Solution specific articles for Water & Environmental Resources.

National Water Center – Machine Learning Approach to Reservoir Operations

Reservoir operation is a complex human-controlled activity that substantially alters surface hydrology. A reservoir operator needs to account for reservoir gains and losses, and other upstream and downstream conditions, in order to make decisions about daily releases that meet operating targets. The National Water Model (NWM) version 2.0 is anticipated to include representation of approximately 5,300 reservoirs. By utilizing a levelpool routing, the current NWM does not represent complex reservoir dynamics, which negatively affects the accuracy of the model’s simulations and forecasts.

Lynker scientists at the NWC are utilizing machine learning techniques based on historical records of release decisions to improve the NWM’s operational forecasting skill. Deep neural networks (i.e., feed-forward backpropagation with three hidden layers and a considerable number of neurons) have been utilized to train the NWM to think like a reservoir operator. K-fold cross-validation and dropout techniques were considered to train generalized models and avoid the overfitting problem, which is commonly ignored in the application of ML techniques in hydrology and water resources fields. There is no need for rule curves to train the model and the proposed technique has the potential to scale well to the national level. Application of the developed models for two study basins, Apalachicola-Chattahoochee-Flint (ACF) and Colorado Headwaters, have shown promising results by mainly utilizing observed previous day release, reservoir inflow, previous day storage and day of the year as inputs to the model.

National Water Center – Geo Intelligence Division (GID)

Lynker leads the Geospatial Intelligence Division (GID) at NOAA’s Office of Water Prediction (OWP) National Water Center (NWC). GID is responsible for taking outputs from the new NOAA National Water Model (NWM) and visualizing these data to make them more useful to stakeholders, including NWS River Forecast Center (RFC) staff at regional forecast offices and Emergency Managers (EMs) on the ground. Both need to rapidly understand and use NWM outputs to save lives, so effective visualization of the NWM data is critical. The Lynker team also develops NWM data visualizations for water managers who are more interested in longer range (e.g. seasonal) water flow forecasts for water supply and planning, and drought monitoring. 

GID has developed an automated real-time system to acquire, post-process, stylize, and publish NWM output data in real-time. Cycling hourly and providing streamflow and land surface conditions forecasts at over 2.7 million locations CONUS-wide, the National Water Model produces billions of data points per day. Because users are unable to efficiently make use of this massive hourly data output without the aid of data post processing and enhanced map viewing, significant development was required for not only the automatic detection of important hydrologic activity across the country, but the visualization of these phenomena across the millions of forecasting locations. Through this project, the Lynker team has developed a variety of data post-processing and optimization techniques for the purpose of visualization. Dr. Graeme Aggett leads the GID team, and Brad Bates continues to be a key technologist supporting this complex data visualization effort.

NOAA’s new National Water Model (NWM) has allowed the National Weather Service (NWS) to expand its hydrologic prediction capabilities from ~3,600 river forecast points to over 2.7 million stream reaches, reaching many previously underserved locations.  However, deriving actionable intelligence from the NWM is challenging because it produces hundreds of gigabytes of data each day.  NWS forecasters need to be able to quickly analyze NWM data before issuing streamflow guidance, thus methods are needed to synthesize and present NWM data in real-time, in such a way that it can easily aid in the decision-support process.  Lynker scientists at the NOAA/NWS National Water Center, located in Tuscaloosa, Alabama, have played a vital role in developing a series of real-time data services that process and visualize NWM output in such a way that it can be used by the NWS’s stakeholders to make quick, informed hydrologic decisions.

Over the past year and half, Lynker has worked with the National Water Center (NWC) to set-up an Enterprise Geographic Information System (EGIS) that has the capability to host dynamic web data services, maps, and applications.  Further, our team at the NWC has implemented a workflow to: intercept the latest NWM data as it becomes available, post-process this data, and ultimately to create dynamic, interactive web-based GIS data services from post-processed NWM data.  These data services provide a diverse array of hydrologic information, for both current and forecasted conditions, about: high flow and floods, low flow and droughts, seasonal anomalies, soil moisture levels, the rate of change in streamflow conditions, and the timing and probability of extreme streamflow conditions.

North Platte River Accounting Program

Lynker staff have provided a number of services to the State of Wyoming resulting in a deep understanding of the hydrologic and institutional factors that govern water resources on the North Platte River. Most of these services provided technical support in the Nebraska v. Wyoming litigation and included development of a water allocation and reservoir operations model of the North Platte River, development of an accounting tool to track balances of accounts in the federal reservoirs, development of a database of historical diversions and operations, and statistical analyses of streamflows on the North Platte River and its tributaries.

Since settlement of the litigation, the extensive knowledge of the river gained by Lynker staff continues to be applied by the Wyoming Board of Control through the North Platte River Accounting Program that Lynker staff developed and maintains and upgrades upon request. For that program, Lynker staff:

1) Developed a custom GUI interface for calculating and reporting natural flow and storage deliveries on the North Platte River between Wyoming and Nebraska irrigators,

2) Designed an application to collect and record river and ditch flow gauging stations and import those readings into a database,

3) Automated flow calculations from USGS gage rating tables and gage heights, and

4) Developed an archiving system to ensure that results are reproducible long into the future.

Eagle River Water Rights Operations Model

Lynker, together with Leonard Rice Engineers, developed a water resources systems model to represent water rights, reservoirs, streamflow conditions and operations by various municipalities in the Eagle River Basin of West-Slope Colorado.  The model provides water resources decision support for the major water providers and users in the basin.   Water rights, key diversions, ski resort snowmaking, reservoir operations, instream flows reaches and other factors were studied in detail.  Model results have been used to guide infrastructure and resiliency planning as well as climate change impact analyses.

South Platte Decision Support Systems Model

Lynker was contracted by the Colorado Water Conservation Board (CWCB) to develop a water rights model for a Colorado Water District under the South Platte Decision Support Systems project. The SPDSS program covers the development of similar water rights models across all of Water District 1 (the South Platte River).  This effort was the most ambitious attempt to date to develop a model that will account for every major water rights operation across the entire South Platte River basin.  All models under SPDSS were developed using the StateMod software, using the Colorado Decision Support Systems (CDSS) Hydrobase dataset as the source for all data.

The model built by Lynker was designed to handle a large number of different water users and systems operations including irrigation demands and subsurface return flows, transbasin diversions from the Western Slope, municipal demands, instream flow rights, reservoir fill and release operations and changed water rights.

The model development process consisted of baseflow development, detailed water rights case research, model calibration and validation, and finally integration into the larger SPDSS system.  Initially the Lynker team developed a dataset of natural flows—what would have happened on the river assuming no human intervention—to serve as a baseline for the model calibration process.  Detailed water rights research was then used to program operating rules within StateMod to simulate historical operations and recreate how history played out in the most accurate way. Modeled diversions, reservoir contents and streamflows were compared to historical data to evaluate the skill of the model at accurately recreating history.  Through this process, the model was built to a consistent standard; the high quality of our work was praised by the client at the CWCB.

Instream Flow Modeling to Support Hydroelectric Plant Permit Application

Pitkin County required technical services to understand both the water rights and instream flow impacts of various operations of a proposed hydroelectric plant on Castle Creek. The project team developed a model to analyze optimal hydroelectric operations based on existing water rights and historic streamflow. Once the model operations had been confirmed, the team developed and ran many test cases to optimize power production while maintaining instream flow. Ideal instream flow targets were based on results of existing R2CROSS analyses as well as stream levels deemed acceptable by the community.

Once the optimization process was completed, it was the job of the project team to effectively communicate the stream impacts to the client. Using the Indices of Hydrologic Alteration (IHA) protocol, details of the changes in streamflow regime were presented in a manner more understandable to the general public.

Lynker leads stream stewardship project through citizen science and adaptive management

Lynker staff developed an adaptive management framework for Lefthand Watershed Oversight Group that serves as an outline to ensure success of stream restoration for Left Hand Creek. Lynker, along with LVBrown Studio, developed a conceptual model which described the changes of Left Hand Creek watershed during floods. This model illustrates how the channel, riparian zones, and components of the ecological community in the creek changes over time. Lynker worked with project stakeholders to develop the Monitoring and Assessment Framework. These stream assessment protocols were used by citizen scientists to evaluate fish habitat, riparian habitat, wildlife, and invasive species. The project ended with a pilot test of the citizen science monitoring protocols including an analysis of results comparing citizen scientist scores to professional scores.

This monitoring and assessment framework serves as a living document for Lefthand leaders to use for future citizen science projects, and an outline for other agencies to develop their own adaptive management programs to ensure continued success of restoration projects.

Lynker Scientists Working with State of Colorado on First-of-its-Kind Climate Risk Analysis Project

Lynker scientists are currently working with the State of Colorado on a first-of-its-kind climate risk analysis project – Future Avoided Cost Explorer (FACE: Colorado Hazards). This statewide study estimates dollar-value damages from future floods, droughts, wildfires across select sectors of Colorado’s economy. The State of Colorado has recognized that these hazards are likely to worsen as the impacts of climate change intensify the severity of the hazards and as population growth increases the number of people at risk.

By helping Coloradans better understand their current and future economic risks from flood, drought, and wildfire, this tool can help stimulate the implementation of smart adaptation strategies and policy frameworks that can improve the prospects for sustainable prosperity in a rapidly changing environment.

View and explore the FACE: Hazards dashboards to understand how changes in global climate patterns can lead to more frequent and intense hazards in Colorado.

Future Avoided Cost Explorer: Colorado Hazards (FACE:Hazards)





World Water Day 2020

Managing water is a growing concern in communities across the US. Many of the states that have projected population growth increases also have higher per capita water use and can expect increased competition for water resources. Forty states expect to have water shortages over the next ten years that are not related to drought, while climate change risk assessment work conducted by Lynker for the State of Colorado and FEMA indicates we will be having longer and harder droughts in Colorado, with profound impacts on sectors of the Colorado economy.

Guaranteed access to clean water is critical in times such as these, where we need it to maintain health services and to sustain our food supply and other sectors of the economy. World Water Day 2020 is Sunday, March 22nd. This year’s theme is Water & Climate Change and how they’re connected. At Lynker our water resource scientists and engineers collaborate closely with our risk assessment and planning team to understand and develop planning tools that help plan for an uncertain future. We are committed to providing data, decision support tools and consulting advice on how local, state and federal agencies can manage this vulnerable resource more effectively and sustainably, now and into the future. In the current coronavirus outbreak situation, Lynker is more committed than ever to our Water Resources Planning mission. Lynker will continue to work hard and adapt to the new circumstances we all find ourselves in to provide our clients the actionable water intelligence they need to make smart decisions while conducting their mission in this ever-changing environment.



Lynker Leads Project Exploring Impacts of Colorado’s Future Floods, Droughts, and Wildfires

Lynker scientists Graeme Aggett & Ryan Spies presented alongside State of Colorado staff on a project Lynker is leading titled “FACE: Hazards” or “Future Avoided Cost Explorer” at the 2020 Colorado Emergency Management Conference in Loveland, CO.

This project explores the impacts of future flood, drought, and wildfire on select sectors of Colorado’s economy. Lynker is working with DHSEM, CWCB, DOLA, and FEMA Region VIII to develop an analysis and visualization tool to help local officials recognize and adapt to these three major hazards facing Colorado.