NIRSpec MOS Observations of NGC 346
Example Science Program #24
The JWST NIRSpec Team provides this set of step-by-step instructions to create multi-object spectroscopy observations of pre-main sequence stars in the star forming region NGC 346.
As a working example of the multi-object spectroscopy mode of NIRSpec, we introduce a science case of extragalactic stellar astronomy. In this example we study the stellar populations of star-forming region NGC 346 in the Small Magellanic Cloud, with emphasis on the pre-main sequence and red giant branch sources using the NIRSpec MOS observing mode.
In this multi-object spectroscopy example we follow all the steps in the MSA Planning Tool (MPT) guide and MPT-related articles. The example starts with the ingestion of a parent catalog of sources and the creation of candidate sets. Using these sets of sources, we explain how to create several plans and we compare the plans and asses their quality. Finally we show how to create an observation and how to access the visit planner. Each step in the process contains links to the MPT articles for more detailed explanations.
We also provide example ETC calculations to bound the sensitivity estimates for the interesting candidate sources in the NGC 346 region.
This article is divided into two main sections:
APT step-by-step instructions
Creating and ingesting a source catalogv
We base this article on the source catalog produced for the star-forming region NGC 346 by Gouilermis, D, et al. (2006). In this paper, the authors share a catalog with more than 99,000 stars with PSF photometry measured using HST/ACS imaging in filters F555W and F814W.
Data for this paper was obtained under HST GO Program 10248 (PI: A. Nota). Figure 1 shows a drizzled data product for this region, using images from two filters: F555W and F814W. The image was built using 3 color channels: red (F814W), blue (F555W), and green (F814W+F555W). Over-plotted are (in red) the 4 MSA quadrants and (in yellow) the NIRSpec IFU aperture to give an idea of the available field of view for obtaining spectroscopy.
Figure 2 is a representation of the [F555W-F814W, F555W] color magnitude diagram with the Vegamag photometry obtained in NGC 346. The authors observed several stellar populations: main sequence (lower and upper), a red giant branch, and a population of pre-main sequence stars.
Sources of interest for the MSA follow up spectroscopy are certainly the red giant branch stars, and the pre-main sequence stars. We identified subsets of those groups and assigned them a weight value in the source catalog. We modified the public NGC 346 catalog to preserve the same ID numbers. We transformed the equatorial coordinates to units of degrees and we removed sources with magnitude error larger than 0.1 mag in either filter. We also added a weight value to each source. The weight values are listed in Table 1. Note that for this study we are mostly interested in obtaining the IR spectra of the pre-main sequence stars, hence the higher weight value.
Table 1. Weight values assigned to the stellar populations
|Source type||Weight value|
|upper main sequence||UMS||10|
|red giant branch||RGB||40|
This catalog is available below.
You may download the catalog and use it in APT to follow our example. For this exercise, we are using APT version 25.4.3.
We start a new observation in the Observation Folder in APT. Under MSA Planning Tool, we import the catalog of sources. The catalog is in white space separated format. Figure 3 shows the window used to ingest the catalog onto MPT. We name the catalog 'ngc346'.
The catalog contains 83,184 sources
Following Figure 4 in the NIRSpec MPT - Catalogs article, we next need to specify the Catalog 'Astrometric Accuracy' and the 'Pre-Image Availability'. We assign it an accuracy of 10 mas because it was generated using HST imaging. We set 'Pre-image availability' to 'not requiered'.
HST/ACS drizzled products generated with images obtained in the past decade deliver an astrometric accuracy of 10 mas or less, depending on the filter. The same applies to HST/WFC/UVIS.
Creating candidate sets
The main scientific goal in this example is obtaining IR spectra of pre-main sequence stars. A secondary goal is to measure some red giant branch stars.
Our primary sources are the pre-main sequence stars, so we create a candidate set called 'pms' that contains only those sources. This is easy to do in MPT by highlighting the parent catalog, and selecting 'New Candidate Set'. A pop-up window will allow us to select a sub-set of sources with a Weight of 100. This candidate set contains 3,319 sources.
We next create a candidate set that contains both types of sources: pre-main sequence stars and red giant stars. In this case, we highlight the parent catalog, and select 'New Candidate Set'. In the pop-up window we select a sub-set of sources with a Weight between 40 and 100. We will call the new candidate set 'pms-rgb'. After clicking 'Make Candidate Set', we find that it has 5,881 sources as shown in Figure 4.
As a "quick look" tool for pre-planning purposes, STScI has developed the General Target Visibility Tool which is a Python command-line tool for calculating visibility of a given target (RA, Dec). This code provides the aperture position angle (APA) for NIRSpec as a function of time.
Using this tool it is possible to obtain ranges of viable angles to use in the MPT Planner.
Please note that the ability to schedule any given target observation is more complex that just its visibility. It also involves the availability of guide stars as a function of time and other observatory constraints.
For any NIRSpec MSA observation, it is highly recommended to experiment with a range of aperture position angles in the MSA planning tool, as opposed to one single fixed angle. This recommendation is made to encourage users to understand the visit and timing overheads between plans created at different angles.
We now move to the Planner pane and create a series of plans.
We assign an Aperture PA value of 120.0. The Primary Candidate List is our list 'pms' and the Filler Candidate List would be 'pms-rgb'. By keeping the primary sources in the Filler list, we increase the chance to observe more pre-main sequence sources.
For the slitlet setup, let us use the default of a 3 Shutter Slitlet and the Entire Open Shutter Area as centering constraint. We set the Dither type as Fixed Dither and use a 5 shutter offset in the dispersion direction, which corresponds to ~1.25 arcsec.
For the Exposure Setup, let us start with a high resolution grating/filter combination G140H/F070LP, NRS IRS2 readout pattern, 10 groups and 2 integrations.
Under Parameters, we check 'Use Weights' and set Number of configuration to 10. This will generate 5 target sets, each with two dithers as explained in the article NIRSpec MPT - Planner. We name this first plan ngc346-01.
Click on Generate Plan, and after a few seconds the plan should be finished. When finished, we land in the Plans pane where plans are reviewed.
Lets go on to make some additional plans first. Returning to the Planner, you will see that the parameters from the last plan are still selected. This was done to make it easier to quickly modify a few values and generate additional plans.
In this plan we change the centering constraint to 'Tightly Constrained' to see the differences with the previous plan. We keep all other parameters the same and call this plan ngc346-02
We now use the same parameters as with plan ngc346-01 but this time we use a finer search grid by changing the Search Step Size from the default 30" to 5". Note how the number of pointings increases considerably. The computation takes a little longer, but it usually results in better performance in terms of the numbers of observed targets. We name this plan ngc346-03.
Now let's return to the same parameters as with plan ngc346-01 but this time add exposure specifications in all four high-resolution grating/filter combinations. This will generate four times the exposures of previous plans (for each of the three gratings). We name this plan ngc346-04.
We now use the same parameters as with plan ngc346-01. For this new plan, we change the default slit shape to a single shutter. We name this plan ngc346-05.
We now use the same parameters as with plan ngc346-01 but add a nod within the slitlet which will create three exposures per pointing. We name this plan ngc346-06.
Users are encouraged to experiment with several configurations in the MPT Planner and then use the MPT Plans pane to asses the quality and completeness of these plans. The "success" of a plan is a function of the spatial distribution of sources in the sky, as well as the properties of the MSA.
Comparing and assessing plans
We now move onto the Plans pane where we can review plan results.
It is interesting to compare the target set size for the 10 configurations in each plan. Table 2 shows this comparison. The number of observed targets does not change much between exposures. However, modifying the shape of the slitlet does increase the number of observed sources as shown with plan ngc346-05. This might be useful in some science cases.
Constraining the centering of sources, also affects the number of sources that are observed. By making the allowed area smaller in each shutter, less sources are observed per configuration as shown with plan ngc346-02.
For proper correction for sky emission, it is advisable to nod in slitlet during the dither setup as we show in the plan ngc346-06. Nodding also increases the number of exposures in which a source is observed.
Table 2. Target size in ten configurations for each plan
|Configuration||Plan name (ngc346-)|
Table 3 compares the number of targets that are in at least one exposure classified according to source type: primary, filler, contaminants, and all targets. These plans have many primary targets observed because the input catalog has a high density (sources per square arcmin). This table is representative of the multiplexing results that can be achieved in this region. Plan number 5 has much higher multiplexing because a single shutter per observed source is used (see description of plan cases above). Figure 6 shows the coverage associated with plan ngc346-06.
Table 3. Number of targets in at least one exposure in the 10 MSA configurations for all six generated plans
|Target Type||Plan Name (ngc346-)|
Aladin is the visualization tool used in MPT. In Figure 7 we display the observed sources from two plans in order to check for differences. It is always useful to display the finder image as one of the Aladin planes.
Creating an observation
We assume that plan ngc346-06 is the best for our science goal and so we use it to create a placeholder observation to be submitted with our proposal. In the Plans pane, we select this plan and click on the 'Create Observation' button. This populates the fields in the observation template. The visits are located in the Tree Editor inside the Observation Folder. Four visits were created as shown in Figure 10.
At the bottom of the 'Visit Planner' view there are three buttons: 'Update Display', 'Reports' , and 'Print'.
Once we update the display, APT will study the availability of the observations and try to schedule the visits. In our example, APT was able to schedule these observations in September 2019. Three reports are available for each visit: namely: the 'Guide Star Availability Analysis for Visit', the 'Sun Roll Analysis for Visit', and the 'Total Roll Analysis for Visit'.
As an example of the graphical output, Figure 11 shows the Sun Roll Analysis.
The APT file containing the plans described above can be found in the link below.
|Example APT file. This is the APT file used to demonstrate the complete example in this article.|
This science use case example was generated using APT version 25.2.3. You may encounter some differences in the user interface if you load the file in the latest version of APT.
ETC step-by-step instructions
The signal to noise calculations are performed using the JWST Exposure Time Calculator. For this science case, we provide a few examples to illustrate how to define sources of interest similar to the ones in Figure 2.
Scene and Sources
From Figure 2, we select two sources: a pre-main sequence source with magnitude 24 mag in the F555W band and a red giant star with F555W = 20 mag. We model these sources with Phoenix Stellar Models in the Source Editor by selecting models "M0V 3750 4.5" and "K5III 4000 1.5". Among other possibilities, users can use a black body spectral energy distribution, or upload their own flux density file.
Here we illustrate the example of the red giant star. Figure 12 shows the ETC Scenes and Sources tab where the Continuum is defined. For the normalization, we assume that this source has a Johnson V magnitude of 20 mag (Figure 13). The plot at the bottom shows the spectrum of the source.
Under the Calculations tab, the user can set the background, the instrument, the detector, and the observing strategy. For the background we use a Medium (the default value), under instrument setup (Figure 14) we select as an example a calculation using the high resolution grating G140H, and NIRSpec filter F070LP with a wavelength range of 0.7 to -1.27 μm. We keep the same slitlet shape as the one used in APT: 3 shutters. See below for a discussion of the MSA Location selection.
For the Detector setup we use a NRSIRS2 readout pattern, 10 groups , 2 integrations, and one exposure as an example. This leads to a total exposure time of ~14 minutes as shown in Figure 15. In this example, these values were selected so that the bright (20 mag) giant source will not saturate in the requested total exposure time.
The number of groups and integrations should be set based on the desired signal to noise and total exposure time. Saturation of bright sources should always be avoided. It is always important to check for warnings in the ETC reports.
Once those parameters have been selected, we perform a calculation and obtain a signal-to-noise ratio (SNR) of 51.44 as shown in Figure 16.
Measured Flux as a function of MSA quadrant
The 4 MSA quadrants are physically separated and the gaps between them block out the light from the source. This affects part of the source spectrum, specially when using high resolution gratings. Figure 17 shows the extracted flux from the source, assuming that the slitlet was placed in the center of each MSA quadrant (Q1, Q2, Q3, and Q4).
Extinction laws comparison
Under the Continuum tab in the Scene and Sources section of the ETC it is possible to assign and extinction law and magnitude for the sources defined within the scene. Here we calculate the signal to noise for a variety of cases that may be useful for the NGC 346 study. Figure 18 shows the available Extinction Laws provided by the ETC. Although NGC 346 is in the SMC, we will compare calculation results for the following extinction laws: LMC Average, LMC 30 Dor, and SMC Bar, with a range of extinction magnitudes in the V bandpass. Table 4 present the results of the calculations.
Table 4. Extinction law comparison
|SED model||Normalization||Groups / integrations / exposures||Extinction law||Extinction magnitude||SNR|
|K5III 4000 1.5||V = 20 Vegamag||10 / 2 / 1||LMC Average||0||51.44|
|LMC 30 Dor||0||51.44|
|M0V 3750 4.5||V = 24 Vegamag||10 / 2 / 4||LMC Average||0||11.52|
|LMC 30 Dor||0||11.52|
The SNR values fall in the same range for a given extinction magnitude, regardless of the chosen extinction law. Note the increase in the number of exposures in the case of the source with spectral type M0. This number was increased in order to obtain a SNR larger than 10 and it is just for demonstration purpose.
The star-forming region NGC 346 in the Small Magellanic Cloud with Hubble Space Telescope ACS observations. I. Photometry.
ACS/WFC Revised Geometric Distortion for DrizzlePac
Published October 4, 2017