SIMLA: Spitzer/IRS Mapping Legacy Archive

• SIMLA is a comprehensive archive of uniformly processed mid-infrared spectral cubes (5.2–38.4 μm) from Spitzer/IRS, covering diverse astronomical sources.
• It employs rigorous data reduction techniques including zodiacal light, dark-current subtraction, and bad-pixel removal to ensure high-fidelity FITS products.
• The archive supports quantitative spectroscopy and legacy science by providing calibrated data for ISM diagnostics, shock studies, and cross-mission research.

The Spitzer/IRS Mapping Legacy Archive (SIMLA) is a comprehensive, uniformly processed data set of mid-infrared (MIR) spectral cubes (5.2–38.4 μm, $R \sim 60$–130) produced from the entirety of low-resolution, mapping-mode fixed-target observations obtained by the Spitzer Space Telescope’s Infrared Spectrograph (IRS). SIMLA encompasses spectral maps for several hundred spatially resolved and unresolved astronomical sources, including galaxies, molecular clouds, supernova remnants, and H II regions. Its workflow incorporates a rigorously constructed pipeline for modeling and subtracting time-variable foregrounds, robust bad-pixel removal, and precise cube assembly, resulting in high-fidelity, ready-to-use FITS data products suitable for quantitative spectroscopic studies and legacy science applications (Donnelly et al., 13 Dec 2025).

1. Scientific Motivation and Scope

Mid-infrared spectroscopy in the 5–38 μm range provides unique diagnostic access to various constituents and physical conditions of the interstellar medium (ISM). Key MIR diagnostics include:

• Polycyclic Aromatic Hydrocarbon (PAH) bands (6–17 μm): Trace the small-grain and aromatic component of ISM dust.
• H₂ rotational lines (e.g., at 28 μm): Directly probe physical and excitation conditions in molecular gas.
• Atomic and ionic fine-structure lines (e.g., [Ne II], [S III]): Diagnose ionized gas, star-formation intensity, and shocks.

SIMLA was developed to fulfill four primary scientific objectives:

1. Systematically harvest the entire Spitzer/IRS low-resolution mapping-mode archive to generate a unified set of 3D spectral cubes.
2. Construct per-cube, depth-optimized backgrounds, crucial for enhancing S/N, particularly in cases lacking dedicated off-source observations.
3. Subtract the time-variable zodiacal foreground using a combination of models and data-driven approaches.
4. Deliver to the community both the fully calibrated cubes and tools for residual offset assessment and adjustment.

SIMLA's astrophysical target set is broad, spanning nearby and distant galaxies (spirals, dwarfs, LIRGs, ULIRGs), Galactic H II regions, photo-dissociation regions, supernova remnants (e.g., Cassiopeia A), molecular clouds, star-forming complexes (e.g., Eagle Nebula), planetary nebulae, photodissociation fronts, and diffuse Galactic sightlines.

2. Instrumental Configuration and Observational Design

The low-resolution modules of Spitzer/IRS collectively map the $$5.2\,\mu\mathrm{m} \leq \lambda \leq 38.4\,\mu\mathrm{m}$$ spectral range with

$$R \equiv \frac{\lambda}{\Delta\lambda} \approx 60 \text{–} 130,$$

where $\Delta\lambda$ is the FWHM of an unresolved line.

The instrumental layout comprises:

• Short-Low (SL) module:
  • SL2: 5.2–7.6 μm, 57″×3.6″ slit, 1.8″ px⁻¹, $R\sim60$–72
  • SL3 (bonus): 7.4–8.7 μm
  • SL1: 7.5–14.7 μm, $R\sim72$–127
• Long-Low (LL) module:
  • LL2: 14.3–21.1 μm, 168″×10.5″ slit, 5.1″ px⁻¹, $R\sim57$–63
  • LL3 (bonus): 19.9–21.1 μm
  • LL1: 20.6–38.4 μm, $R\sim63$–126

Spatial and spectral sampling in assembled cubes directly reflect the slit width, pixel scale, and observer-specified map step. Fully sampled maps typically employ slit steps of $≤\frac{1}{2}$ slit width, providing spaxels of 1.8″×1.8″ (SL) or 5.1″×5.1″ (LL). Spectral grid spacing ensures at least two slices per native $\Delta\lambda$.

3. Data Set Composition and Content

The initial SIMLA release (SIMLA-1) includes all low-resolution mapping-mode Area-Observation Requests (AORs) from Spitzer/IRS, excluding moving-target Solar-System programs. The archive comprises:

• Several hundred unique objects
• Over 5000 hr of mapping
• Sky coverage: 2.94 deg² total (0.9 deg² SL, 2.34 deg² LL)

For each AOR, up to three spectral cubes (orders 1, 2, and bonus 3) per module are produced. Typical cube properties are:

• Spatial extent: tens to hundreds of arcseconds per side
• Spatial grid: 1.8″ or 5.1″ per pixel
• Spectral resolution elements: ~60–130
• $\sim$1000 wavelength samples (module dependent)

AORs are selected based on processing through the Spitzer/IPAC S18.18.0 pipeline and stratified by module, RAMPTIME, and LL-bias era (pre/post campaign 45).

4. Data Reduction and Calibration Pipeline

SIMLA addresses foregrounds and detector systematics in Basic Calibrated Data (BCD) through a multi-component subtraction and bad-pixel mitigation workflow:

4.1 Zodiacal-Light Subtraction

A time- and pointing-dependent zodiacal emission model (ZEM; Kelsall et al. 1998) is evaluated at each AOR. Given that pipeline “super-darks” encode the zodiacal level at the Continuous Viewing Zone (CVZ), the zodiacal correction for each AOR is

$$\mathcal{Z}_{k,ij}(\lambda) = W_{ij}^{-1}\left[ Z_k(\lambda) - a(t)\,\widehat{Z}_{\rm CVZ}(\lambda) \right],$$

with $Z_k$ as the ZEM spectrum, $\widehat{Z}_{\rm CVZ}$ the normalized CVZ model, $a(t)$ a campaign-dependent amplitude, and $W_{ij}^{-1}$ a projection operator onto 2D BCD frames.

4.2 Baseline Dark-Current Subtraction

Residual dark-current patterns, including model residuals, are binned by ZEM intensity: $$s_{z',ij} = \mathrm{median}\left[d_{n,ij}-\mathcal{Z}_{n,ij}\right] \quad \text{for}\;z_L<z_n<z_H,$$ and interpolated to provide an AOR-specific baseline $S_{k,ij}$.

4.3 Time-Variable Pixel Pedestal Correction (“Shard” Stacking)

Wavesamp regions are divided into five spatial “shards.” Off-source shards are selected by:

• WISE W3 surface-brightness $|I_{\rm W3}|<0.1\,\mathrm{MJy/sr}$
• BCD-extracted median shard intensity $|I_{\rm shard}|<2\,\mathrm{MJy/sr}$ post-subtraction of $\mathcal{Z}+S$
• Zodiacal intensity $|\Delta z|<10\,\mathrm{MJy/sr}$ and time $|\Delta t|<10\,$days matched to the target

Up to 50 shards are mean-combined with $1.5\sigma$ clipping to yield $D_{k,ij}$.

The total background subtraction applied is

$$B_{k,ij} = \mathcal{Z}_{k,ij} + S_{k,ij} + D_{k,ij},$$

subtracted from each BCD frame.

4.4 Bad-Pixel Detection and Removal

CUBISM’s global pixel-diversity flagger rejects pixels exceeding $4\sigma$ deviations in more than 40% of contributing BCD fragments, exploiting mapping redundancy for robust, automated outlier removal.

5. Cube Assembly and Calibration

CUBISM (Smith et al. 2007) reconstructs regularly gridded $(\alpha,\delta,\lambda)$ cubes by:

1. Accounting for wavelength-dependent slit rotation and centroids.
2. Fractionally distributing each BCD’s pixel fluxes into appropriate spaxels (“3D drizzle”).
3. Propagating input error estimates to build both flux and 1$\sigma$ uncertainty cubes.

Automatic calibrations include:

• Wavelength solution application from IRS calibration files
• Extended-source flux calibration via slit-loss corrections
• No ad-hoc smoothing; final resolution and S/N are set by native detector sampling and mapping redundancy (Donnelly et al., 13 Dec 2025).

6. Validation and Quality Assessment

SIMLA cubes undergo quantitative quality checks:

• Dark-region spectral extraction: Spectra from “dark” spaxels are tightly distributed around 0 MJy sr⁻¹, with rms $\lesssim 0.1$ MJy sr⁻¹, indicating effective background subtraction.
• WISE W3 photometry comparison: Synthetic W3-band fluxes from cubes, when compared to matched 5″-radius measurements from WISE W3 maps, agree to $\lesssim 20\%$ above 0.1 MJy sr⁻¹. Below this flux, results are treated with caution due to faintness limits.
• Noise characterization: Standard deviations of dark spaxels, binned by RAMPTIME and spectral order, establish module-specific noise floors (typically 0.02–0.2 MJy sr⁻¹) and are logged in cube metadata.

7. Data Products, Accessibility, and Impact

For each mapping-mode AOR (by order: SL2, SL3, SL1, LL2, LL3, LL1), SIMLA-1 releases:

• Multi-extension FITS files:
  • Primary HDU: flux cube [MJy sr⁻¹]
  • “UNC”: 1$\sigma$ uncertainty cube
  • “MSK”: bad-pixel mask
• Dark-mask files delineating spaxels for zero-level offset assessment
• Associated README/ASCII headers (AOR ID, program details, observing parameters, background/depth, $\sigma$-clipping settings, etc.)
• Metadata compliant with IRSA/Spitzer standards (WCS, MJy sr⁻¹ units, astrometry, wavelengths in μm)

Access is provided via the NASA/IPAC Infrared Science Archive [https://irsa.ipac.caltech.edu/data/SPITZER/SIMLA/], supporting searches by AOR ID, target, coordinates, or proposal identifier.

SIMLA facilitates rapid, contextualized MIR spectral studies for hundreds of Galactic and extragalactic targets. It provides a foundation for ISM dust analysis, shock and PDR studies, extragalactic star-formation calibration, and cross-mission data-mining (e.g., Herschel, JWST), as well as template creation for machine-learning applications. Complete documentation enables the community to reproducibly modify any reduction step, ensuring long-term scientific utility in the JWST era and beyond (Donnelly et al., 13 Dec 2025).