Improving Data Anomaly Detection and Forecasting in the US Army Corps of Engineers Reservoir Sedimentation Information (RSI) Database | W81EWF 23 SOI 0007

Frequently Asked Questions (FAQs)

What agency is offering this grant opportunity?

This opportunity is offered by the U.S. Department of Defense through the U.S. Army Engineer Research and Development Center (ERDC).

What is the main goal of this cooperative agreement?

The main goal is to improve data quality controls and forecasting capability within the U.S. Army Corps of Engineers (USACE) Reservoir Sedimentation Information (RSI) database. The work centers on expanding anomaly detection and prediction methods for reservoir sedimentation surveys and then demonstrating whether the models can credibly forecast future storage capacity loss.

What kind of work is being funded (research, operations, tool development)?

The effort is described as applied research and development that supports planning and decision-making. The base work focuses on dataset expansion and forecasting validation, and an optional task would move toward an implementable toolset for routine analyst workflows.

How many awards does the government expect to make?

The government anticipates making a single award.

What is the maximum funding amount available?

The award ceiling is $60,000.

What is the funding instrument?

The opportunity is structured as a cooperative agreement, which indicates substantial involvement or collaboration with the government during performance.

What is the Funding Opportunity Number (FON)?

The Funding Opportunity Number is W81EWF 23 SOI 0007.

What is the CFDA number for this opportunity?

The CFDA number listed is 12.630.

When was the opportunity posted and when did it close?

The opportunity was posted on May 26, 2023. The original closing date was July 26, 2023.

Who is eligible to apply?

Eligibility is listed broadly as "Others," with additional details referenced in the full eligibility text.

What database is the work focused on?

The project is focused on the USACE Reservoir Sedimentation Information (RSI) database.

What is being improved in the RSI database?

The work aims to strengthen RSI data quality control and forecasting by extending preliminary survey flagging, autonomous anomaly detection, and prediction model development across a much larger portion of the reservoir inventory.

What is the purpose of anomaly detection in this project?

Anomaly detection is intended to help identify questionable or unusual sedimentation survey data (preliminary survey flagging) so RSI can maintain stronger data quality controls as the dataset expands.

What is Base Task 1 about?

Base Task 1 expands the underlying dataset used for analysis. Prior work included only USACE and U.S. Bureau of Reclamation (USBR) reservoirs that had three or more sedimentation surveys recorded in RSI. Under this award, the analysis is broadened to include every USACE and USBR reservoir in RSI with at least two surveys.

How many reservoirs are expected to be added under Base Task 1?

The expansion adds 133 reservoirs immediately, with the likelihood of additional reservoirs being included as new data are incorporated during summer 2023 updates.

Why does the expanded dataset require additional technical preparation?

To support consistent capacity analysis and technically sound comparisons across surveys, the performer will need to quantify reservoir pool elevations and compute vertical datum correction factors for the additional reservoirs.

What is the role of reservoir pool elevations in the analysis?

Reservoir pool elevations are needed for consistent capacity analysis across sedimentation surveys.

Why are vertical datum correction factors needed?

Vertical datum correction factors are needed so comparisons across multiple surveys are technically sound, particularly when survey elevations or reference datums differ.

Does the project involve coordination with the U.S. Bureau of Reclamation (USBR)?

Yes. The notice highlights interagency coordination. Because USBR is developing a separate database and plans to eventually connect the two systems, the project is expected to require coordination with a USBR collaborator to access the most current USBR survey information in the near term.

What is Base Task 2 about?

Base Task 2 focuses on proving forecasting feasibility in a way USACE can trust and use. The project compares reservoir capacities forecasted by statistical and machine learning methods against capacities estimated through traditional extrapolation of known sedimentation rates.

What forecasting methods are specifically mentioned?

The notice references ordinary least squares (OLS) approaches and machine learning predictive models, along with comparisons to traditional extrapolation of known sedimentation rates.

How will forecasting performance be evaluated?

Forecasted reservoir capacities from statistical and machine learning approaches will be compared head-to-head against capacities estimated via traditional extrapolation of known sedimentation rates. This comparison is intended to validate the proof of concept for predicting sedimentation and storage loss.

Why is the comparison to traditional extrapolation included?

The comparison is intended to demonstrate whether newer statistical and machine learning approaches can credibly forecast future storage capacity loss relative to methods based on extrapolating known sedimentation rates, helping establish trust and usability for USACE decision-making.

Is there an option for additional work beyond the base tasks?

Yes. If early work demonstrates sufficient progress and strong results in early FY24, the government may exercise an option for additional work in FY25.

What does the optional task include?

The optional task shifts from research validation to an implementable toolset. It calls for an automated ArcGIS Pro Add-In that can estimate reservoir capacity loss from the dam construction date to the present and then project future losses using the forecasting outcomes from Base Task 2.

What inputs would users provide to the ArcGIS Pro Add-In?

Users would provide a minimal set of inputs: dam location, construction date, and original volume capacity.

How are the remaining parameters expected to be generated in the tool?

The remaining model parameters are expected to be derived through automated GIS analysis.

What kind of visualization is expected in the optional toolset?

The optional toolset is expected to include a water-supply-focused visualization component.

What is the "stoplight-style ranking" mentioned in the option?

It is a simple ranking approach to categorize reservoirs based on selected RSI database fields, intended to help users quickly triage reservoirs that may be higher risk or require attention.

Is integration with other databases expected?

Yes. The optional tool is expected to be built in a way that supports integration with the RIMORPHIS geomorphic database.

Does the project include plans for public-facing hosting?

Yes. The team would collaborate with USACE and other federal partners to determine a permanent, public-facing hosting arrangement for the database or related outputs.

What decisions or outcomes is this work intended to support?

The work supports USACE planning and decision-making by helping anticipate the economic, environmental, operational, and social consequences of reservoir sedimentation, particularly through improved forecasting of storage capacity loss.

What is the relationship between the base tasks and the optional task?

The base tasks function as a proof and validation phase (expanded data quality controls, anomaly detection, and forecasting feasibility). The optional task is positioned as a next step toward an operational, automated tool analysts can use routinely, and it may be exercised only if the base work shows strong early results.