About this Article

 Ori Fox (ofox[at]stsci.edu)

Published February 16, 2026

As the launch of the Nancy Grace Roman Space Telescope approaches, preparations for delivering mission data to the astronomical community continue to advance. Roman’s wide-field imaging and spectroscopy will enable transformative science from the solar system to the high-redshift universe. The Roman Wide Field Instrument (WFI) will include observing programs under two main categories. The Core Community Surveys (CCS), which include the High-Latitude Time-Domain Survey (HLTDS), High-Latitude Wide-Area Survey (HLWAS), and Galactic-Bulge Time-Domain Survey (GBTDS), are designed to address foundational questions that define the mission’s science requirements. General Astrophysics Surveys (GAS) programs may address any area of astrophysics and already include the community-defined Galactic Plane Survey (GPS). How these data are processed, curated, and released will influence the pace and quality of scientific discovery.

All Roman data will be publicly available with no proprietary period for both CCS and GAS. To maximize the scientific utility of Roman data, the mission will deliver a range of WFI data products and data releases to the community. This is a high-level overview of STScI-specific data delivery plans, which are geared toward WFI imaging from the HLTDS, HLWAS, and GAS, including the GPS. IPAC will generate products specific to both microlensing data from the GBTDS and slitless spectra from all surveys. Additional data products and releases for the CCS will come from the Project Infrastructure Teams (PITs). The PITs are community-led groups, selected through the NASA-ROSES program, that are currently building and maintaining essential software, simulations, and data analysis tools tailored to key Roman science objectives.

A Framework Designed for Scale and Scientific Impact

Roman will generate data at a scale and cadence that are exceptional for a space-based observatory. Enabling the community to use these data efficiently — and with confidence — requires a data delivery strategy that balances rapid access with long-term stability. To meet this challenge, the Roman Mission at NASA Goddard Space Flight Center and the science centers at STScI and IPAC are developing a structured framework for data processing and releases that is explicitly designed to maximize Roman science while considering Roman’s large data volume.

At the core of this framework is a hierarchy of data products organized into increasing levels of processing (levels 1 to 4, known as L1 through L4). As data move up this hierarchy, they become more calibrated, combined, and scientifically enriched, progressively supporting more complete analyses. Within this framework is a deliberate distinction between prompt data availability and stable, version-controlled data releases. Prompt products provide the community with rapid access to calibrated, new observations soon after they are acquired, supporting early exploration, transient science, and quality assessment. In contrast, periodic formal data releases provide uniformly reprocessed, strategically grouped, well-validated products intended for long-term scientific use. This structure allows users to choose the products best suited to their science, from early inspection and time-critical work to precision measurements that require uniform, deeply processed datasets.

This distinction is motivated by practical as well as scientific considerations. Roman’s data volume makes it infeasible to continuously reprocess the full archive every time calibration reference files or pipeline algorithms are updated. Furthermore, there is an explicit need for scientifically motivated, well-defined data releases. For this reason, Roman adopts an approach that ensures the community receives both timely access to new data (i.e., prompt products) and stable, high-quality products that can serve as reliable foundations for scientific discovery (i.e., data releases).

Prompt Products from STScI

Shortly after observations are downlinked, Roman data will be processed through the STScI calibration pipeline, RomanCal, to generate prompt data products. These products will be made available to the community within days of the observations and provide the earliest access to Roman observations. 

• Prompt Level 1 (L1) products consist of uncalibrated detector ramps and preserve the raw instrumental information needed for detailed inspection and diagnostic work.
• Prompt Level 2 (L2) products will be calibrated, two-dimensional images of individual WFI exposures, delivered within approximately two days of downlink. These products will be processed using the best available calibration information at the time, but do not wait for subsequent calibration updates, enabling timely access at the cost of long-term uniformity.
• Prompt Level 3 (L3) products build on the L2 images, combining exposures taken within a single visit, and will follow within approximately five days of downlink. These images will be geometrically rectified, resampled onto a common pixel scale, and projected onto a standardized sky grid, allowing for straightforward visualization and analysis.
• Prompt Level 4 (L4) products build on Level 3 images and will be released by STScI in the form of single-band source catalogs, and follow within approximately seven days of downlink. L4 products include basic measurements such as positions, photometry, morphological parameters, and segmentation maps.

Together, these prompt products are intended to support early science exploration, time-sensitive investigations, and assessment of data quality. Please note that descriptions of L3 and L4 data products from STScI do not apply to the GBTDS or spectroscopic mode observations.

Data Releases from STScI

In contrast to prompt products, data release products will be generated through planned, periodic, uniform reprocessing campaigns designed to provide stable, mission-quality datasets for the community. During a data release, all relevant observations will be reprocessed using a consistent and optimized set of calibration reference files and pipeline software. L2 data release products will supersede earlier versions of L2 calibrated exposures (whether they were generated by prompt products or data release processing) and will be overwritten to manage archive complexity and storage, following established practices from missions such as Hubble and JWST. Higher-Level data release products, however, will be preserved.

L3 data release images will combine data across multiple visits to produce deeper, more uniform stacks for each filter, while L4 data release catalogs will merge measurements across all identified filters and observations to create deeper, multi-band source catalogs that may include derived quantities such as photometric redshifts.

Once released, L3 and L4 data release products will be versioned and maintained as stable references for scientific analysis. All official Roman data products, regardless of origin, will be distributed through the Mikulski Archive for Space Telescopes (MAST) with catalogs accessible through both traditional database interfaces and modern, cloud-optimized formats. All Roman data in MAST, including catalogs, will also be accessible through the Roman Research Nexus, a science ecosystem that provides a cloud-based computing space for the science community to explore, access, and analyze Roman’s petabyte-scale data. (Learn more about the full suite of resources and simulated data on the Nexus.)

Data releases are intended for long-term scientific use. In the case of the HLWAS, for example, prompt products will allow early exploration of newly observed fields as the survey progresses, while planned data releases will reprocess all HLWAS observations obtained over a defined period using a consistent set of calibration reference files and pipeline versions. This reprocessing will produce deeper, multi-visit L3 image stacks and corresponding, multi-band L4 catalogs that will be uniform across the survey area and suitable for precision cosmological analyses. Similar strategies apply to other surveys, such as the HLTDS, GPS, and the GAS programs, with release boundaries chosen to reflect both scientific coherence and operational practicality. Ultimately, this layered approach ensures that Roman data will be both rapidly accessible and scientifically robust, enabling a wide range of investigations throughout the mission.

Planning and Releasing Roman Data

Roman’s data release schedule will be synchronized with survey execution, calibration readiness, and pipeline developments. Decisions about the timing and composition of data releases necessarily involve trade-offs. For example, between releasing data earlier with less mature calibrations or waiting longer to achieve greater uniformity and depth. These decisions will be evaluated with input from the community (e.g., PITs) through established coordination and governance structures, ensuring that scientific priorities, calibration readiness, and operational constraints will be weighed transparently. These discussions will also draw on lessons from prior missions such as JWST, where community readiness and buy-in was critical for success.

Data Products Delivered by Roman Mission Partners

The Roman Mission has a distributed ground system, which means that the full suite of available Roman WFI data products will come from multiple partners. The Science Support Center at IPAC is responsible for the extraction and calibration of slitless spectroscopy observations as well as the high-level processing of the GBTDS data. This means that in addition to the STScI data products there will be L3 and L4 products defined and generated by IPAC. This includes all L4 spectral products, like calibrated, extracted spectra and catalogs, and higher-level data products for the GBTDS.

The Roman Mission is also supported by funded science teams, who will create data products tailored to enable specific scientific analyses. User-contributed high-level data products, including those produced by the PITs, are designated as Level 5 (L5). Examples include, but are not limited to, transient detection and alerts, Type Ia supernova light curves, microlensing light curves, weak lensing cosmic shear measurements, and value-added galaxy catalogs tailored to Roman slitless spectroscopy.

The Ready, Set, Roman! Training Series will host two webinars dedicated to understanding PIT deliverables on Tuesday, February 17 at 1 p.m. Eastern and Tuesday, February 24 at 1 p.m. Eastern. Both recordings will be hosted on the NancyRomanSci YouTube channel.

The timing and availability of data products and formal data releases from mission partners will be coordinated with STScI data releases.

Looking Ahead

Roman’s processing and release framework is now well defined, with workflows established and the focus shifting toward execution and coordination across mission teams. As launch draws closer, STScI continues to refine our roadmap for delivering high-quality, science-ready data so that the astronomical community can fully exploit Roman’s unprecedented datasets. Stay tuned for future updates about pipeline milestones, release schedules, and early science planning activities in upcoming articles about Roman in the STScI newsletter.

Resources

• Overview of WFI data levels and products
• Additional information about STScI’s WFI prompt products
• Full details about STScI’s WFI data release plans
• Overview of Roman WFI Data Pipelines, including imaging, spectroscopy, and GBTDS processing
• Information about the data products to be delivered by the Project Infrastructure Teams. Find the PDF on this page under the heading Key Resources, within the subheading PIT & WFS Data Products.