Data Reduction for Science | DE FOA 0002501

Opportunity Information: Apply for DE FOA 0002501

The Department of Energy (DOE) Office of Science, through the Advanced Scientific Computing Research (ASCR) program, is seeking grant proposals for research on data reduction methods that make it easier to work with extremely large scientific data sets. The central problem the opportunity targets is that modern observations, experiments, and simulations generate data faster than the scientific community can realistically store, move, stream, analyze, or archive it. Much of this raw data contains redundancy or low-value detail, while the scientifically meaningful signals are often buried inside. The program’s goal is to support high-impact advances in algorithms, techniques, and end-to-end workflows that reduce data volume while still preserving the information scientists actually need.

A key theme of the announcement is that data reduction is already happening in many fields, but DOE wants the next step: approaches with stronger mathematical and computational foundations that can be trusted. The solicitation explicitly mentions common reduction strategies such as compression, reduced-order modeling, experiment-specific triggers, filtering, and feature extraction, but emphasizes that future work must go beyond ad hoc or purely application-tuned methods. Proposed techniques should come with rigor around what is preserved, what is discarded, and how well quantities of interest and scientific constraints are maintained. Just as importantly, the reduced data products and algorithms need to fit into real scientific workflows in ways that inspire confidence that the essential science is not being lost.

DOE frames this as a cross-cutting need across a wide range of mission-relevant domains, including (but not limited to) light sources, accelerators, radio astronomy, cosmology, fusion, climate, materials, combustion, the power grid, and genomics. Because the same underlying data challenges appear in many of these areas, ASCR is specifically interested in solutions that generalize across multiple applications rather than narrowly targeting one experiment, one instrument, or one simulation code. In fact, the opportunity notes that pre-applications focused on only a single science application may be discouraged from moving forward to a full proposal. Preference may be given to pre-applications that provide reduction estimates for at least two different science applications, reinforcing the expectation of broad relevance and transferability.

The technical directions are guided by a DOE-hosted virtual workshop from January 2021 titled "Data Reduction for Science," which produced four priority research directions (PRDs). The first PRD centers on developing effective algorithms and tools that scientists can trust for both accuracy and efficiency, reflecting the need for methods that are not only fast but also reliable and verifiable. The second PRD focuses on progressive reduction approaches, meaning techniques that can reduce and prioritize data in stages so that the most valuable information can be streamed or accessed first, which is especially important when bandwidth or time-to-insight is a limiting factor. The third PRD targets methods that preserve features and quantities of interest with quantified uncertainty, highlighting that reduction should not just shrink files but should also communicate what confidence scientists can place in downstream inferences made from reduced representations. The fourth PRD addresses mapping data reduction to new computing architectures and emerging use cases, recognizing that HPC systems, edge devices, and heterogeneous architectures change what is feasible and what is optimal for reduction performed in situ, in transit, or during analysis.

In terms of what DOE wants to fund, the announcement prioritizes applied mathematics and computer science research that directly advances one or more of these PRDs. The solicitation is not intended to fund incremental extensions of existing projects; instead, it calls for significant innovation and viable strategies aimed at hard, high-impact problems in scientific data reduction. Proposals are expected to contribute broadly applicable knowledge, frameworks, and understanding that can inform future scientific advances, rather than producing a one-off pipeline tuned to a particular dataset. Cross-cutting concepts such as scientific machine learning and trustworthiness are explicitly called out as potential focal points, as long as they serve the larger objective of dependable, generalizable reduction that enables efficient scientific discovery.

Administratively, this is a discretionary DOE Office of Science funding opportunity (DE-FOA-0002501) in the science and technology R and D category, using the grant instrument. Eligibility is listed as unrestricted (open to any type of entity), subject to any additional eligibility language in the full announcement. The opportunity was created on April 15, 2021, with an original closing date of June 4, 2021. DOE anticipated making around 12 awards, with an award ceiling of $800,000 per award (as stated in the provided source details).

• The Department of Energy - Office of Science, Office of Science in the science and technology and other research and development sector is offering a public funding opportunity titled "Data Reduction for Science" and is now available to receive applicants.
• Interested and eligible applicants and submit their applications by referencing the CFDA number(s): 81.049.
• This funding opportunity was created on Apr 15, 2021.
• Applicants must submit their applications by Jun 04, 2021. (Agency may still review applications by suitable applicants for the remaining/unused allocated funding in 2026.)
• Each selected applicant is eligible to receive up to $800,000.00 in funding.
• The number of recipients for this funding is limited to 12 candidate(s).
• Eligible applicants include: Unrestricted (i.e., open to any type of entity above), subject to any clarification in text field entitled Additional Information on Eligibility.

Apply for DE FOA 0002501

[Watch] Creating a grant proposal using the step-by-step wizard inside the applicant portal: