Data Paper Zone II Versions EN1 Vol 3 (2) 2018
Data of the MWISP Sky Survey (2011 – 2017)
: 2018 - 01 - 05
: 2018 - 03 - 21
: 2018 - 06 - 26
2406 11 0
Abstract & Keywords
Abstract: Using the PMO 13.7 m millimeter-wave radio telescope located in Delingha and a nine-beam Superconducting Spectroscopic Array Receiver (SSAR) system, we preformed a large-scale survey called MWISP project (Milky Way Imaging Scroll Painting) in the 12CO (J=1–0), 13CO (J=1–0), and C18O (J=1–0) lines along the Galactic Plane. The survey covers the Galactic region of -10°≤L≤250° and -5°≤B≤5°, about 2600 square degrees of sky. This dataset assembles all the survey data acquired during 2011 and 2017. Its rms noise level, which is about 0.5 K for 12CO (J=1–0) and 0.3 K for 13CO (J=1–0) and C18O (J=1–0), is uniform across the whole covered area. This sensitivity basically satisfies the requirements of astronomical research.
Keywords: millimeter wave; sky survey data; Galactic structure
Dataset Profile
Chinese title2011~2017年“银河画卷”巡天数据集
English titleMWISP data (2011 – 2017)
Data corresponding authorLiu Liang (, Yang ji (
Data authorsLu Dengrong, Sun Jixian,Yang Zherui, Gao Na, Liu Liang, Yang Ji
Time range2011 – 2017
Geographical scopeGalactic region of -10°≤L≤250° and -5°≤B ≤5°, covering a sky area of about 2600 square-degrees.
Spatial resolution50 arcsecs (grid of 30 arcsecs)Data volume20 TB
Data format*.bur, *.fits
Data service system<>;
Sources of fundingThe "integration and sharing of scientific and technological data resources" project of the Chinese Academy of Sciences 12th Five-Year Program (XXH12504-2-01)
Dataset compositionThis dataset consists of data of 12CO (J=1–0), 13CO (J=1–0), and C18O (J=1–0) lines toward the Milky Way. Data for each sky area are stored separately across the Galactic region of -10°≤L≤250° and -5°≤B≤5°. It has a total data volume of 20 TB.
1.   Introduction
Molecular gas is an important component of the universe. Studies have shown that molecular gas in the Milky Way accounts for about 1% of the mass of baryon matter. Molecular gas is the site of star formation and a key link of material circulation in the universe. It plays a very important role in the evolution of the Milky Way and even of the entire universe. Therefore, understanding the properties of molecular gases, such as their structure, distribution, and motion, is essential for studying the origin of the universe, as well as the origin and evolution of galaxies.
CO lines survey is the most effective way of studying molecular gas of the Milky Way. Observations of CO molecules provide a good approach to investigating molecular gas distribution in the Milky Way. In our study, a 13.7m millimeter-wavelength telescope is used to perform 12CO (J=1–0), 13CO (J=1–0) and C18O (J=1–0) observations toward the Milky Way by using a multi-beam receiver on OTF observation mode. The project is named “Milky Way Imaging Scroll Painting” (MWISP). With multi-line mapping, moderately-high spatial resolution, and large-scale spatial coverage, the MWISP CO survey leaves an important legacy for astronomical research.
2.   Data collection and processing
Data collection of this study relies on China’s only observing equipment in the millimeter wave band, which is also the most advanced in the world among those of the same aperture. In 2010, a national major imaging equipment “SSAR” was installed on the telescope and the On-the-Fly (OTF) mode was developed. The telescope was then used to conduct large-scale surveyson the three CO lines toward the Galactic plane between b= ±5 degrees. So far, an area of 1250 square degrees has been completed from 2011 to 2017.
The standard chopper wheel calibration method was used in the molecular line observation performed by the 13.7m millimeter-wavelength telescope. The temperature scale obtained therefrom was “antenna temperature” (i.e., TA* )1 with atmospheric absorption corrected and ohmic loss compensated. For extended molecular cloud sources, the temperature scale usually needs to be further corrected through the main beam efficiency η of the telescope, in order to obtain “radiation temperature” (i.e., T R* ) comparable to that observed by other telescopes. This temperature scale stands for the convolution of the ideal main beam of the telescope with the bright temperature distribution of the source space. Raw data of the dataset were corrected for beam efficiency according to the formula \({\mathrm{T}}_{\mathrm{R}}^{\mathrm{*}}=\frac{{\mathrm{T}}_{\mathrm{A}}^{\mathrm{*}}}{{\mathrm{\eta }}_{\mathrm{m}\mathrm{b}}}\), where ηmb is the result of the main beam efficiency of the antenna obtained from pitch tests on different antennas by using the von Hoerner–Wong structural function efficiency formula whose physical meaning is clearer.2
\[\mathrm{f}\left(el\right)=\mathrm{A}×{\mathrm{e}}^{-{\left(4×\frac{\mathrm{\pi }}{4}\right)}^{2}×\left\{{\mathrm{h}\mathrm{z}}^{2}×{\left[\mathrm{sin}\left(el\right)-\mathrm{s}\mathrm{i}\mathrm{n}\left({\mathrm{e}\mathrm{l}}_{0}\right)\right]}^{2}+{\mathrm{h}\mathrm{h}}^{2}×{\left[\mathrm{cos}\left(el\right)-\mathrm{c}\mathrm{o}\mathrm{s}\left({\mathrm{e}\mathrm{l}}_{0}\right)\right]}^{2}\right\}}\]
3.   Sample description
The data are stored in BUR and FITS formats. BUR files can only be read, processed and analyzed by the GILDAS/CLASS software package developed by the IRAM software group. These files are named in the format “source_MMddhhmmU[L].bur”, where “source” stands for observation target, “MMddhhmm” stands for month, day, hour, minute of the observation, “U [L]” stands for observation results of the upper[lower] sideband. For example, if observation of the target source “0845+025” begun on Dec 31 (at 19:23), the observation files are named “0845+025_12311923U.bur”and“0845+025_12311923L.bur”. In general, a target has 2 or 3 (or even more) groups of files observed at different time periods, which can be used to improve the signal-to-noise ratio of the observed data.
FITS files are data files with an internationally common format generated by processing and merging celestial BUR files of the same target. Table 1 shows the header of a FITS file. As it is an internationally common format, data of a FITS file can be processed and analyzed by using a more generic software package. FITS files are named in the format “sourceU[L][L2].fits”, where“source” stands for name of the observation target, “U [L][L2]” for 12CO (J=1–0), 13CO (J=1–0) and C18O (J=1–0) observation data, and each file is 514MB. FITS files of 12CO (J=1–0) are obtained by processing and merging all the files named “source_MmddhhmmU.bur”, and those of 13CO (J=1–0) and C18O (J=1–0) are obtained by processing and merging all the files named “source_MmddhhmmL.bur”.
We take the observation target “0845+025” as an example. As Figure 1 shows, the source was observed on April 5 (at 13:30), December 30 (at 19:26) and December 31 (at 19:23). Six BUR files were obtained, among which three were for upper sideband and three for lower sideband. The files were then processed and combined according to the two sidebands. Three FITS files were produced which corresponded to the three CO isotopic molecular lines of 12CO (J=1–0), 13CO (J=1–0) and C18O (J=1–0), respectively.

Fig.1   BUR files for the target source “0845+025”
Table 1   Header of a FITS file
BITPIX-32Number of digits used to represent each pixel’s value
NAXIS4Number of axes in the image
NAXIS191Number of pixels along the X-axis
NAXIS291Number of pixels along the Y-axis
NAXIS316384Number of a spectrometer’s channels, that is, the number of data points that make up a spectral line
DATAMIN-2.09E+03KMinimum value in the array
DATAMAX9.10E+03KMaximum value in the array
BUNITK (Ta*)KIntensity
CRVAL18.55E+01degreeReferential X-axis coordinate
CDELT1-8.33E-03degreeIncrement of X-axis coordinate
CRPIX14.60E+01Pixel position of X-axis
CRVAL25.00E+00degreeReferential Y-axis coordinate
CDELT28.33E-03degreeIncrement of Y-axis coordinate
CRPIX24.60E+01Pixel position of X-axis
CRVAL30.00E+00m/sValue of velocity at referential pixel
CDELT31.59E+02m/sIncrement of velocity
CRPIX39.00E+03Referential pixel position of velocity
OBJECT0855+050Target’s name
GLAT8.55E+01degreeGalactic longitude
GLON5.00E+00degreeGalactic latitude
LINE12 CO(1-0)Class of spectral line
VELREF0.00E+00Referential velocity
DATE2012-12-03Date when the FITS file is generated
ENDEnd mark
4.   Quality control and assessment
To eliminate abnormal data such as extraordinarily high values of individual channels, fluctuating baselines, and interfering signals, a series of algorithms and measures were applied for data quality enhancement. Values of a bad fixed channel would be replaced by an average value of its 2 or 3 adjacent channels. As for random bad channels, if the channel’s value is greater than 300K or less than -300K, or 50K greater or less than its previous and following channels, the value of the channel would be replaced with that of its adjacent channel. The programming code is as follows:
Dl=ry[i] - ry[i-1];
Dr=ry[i] - ry[i+1];
if(((Dl > 50.0 )&&(Dr > 50.0 )) || ((Dl < -50.0 )&&(Dr < -50.0 )) || fabs(ry[i])>300)
ry[i] =(ry[i+1] + ry[i-1])/2.0 ;
During the observation, a few groups of data were of fluctuating baseline or with too much noise (Fig.2). In this case, we would compare them with the rms values calculated through theoretical formula (2). If the noise was greater than the theoretical value, the spectral data would be removed.

Fig.2   Example of a spectrum with bad baseline
In addition to automatic assessment and elimination through computer processing, the results were also manually checked and validated to ensure that the noise of 12CO (J=1–0) data was less than or equal to 0.5 K, and those of 13CO (J=1–0) and C18O (J=1–0) data were less than or equal to 0.3K.
5.   Value and significance
The Columbia Sky Survey is by far the most complete sky survey in the world. It uses two 1.2-meter telescopes to perform 12CO (J=1–0)line (115 GHz) unbiased patrols on the Galactic Plane and the results have a significant impact on the international astronomical community. However, the data it produces do not have a high resolution. As the sky areas are not fully sampled, the survey fails to capture many molecular cloud features of the sky. In addition, it only performs 12CO (J=1–0)observations. As self-absorption might occur due to the large optical depth of 12CO (J =1–0)in high-density regions, there might be errors in estimating the properties of the molecular gas (sometimes the error could be as large as an order of magnitude). Finally, the small aperture of the 1.2m telescope cannot ensure the detection of gas content with weak emission. Therefore, it is extremely necessary to enhance such observations to detect weaker, finer components.
Another influential sky survey plan is the Five College Radio Astronomical Observatory (FCRAO), which uses a 14 m aperture telescope with about 46arcsecs for 12CO (J = 1–0)observation. While it uses greatly-improved resolution and sensitivity compared with Columbia, the survey only targets at specific areas: some observations only cover 12CO (J=1–0) line, while others might only include 13CO (J=1–0) line but not C18O (J=1–0) line. It is not a complete sky survey in this sense.
MWISP has several advantages in studying the Galactic molecular gas, such as multispectral observation, high spatial resolution, uniform sensitivity, and large-scale spatial coverage.

Fig.3   A comparison of observations by MWISP 13.7m with those by Colombia 1.2m and FCRAO 14m
This project was supported by the Purple Mountain Observatory and the Computer Network Information Center of Chinese Academy of Sciences. The authors appreciate the Milky Way Imaging Scroll Painting project team for their help and contribution.
Ulich BL & Haas RW. Absolute calibration of millimeter-wavelength spectral lines. The Astrophysical Journal Supplement Series 30(1976): 247.
von Hoerner S & Wong WY. Gravitational deformation and astigmatism of tiltable radio telescopes. IEEE Transactions on Antennas and Propagation 23(1975): 689 – 695.
Scientific Database Office of the Chinese Academy of Sciences. A Collection of Papers on Scientific Databases and Information Technologies. Beijing: China Science and Technology Press, 2012: 64 – 69.
Wells DC, Greisen EW & Harten RH. FITS—A flexible image transport system. Astronomy & Astrophysics Supplement Series 44(1981): 363W.
Hanisch RJ, Farris A, Greisen EW et al. Definition of the Flexible Image Transport System (FITS). Astronomy & Astrophysics 376 (2001): 359 – 380.
Dame TM, Hartmann D & Thaddeus P. The Milky Way in molecular clouds: A new complete CO survey. The Astrophysical Journal 547(2001): 792 – 813.
Ke D & Zhao Y. An image transport system and its FITS basic format. New Technology of Library and Information Service (1994): 22 – 27.
Data citation
1. Lu D, Sun J, Yang Z et al. MWISP data (2011 – 2017). Science Data Bank. DOI: 10.11922/sciencedb.570
Article and author information
How to cite this article
Lu D, Sun J, Yang Z et al. Data of the MWISP Sky Survey (2011 – 2017). China Scientific Data 3(2018). DOI: 10.11922/csdata.2018.0001.zh
Lu Dengrong
data processing, database design and maintenance.
Associate Professor, research area: radio astronomy technology.
Sun Jixian
telescope controlling software development, database design and maintenance.
Associate Professor, research area: radio astronomy technology.
Yang Zherui
Web development, data publishing and sharing, data paper writing.
Assistant Professor; research area: computer science.
Gao Na
equipment maintenance, database maintenance.
Assistant Professor; research area: computer science.
Liu Liang
project application and organization.
Professor; research area: computer science.
Yang Ji
project application, organization and guidance.
Professor; research area: molecular cloud and star formation.
The "integration and sharing of scientific and technological data resources" project of the Chinese Academy of Sciences 12th Five-Year Program (XXH12504-2-01)
Publication records
Published: June 26, 2018 ( VersionsEN1
Released: March 21, 2018 ( VersionsZH2
Published: June 26, 2018 ( VersionsZH3