Figure 1. Setting up the source in the ETC
Defining the source spectrum in the ETC interface.
The 2MASS J-band magnitude of GJ 1214 is 9.750, so we will normalize to this wavelength as a Bessel J Vega magnitude in the Renorm
tab in the ETC (Figure 2).
Figure 2. Normalizing source flux
Normalizing the source flux to the known infrared J-band magnitude.
We will not add any spectral emission lines, we will keep the ETC default source shape as a point source to model GJ 1214, and not include any offsets because we are only interested in this single star (do not need to build up an ETC scene). Hence, we leave the Lines
tabs to their default values. The lower ETC panel shows the renormalized M 5 dwarf spectral energy distribution for the calculation (Figure 3).
Figure 3. View of the source
The view of the ETC source simulation for GJ 1214.
Creating the NIRSpec calculation
The simulated spectrum of GJ 1214 is now used in the Calculations tab to estimate signal to noise ratio (SNR) in a NIRSpec BOTS calculation. The NIRSpec BOTS mode uses the 1.6" × 1.6" square aperture, which is one of the Fixes Slits (FS), so a NIRSpec Fixed Slit calculation in the ETC can be used.
The first step is to create a NIRSpec fixed slit calculation in the Calculations tab. Select GJ 1214 as the source in the Scene tab. The initial calculation gives a warning because the default ETC parameters saturate on GJ 1214. For the background, leave the default Medium option in the Backgrounds tab. This bright star will be limited by Poisson noise from the source, background flux level will be negligible in the calculation. In the Instrument Setup tab, select the S1600A1 slit (1.6" x 1.6" square aperture). This is the default slit that is used in all NIRSpec BOTS observations.
Now, we are going to run some test calculations to look at the settings that can be used for GJ 1214. The goal of the science is to observe broad molecular absorption features by binning spectra, as well as narrow molecular absorption features if they exist. NIRSpec's high spectral resolution R = 2700 grating settings would allow study of narrow features. We will start off by selecting the G140H/F100LP grating+filter combination. This selection gives an immediate error, because the wavelength range in the Strategy tab is out of bounds of this grating+filter selection. Updating the measurement wavelength in the Strategy tab clears this error. (The measurement wavelength is the wavelength where the signal-to-noise calculation is presented).
Figure 4. ETC calculation error
Configuring the instrument to a new NIRSpec spectral grating+filter setting gave an error because the default 'wavelength of interest' in the strategy section was outside of the allowed range. Changing the wavelength of interest clears this error.
In the detector settings, we need to change the default parameters to search for sensitivity in the GJ 1214 calculation. The default FULL FRAME detector readout
readout pattern cause very hard saturation in this bright star, even for the minimum of two groups
. We select the S1600A1
, which will read out the full detector in the spectral direction and no signal will be lost by subarray truncation. Next we change the readout pattern to the most rapid option NRSRAPID
, and decrease the number of groups to the minimum of 2. This will allow us to test the acquired SNR in the fastest readout setting that can observe the full spectrum in the G140H/F100LP
instrument spectral setting.
Returning to the Strategies tab, change the default to extract a spectrum centered on the source, with extraction radius set to 0.3”, and use a background aperture that is 0.3" wide. Now that all of the calculation parameters are reset for our use case, run the test calculation by clicking the ‘Calculate’ button to the lower right.
At our target wavelength of 1.15 μm, the SNR is ~50, and the star is in no danger of saturating. The lower calculation plots of results show the aperture-summed source flux as a function of wavelength, and reveals that this calculation gives ~8000 e–/s, which is in no danger of saturating the detector in the short 2.7 s integration time.
Figure 5. Calculation results
Calculation results for a single ETC integrations.
Now return to the Detector Setup
tab and increase the number of groups to create longer integrations. Higher SNR will typically be achieved when the maximum amount of flux is accumulated within a single integration without saturating. After running several test calculations, we find that an integration using NRSRAPID
with 13 groups will accumulate the maximum signal per integration without saturating. This gives an integration time of 12.62 s and a signal-to-noise of ~179 at 1.15 μm in this calculation.
This use case seeks to acquire spectra over the full 1–5 μm wavelength range by observing three transits using the NIRSpec high resolution grating settings. Copying the calculation and changing the settings to the G235H/F170LP filter, changing the Strategy wavelength to 2.45 and decreasing the groups to 12 shows a comparable SNR of 172 in 11.7 s at this setting. Repeating for the G395H/F290LP gives SNR = 174 in 17 groups for 16.2 s at 3.95 μm in one integration with no detector saturation. See Table 1 for a summary of the spectral settings, SNR and integration times for the three observations investigated here.
Figure 6. Results for multiple NIRSpec calculations
Calculation results for the three NIRSpec grating+filter setting calculations.
Table 1. Instrument settings and ETC results for single integration calculations.
|Instrument setting||Calculation wavelength (μm)||Integration time (s)||SNR per integration|
These calculations are for the SNR in a single integration. Because the observation is in the photon-dominated regime and Poisson statistics apply, the SNR will increase as the square root of the number of integrations. The SNR will also increase as the square root of the number of spectral pixels binned into a measurement resolution element. The ETC presents calculations that are for a single spectral pixel.
NIRSpec BOTS calculations in the ETC should be performed for a single integration. BOTS SNR calculations are currently implemented as a FS calculation. Multiple FS integrations are co-added in the ETC to determine the overall SNR, but for BOTS each integration must achieve the SNR required for the science. For BOTS mode, SNR should increase as the square root of the number of integrations since each integration represents an independent measurement.
The transit of GJ 1214b has a peak depth of approximately 0.16% of the source flux, and the transit duration spans approximately 1 hour from beginning to end (T14 = 1 hour; see Figure 1 in the science description). For the G140H/F100LP integrations of 12.62 s, approximately 290 integrations would be necessary to span the time of the transit. Binning in time would give a combined SNR over the transit of 3048. Further binning spectral elements (by 100 pixels) can increase the combined SNR to over 30,000, which corresponds to ~33 parts-per-million accuracy in the transit sensitivity.
|If the SNR in your NIRSpec calculation is 0.00 but all calculation parameters seem to be otherwise okay, check to make sure that the calculation wavelength in your Strategy tab is not located within the gap that lies between the two NIRSpec detectors. This is evident in the results plots in the lower ETC panels.|
Testing NIRSpec target acquisition
After the science spectral calculation is carried out, information on target acquisition needs to be investigated. Target acquisition is performed with the same aperture, the S1600A1 slit using the WATA method. In the Calculations tab, add a new calculation by selecting NIRSpec and Target Acquisition. In the Scene* tab in the panel to the right, select GJ 1214 . In the Backgrounds tab, select the background option to be identical to that of the science calculation (Medium).
In the Instrument Setup tab, the Acs Mode should be WATA, which corresponds to the acquisition through the 1.6" × 1.6" wide aperture. The narrowest NIRSpec target acquisition filter option is the F110W filter, this is used for GJ 1214. In the Detector Setup tab, select the SUB32 subarray, which is the default for acquisition in the BOTS observing mode. Select the readout pattern NRSRAPID, which is the fastest readout available with the shortest exposure time 0.06 s. The remaining default parameters in the Detector Setup tab and the Strategy tab do not need to be changed for the target acquisition calculation.
Calculating the sensitivity for this target acquisition observation shows that the central pixel of GJ 1214 will be saturated in these settings. This calculation was constructed using the narrowest filter for TA and the fastest subarray for readout, resulting in the brightest calculation possible for the wide aperture target acquisition. Because GJ 1214 is found to saturate in the central pixel, the target acquisition executed with these settings may not be accurate to better than 100 mas. A fainter offset star could be used to carry out the target acquisition, but relative offset coordinates between GJ 1214 and the offset target must be known to better than 100 mas to improve TA accuracy in this manner.