the atmospheric emission. The 'deatmosphere' program f i r s t used each beam's gain calibration information to scale (see section 6.1) the signal information from units of Kelvin to units of Jansky per beam area. The program then applied one of the algorithms described in section 8.5 or 8.6 to remove the atmospheric emission. In order to display this data a set of contour plotting u t i l i t i e s was developed. These u t i l i t i e s displayed the data upon either a plotter or upon any Tektronics 4010 compatible graphics terminal. These programs read the data to be plotted in a standard format of 160 x 160 arrays. To produce these smaller arrays either the condensing or convolution program was used. The convolution program read i t s data in the larger array (160 x 480) format and convolved this data with a normalized (see section 8.3) isotropic gaussian of specified half width. It then wrote arrays of 160 columns and 160 rows. The written array then contained columns separated by half a half power beam width (a Nyquist interval). Each row represented an elapsed time of 0.6 seconds, which corresponded to a change in declination of slightly 64 less than a Nyquist interval. The condensing program read and wrote arrays in exactly the same format as the convolution program. The only difference was that the condensing program averaged 3 consecutive 0.2 second samples for each 0.6 second sample of the smaller array. This far faster program was useful for obtaining quick interim pictures in the development of maps. 65 8-3 The C o n v o l u t i o n Program The a p p l i c a t i o n of the c o n v o l u t i o n program was r e q u i r e d i n o r d e r t o reduce the e f f e c t s of the s h o r t term g a i n v a r i a t i o n s and n o i s e . In t h i s s e c t i o n the e f f e c t s and uses of t h i s program are d e s c r i bed. The e f f e c t of the c o n v o l u t i o n was to i n c r e a s e the apparent beam h a l f width s l i g h t l y . To show t h i s we c o n s i d e r the f o l l o w i n g example < For n o t a t i o n a l s i m p l i c i t y , the example i s one di m e n s i o n a l [ i n z1, however, a l l the s t e p s are e q u a l l y v a l i d f o r any number of d i m e n s i o n s ) . D e f i n e : ( n e g l e c t i n g n o r m a l i z a t i o n s ) B(z) - B r i g h t n e s s of the sky at the p o s i t i o n z. F - An o p e r a t o r r e p r e s e n t i n g the F o u r i e r t r a n s f o r m o p e r a t i o n . # - An o p e r a t o r r e p r e s e n t i n g the c o n v o l u t i o n o p e r a t i o n . S(z) - S i g n a l r e c e i v e d with the beam i s c e n t r e d upon p o s i t i o n z. R(z) - A g a u s s i a n beam model, R(z) = exp( - ( Z / K ) 3 5 ) where: (2« x /ln2) i s the f u l l width at h a l f maximum of the beam. G(z) - The g a u s s i a n a p p l i e d by the c o n v o l u t i o n program, G(z) = exp( - ( z / j ? ) 2 ) where: (2j9 x y\ n2) i s the f u l l width at h a l f maximum of the g a u s s i a n . B e f o r e the a p p l i c a t i o n of the c o n v o l u t i o n program we had the s i g n a l : S(z) = B(z) * R(z) A f t e r the a p p l i c a t i o n of the program we had the new s i g n a l N ( z ) , N(z) = < B(z) * R(z) ) * G(z) By the use of the c o n v o l u t i o n theorem: FL N(z>3 = Fl B(z) * R(z) 3 x FL G(z) 3 = Fl B(z) 3 x Fl R(z) 3 x Fl G(z) 3 but FL R(z) 3 « («^n) x exp< - k a x « z / 4) 66 and Fl G(z) 3 = ( J9/TT) x exp < -k z x Px I 4) so that, Fl N<z) 3 = /=•£ B(z) 3 x («j9n) x exp ( -k z x ff* / D where o"z = + a^s thus, N(z) = iK?SW7c) x B<z) * exp( -(z/cT)*) Since the beam half width was about 3.5 arc minutes and the convolution program used a gaussian of 2.5 arc minutes, the resulting signal, N<z), was that which would have been produced (neglecting normalizations) had the telescope had a 4.3 arc minute beam half width. The convolution program normalized the gaussian with which it convolved by dividing by the weighting integral: W(z) = exp( -((z - Zi)/j9)=) x D ( Z i ) dzx = (if f u l l y sampled) where: D(z) = 1 if the telescope took data in the region containing z D(z) = 0 otherwise This normalization allowed the program to deal with missing data by introducing a form of extrapolation. When the normalization integral was large, there was l i t t l e extrapolation. When this integral was less than a specified amount, the program set the point about which the integral was evaluated to the flag value of -999.0. In this manner the program flagged regions which were not adequately sampled. For the maps in this test region, a c r i t i c a l value of 10.0 was used. This value would only have been exceeded if Nyquist sampling had been maintained. No points in the test region were flagged as undersamp 1ed. The normalization affected the intensities. The intensities 67 a f t e r the c o n v o l u t i o n stage were g i v e n by: K z ) = N(z)/W(z) With N y q u i s t sampling W(z) = B^Wf thus: K z ) = ( («J9/TT)/cT) x (J9/TI)-* x B(z) * exp( -<z/<5)a) <«/cT) x B(z) * exp< - ( z / f f ) 3 5 ) For r e g i o n s of extended s t r u c t u r e s such t h a t B(z) = B 0, I ( z ) = («/cT) x Bo x o"-/if = BOK^TT w h i l e the o r i g i n a l s i g n a l ( b e f o r e c o n v o l u t i o n ) was: S(z) = B(z) * exp( -(z/«) z) = BO«-^TI S i n c e I ( z ) / S ( z ) = 1, f o r such extended s t r u c t u r e t h e r e w i l l be no e f f e c t . For u n r e s o l v e d ( p o i n t ) s o u r c e s the i n t e n s i t i e s were: I ( z ) = («/cT) x B o S ( z - Z o ) * exp( - ( ( z - z 0 ) / cl) ») = B 0(K/CT) x exp ( -( ( z - z 0 ) / ( J ) z ) as compared t o the o r i g i n a l s i g n a l o f : S(z) = Bo x exp( -(z/«) a) Thus, f o r p o i n t s o u r c e s , the peaks a f t e r the a p p l i c a t i o n of the c o n v o l u t i o n program were a f a c t o r of («/c7) ( f o r each dimension) lower than those of the o r i g i n a l s i g n a l . For p o i n t s o u r c e s i n two dimensions (E-W, N-S) the p l o t t e d i n t e n s i t i e s s h o u l d be m u l t i p l i e d by the c o r r e c t i o n f a c t o r : (<xBW, « n a a r e the beam h a l f widths) C = ( « E W « M 0 ) ~ X ( ( « E W Z + 2 . 5 * ) ( « N O z + 2 . 5 Z))o-= ~ 1.53 f o r the t e s t r e g i o n . The u n c e r t a i n t y i n C due t o v a r i a t i o n s i n K b u , * n s from beam to beam ( s e v e r a l per cent) was s m a l l e r than the u n c e r t a i n t y i n g a i n (ten per cent) and was t h e r e f o r e n e g l e c t e d . 68 8.4 Base Line Removal As mentioned above, the gain variations were erratic. An excellent technique -for removing the linear portion of the gain variation artifacts was to remove base lines -from the data immediately after the gridding process. Although this removed a l l galactic structure which was linear on a scale of 2.6 degrees (such as the galactic background emission), i t also removed the linear components of the atmospheric and spillover emissions as well as the spurious signal produced by the linear component of the gain variation. The base lines were estimated in the following manner. A median (of intensity) was determined in the top and bottom f i f t h (100 points) of each column of 480 rows. The line drawn between these two points is referred to as a base line. By subtracting the line defined by these two medians, the base line was determined and removed. Although i t was obvious that i t was necessary to subtract base lines, i t was not clear whether i t was best to subtract them before or after the removal of the atmospheric emission. To il l u s t r a t e the difference between the two p o s s i b i l i t i e s , consider the situation in which the feeds were tracked across a small (less than 1 degree across) plateau of emission located at the centre of the map (such that feed n was not tracking the plateau), with smooth gain variations affecting the 14 channels. 69 Def i ne: Si - apparent signal in -feed i Ti - true signal in feed i ( with T„ = O ) A,C,D,E... - the various terms describing the variation of the atmospheric emission with declination. Fi - the linear term of the gain variation in feed i z - change in declination from the start of the scan Then, the raw data (before the removal of base lines and atmospheric emission) consisted of the six signals (of feeds tracking the plateau) : Si = Ti + Az + Fiz + C + Dz* + Ez 3 + (i 5* n) with the signal of the feed which was not tracking the plateau: S„ = Az + F„z + C + Dz 2 + Ez 3 + In the original method of removing the base lines, the atmospheric emission was f i r s t removed and then the base lines were estimated and removed. Consider now the effects of such a procedure. To remove the atmosphere emission, the minimum signal was subtracted from a l l the others. The minimum signal should have been S„ (the only feed not tracking the source) but a gain variation could have made any other feed the minimum. Let us assume that such a gain variation occurred. This gain variation resulted in the wrong signal (designated S m) being subtracted from a l l the others, thereby giving: Si = Ti - J„ * (Ft - F m ) z (i s* n) S„ = - T m + (Fi - F„)z The base lines were then removed. Let BCS tl represent the result of applying the base line removal algorithm to the signal of feed i . Applying the algorithm to the seven signals: BC SiD = BE Ti - T m + (Ft - F m)z ] Since the base line removal algorithm was linear: 70 BC S t3 = BC T» - T m + ( F i - F m ) z 3 = BC T i 3 - BC T m3 + BC< F, - F m )z3 = BC T i 3 - BC T m3 The base lines were determined well away -from the structure, so, BCTi3 = T i . Thus giving the -final signals: BC S t3 - Tt - T m Since the above equation represents the subtraction o-f structural signals ( T i ) , this procedure has given rise to structure removal and has generated spurious structure. The atmospheric emission and base line removal procedure was improved by applying the base line removal stage -first. Consider now the effects of this improved procedure. Since the base line removal algorithm was linear: BCSi3 = BCTt +• F i z + C + Az + D z z + E z 3 + ....3 = BCTt3 + BCF»z3 + BCC + Az + Dz* + E z 3 + 3 but BCF»z3 = 0 , and BCT43 = Tt (as above) so that BCSt3 = T, + BCC + Az + Dz 2 + Ez 3 + ... 3 and BCS„3 = BCC + Az + D z a + E z 3 + ... 3 When the data contained a large and rapidly varying atmospheric signal, the BCSil tended to look rather non-physical. However, by this method BCSn3 was preserved as the minimum (since the T i were never negative). In the subsequent atmospheric emission removal stage the minimum of the BCSi3 was subtracted from the other BCSt3. By subtracting the minimum, BCS n3, from each of the seven signals B C S i 3 , the set of signals: BCSt3 - BCS„3 = T t 71 were constructed. These signals represent the true signals without the spurious emissions. In summary, the base lines were removed be-fore the application of the atmospheric emission removal algorithm. If they were not removed f i r s t then the gain variations confused the atmospheric emission removal algorithm. In practice i t was found that the use of this procedure improved the sensitivity of the maps by a factor of two or more. As an example, consider figure 57. The downward trend of the data was due to the slow decrease in atmospheric emission from left to right. (It is apparent that the emission must have been due to atmospheric emission when one considers that the data of the nearby scan shown in figure 56 had no such trend. This nearby scan was less than half a degree away in right ascension and covered the same declination range as the scan shown in figure 57) Since channels 10 and 11 were on the same feed, the widening gap seen in figure 57 (from left to right) was due to a gain variation. When base lines were removed, the slow gain variation seen in channel 10 (of figure 57) was removed as well as the steady downward trend visible in a l l channels. After the base line removal the data resembled the data shown in figure 56, essentially f l a t and with most of the atmospheric emission and gain variation effects removed. The f i t t i n g of base lines to the offending channel's signal removed the typical gain variations down to the 25 mK level. These relatively extreme variations were quite rare. Because of this rarity, the gain variations were susceptible to smoothing 72 operations. After the data were convolved with an isotropic gaussian with a half power beam width of 2.5 arc minutes, the variations were smoothed to about half their original magnitude. Convolving with a gaussian Df half width much smaller than that of the telescope did not greatly affect the resolution (see section 8.3), the main effect was to smooth out much of the noise and spurious gain variations (see section 6.2). Thus, i t was possible to map extended structure down to sensit i v i t i e s of about 25 mJy/Beam. The net effect of baseline subtraction was to reduce the effects of gain variations and spillover emission. (Although base line removal often removed much of the atmospheric and galactic background emissions, these emissions were wholly removed by the subsequent application of the atmospheric emission removal algorithm.) Since the gain variation artifacts were so severe, base line removal from the raw data led to a great improvement in sens it i v i ty. 73 8.5 Some Simple Atmospheric Emission Removal Methods Several methods -for removing the non-linear components of the atmospheric emission were developed and applied. These methods were a l l presented with the same problem of how to distinguish atmospheric emission from the signal due to galactic structure. The characteristics of atmospheric emission were that i t varied slowly and was seen by a l l feeds to an equal extent (see figure 59). The problem was that whatever c r i t e r i a were applied to remove the atmospheric emission also removed that component of the galactic structure that f u l f i l l e d the same c r i t e r i a . The feed average subtraction was the f i r s t method developed. This method serves as a starting point for understanding the more complicated methods to follow. In this method, from each feed's signal was subtracted the average of a l l seven signals. The strong point of this method was that i t was relatively immune to signal variations due to noise, gain variations and spurious emission. The great weakness of this method was i t s lack of selectivity. As an example, when any beam tracked across an strong isolated point source the resultant map showed a feed pattern of six weak negative regions and one strong positive region. This method tended to remove structure strongly. For instance, i f a l l seven beams tracked a plateau of emission then a l l the structure was removed. In order to prevent the artifact producing behaviour mentioned above, the feed near-mode subtraction method was developed. For this method, an average and a standard deviation were found for the signals of the seven feeds. A second pass 74 average was then made up of a l l the signals which differed from the f i r s t average by less than a specified number of standard deviations. This procedure excluded from the second pass average the values of any feeds which tracked a strong source. By subtracting from a l l the feeds this second pass average, the atmospheric emission was removed. The artifacts produced by this method were weak and easily recognised. If most of the feeds were upon an extended source, (for simplicity assume that the extended source was a plateau) then those that were off the source were assigned a negative value. These artifacts were usually shallow negative regions in the vi c i n i t y of an extended structure. The method did not produce the artifacts seen with the feed average subtraction method about an isolated strong point source. In the feed minimum subtraction method the smallest signal seen by any of the seven feeds was subtracted from a l l the feeds. The assumption was made that the atmospheric emission contribution to the signal was equal in a l l feeds and that the minimum signal had no contribution from galactic structure. This method worked f a i r l y well but subtracted galactic structure which appeared in a l l feeds at once. Any galactic structure whose diameter was greater than about 16 arc minutes was likely to suffer the removal of valid structure. This method subtracted far less structure than either of the earlier methods. An unfortunate side effect of this method was that i t tended to increase the effective noise. By choosing the minimum we were most sensitive to the noise. This term, effective noise, refers 75 not only to the rms noise but also to the low frequency noise produced by gain variations. With base line removal and a convolution stage this was not an important point but had base lines not been removed or had the map size been considerably larger then the use of this method may have been precluded. A method much like that used in the 21 cm survey work of Condon and Broderick was also implemented. Consider now the set of data points from one feed alone. By subtracting from each point the median value determined from the neighbourhood about that point, the slowly varying atmospheric emission was removed. This method also worked f a i r l y well. It tended to remove structure which spanned the same declination range as did the neighbourhood. For example, with a neighbourhood of 100 points (which span a declination range of 0.66 degrees), any component of the galactic structure which was about half a degree or larger in extent in declination, suffered the removal of valid structure. ' Each of the above methods estimated the signal due only to atmospheric emission and then subtracted this estimate, the reference signal, from a l l the feeds. The common factor in these methods was the compromise between structure removal and sensitivity. When the region over which the reference signal was determined was much larger than the size of any extended structure in the f i e l d , these methods did not remove valid structure. However, as the region was made larger, the uncertainty in this reference signal also became larger. This uncertainty arose from gain variations and variations in the spurious emissions. 76 8.6 The Hybrid Method: The hybrid method o-f atmospheric emission removal was a combination of the last two methods. The f i r s t step of this method was the removal of base lines. Then, for the signal of each feed the running medians were calculated in a neighbourhood of specified size. The smallest of these medians was then subtracted from a l l the feeds. The strength of this approach was that the region over which the reference signal (the smallest of the medians) was determined, was large (usually much larger than the regions used by the methods mentioned in the last section). Moreover, this region was chosen in such a way that the uncertainties due to gain variations and variations in spurious emissions were minimised. By varying the size of the neighbourhood used to calculate the medians, the amount of structure removal was controlled. If the neighbourhood extended across 1 degree, then any structure smaller than half the neighbourhood, 0.5 degrees, was left intact. Of course this was only a relevant consideration for sources larger than the angular separation of the beams in the sky, about 16 arc minutes. Any structure smaller than this size could not have been tracked simultaneously by a l l seven feeds. For such a structure there would therefore have been no structure removal. As the neighbourhood was extended, the sensitivity of the resultant map decreased due to fluctuations in the atmospheric emission and gains. For example, when a neighbourhood of 1 degree was used, then any atmospheric emission variations on a scale much less than 1 degree appeared as real structure. Such maps had a 77 much larger uncertainty in their intensity levels. Weak negative regions appeared in the regions of the map where variations in the atmospheric emissions (or gain variations) were too strong and erratic for a neighbourhood o-f the given size to have suppressed them. The typical depth of the negative regions is approximately the minimum uncertainty in the intensities. (For the total uncertainty the additional effects of the 10 per cent uncertainty in the gain curves should be considered, see section 6.1) Editing of the data permitted the sensitivity of the hybrid method to be improved. This editing was done with a series of base line removal u t i l i t i e s which allowed the operator to remove slow variations due to gain changes or varying atmospheric emission. These u t i l i t i e s had l i t t l e effect upon the performance of the hybrid method when applied with a small neighbourhood. However, when a large neighbourhood was used the u t i l i t i e s were required in order to obtain sensitivity comparable to that given with a small neighbourhood. The hybrid method preserved structure of maximum sizes ranging from 16 arc minutes to about 1 degree. It provided an indication, in the form of negative bowls, of i t s sensitivity. 78 9. Results The vast majority o-f this work was per-formed using only the north bound data. The test region was also mapped using the south bound data. These south bound data were so heavily corrupted by noise that the resultant maps were only included here for completeness, (see Appendix C) The -following discussion deals with the north bound mapping of the test region. The hybrid method was applied at the two extremes of its a b i l i t y . For a neighborhood (n) of 5 points (a half power beam width corresponds to about 7 points), the method produced i t s greatest sensitivity but also removed the most structure. For a neighbourhood of 200 points the sensitivity was poorer. The feeds tracked the supernova remnant of 6109 in the worst possible manner they were centred upon i t . Since the set of seven beams was simultaneously tracking the supernova remnant when so centred, with a small neighbourhood the structure of the remnant was strongly removed. The test region was therefore well suited for these demonstrations of structure removal. The main measure of structure removal used here was the antenna temperature at reference point 1 (figure 60) in the hemispherical structure of G109. An existing map indicated an intensity of about 125 mJy/Beam for this point (Braun 1981)(see f igure 60). For each of the following examples, the hybrid method was applied with the specified size of neighbourhood. The resulting data were then convolved with a gaussian of half-width of 2.5 arc mi nutes. 79 RIGHT ASCENSION(HRS. .MINS. .SECS. ) Figure 60. A map of the test region produced by Braun (1981). The output o-F the MEM was convolved with a gaussian of half widths 3.9 and 3.0 arc minutes in declination and right ascension respectively. (Convolved with the telescope beam) Contours are labelled in mK at the intervals: 15, 30, 50, 70, 100, 120, 200, 300, 500, 700 mK which correspond to: (conversion factor of 0.8 K/(Jy/Beam)) 19, 38, 63, 88, 125, 150, 250, 375, 625, 875 mJy/Beam 80 The map shown in figure 61 was produced by applying the hybrid method with a neighbourhood o-f 5 points. Considerable structure has been removed. The intensity at reference point 1 was approximately half the intensity given by Braun 1981 (figure 60). The same map without contour labelling has been replotted as figure 62. By specifying a neighbourhood of 200 points, the base lines were determined over a declination range more than twice as large as the supernova remnant. The method therefore did not remove structure from the plateau-like remnant (see figure 63). Reference point 1 has an intensity of about 125 mJy/Beam in agreement with the published map. By extending the neighbourhood so far, the method was very vulnerable tD gain variations and atmospheric emission variations. The application of the enhanced base line removal u t i l i t i e s (described in section 8.6) was required. This vulnerability (weakened after the application of the above u t i l i t i e s ) can be seen by the relatively frequent occurrence of deep negative regions in figure 63 as compared to figure 61. There was strong atmospheric emission upon the day that the data in these regions were taken. The map of figure 63 was replotted without contour labels as figure 64. Braun (1981) found the peak of S152 (shown in figure 60 as reference point 2) to be 0.753 K which corresponds to 0.941 Jy/Beam. This agrees very well with the values from figures 61 and 63 of 0.937 and 0.952 Jy/Beam respectively. The map produced by Braun (figure 60) has approximately the same resolution as the maps of figures 61-64. 81 Figure 61. A map o-f the test region produced by the hybrid method with a neighbourhood of 5 points (structure has been removed). Contours are at -50, -25, 25, 75, 125, 200, 500 mJy/Beam with corresponding labels of -2, -1, 1, 2, 3, 4, 5. The galactic longitudes of 1=108, 109 and 110 degrees and latitudes of b = -2, -1 and 0 degrees are plotted also. 82 Figure 62. This plot is that of figure 61 except that the contours are not labelled. A map of the test region produced by the hybrid method with a neighbourhood of 5 points (structure has been removed). Contours are at -50, -25, 25, 75, 125, 200, 500, 900 mJy/Beam. The galactic longitudes of 1=103, 109 and 110 degrees and latitudes of b= -2, -1 and 0 degrees are plotted also. 83 Figure 63. A map o-f the test region produced by the hybrid method with a neighbourhood of 200 points. Contours are at -50, -25, 25, 75, 125, 200, 500 mJy/Beam with corresponding labels of -2, -1, 1, 2, 3, 4, 5. The galactic longitudes of 1=108, 109 and 110 degrees and latitudes of b= -2, -1 and 0 degrees are plotted al so. 84 Figure 64. This map is that o-f figure 63 except that the contours are not labelled. A map o-f the test region produced by the hybrid method with a neighbourhood o-f 200 points. Contours are at -50, -25, 25, 75, 125, 200, 500, 900 mJy/Beam. The galactic longitudes of 1=108, 109 and 110 degrees and latitudes of b = -2, -1 and 0 degrees are plotted also. 85 The agreement between figure 64 and figure 60 is very good. Within the uncertainties set out in the calibration section, the maps show the same peak intensities and extended structure. In figure 65 the map produced for figure 61 is replotted with much smaller contour levels. The lowest positive contour level is 10 mJy/Beam. An interesting structure is seen extending from G109 towards the point 1=109.5, b= -1.5 . Although this structure is very weak, i t s occurrence in a region of l i t t l e uncertainty (no nearby negative regions) and i t s extent argue in favour of its validity. An additional point of interest is that this structure lies opposite the structure which extends from G109 approximately along the line 1=109. The map of figure 63 is replotted with much smaller contour levels as figure 66. (The contour levels are those of figure 65.) Since this figure was produced with the hybrid method with a neighbourhood of 200 points, the extended structures seen in figure 65 should not be attenuated. The negative regions are more frequent and intense, than in figure 65, thereby indicating that the intensity uncertainties are larger. The weak structure seen in figure 65 is now far more extensive and is less attenuated. Using the point source correction factor (C as described in section 8.3) the peaks of figure 61 are (within 10 per cent): Peak Jy 1 (degrees) b (degrees) 1 1.43 108.76 -0.95 2 0.57 108.37 -1.05 3 0.96 107.95 -1.51 4 0.40 108.96 -1.22 It has been assumed that each of these peaks arises from one isolated point source. This is an invalid assumption in the case 86 Figure 65. A map of the test region produced by the hybrid method with a neighbourhood of 5 points. Contours are at -35, -15, 10, 30, 50, 125, 175, 250, 500 mJy/Beam with corresponding labels of -2, -1, 1, 2, 3, 4, 5, 6, 7. The galactic longitudes of 1=103, 109 and 110 degrees and latitudes of b= -2, -1 and 0 degrees are plotted also. 37 Figure 66. A map of the test region produced by the hybrid method with a neighbourhood o-f 200 points. Contours are at -35, -15, 10, 30, 50, 125, 175, 250, 500 mJy/Beam with corresponding labels of -2, -1, 1, 2, 3, 4, 5, 6, 7. The galactic longitudes of 1=108, 109 and 110 degrees and latitudes of b= -2, -1 and 0 degrees are plotted also. 88 o-f peak 1 which is composed o-f the HII regions S152 (with an intrinsic size o-f 40 arc seconds) and S153 (Gregory, Fan 1 man 1981). Hughes et a l . -found that the spectrum o-f the remnant o-f G109 was well described by a spectral index of 0.5. Their f i t to the spectral data predicts a flux density of 9.1 Jy at a wavelength of 6 cm. From figure 66 the integrated flux of the remnant G109 (of a l l points inside the contour of 50 mJy/Beam, this includes only the hemispherical portion of the structure) was 8.7 Jy (within 10 per cent). This value is in excellent agreement with the spectral data and f i t presented by Hughes et a l . The convolution program produced maps with points spaced slightly closer than the Nyquist spacing. This resulted in the apparent fluxes being too high. This effect has been corrected for. 89 10. SUMMARY AND CONCLUSIONS A method has been developed that will permit the production o-f an atlas of galactic radio emission from the Patrol's 1986 survey data. The sensitivity and quality of the maps so produced are roughly comparable to those produced with the MEM method by Braun (1981). Based upon the results of this work I recommend the use of the Hybrid method with a neighbourhood of 5 samples. The maps from this method should be contoured with smallest intervals of about 15 to 50 mJy/Beam (depending upon data quality). By way of comparison, the best MEM map produced by Braun (1981) had a smallest contour spacing of 19 mJy./Beam. The sacrifice made in sensitivity and time in order to retain a l l structure in the supernova remnant of G109 is high. Maps made with such a neighbourhood (200 points) are subject to variations in gain and spurious emissions that must be removed by editing in order to achieve a sensitivity comparable to that achieved with a neighbourhood of 5 points. These large neighbourhood maps take about 1 day to generate. It is therefore reasonable to map the survey region with a maximum of sensitivity and a small loss of structure. The atlas so produced can be followed up by individual mappings of regions of interest using a larger neighbourhood. Such small neighbourhood maps take about 1 hour to generate. If extended structure is present then the next mapping could use a neighbourhood large enough to retain the structure. During the second mapping, the operator would have a clear idea of what he was trying to retain. 90 This method will now be developed into the program for the production run of the 1986 6 cm atlas. This atlas will complement in coverage the surveys of Altenhoff et al (1978) and Haynes et al (1978) and will be several times more sensitive. 91 References Haslam, C.G.T., Salter C.J., Stoffel, H., Wilson, W.E., 1 9 8 2 Astron. 8c Astrophys. Suppl. Ser. 4 7 , 1 - 1 4 3 Berkhuijsen, E.M., 1 9 7 2 : Astron. tk Astrophys. Suppl. Ser. 5, 2 6 3 Reich, W., 1 9 8 1 : Astron. & Astrophys. Suppl. Ser. 4 8 , 2 1 9 Condon, J.J., Broderick, J.J., 1 9 8 5 The Astronomical Journal 9 0 , 1 2 - 2 5 4 0 Haynes, R.F., Caswell, J.L., Simons, L.W.J., 1 9 7 8 Australian J. of Physics Astrophysical Suppl. 4 5 - 3 5 1 Altenhoff, W.J., Downes, D., Pauls, T., Schraml, J., 1 9 7 8 Astron. Bt Astrophys. Suppl. 3 5 . 2 3 Nityananda, R. , Narayan, R. , 1 9 8 2 : J. Astrophys. & Astr. 3, 4 1 9 Cornwell, T., 1 9 8 3 : VLA Scientific Memorandum 1 4 9 Steer, D.G., Dewdney, P.E., Ito, M.R., 1 9 8 4 Astron. & Astrophys. 1 3 7 . 1 5 9 Tiuri, M.E., 1 9 6 6 in 'Radio Astronomy', Kraus, J.D., McGraw-Hill Altenhoff, W., 1 9 6 8 , in 'Interstellar Ionised Hydrogen', Terzian, Y. Fisher, J.R., Payne, H.E., 1 9 8 2 : NRAO Engineering Memo No. 1 4 8 Braun, R. : MSc 1 9 8 1 , Physics, University of British Columbia Picha, J. : MSc 1 9 8 6 , Physics, University of British Columbia Tor Hayfors, 1 9 7 6 in ' Methods of Experimental Physics ', Vol 1 2 , Part B ed. M.L. Meeks, Academic Press Gregory, P.C, Taylor, A.R., 1 9 8 1 , Astrophysical Journal 2 4 8 . 5 9 6 Gregory, P.C, Taylor, A.R., 1 9 8 6 , Astronomical Journal 9 2 , 3 7 1 Gregory, P.C, Fahlman, G.G., 1 9 8 1 , Vistas in Astronomy, 2 5 , 1 1 9 Taylor, A.R., Gregory, P.C, 1 9 8 3 , Astronomical Journal 8 8 . 1 7 8 4 Hughes, V.A., Harten, R.H., Costain C.H., Nelson, L.A., Viner,M.R. 1 9 8 4 : Astrophysical Journal 2 8 3 . 1 4 7 9 2 APPENDIX A: The Data Format All data on tape is kept in i t s original -Format. The C definition of this format follows. /* DATA DEFINITIONS FOR SECTION C - PERMANENT DATA STORAGE */ /* data ttdef i ne ttdef i ne ttdef i ne ttdef i ne /* ttdef i ne /* ttdef ine ttdef i ne ttdef i ne ttdef i ne ttdefine ttdefine block array sizes */ FSDATASIZE 8192 /* tt data values fast sample */ C0NTDATASI2E 540 /* * data values continuum */ CONTPOSCOOR 2 /# tt position coord continuum *7 CONTDATASET(N) (CONTDATASIZE/(N+2)) tt data sets continuum #/ FSDATASET(N) <FSDATASIZE/N) tt data sets fast sample #/ ACCHAN ACCNTR SADUMP PHABLKCNT DES4BLK DES16BLK 384 /•* tt channels auto-corr *•/ 8 /# tt counters auto-corr •*/ 1024 /* tt data points signal avg */ 4 /* tt phase blks in header */ 4 /# tt descrip blks in header #/ 16 /•* tt descrip blks in header -X/ /* STANDARD HEADER DATA STRUCTURES */ /•X Basic Information -X/ typedef struct C float head_len, data_len, scan; char obs_idC83, observer!163, telescop C83, proj _ i d I 83, sourceC163, D b s _ m D d e t 83, frontendE83, backend[83, prec isC83, sparet163{ > basiclnfo; /•X length of header #/ /* length of data */ /# scan number */ /# observer's i n i t l s * / /•X observer's name #/ /# telescope descrip -X7 /•X project identific #/ /•* source name */ /*• type data obs mode */ /•X- front end descptor #/ /•X- backend descriptor #/ /*• data precision *•/ /•X telescope parameters #/ typedef struct { float x_point, y_point, ux_poi nt, uy_point, /# az/ra pntng corr -X-/ /# el/dec pntng corr*-/ /•X usr az/ra pntg corr*/ /•* usr el/dec pntg cor*/ 93 ra_pt_con[33, /* R.A. pntng coefs #/ dec_pt_conC43, /# DEC coe-f f icients */ lev_pt_con[43, /# level pointing coefficients */ extra_pt_con[103 , /* spare pointing coefficients */ orient, /-X- box/secondary orientation *7 focus_r, /*• axial focus #/ focus_ns, /# north-south focus *7 f o c u 5 _ l ; /# east-west focus #/ char sparer. 163; > teleParam; /# observing parameters #/ typedef struct C float ut_date, /# universal time date *7 ut, /*• universal time -X7 1st, /* local sidereal time */ no_rchan,/# number of receiver channels *•/ no_swvar, /# number of switching variables */ no_phase5, /# number of phases per cycle */ cycl_len, /•* length of cycle •*•/ no_cyclej /•X- number of cycles per integration */ char spareC163; } obsParam; /* positions */ typedef struct { float epoch, x_source, y_source, x_ref, y_ref, epoc_ra, e p D C _ d e c , gal_long,/# gal_lat, /# az, e l , i nd_x, ind_y, origi nl, origin2, origin3} char c o D r d_cdC83, spareC24 3 j > pos; /* epoch */ /•X- commanded source x *•/ /# commanded source y #/ /•X- commanded ref x *•/ /•* commanded ref y #/ /•X- commanded epoch r a #7 /*• commanded epoch dec */ commanded galatic longitude #7 commanded galatic latitude $7 /•X- commanded azimuth *7 /•* commanded elevation *7 /# indicated x position *7 /•X- indicated y position *7 /* descriptive origin */ /•X- coordinate system code *7 /# environment •*/ typedef struct { float tamb, /-X- ambient temperature -X-/ 94 > env; char pressure, humi di ty, refrac, tau_h2o, t_h2o, tau_o2, t_o2, dew_pt, mmh2Di spareC163; /•* pressure *7 /*• relative humidity */ /•* index of refraction *7 /* H20 opacity */ /* H20 temperature */ /* 02 opacity */ /# 02 temperature #/ /* dew point #7 /* mm H20 *7 /* map parameters *7 typedef struct { float 5can_ang, /•* map scanning angle */ bfwhm, /# beam fullwidth at half max #7 xzero, /*• x position at zero #/ yzero, /* y position at zero #7 delta_x, /* delta x/x_rate *7 delta_y; /* delta y/y_rate */ char no_ptst83, /•* number of points #/ no_xptsC83, /# number of x points #7 ho_ypts[83, /# number of y points *7 x_cell0C83, /* starting x grid c e l l *7 y_cellOC83, /* starting y grid c e l l */ frameC83, /# xy reference frame code *7 sparer.163; > mapParamj /# observing dependent parameters *7 typedef struct { float t_source, /* source temperature #7 t_rms, /* RMS of mean #/ rvsys, /# velocity correction *7 velocity, /* velocity wrt reference *7 off_scan, /#• off scan number #7 badch_v; /# bad channel value *7 char type_calC83, /# type of calibration *7 cal_desC83, /# calibration description *7 vel_deft83, /* velocity definition & ref *7 spareC24J; } obsDeParam; typedef struct { float ap_eff, /•* antenna aperture efficiency #/ beam_eff, antenna beam efficiency tETA(MB)3 *7 ant_gain, /*• antenna gain #7 etal, /* rear s p i l l & scat efficiency *7 etafss;/*- forward s p i l l k scat efficiency */ char spare(83; > engParamj /•* telescope dependent parameters <NRA0_GB) #/ 95 typedef struct { float 11, l l f 1, 1 If 2, 12, 12f 1, 12f 2, la, lb, Ic, Id, 1ev_corr, p_valsC3], rho, theta, pos_type, /# ULO primary frequency */ /# ULO secondary frequency 1 #7 ULO secondary frequency 2 #/ /* ULO primary frequency *7 /# ULO secondary frequency 1 *7 /# ULO secondary frequency 2 #/ /# IF frequencies #/ /# level correction #/ /# pointing fudge *7 /* rho */ /* theta */ /* H316 command block type */ obs_number, /# observers I.D. number converted from obs_id */ fspareC063, AD_CalC163; /# in Volts */ char cf_formC643,/# center frequency formula *7 spare!163 5 } teleDeParamJ /# phase block -XV typedef struct { float var_val; /# variable value */ char var_desC83, /# variable description *7 phas_tb£83; /* phase table */ > phaBlk; /•X- descriptor block for each receiver channel #/ typedef struct i float obs_freq, /# observed frequency #/ rest_freq, /# rest frequency #/ freq_res, /•* frequency resolution #/ bandwidth, /*• rx channel bandwidth #/ t_rx, /* receiver temperature */ t_cal, /*• calibration temperature #/ s_tsys, /-X- source system temperature •*/ r_tsys,/* ref system temperature */ ref_pt, /*• reference point *•/ xO, /* x value at reference point #/ deltax, /* delta x */ int_time, /# integration time */ no_int, /* number of integrations *•/ opacity; /# opacity #/ char pol_codeC83J /* polarization code #/ float rho, /*• feed distance from center #/ theta; /-X- feed direction from center */ > desBlk; typedef struct stDrl6Header basiclnfo b i ; 96 teleParam tp; obsParam Dp; pos pi env e; mapParam mp; obsDeParam odp; engParam ep; teleDeParam tdp; phaBlk pbC43; desBlk dbC1635 > stl6Head; /* STRUCTURES FOR ACQUIRED DATA */ typedef struct { float posCoorC2], chant 1615 } contl6SetS! /* BUFFER DEFINITIONS - FULL REC DEFS - HEADER PLUS DATA /#• Continuum data are the integrated output power from # wideband radiometers. A block of continuum data # contains a standard header format plus 540 4-byte # real words in one of four formats. #/ /# 5: continuum - 16 channel *•/ typedef struct { stl6Head header; contl6SetS contl6Sett30D; > contl6BufS; 97 APPENDIX B: Translation of Floating Point Formats The following program is used to convert the IEEE standard floating point format data (written by the Green Bank Masscomp) to the Dec floating point standard. void xlf<x) float #x? /* translate IEEE format to Vax format */ { unsigned char #b,cj b = (unsigned char #) xj /# convert pointer to byte size */ c = (<*b) Be 127) + l ; /* mask off sign bit and then add 1 */ if (c>127) {printf("exponent is too large., doom is upon us\n"); exit(98);> c += ((*b> & 128)t /* put back the sign bit */ #b = *(b+l); /# swap sign-exponent-mantissa bytes*/ *(b+l) = c; c = *(b+2); *(b+2) = *(b+3); *(b+3) = c; /* swap mantissa bytes */ > 98 APPENDIX C: South Bound Mapping The vast majority of this work was done using the north bound data. Although the calibration work -for the south bound mapping was not as complete as -for the north, a l l the programs used in the north bound mapping have counterparts for the making of maps with south bound data. In the following figure the supernova remnant of G109.1-1.0 was mapped using only south bound data. These data were heavily corrupted by the same type of interference as seen in figure 58. The resultant map is of poor quality and was only included here for completeness. 99 A map of the test region produced by the hybrid method with a neighbourhood of ZOO points (no structure has been removed). Contours are at -50, 50, 125, 200, 500, 1000 mJy/Beam. (Galactic coordinates used) 100 

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
http://iiif.library.ubc.ca/presentation/dsp.831.1-0096860/manifest

Comment

Related Items