Skip to main content

Updated Hubble Spectroscopic Legacy Archive (HSLA) Includes More Robust Spectra

STScI Newsletter
2025 / Volume 42 / Issue 02
View All Newsletter Articles

About this Article

John Debes (debes[at]stsci.edu) and the HSLA team

Published November 19, 2025

Illustration shows large text, HSLA, above smaller text, Hubble Spectroscopic Legacy Archive. Jagged rainbow-colored lines appear behind HSLA. The background is gray with various astronomical targets in black.
A new version of the Hubble Spectroscopic Legacy Archive (HSLA) automatically defines individual targets, groups multiple observations of a single target into associations, and produces a classification for each target, making it significantly faster to query the data and refine analyses.

The new Hubble Spectroscopic Legacy Archive (HSLA) provides scientifically validated coadded spectra (coadds) of individual targets that have been observed by the Cosmic Origins Spectrograph (COS) and the Space Telescope Imaging Spectrograph (STIS) over their operating lifetimes. HSLA uses data available in the Mikulski Archive for Space Telescopes (MAST) and automatically produces coadds whenever new data become publicly available or when there are newly recalibrated datasets.

A key feature of the new HSLA is that it automatically defines individual targets, groups multiple observations of a single target into associations, and produces a classification for each target. Target associations make use of the dataset coordinates accounting for proper motions, and use SIMBAD, NED and the Phase II observing proposals to determine which datasets should be associated with each unique target. Then, using the SIMBAD, NED, or Phase II keywords, a detailed classification is determined for an object to aid in the spectroscopic study of classes of astrophysical objects.

The classifications consist of three tiers of detail, mapped to the closest Unified Astronomy Thesaurus (UAT) concept. For example, the HSLA target Markarian 817 is classified at Tier 1 as a Galaxy, Tier 2 as an Active Galaxy, and at Tier 3 as a Seyfert. It corresponds to the UAT object Seyfert Galaxies (1447). The HSLA has spectra for over 3,000 individual stars and over 2,000 individual galaxies. A full listing of classifications is available on the HSLA website.

For each individual target, HSLA also provides a human-readable metadata file with key information that can be used in searches and further exploration of the data. The metadata file includes the target name and unique identifiers, the target coordinates in J2000, information about the target's name and classification information, redshift or radial velocity information, and a summary of the programs and instrument modes that are included in the target association.

HSLA data products, including quicklook coadds (_aspec files), coadded single grating products (_cspec files), metadata files (_metadata files), and code output logs (.trl files) are available on the MAST Portal, HST Mission Search Form or via astroquery.

Detailed Use Case Examples to Preview a Range of Unique Search Options

The HSLA website includes details about which option is preferred for different use cases. Since the HSLA is fully automated, it will be updated routinely as new HST spectroscopic data are taken or if data are reprocessed with improved calibrations. Data access may change with time and users are encouraged to visit the HSLA webpage for the latest information on how to access data.

An example aspec file for β Pictoris is shown in Figure 1. β Pictoris is a well-known nearby A-star that possesses a debris disk and two giant planets. The stellar spectrum shows evidence for both stable and time-variable absorption features due to exocomets that break up and evaporate. The aspec file covers multiple COS and STIS gratings from many individual observing programs and over four orders of magnitude in flux.

A HSLA spectrum of Beta Pictoris. The y-axis is labeled for energy per second per unit area per unit wavelength (flux) from ten to the negative fourteenth to ten to the negative tenth. The x-axis is labeled in wavelength (angstroms), from 1,500 to 3,000. The jagged lines start at bottom left, rising almost straight up around 1,500 angstroms, largely remaining along the top to 3,000 angstroms, with several significant valleys near 2,500, which are spectral features. There are three labeled regions: G130M (1,000 to 1,400 angstrom), E140H (1,400 to 1,700 angstrom), and E230H (1,700 to 3,000 angstrom).
Figure 1: Full UV SED of β Pictoris derived from the HSLA quicklook aspec file that includes three separate COS and STIS gratings. Abundant stellar and circumstellar absorption features are seen in the spectrum.


The HSLA products work best for astrophysical sources that have compact angular sizes and are not variable, which ensures flux continuity across different gratings and apertures over the lifetime of COS and STIS. In those cases, the coadditions should match each instrument’s requirements for flux and wavelength repeatability and accuracy. Solar System objects are currently not included in the HSLA.

The automatically generated HSLA products do not fit all science cases or requirements, so an integral part of the project is to provide user-friendly access to custom coaddition for situations when more individualized approaches might need to be taken. The HSLA project has a notebook that introduces users to custom coaddition of multiple HST observing programs and another that introduces users to the multiple COS lifetime positions and how those may impact the line spread function of their spectra. An example from the HSLA Introduction Notebook is shown in Figure 2.
 

Two graphs, one on top of the other. Both have the y-axis label S/N Ratio and x-axis label Wavelength (Å). The top graph shows HSLA Default results, and the bottom Custom HSLA results. Both have a legend at bottom left: STIS G140L appears blue, STIS G230L appears orange, and HSLA appears green. Both show the same irregular X-shape. On the HSLA Default results, the line that begins toward the top left is blue at top left and green to just below center. The second line begins at bottom center in orange and remains so as the line rises past the second line, later turning green as it rises toward top right. At bottom, the line that begins toward the top left on the Custom HSLA results also begins in blue, and is green until it meets the second line, when it changes to blue. The second line begins at bottom center in orange, switches to green where the lines meet, and also rises toward the top right.
Figure 2: Results from HSLA Default and Custom HSLA searches. At the top, signal-to-noise (S/N) is achieved by each of the two gratings (orange and blue curves) that participated in a quicklook aspec spectrum (green curve). The default grating priority results in lower S/N data being used between 1,650 and 1,720 Å. The bottom spectrum shows user-adjusted grating priority table changes. The transition wavelength for a custom aspec spectrum ensures the higher S/N data are used in the overlap region. This is one of several ways that users can customize their queries. Full instructions are in the HSLA Introduction Notebook. This step-by-step guide allows users to create custom coadded datasets to tailor spectral data combinations to fit their unique research needs.

How to Access HSLA and Refine Your Queries

HSLA is seamlessly integrated into the Hubble MAST search form and the MAST Portal (see Figures 3 and 4) as a standard component of data downloads from the archive. Researchers looking for HSLA data may access them by using the Hubble Search form to select individual objects or discover objects of different classifications.
 

A screenshot of a web page with three large rows. Top row labeled MAST Advanced Search. Below that a button reads, search; a button reads, export table; text, records found 3,891. Second row, labeled Applied Filters, is a button: Clear All; and three filters in bubbles reading: Provenance Name: HSLA, Project: HSLA, and Target Classification: *Tier1=Star*. Bottom row has three columns. At left a box named column has options for filters: Object Name or Position, Observation Type, Mission, Provenance Name, Instrument, Project (selected with a check mark), Filters, Waveband, Target Name, and Target Classification (also selected). The center column, named Project, text, HSLA. Below are results: TESS (3,821,302), K2 (1,693,051), PS1 (998,018), HST (918,292), Kepler (703,547). Below, text, show 91 more options. The right column is Target Classification, which has *Tier1=Star* in the field. Below, text: This column must be searched via the text box at the top of this panel.
Figure 3: A snapshot from the Hubble MAST Portal showcases the integration of the Hubble Spectroscopic Legacy Archive (HSLA) products using the Advanced Search page. In this example the filters for “Project” and “Target Classification” are used to select all stars in the HSLA, a total of 3,891 unique targets. For more instructions about how to perform searches, visit the HSLA webpage.
Screenshot of results from a Hubble MAST Portal query. At far left is the word filters, which runs vertically. Just to the right are two tabs. The First, which is displayed in full below reads, List View. The second reads, Album View. Within the List View are five of the search results. They appear in rows labeled Target Names, Target Classifications, and Observation ID. Each target has three classifications and a single observation ID. At the top, there are additional options within the search form to narrow the results.
Figure 4: Results from the search run in Figure 3. All results appear with the “Target Name,” “Target Classification,” and “Observation ID.” The Target Classification column provides three methods of identifying classes of targets.

The MAST Portal is also useful for searching for different classes of objects by using the Advanced Search page, setting Project=”HSLA” and using the Target Classification column to filter desired object types. Detailed instructions are available at the HSLA webpage. The most efficient approach to downloading a large number of spectra (such as getting all the spectra of objects classified as “Stars”) would be to use astroquery, a Python package that is part of astropy. Detailed instructions about how to use astroquery for large numbers of targets are present in the HSLA introductory Notebook in the STScI Jupyter Notebook repository.

Key Takeaways

  1. The new HSLA includes almost all STIS and COS spectra that have been extracted by the default pipeline.
  2. HSLA has been updated to provide more efficient and effective search tools for users.
  3. Users have a flexible array of options to search for individual targets or classes of astrophysical objects, and may create custom products with user-friendly Jupyter notebooks.
  4. The new HSLA is automated and will include new observations for both COS and STIS as data become public.

For comprehensive details about HSLA, please refer to the HSLA webpage or HSLA Instrument Science Report.

 

 

Share This Page

This site is protected by reCAPTCHA and the Google

Contact our News Team 

Contact our Outreach Office