astro-legacy-archive/iras-low-resolution-all-sky-maps

IRAS Low-Resolution All-Sky Maps

[image]

Galactic Aitoff equal-area maps assembled from the IRAS 0.5 degree beam Zodiacal Observation History data.

Data structure

The source serves two Galactic fields for each of three HCONs and four IRAS bands. Every configuration contains one complete raw 649 by 325 FITS integer image.

The holding has 24 source-named configurations from 24 served files. Each Parquet row is one FITS axis-2 scanline and its fixed-size list is FITS axis 1, so stacking the 325 rows gives the source 325 by 649 map without transposition. The raw signed 32-bit integers and the source FITS cards are stored together. FITS BSCALE, BZERO, BUNIT, and BLANK cards are recorded in the structured decode record but are not applied to stored raw values.

Configuration grammar

ALLSKY_Gx_HnBm.INTE uses GA for the Galactic anticentre field (OBJECT=ACENTERn) and GC for the Galactic-centre field (OBJECT=CENTER n). H1 and H2 are interwoven hours-confirming surveys separated by about 1–2 weeks; H3 was observed after the first six months. B1–B4 select 12, 25, 60, or 100 µm.

Physical interpretation

Mask the configuration's image.decode.blank.value, then apply physical = BZERO + BSCALE * raw; image.decode.bunit.value gives the output unit (Jy/sr). The physics record gives each configuration's wavelength, 0.5° sampling, Galactic Aitoff projection, reference pixels, reference coordinates, instrument, and source beam values.

Embedded map metadata

image.axes is [649, 325, 1]; flattening the list column in row order and reshaping to reversed FITS axes gives (1, 325, 649). FITS CRPIX values refer to one-based pixel centres: for a zero-based NumPy [row, column], subtract CRPIX1 from column + 1 and CRPIX2 from row + 1, then use the pinned Aitoff page's reverse equations with centre longitude CRVAL1. The legacy ATF projection code is not accepted by Astropy WCS. Both spatial CDELT values are -0.5°, so increasing array indices decrease Galactic longitude and latitude; GA and GC have centre longitudes 180° and 0°.

How to use

Install in an isolated environment with python -m venv .venv && .venv/bin/pip install datasets huggingface_hub pyarrow.

import numpy as np
from datasets import load_dataset
data = load_dataset("astro-legacy-archive/iras-low-resolution-all-sky-maps", "ALLSKY_GA_H1B1.INTE", split="train")
raw = np.asarray(data["ALLSKY_GA_H1B1.INTE"], dtype=np.int32)
physical = np.where(raw == -2000000000, np.nan, 0.0 + 185499.9 * raw)
print(raw.shape, raw.dtype, float(np.nanmin(physical)), float(np.nanmax(physical)))

(325, 649) int32 12613993.2 237439872.0

import json, pyarrow.parquet as pq
from huggingface_hub import hf_hub_download
path = hf_hub_download("astro-legacy-archive/iras-low-resolution-all-sky-maps", "ALLSKY_GA_H1B1.INTE.parquet", repo_type="dataset")
meta = json.loads(pq.read_metadata(path).metadata[b"grid_map"])
print(meta["image"]["decode"].get("bunit", {}).get("value"), meta["image"]["decode"].get("blank", {}).get("value"))

JY/SR -2000000000

Interpretation notes

The product page divides the standard Zodiacal Observation History data into three separate sky coverages and two fields per coverage. The standard ZOHF averaged all detectors in a band into a 0.5°-wide fan beam, then applied an eight-second boxcar average to produce a 30′ by 32′ effective beam. This is distinct from the maps' 0.5° pixel sampling.

This historical version 1.0 product was released on 1984-11-26. It is a low-resolution, zodiacal-analysis product rather than a compact-source photometry product. Blank pixels remain as the per-configuration source code and are not Parquet nulls. The pinned product and HCON pages publish no calibration-uncertainty, weighting, or coaddition rule, so analyse H1–H3 as distinct coverages unless another source authority supplies one; use each field's WCS at the Galactic seam and poles.

Preview provenance

The grid preview masks BLANK, applies BZERO + BSCALE * raw, and uses the 2nd–98th finite-value quantiles with the magma colour map. From the repository root at commit b017ad685b638da566f3f9c42398aa75af65ecb2, rebuild it with:

python tools/render_iras_preview.py grid data/iras-low-resolution-all-sky-maps/ALLSKY_GA_H1B1.INTE.parquet datasets/iras-low-resolution-all-sky-maps/preview.png --title "IRAS Low-Resolution All-Sky Maps"

Provenance, rights, and citation

The source authorities below define the released maps, standard-ZOHF beam, HCON timing, Aitoff geometry, and canonical citation.

the IRAS project produced the data. LAMBDA, part of HEASARC at NASA Goddard Space Flight Center, hosts this copy. The linked HEASARC page is a data-use policy, not a separately named formal licence. It states: “HEASARC materials are all available freely for your use.”

Requested acknowledgement (exact text):

We acknowledge the use of the Legacy Archive for Microwave Background Data Analysis (LAMBDA), part of the High Energy Astrophysics Science Archive Center (HEASARC). HEASARC/LAMBDA is a service of the Astrophysics Science Division at the NASA Goddard Space Flight Center.

Please cite the *Infrared Astronomical Satellite (IRAS) Catalogs and Atlases*, vol. 1, *Explanatory Supplement*, 1988, ed. C. Beichman et al., NASA RP-1190 (Washington, DC: GPO).

Source-backed documentation for the statements above:

• —source product description and scientific caveats — SHA-256 e34a1d8c1a3897f0aacf8abe72b7a0ccfa2d7ed654c98d49b9355f2111d15001 checked 2026-09-02.
• —standard ZOHF beam construction and effective resolution — SHA-256 41d511a29d8556d25703e36e8261d0a5dc865f58c6847822604eaa70ae20d69f checked 2026-09-02.
• —HCON confirming-survey timing — SHA-256 3bccf3b1a05fe9fba7b6a5dc29ba826c9d467e8fa5613f024335482b66c64de9 checked 2026-09-02.
• —Aitoff projection geometry and map pixel scale — SHA-256 3eb545f2f72cd657106f95c1bcb15616689a3d152fd00d309ef618bd5fa7650d checked 2026-09-02.
• —online IRAS Explanatory Supplement — SHA-256 7b9f5021b719ac6a1881d9c518974a84d419e7c9eae691873fb1b2c4927ee375 checked 2026-09-02.

Errata and requests

To report an erratum or ask a question, use the dataset Community tab. Confirmed corrections are made in the source conversion tools and published by rebuilding the dataset.