Frequently Asked Questions

1. Why is the signal to noise ratio (S/N) from LAMOST catalog or the .fits file sometimes zero?

In the following two cases, the S/N is zero. 

(1) The calculated S/N is 0 with the formula of “S/N = Flux * (Inverse Variance) ^ 0.5”.

(2) When we estimate S/N of a range of a spectrum, if the number of pixels with “ormask = 1” or “Inverse Variance = 0” in the range is over the half of the total number of this spectral range, the S/N is artificially set to zero.


2. When you search data on LAMOST data release website (for example, the website of, you can see that there are over ten ways to obtain data, and there is a button of “Search” below each retrieval method. What is the logical relationship between the “Search” buttons? Which “Search” button can be used to submit when over one retrieval way are used?

The relationship between the “Search” buttons on LAMOST data retrieval website is logic “AND”, you can use any “Search” button to submit if over one retrieval way are used.


3. Dose the new version of LAMOST data (such as DR7) include all data of previous years (such as DR6)?

Below is the data release policy of LAMOST.

(1) Quarterly data release does not include data of previous quarters.

(2) Domestic and international data release include all data from pilot survey.


4. What is the method of the atmospheric parameters of LAMOST medium-resolution spectra?

The LASP software and CNN algorithm are used to estimate atmospheric parameters for LAMOST medium-resolution spectra, respectively, and the two results are all published in the catalog of medium-resolution spectra. The LASP software also provides uncertainties of atmospheric parameters, which not only include the algorithm error but also consider the errors introduced during the process of observation.

On data release website, there is a document named “Release Note” in “DOCUMENTS” menu shows the analysis results of atmospheric parameters for both low-resolution and medium-resolution spectra, such as the distribution of atmospheric parameters in different S/N bins, the comparison results of atmospheric parameters for common stars of LAMOST and SDSS or LAMOST and APOGEE.


5. What is the unit of LAMOST spectral flux?

Relative flux calibrations have been performed for LAMOST low-resolution spectra, which change the shape of spectral continuous with flux standard stars, but no flux calibration has been done for medium-resolution spectra. Therefore, fluxes of LAMOST spectra have no units no matter low resolution or medium-resolution.   


6. Why does the “subclass”from LAMOST catalog include other values except spectral types, for example “double star”?

The“subclass”field provides the classification results when match LAMOST spectra with the template spectra, and it also includes other classification results except the spectral Harvard classification such as luminosity type, “double star” and “carbon star”. Therefore, do not equate the value of“subclass”field with the Harvard classification.


7. Why the row number of the “LAMOST MRS General Catalog” is larger than the total number of Medium-resolution non time-domain and time-domain data given on the Home page of data release website?

In the “LAMOST MRS General Catalog”, a single exposure spectrum is an entries, which is the same for a coadded spectrum, and the different of total number mentioned in this problem mainly lie in the statistical method on website is different from the “LAMOST MRS General Catalog”. On the website, medium-resolution spectra are divided into two types, i.e., time-domain data and non time-domain data, and different statistical methods are used for the two types of spectra.

(1) time-domain data: the total number on website does not include the number of coadded spectra, and only record once for the red and blue spectra in a single exposure for each target.

(2) non time-domain data: only record the number of coadded spectra for each target.


8. What is the calculation method of the “mjd” and “lmjd” fields?

“mjd” is the modified Julian day, and the “lmjd” is the local modified Julian day.

mjd = [jd-2400000.5],where square brackes represent round down.

lmjd = [mjd + 5/6],where square brackets represent round down.

So, “lmjd” of LAMOST is one more than “mjd”.

Note that “MMMMM” in the name of .fits file (for example, “med-MMMMM-YYYY_spXX-FFF.fits”) is “lmjd” not “mjd”.


9. How to obtain observation time of LAMOST medium-resolution spectra? (To the minute)

(1) For coadded spectra, LAMOST provides the local modified Julian minute (lmjm) list, which records the starting time of each exposure, and you can obtain the it with the following method:

The first and second extensions provide the blue and red coadded spectra, respectively, and the “lmjmlist” field in their header files gives the lmjm list, and the lmjm of each exposure in “lmjmlist” is separated by “-”.

(2) For single exposure spectra, LAMOST provides the lmjm of starting time of each exposure, and you can obtain it with the following methods:

i. The extensions after the second extension in the .fits file of medium-resolution spectra store the single exposure data, and the number in the name of each extension after the second extension is the lmjm of the starting time of this exposure. '

ii. The “lmjm” fields in each extension after the second extension is the lmjm of starting time of this exposure.

iii. In LAMOST MRS General Catalog and LAMOST MRS Parameter Catalog, the “lmjm” field provides lmjm for single exposure spectra.


10. What is the difference between the “mobsid” field and the “obsid” field in the parameter catalogs of LAMOST Medium-Resolution Spectroscopic Survey ?

The “mobsid” field is the unique ID of each medium-resolution spectrum, and the “obsid” field is the ID of each observation for each target. For a single exposure spectrum and a coadd spectrum, the “mobsid” is different. For a single exposure spectrum, the “mobsid” equals to “obsid” + “lmjm” + “band”, and it equals to “obsid” + “band”. In “mobsid”, “lmjm” is the local modified Julian day, and “band” has two values of B and R, which represents it is a blue band spectrum or a red band spectrum.


11. In what cases, may the spectral flux be unreliable ?

In a FITS file, there are also inverse variance, andmask and ormask except flux for each wavelength. Generally speaking, the flux is likely not reliable when one of the below cases happens, and the removal of corresponding spectral points is recommended.

i. the inverse variance equals to 0

ii. the flux is less than 0

iii. the andmask is not 0 or the ormask is not 0 for a wavelength


12. After September of 2018, what is the naming rule of the “planid” field in the medium-resolution catalogs ?

i. The first class:the medium-resolution plan

   a. The naming rule: NT/TDhhmmssN/SddmmssXnn

           NT = Non time domain

           TD = Time domain

           N/S = North/South

           The third character from the end represents the sub-project name of LAMOST Medium-Resolution Spectroscopic Survey,  and it has the following eight values:

            if X = K:  it represents the Kepler region survey

            if X = B:  it represents the binary survey

            if X = M:  it represents the Milky Way survey

            if X = T:  it reprensents the TESS follow-up survey

            if X = S:  it represents the star forming region survey

            if X = C:  it represents the open cluster survey

            if X = N:  it represents the Galactic nebulae survey

            if X = H:  it represents the subclass of the Kepler plan

            The last nn: the assigned number of a plan

  1. Examples:

          TD193637N444141K01 and NT112445N020547M01

ii. The second class: the high-resolution test plan

    a. The naming rule: HIPXXXXX+HNNNN +AB NN

           HIPXXXXX: the name of the center star of a plan

            HNNNN: the name of the fiber unit

            AB: the project ID

            NN: the assigned number of a plan

    b. Examples:

          HIP39189H272201 and HIP31511H192101

iii. Other cases

This class is generally internal instrument test plan, and the number of such plans is very few. These plans were named with no strict rules, and their data are usually not published. If their data are found to have potential application value, they will be published together with the scientific observation data. If you have any questions about those data, please contact with us with