In-orbit Performance of the Near-Infrared Spectrograph NIRSpec on the James Webb Space Telescope

TL;DR

This paper assesses the in-orbit performance of JWST NIRSpec from commissioning through early science, detailing hardware performance (detectors, wheels, MSA), observatory factors (OTe quality, thermal background, ACS), and scientific capabilities (PCE, calibration, wavelength accuracy, time-series stability, TA). It reports excellent sensitivity and calibration readiness across modes, with the instrument meeting or exceeding pre-launch expectations, while also outlining remaining pipeline challenges and operational considerations for planning NIRSpec programs. The results underscore NIRSpec as a transformative, high-sensitivity near-infrared spectrograph with a robust calibration framework and precise target acquisition capabilities. The study provides practical guidance for observers, including considerations for saturation risk, MSATA planning, and catalog anchoring, to maximize scientific return in Cycle 1 and beyond.

Abstract

The Near-Infrared Spectrograph (NIRSpec) is one of the four focal plane instruments on the James Webb Space Telescope. In this paper, we summarize the in-orbit performance of NIRSpec, as derived from data collected during its commissioning campaign and the first few months of nominal science operations. More specifically, we discuss the performance of some critical hardware components such as the two NIRSpec Hawaii-2RG (H2RG) detectors, wheel mechanisms, and the micro-shutter array. We also summarize the accuracy of the two target acquisition procedures used to accurately place science targets into the slit apertures, discuss the current status of the spectro-photometric and wavelength calibration of NIRSpec spectra, and provide the as measured sensitivity in all NIRSpec science modes. Finally, we point out a few important considerations for the preparation of NIRSpec science programs.

Figures (8)

• Figure 1: Left: The optical path through the NIRSpec instrument. The dispersion direction is out of the page. Right: CAD rendering of NIRSpec with its major mechanisms identified. The dimensions of the NIRSpec optical assembly are 1.9 m $\times$ 1.4 m $\times$ 0.7 m, with a total mass of 196 kg.
• Figure 2: Median total noise signal (in e$^-$) as a function of effective integration time for the two NIRSpec detectors and three readout modes: IRS$^2$, traditional full frame (TRAD), and ALLSLITS subarray.
• Figure 3: NIRSpec point-source continuum sensitivity in MOS/FS (left) and IFS (right) mode, derived from in-orbit measurements and assuming a well-centered source in a microshutter or an IFU slice, respectively. The sensitivity for the S200 slits in FS mode is similar to the one in MOS mode. The plots show, for each disperser and as a function of wavelength, the flux required to reach $\rm S/N=10$ per spectral pixel in 10,000s. More specifically, the computations assume 10 NRSIRS2RAPID exposures of 1006.7 s each (70 groups of 1 frame), using the methodology described in Appendix A of jakobsen22.
• Figure 4: The spectra of standard stars 1808347 (left) and P177D (right), extracted with the NIRSpec-internal reduction pipeline, and compared to the CALSPEC model to verify the flux calibration. Both spectra were taken through the S1600A1 slit using the F290LP/G395H configuration. The bottom frames show the residuals. The observations of P177D were obtained without a target acquisition step, and the systematically lower flux compared to the model is most likely due to imperfect centering in the S1600A1 aperture, which can easily cause aperture losses of 1-2%.
• Figure 5: Wavelength-calibrated spectra of the spectrophotometric standard star 1808347, obtained in FS mode through the S200A1 slit (see Table \ref{['tab:modes']}), and using the medium-resolution gratings in three of the four spectral configurations (F100LP/G140M is not shown). These data were obtained during the NIRSpec commissioning program 'Spectrophotometric Sensitivity and Absolute Flux Calibration' (PID1128).