Cooperative Agreement for CESU-affiliated Partner with Great Plains Cooperative Ecosystem Studies Unit | G23AS00044

Frequently Asked Questions (FAQs)

What is this funding opportunity about?

This is a US Geological Survey (USGS) cooperative agreement focused on advancing camera-based water monitoring under the Water Mission Area (WMA) Next Generation Water Observing System (NGWOS). The funded work centers on using still-camera imagery together with image analysis and artificial intelligence or machine learning (AI/ML) to produce practical water data products, especially measurements such as water stage and discharge.

Which USGS program is sponsoring this work?

The opportunity is tied to the USGS Water Mission Area (WMA) Next Generation Water Observing System (NGWOS), which is aimed at improving and modernizing water observing capabilities, including approaches that can scale across sites and produce reliable, comparable data.

What specific technical approach does the opportunity emphasize?

The emphasis is on research and development using still-camera imagery combined with image analysis and AI/ML methods to derive usable water monitoring products. The description highlights applications such as extracting water stage and discharge from imagery.

What problem is USGS/NGWOS trying to solve with this award?

NGWOS is addressing a core bottleneck for machine learning in camera-based water monitoring: strong AI/ML models generally require large collections of fully labeled training images, but the community lacks consistent, widely adopted practices for capturing, documenting, storing, sharing, and processing imagery into training datasets. This fragmentation makes it harder to build robust models and to reuse data across projects.

Is the goal to build a single machine learning model?

No. The project goal is described as broader than developing a single model or one-off dataset. The intent is to create repeatable, community-useful practices and workflows that enable consistent creation of labeled imagery and training data needed for AI-driven, camera-based water observations.

What are the main themes of the opportunity?

A central theme is standardization and workflow development. The opportunity points to the need for more consistent end-to-end practices so that imagery can be turned into usable training datasets and data products in a way that can be adopted by NGWOS and potentially the wider water monitoring community.

What areas are identified as inconsistent across the current community?

The opportunity highlights multiple areas where inconsistency is limiting progress:

• Image capture protocols (how cameras are set up and how images are collected)
• Metadata standards (what descriptive information is recorded with each image so others can interpret and reuse it)
• Data storage and delivery (how datasets are archived, organized, and made accessible)
• Image processing and ML workflows used to build training datasets from imagery

What kind of deliverable is USGS looking for?

The description indicates the work is meant to produce a functional, standardized image processing workflow that can be adopted by NGWOS. The broader objective is to reduce fragmentation and make camera-based monitoring more scalable and reliable by enabling consistent creation of labeled imagery and training datasets.

What is meant by a "standardized image processing workflow" in this context?

Based on the information provided, it refers to an end-to-end approach that connects image capture and documentation to storage, sharing, processing, and preparation of labeled imagery that can be used to train AI/ML models. The focus is on repeatability and consistency so that datasets and methods can be reused across sites and projects.

What is the assistance mechanism for this opportunity?

The assistance mechanism is a cooperative agreement.

How is a cooperative agreement different from a typical grant for this opportunity?

The posting indicates that a cooperative agreement typically means USGS expects substantial involvement during the project. Examples provided include collaboration, iterative technical input, or coordination on standards and deliverables, rather than a fully hands-off funding relationship.

What framework is this award issued under?

This award is issued under the Cooperative Ecosystem Studies Unit (CESU) framework, specifically tied to the Great Plains CESU.

Who is eligible to apply?

The opportunity is restricted to eligible organizations that are already participating partners within the Great Plains CESU Network.

If an organization is not a Great Plains CESU partner, can it apply?

Based on the posting, eligibility is restricted to existing participating partners in the Great Plains CESU Network. No additional eligibility pathway is described in the provided information.

What is the opportunity number?

The opportunity number is G23AS00044.

What is the CFDA number listed for this opportunity?

The CFDA number is 15.808.

How is this opportunity classified?

It is listed as a discretionary funding opportunity in the science and technology and other research and development category.

What is the maximum award amount (award ceiling)?

The posting indicates an award ceiling of $275,000.

When was the opportunity created and when did it close?

The opportunity was created on January 4, 2023, with an original closing date of February 6, 2023.

What water measurements are specifically mentioned as target products?

The opportunity specifically calls out practical water data products, especially water stage and discharge, derived from still-camera imagery using image analysis and AI/ML methods.

Why does metadata matter for camera-based water monitoring in this opportunity?

The posting identifies metadata standards as a key gap. Recording consistent descriptive information with each image helps others interpret the imagery correctly and reuse it for training datasets and analysis, which is important for building scalable, reliable AI/ML approaches.

Why does the opportunity focus on storage, delivery, and sharing of image datasets?

The description notes that inconsistent approaches to archiving, organizing, and making datasets accessible is holding the field back. Standardized storage and delivery practices support reuse of imagery and labels, reduce duplication, and make it easier to build larger training datasets.

How does this opportunity relate to building AI/ML training datasets?

The opportunity directly targets the end-to-end workflow needed to create consistent, labeled imagery and training data. The underlying rationale is that robust AI/ML depends on large collections of fully labeled images, and current inconsistency in how images are captured and handled makes it difficult to assemble such datasets at scale.

What broader impact is USGS aiming for beyond NGWOS?

The posting indicates the workflow is meant to be adopted by NGWOS and, ideally, by the broader water monitoring community, helping reduce fragmentation and improving scalability and reliability of camera-based monitoring methods.