Frequency Calibration of Spectral Observation System of the TM65m Radio Telescope

CHINESE ASTRONOMY AND ASTROPHYSICS Chinese Astronomy and Astrophysics 40 41 (2016) 578–589

Frequency Calibration of Spectral Observation System of the TM65m Radio Telescope LI Juan1,2

WU Ya-jun1,2

QIAO Hai-hua1,2,3

WANG Jun-zhi1,2

ZUO Xiu-ting1,2 1 2

Shanghai Astronomical Observatory, Chinese Academy of Sciences, Shanghai 200030

Key Laboratory of Radio Astronomy, Chinese Academy of Sciences, Shanghai 200030 3

University of Chinese Academy of Sciences, Beijing 100049

Abstract In order to carry out the spectral observation with the TM65m radio telescope, the frequency calibration and test of DIBAS (Digital Backend System) are performed, it is found that it has a good performance. First, by injecting the PCAL signals, the frequency resolution, frequency drift and the stability of frequency interval between two spectral lines of the DIBAS backend are measured. It is found that in one hour, the maximum frequency drift of a single spike is 0.03 channel, the maximum fluctuation of spike interval is 0.05 channel. Then, by the observations on the H2 CO maser and absorbtion lines of massive star formation regions, and the comparison with the results observed by the GBT (Robert C. Byrd Green Bank Telescope), it is shown that the results of frequency calibration are correct. Finally, by the OH maser observations in more than one hour toward W3(OH), and the methanol maser observations in more than 5 hours, it is found that the spectral profiles keep consistent, and the observational noise is consistent with the theoretical value, indicating the stability and reliability of the frequency calibration program. Key words techniques: photometry—spectral line: identification—radio lines: ISM—ISM: molecules †

Supported by 973 Project (2012CB821800) and National Natural Science Foundation (11103006,

11173051) Received 2015–03–23; revised version 2015–05–21  

A translation of Acta Astron. Sin. Vol.56, No.6, pp. 648–657, 2015 [email protected]

0275-1062/16/$-see front matter © 2016 Elsevier B.V. All rights reserved. doi:10.1016/j.chinastron.2016.10.005

LI Juan et al. / Chinese Astronomy and Astrophysics 40 (2016) 578–589

1.

579

INTRODUCTION

The TM65m radio telescope of Shanghai Astronomical Observatory is located in the Songjiang District of Shanghai. With an aperture of 65 meters, it is a fully steerable radio telescope of altazimuth mounting. Its total weight attains 2700 tons, and the designed highest working frequency is 43 GHz. The antenna of the telescope adopts high-precision solid panels, and it is equipped with an active surface regulation system. Its pointing accuracy is better than 15 . At present, it can work on the L(1.25∼ 1.75 GHz), C(4.0∼8.0 GHz), S/X(2.15∼2.45/8.2∼9.0 GHz) and Ku(12.0∼18.0 GHz) bands, and will soon be equipped with new receivers at the K(18.0∼26.5 GHz), Ka(30.0∼34.0 GHz) and Q(40.0∼46.0 GHz) bands[1−4] . The observation of molecular spectral lines is one of the principal tasks of the TM65m radio telescope, its rather broad frequency coverage makes it able to observe the 21 cm hydrogen line, hydroxyl maser, methanol maser, water maser, ammonia molecule, silicon monoxide molecule, formaldehyde absorption line, and the recombination lines of the hydrogen, helium and carbon ions, and so on. Using these spectral lines a lot of work may be done to study such as the star formation and chemical evolution in the Galaxy, the large-scale structure of the Galaxy, etc. By means of the velocity information hinted by the spectral lines we can study the dynamics of molecular clouds. Table 1 lists some relatively strong spectral lines which are under the frequency coverage of the TM65m radio telescope. Table 1 List of the rest frequencies of molecular/atom emission lines that can be observed with the TM65m radio telescope Name H OH H2 CO CH3 OH

Rest frequency/GHz 1.420 1.612/1.665/1.667/1.720, 4.750/4.765, 6.030/6.035 4.83/14.5 6.7/12.2...

HC3 N

18.196/27.294/36.392...

HC5 N

7.988/15.976...

NH3

23.694/23.723/23.870

H2 O

22.235

CS

48.991

SiO

42.424

In general, there are 5 important parameters for the observations of spectral lines of interstellar molecules: the line intensity, the line-center velocity, the velocity range of spectral line, the half-power line width, and the line profile. Because of the earth’s rotation, the velocity components of different stations along the line-of-sight direction are different, and vary with the time. In order to homogenize the observations all around the world,

580

LI Juan et al. / Chinese Astronomy and Astrophysics 40 (2016) 578–589

it is required to correct this effect, that is to say, for the observations made at different stations at different times, the Doppler correction on the observed data should be made. The performance of the spectrometer, such as the spectral response, spectral resolution, are also very important for the measurements of above-mentioned spectral line parameters[5] . In this paper, firstly, the composition of the DIBAS backend system of the TM65m radio telescope, and the mode of spectral line observation are introduced; then the principle of frequency calibration of the spectral backend system is specified; finally, the result of the measurement on the spectrometer response with the PCAL signal and the results of the test line observations with the DIBAS backend after the frequency calibration are given. 2.

SPECTRAL BACKEND OF THE TELESCOPE

The spectral observation of the TM65m radio telescope is carried out with the DIBAS backend, which is also the pulsar backend of the TM65m radio telescope. 4 parts, namely, the data acquisition, data handling, data storage and observation control make up all the device (refer to Fig.1.), which is developed by the Shanghai Astronomical Observatory in cooperation with the National Radio Astronomy Observatory (NRAO) of the United States. The DIBAS backend supports 29 different modes of spectral observations, which can be classified as the single-window mode and multi-window mode; and the former can be further classified into the broad-band mode and narrow-band mode. The modes 1,2,3 belong to the broad-band mode of single window; they can cover very broad frequency range, the broadest bandwidth attains 1.5 GHz, the largest channel number is 16 384 and the highest frequency resolution is 61 kHz (corresponding to the velocity resolution of 2.7 km·s−1 in the C band). Several spectral lines may be simultaneously observed with the broad-band mode, thus the observation efficiency is considerably raised. The modes 4∼19 belong to the narrowband mode of single window, which are mainly used to implement the observations of high frequency resolution, with the highest frequency resolution up to 0.02 kHz (corresponding to the velocity resolution of 0.000 9 km·s−1 in the C band). The modes 20∼29 belong to the multi-window mode, which have 8 windows of 24 MHz or 16 MHz bandwidth, the highest frequency resolution attains 0.24 kHz (corresponding to the velocity resolution of 0.01 kms in the C band). These windows can be put on any positions in the whole receiving bandwidth, through which several spectral lines may be simultaneously observed with a very high frequency resolution. In brief, the DIBAS backend can provide a broad frequency coverage, and the observational modes of high frequency resolution, it has greatly enhanced the capability of the TM65m radio telescope in the spectral observation at the centimeter waveband, and will help observers to perform more efficiently the astronomical observation with the TM65m radio telescope, so that they may obtain the maximum scientific products from their observed data. Table 2 gives the window number Nband , bandwidth BW , channel number Nchan , frequency resolution Δf , and the corresponding velocity resolution Δv for the spectral modes 1∼29 of the DIBAS backend. As shown in Table 2, Mode 6 has only

LI Juan et al. / Chinese Astronomy and Astrophysics 40 (2016) 578–589

581

one window of 187.5 MHz bandwidth, with the 131072 channels the frequency resolution is 1.4 kHz, and the corresponding velocity resolutions are about 0.25, 0.06, 0.05, 0.02, and 0.01 km·s−1 at the L, C, X, K, and Q bands, respectively, while Mode 20 has 8 windows of 23.44 MHz bandwidth, each window has 4096 channels, their frequency resolution is 5.7 kHz, and the corresponding velocity resolutions are about 1.01, 0.26, 0.19, 0.076, and 0.038 km·s−1 at the L, C, X, K, and Q bands, respectively.

Fig. 1 The DIBAS backend

3. 3.1

FREQUENCY CALIBRATION

Doppler Calibration

3.1.1 The local standard of rest Because of the Earth rotation, the components of the motions of various stations in the direction of line-of-sight direction are different, and vary with the time. In order to correct this effect, it is necessary to deduce the motion of each station onto the local standard of rest (LSR). The analysis of stellar spectra (mainly the A∼G spectral types) indicates that the Sun has a systematic motion relative to its adjacent stars. So, the LSR is defined as a point in accordance with the Sun’s position which revolves with a local circular velocity around the Galactic center. In the LSR system the Sun moves with a velocity of 20 km·s−1 towards the direction of right ascension 18h and declination 30◦ (1900)[6] .

582

LI Juan et al. / Chinese Astronomy and Astrophysics 40 (2016) 578–589

Table 2 The parameters of spectral observation modes Mode

Nband

BW /MHz

Nchan

Δv /(km · s−1 )

Δf /kHz

45 GHz

22.5 GHz

9 GHz

6.7 GHz

1.7 GHz

1

1

1500

1024

1465

9.8

19.5

49

66

259

2

1

1500

16384

92

0.6

1.2

3.1

4.1

16.2

3

1

1000

16384

61

0.4

0.8

2.0

2.7

10.8

4

1

187.5

32768

5.7

0.04

0.08

0.19

0.26

1.01

5

1

187.5

65536

2.9

0.02

0.04

0.10

0.13

0.51

6

1

187.5

131072

1.4

0.01

0.02

0.05

0.06

0.25

7

1

100

32768

3.1

0.02

0.04

0.1

0.14

0.55

8

1

100

65536

1.5

0.01

0.02

0.05

0.07

0.26

9

1

100

131072

0.8

0.005

0.01

0.03

0.036

0.14

10

1

23.44

32768

0.7

0.005

0.009

0.023

0.031

0.124

11

1

23.44

65536

0.4

0.003

0.0053

0.013

0.018

0.07

12

1

23.44

131072

0.2

0.0013

0.0026

0.0067

0.009

0.035

13

1

23.44

262144

0.1

0.0007

0.0013

0.0033

0.0045

0.018

14

1

23.44

524288

0.05

0.00035

0.00065

0.00165

0.00225

0.009

15

1

11.72

32768

0.4

0.003

0.0053

0.013

0.018

0.07

16

1

11.72

65536

0.2

0.0013

0.0026

0.0067

0.009

0.035

17

1

11.72

131072

0.1

0.0007

0.0013

0.0033

0.0045

0.018

18

1

11.72

262144

0.05

0.00035

0.00065

0.00165

0.00225

0.009

19

1

11.72

524288

0.02

0.00013

0.00026

0.00067

0.0009

0.0035

20

8

23.44

4096

5.7

0.038

0.076

0.19

0.26

1.01

21

8

23.44

8192

2.9

0.02

0.04

0.1

0.13

0.51

22

8

23.44

16384

1.4

0.01

0.02

0.05

0.06

0.25

23

8

23.44

32768

0.7

0.005

0.009

0.023

0.031

0.124

24

8

23.44

65536

0.4

0.003

0.0053

0.013

0.018

0.07

25

8

15.625

4096

3.8

0.025

0.051

0.13

0.17

0.67

26

8

15.625

8192

1.9

0.013

0.025

0.063

0.085

0.34

27

8

15.625

16384

0.95

0.006

0.013

0.032

0.043

0.17

28

8

15.625

32678

0.48

0.0032

0.0064

0.016

0.021

0.085

29

8

15.625

65536

0.24

0.0016

0.0032

0.008

0.011

0.042

3.1.2

The Doppler correction

In order to convert the velocity into the LSR system, it is required to correct the effects of 4 kinds of motion, namely, the line-of-sight component of the velocity of the Sun relative to the LSR, that of the velocity of the barycenter of the Earth-Moon system relative to the Sun, that of the velocity of the Earth’s center relative to the barycenter of the Earth-

LI Juan et al. / Chinese Astronomy and Astrophysics 40 (2016) 578–589

583

Moon system, and that of the velocity of the observer relative to the Earth’s center. The algorithm of calculating all these components is introduced as follows, and here the doctorial dissertation of Dr. LI Gang of the Shanghai Astronomical Observatory has been taken as our primary reference[7] . (1) Vsun , the line-of-sight component of the velocity of the Sun relative to the LSR Let (α, δ) be the equatorial coordinates of the radio source at the epoch of observation; (R, D), those of the Sun’s vector, then, ⎡ ⎤ cos D cos R ⎢ ⎥  = 19.5 × ⎢ cos D sin R ⎥ , and the unit vector on the direction of the radio source is V ⎣ ⎦ sin D ⎡ ⎤ cos δ cos α ⎢ ⎥  = ⎢ cos δ sin α ⎥ , so, Vsun = −V  ·S  , here the velocity is assumed to be negative for the S ⎣ ⎦ sin δ motion towards the radio source. (2) VEM , the line-of-sight component of the velocity of the barycenter of the Earth-Moon system relative to the Sun As shown in Fig.2, S stands for the Sun; ABCD, the orbit of the barycenter of the Earth-Moon system; E, the barycenter position, r, the radius vector from the Sun to the barycenter; ω, the heliocentric ecliptic longitude of the perihelion; f , true anomaly; let a,e be the semi-major axis and the eccentricity of the orbit, respectively; n, the mean angular velocity of the barycenter around the Sun, then, the velocity of the barycenter na can be decomposed into V1 = √1−e , the component perpendicular to the radius vector r, 2 nae and V2 = √1−e2 , the component perpendicular to the axis AC. Let (λ, β) be the ecliptic coordinates of the radio source at the epoch of observation, then, VEM = −V1 cos β[cos(90◦ + ω + f − λ) + e cos(λ − ω − 90◦ )] = V1 cos β[sin(ω + f − λ) + e sin(ω − λ] , in which the relations of ω, f with the basic angular arguments of nutation series are as follows: ω = F + Ω − D − l − π , f = fM + (2e − 0.25e3 ) sin fM +





5 2 1 4 13 e − e sin(2fM ) + e3 sin(3fM ) + . . . , 4 8 12

(1) (2)

in which fM = l is the mean anomaly. (3) VE , the line-of-sight component of the velocity of the Earth’s center relative to the barycenter of the Earth-Moon system Its calculation principle is fully analogous to that of VEM , thus only is given the formula: VE =

n e  √ cos β  [sin(ω  + f  − λ ) + e sin(ω  − λ )] , 81.3 1 − e2

584

LI Juan et al. / Chinese Astronomy and Astrophysics 40 (2016) 578–589

in which, all the variables are correspondent to the lunar orbit, e is the orbital eccentricity; n , the mean angular velocity of the Moon; a , the semi-major axis of the orbit; ω  , the mean longitude of the Moon’s perigee; f  , true anomaly; 81.3, the ratio of the Earth’s mass over the Moon’s mass; (λ , β  ), the coordinates of the source relative to the Moon’s path. The  relations of ω  ,f  with the basic angular arguments of nutation series are ω  = F − l, fM = l,   and the relation between the true anomaly f and the mean anomaly fM is analogous to Equation (2). V1

B

V2

r

E

C

a

S

D

f ω

A

r'

Fig. 2 The sketch map for the orbit of the barycenter of the Earth-Moon system. S stands for the sun. ABCD stands for the orbit of the barycenter of the Earth-Moon system. E is the position of the barycenter.  r is the radius vector between the Sun and the barycenter of the Earth-Moon system. ω is the heliocentric ecliptic longitude of perihelion. f is the true anomaly. a is the semi-major axis of the orbit[7] .

(4) Vobs , the line-of-sight component of the velocity of the observer relative to the Earth’s center Let φ and h be the geodesic latitude and altitude of the observer, φ is transformed into the geocentric latitude φ by: φ = φ − a sin(2φ ) + b sin(4φ ) − c sin(6φ ), in which a = 692.743 , b = 1.163 , c = 0.0026 . The geocentric distance of the observer is R = A + B cos 2φ − C cos 4φ + h, in which A = 6367489.8 m, B = 10692.6 m, C = 22.4 m. Let L be the local sidereal time at the observational time, then, Vobs = −ωR cos φ cos δ sin(α − L) = ωR cos φ cos δ sin(L − α), in which ω = 7.292 × 10−5 rad · s−1 is the angular velocity of the Earth’s rotation. It leads to V = Vsun + VEM + VE + Vobs .

585

LI Juan et al. / Chinese Astronomy and Astrophysics 40 (2016) 578–589

In the above algorithm, both the perturbations of planets on the Earth’s orbital motion and the Sun’s motion relative to the barycenter of the solar system have not been taken into consideration, which results in an about 0.02 km · s−1 error of V . The contribution to V from different velocity components is shown in Table 3. Table 3 Contributions to V from different velocity components Velocity component

Velocity magnitude /(km · s−1 )

Velocity of the source relative to LSR (VLSR )

-

Velocity of the Sun relative to LSR (Vsun )

20

Velocity of the barycenter of the Earth-Moon system relative to the Sun (VEM )

30

Velocity of the Earth relative to the barycenter of the Earth-Moon system (VE )

0.1

Velocity of the station relative to the Earth (Vobs )

0.5

Planetary perturbation of the Earth orbit

0.013

3.2 Test of PCAL Signal Injection When the single-frequency PCAL signal is injected to the input terminal of the spectrometer, at its output terminal a spectrum composed of a series spikes will be observed. By measuring the half-power widths of these spikes, the spectral features of the spectrometer can be obtained. Prof. ZHENG Xing-wu of Nanjing University and Dr. LEI Cheng-ming have carefully measured the spectroscopic performance of the acoustic-optic spectrometer of the 13.7m telescope of Purple Mountain Observatory with this method [5] . 3.2.1 Width of spectral line In August 2014, different single-frequency signals with a frequency interval of 1 MHz are successively injected to the input terminal of the spectrometer within the integration time under the DIBAS mode 24 to observe the response of the spectrometer. The mode 24 provides with 8 windows, each window possesses the 23.44 MHz bandwidth and 65 536 channels, from which a spectrum composed of 243 spikes can be observed, as shown by the left panel of Fig.3. To measure the spike’s half-power width by Gaussian fitting, the spectral resolution of the spectrometer is obtained. The Gaussian fitting curve of a typical spike in the spectrum is shown in the right panel of Fig.3. From which we can find that the half-power width of the spike is 1 channel without broadening nor sidelobes, in other words, the the spectral resolution of DIBAS is just the same as the channel interval, while the error of spectral resolution is far less than the channel interval. And the spectral line profile keeps invariable in the whole receiving system.

586

LI Juan et al. / Chinese Astronomy and Astrophysics 40 (2016) 578–589

8300

Frequency /MHz 8305 8310 8315

Frequency /MHz 8306.049 8306.05 8306.051 8306.052

8320 20

15

15 TA* /K

TA* /K

20

10

10

5

5

0 20000 40000 Channel number

60000

0 23565

23570 23575 Channel number

Fig. 3 Left: the response of DIBAS to a PCAL signal. The PCAL signals with an frequency interval of 1 MHz are input to the X-band IF input terminal under the DIBAS mode 24. This figure gives the spectrum of one window. Right: the Gaussian fitting for a spike in the spectrum. The solid line stands for the observed result, while the dashed line stands for the Gaussian fitting curve.

3.2.2

Test of frequency shift

In order to test the frequency shift, the PCAL signals are injected for 1 hour with the above-mentioned method to observe the response of the spectrometer. The observation is made of the PCAL ON and PCAL OFF two states, each lasts for 1 minute to get 30 scans. The centroid positions of the 12th spike in the every scan are obtained by Gaussian fitting. Fig.4 (left) shows the shift of the centroid position of the 12th spike with the time, we can find that the shifts are concentrated at 0 or so. Within 1 hour the maximum amplitude of variation is 0.03 channel, about 0.01 kHz. For ordinary measurements, such a shift is negligible. 3.2.3

Test of the stability of spectral line interval

In order to test the stability of the interval between two spectral lines, the interval between the 5th and the 20th spikes in the spectrum of the every scan is measured with CLASS. Fig.4 (right) shows the shifts of the interval between the 5th and the 20th spikes, which are still concentrated around 0. The maximum fluctuation of the interval between two spikes is 0.05 channel, about 0.02 kHz. The shift of spike interval is far less than the channel width, so in ordinary measurements such an effect may be ignored. 3.3 3.3.1

Test of Frequency Calibration Observations of H2 CO absorption line and radio recombination lines

On 21st March 2014, the observations of the H2 CO absorption line and hydrogen recombination line were performed towards a few star formation regions at the C band of the TM65m radio telescope, and the DIBAS mode 6 was adopted by the spectral terminal. The observation adopted the position switching mode, in which both the times of ON SOURCE and OFF SOURCE were 2 minutes. Fig.5 (left) shows the spectrum of the H2 CO absorption line of IRAS00338. In order to raise the signal-to-noise ratio the velocity resolution is smoothed to 0.36 km·s−1 , and the velocity at the flux peak position is −17.41(0.04) km·s−1 . That

587

LI Juan et al. / Chinese Astronomy and Astrophysics 40 (2016) 578–589

0.1

1

0.05

0.5 Frequency drift

Frequency drift

obtained by Araya et al. [8] in the observation toward the same source with the American GBT telescope is −17.37 (0.01) km·s−1 . A comparison between the ours and theirs indicates that both results are consistent within tolerance. Fig.5 (right) displays the spectrum of the H2 CO maser lines and absorption line of the massive star formation region NGC 7538. This source exhibits both the H2 CO maser lines and H2 CO absorption line, the velocities at their flux peaks are −60.06(0.02) km·s−1 , −57.84(0.02) km·s−1 , and −55.63(0.31) km·s−1 , respectively. The results obtained by Araya et al. [8] from their observation of NGC 7538 are −60.12(0.01) km·s−1 , −57.91 (0.01) km·s−1 and −55.95(0.05) km·s−1 . The differences of velocity in the maser portion are 0.06 km·s−1 and 0.07 km·s−1 , respectively, and they are far less than the velocity resolution. The velocities of the absorption component are consistent within tolerance. As shown from the above results, the frequency calibration made on the data is correct.

0

−0.05

−0.1 0

−0.5

channel number−32983

10

20 scan

30

0

channel number−41943 40

−1 0

10

20 scan

30

40

Fig. 4 Left: the frequency shift of the 12th spike within one hour. The horizontal axis gives the scan number, while the vertical axis gives the channel number minus 32983. Right: the shift of the interval between the 5th and 20th spikes within one hour. The horizontal axis gives the scan number, while the vertical axis gives the value of channel number minus 41943.

3.3.2 Observation of OH maser On 17th April 2014, the hydroxyl maser of W3(OH) was observed for more than 1 hour under the DIBAS spectral mode 24 with the velocity resolution of 0.06 km·s−1 . Totally 31 scans were observed, and the ON SOURCE time and OFF SOURCE time for each scan were 1 minute. Fig.6 (left) shows the spectra of the first scan (black), of the last scan (red) and of the average of all scans (blue). We can find from this figure that the line profiles of these 3 spectra are consistent, the flux peaks of the first scan, last scan, and the average line profile are all positioned at the channel 32175. The amplitude difference between the first and the last scans is 6%, which means the variation of the system gain. The Root Mean Square (RMS) noise of the first scan is 0.026, the RMS noise of the average of all the 31 scans should be nearly 0.026 5.6 = 0.0046. In fact, the RMS noise of the spectrum obtained from an integration of all the 31 scans is 0.006. So, it may be recognized that our observational

588

LI Juan et al. / Chinese Astronomy and Astrophysics 40 (2016) 578–589

values are in accordance with the theoretical ones. Frequency /MHz 4830

4830.2

4829.8

0.06

0

Frequency /MHz 4830.7 4830.6 4830.5 4830.4

TA* /K

TA* /K

0.04 0.02

0.02 0

0.04

0.02 30

20 10 Velocity /(km·s1)

70

40

60 50 Velocity /(km·s1)

Fig. 5 Left: the H2 CO absorption line of IRAS00338 observed with the DIBAS mode 6 on 21st March 2014, with the ON-SOURCE time of 2 min. Right: the H2 CO maser and absorption lines of NGC 7538 observed with the DIBAS mode 6 and the ON-SOURCE time of 2 min.

On 14th July 2014, the methanol maser of W3(OH) was observed under the DIBAS spectral mode 6 with the velocity resolution of 0.13 km·s−1 . Totally 6 scans were observed, one hour for each scan. Fig.6 (left) shows the spectra of different scans, which are plotted in different colors. We can find from this figure that the spectral line profiles of different scans are identical, the Doppler correction, therefore, is reliable and stable during the observations of more than 5 hours. 32100

Channel number 32150 32200 32250

6669.6

32300

30

Frequency /MHz 6669.5

6669.4

100 TA* /K

TA* /K

20

50

10 0 50

45 Velocity /(km·s1)

40

0 50

45 Velocity /(km·s1)

40

Fig. 6 Left: the spectrum of 1665 MHz OH maser of W3(OH) observed with the DIBAS mode 24 in April 2014. 31 scans were observed in total, with the ON-SOURCE time of 1 min for each scan. The black, red, and blue lines show the spectra of the first scan, last scan, and averaged scan, respectively. Right: the spectrum of 6.7 GHz methanol maser of W3(OH) observed with the DIBAS mode 6 in July 2014. The duration of each scan is one hour. The spectra of different scans are plotted in different colors.

4.

CONCLUSION

The frequency calibration and test for the DIBAS spectral backend of the TM65m radio telescope of Shanghai Astronomical Observatory are implemented, the obtained results in-

LI Juan et al. / Chinese Astronomy and Astrophysics 40 (2016) 578–589

589

dicate its nice performance. The test is composed of (1) the test of the spectral resolution, frequency shift and the stability of intervals between two spectral lines of the spectrometer by using PCAL signals; (2) the observation of formaldehyde absorption lines of 2 star formation regions; (3) the long-time observation of the hydroxyl maser and methanol maser of W3(OH). The results show that (1) the spectral resolution of the DIBAS backend is in accordance with the channel interval. The maximal amplitude of variation of the frequency shift of a single spike is 0.03 channel, and the maximal fluctuation of the interval between two spikes is 0.05 channel; they are all far less than the channel width and can hardly affect normal observations. It means indirectly that the frequency output of the PCAL instrument is stable; (2) after the frequency calibration, the velocities at the line peaks are in accordance with the previous observational results within the error limits; (3) after the frequency calibration, the profile of the maser line keeps stable in several hours, the flux peak falls on the same channel, and the observational noise level is consistent with theoretical value. The intensity calibration will be further made, so that the TM65m radio telescope can carry on astronomical spectroscopic observations. ACKNOWLEDGEMENTS Heartfelt thanks to Prof. ZHENG Xing-wu for his disinterested help in the adjustment of the TM65m radio telescope and the attentive advice in the writing of this paper. Thanks to Dr. SUN Ji-xian of Qinghai Station of Purple Mountain Observatory and Prof. Kalken of Xinjiang Observatory for their hearty support and help to our work. References 1

Wang J. Q., Yu L. F., Zhao R. B., et al., Acta Astronomica Sinica, 2015, 56, 63

2

Wang J. Q., Yu L. F., Zhao R B, et al., ChA&A, 2015, 39, 394

3

Yu L. F., Wang J. Q., Zhao R. B., et al., Acta Astronomica Sinica, 2015, 56, 165

4

Yu L. F., Wang J. Q., Zhao R. B., et al., ChA&A, 2015, 39, 524

5

Zheng X. W., Lei C. M., Acta Astronomica Sinica, 1997, 38, 446

6

Rohlfs K., Wilson T. L., Tools in Radio Astronomy, tranlated by JIANG Bi-wei, Beijing: Beijing Normal University Press, 2008, 11

7

Li G., Method of VLBI Observations of Celestial Maser Sources, Shanghai: Shanghai Astronomical Observatory, Chinese Academy of Sciences, 1986

8

Araya E., Hofner P., Goss W. M., et al., ApJ, 2007, 170, 152