Title: Enhanced Forecast Design Through Experimental Gaming and Social Impact Assessment of Connected River and Floodplains

Principal Investigator: Scott Merrill

Research Team: Christopher Koliba, Trisha Shrum, Beverley Wemple, Brendan Fisher, Taylor Ricketts

Insitution(s): University of Vermont, University of Kansas

^{Abstract: Extreme water hazard events are increasing in magnitude and frequency, putting individuals, communities and emergency responders at risk, and forcing people to quickly decide how to respond, what information to use and communicate. Our overarching goal is to understand human behavioral aspects associated with current NOAA and National Water Model products that are related to flood water hazards. Our objectives are to better understand flood hazard mitigation and response behaviors to allow for optimization of communication strategies that will allow for more resilient communities, build trust in NOAA products, and help protect infrastructure and lives. Goal 1: Broaden capacity for forecast design by improving the ability of first responders to understand how forecasts will be understood and used by individuals and communities and by providing insights to inform longer-term community planning. The pursuit of this goal provides a significant contribution to CIROH’s efforts to advance social, economic, and behavioral science to improve water prediction and forecast uses. Goal 2: Contribute to the development of the CIROH Enhanced Forecast Design Center by evolving forecast platforms that consider the heterogeneity of risk perception and behavior, and decision heuristics relative to water hazard mitigation that may be integrated into National Water Model products. Objective 1: Gauge public awareness and uses of water hazard forecasting products, including factors impacting variability of risk perception and efficacy of risk communication. Objective 2: Create a platform to test changes to risk perception and communication resulting from alternative forecast design parameters and treatments through a modular experimental gaming platform. Objective 3: Improve integration between stakeholder input and forecast design using representative regional empaneled focus groups. Objective 4: Integrate river and floodplain connectivity to better estimate social impacts and costs using an ecosystem services framework. Accomplishing these goals and objectives will provide actionable, impact-based decision support to NOAA and CIROH scientists that will enhance our ability to develop effective communication strategies, enhance information transfer and understand barriers and opportunities associated with use of National Water Model products.}_{https://water.w3.uvm.edu/scripts/bib/journal_article/query2.php?id=1}

 
Title: Leveraging emerging sensing technology and machine learning to improve and expand hydrological forecasting to hyper-local scales with NWM-coupled adaptive sensor networks 

Principal Investigator: Andrew Schroth

Research Team: Mirce Morales, Beverley Wemple, Jamie Shanley, Jarlath Oneil-Dunne 

Insitution(s): University of Vermont, USGS

^{Abstract: Much of the northeastern U.S. is dominated by montane headwater catchments, and recent flooding events illustrate the need for accurate and timely forecasts for such systems. However, model forecast accuracy is often reduced in mountainous regions due to sparse gaging, complex topography, and spatially heterogeneous rainfall/runoff patterns. This project aims to improve the performance and expand the capacity of the National Water Model (NWM) forecasting in montane headwater catchments by achieving three main objectives: 1) assess the NWM performance in montane headwaters, which will improve our understanding of the combinations of geophysical and hydro-climatic forcings that govern streamflow at the subwatershed scale during different events and across seasons, and develop a machine learning correction algorithm that improves flow forecasts in these systems that could be operationalized, 2) deploy a distributed water level sensing network in focal Vermont watersheds in different river corridor environments upstream and downstream of gages that will allow us to use the NWM to forecast water level across different reach environments, and 3) predict water levels at different locations within each focal watershed based on the high-frequency water level data and NWM flow forecasts using machine learning. Upon completion of this project, users and managers of streams in montane watersheds will be able to easily access hyper-localized water level forecasts based on short-range NWM discharge predictions that are post-processed via adaptive sensing and machine learning models. Ultimately, we intend to develop an operational workflow that would allow other communities across the country to improve the performance of these forecasts and leverage relatively low-cost sensor technologies to provide distributed NWM-derived water level forecasts across environments and infrastructures of concern. Improving (through correction algorithms) and expanding (through distributed water level forecasting) the forecast capacity at these sites and providing a template for others to do so will improve operational workflows and extend water resources predictions, capabilities, and applications. Furthermore, the approaches developed here will be particularly suitable for the modular and model-agnostic environment envisioned for the NWM Next Generation Water Resources Modeling Framework (NextGen)._{https://water.w3.uvm.edu/scripts/bib/journal_article/query2.php?id=2}}

 
Title: Coupling novel low cost spatially distributed nutrient sensors and National Water Model output to forecast nutrient loading and inform state implementation of EPA mandated nutrient reduction targets—The Lake Champlain Basin Test Bed
 
Principal Investigator: Andrew Schroth

Research Team: John Kemper, Kristen Underwood, Scott Hamshaw, Jamie Shanley, Raju Badireddy, Monireh Dehabadi, Severin Schneebelli, Jarlath O’Neil-Dunne

Insitution(s): University of Vermont, USGS, Purdue University 

^{Abstract: While not included in the current version of the National Water Model (NWM), there is vast potential and associated demand to expand the model’s capacity to forecast water quality. Here, we are focused on leveraging NWM flow forecasts to force novel nutrient loading forecasts for select basins within the Lake Champlain Basin. This is a particularly relevant use case, as the Lake Champlain Basin has been mandated to reduce phosphorus loading through the Total Maximum Daily Load (TMDL) framework to improve impaired waters as mandated by Section 303a of the Clean Water Act. To develop these forecasts, we will primarily utilize long-term and high frequency concurrent flow and phosphorus concentration time series and machine learning algorithms to develop robust predictive models of nutrient loading.  Initial studies will focus on developing models in two small watersheds of distinct landcover monitored with sensors by our group since 2014 that have high frequency observational time series (nutrient concentration and flow measurements taken every 15 minutes). We will then expand our models to larger watershed systems within the basin that have flow measurements by USGS gages and long-term phosphorus monitoring by the Vermont Department of Environmental Conservation (grab water samples going back to 1991). These will constitute the first NWM-forced nutrient loading models. As loading model development is ongoing, we will also develop low-cost electrochemical phosphate sensors to be distributed spatially within our focal watershed across different phosphate source environments. Ultimately, we plan to use these distributed sensor data with NWM forcing and machine learning to forecast not only the riverine load of phosphate in test case watersheds during storms, but also the source environments of the phosphate. Beyond research papers and presentations, one of the most significant outcomes of this work will be providing a template for others to use the NWM with water quality monitoring data to produce water quality forecasts across the country. We intend to develop an approach that will be fully compatible with NWM Nextgen that is also NWM version agnostic, thus being a useful long-term Research to Operations tool that expands the operations context and community. Given that there are over 50,000 impaired waters governed by TMDLs in the United States, the potential for this research to expand the operations community and NWM utility is substantial.}  _{https://water.w3.uvm.edu/scripts/bib/journal_article/query2.php?id=3}

 
Title: Post-processing NWM output with spatially distributed turbidity sensing to forecast turbidity loading and source for reservoir operation management

Principal Investigator: Andrew Schroth

Research Team: John T. Kemper, Kristen L. Underwood, Jarlath O’Neil-Dunne, Scott Hamshaw, Jamie Shanley 

Insitution(s): University of Vermont, USGS 

^{Abstract: Reservoir operations and management increasingly extend beyond considerations of water volume and water level to include water quality, especially in situations where reservoir water is unfiltered or treatment of a particular contaminant is onerous. A primary concern regarding reservoir water quality is often turbidity – a measure of water clarity that is chiefly impacted by how much sediment and other material is suspended in the water column – which can impede reservoir operations when certain levels are exceeded. Because the National Water Model (NWM) has been shown to have substantial utility for reservoir operations by providing flow forecasts that inform anticipation of future water volumes, it is sensible to leverage this forecasting capability to provide insight into future turbidity levels. Additionally, many prior studies have suggested that turbidity is primarily influenced by water discharge and other hydrologic parameters forecasted by the NWM (such as rainfall) as well as easily obtainable watershed characteristics (such as geology), indicating turbidity prediction may be readily achievable by coupling the hydrologic forecasting capability of the NWM to empirical models of turbidity production. In this project, we employ such an approach in the Esopus Creek catchment in the Catskills Mountains of New York State, which drains to the Ashokan Reservoir of New York City water supply system and has been extensively monitored for the past decade. We build off prior research to construct a machine learning-based model of turbidity as a function of antecedent conditions, storm hydrology, and watershed characteristics. In initial testing, this model, which leverages the distributed, high-resolution sensor network present in the Esopus watershed, outperforms the current model used by the New York City Department of Environmental Protection (NYC DEP) in reservoir operations. These preliminary results support the utility of our proposed approach and suggest that machine-learning models built on understanding of watershed processes can be fed NWM forecast products to successfully anticipate future turbidity loading. Work in the second and third year of the project will continue to fine-tune turbidity models with additional site-specific data from the Esopus sensor network to further improve forecasting capabilities and enhance forward-thinking reservoir operations. In particular, the next steps will aim to forecast not only turbidity levels, but also anticipate where in the watershed turbidity will be produced. This type of source-specific forecasting has substantial importance for both operational planning and management efforts (e.g., erosion mitigation) to suppress sediment loading to Esopus streams. Overall, results of this project will emphasize the capabilities of the NWM to extend beyond hydrologic forecasts and provide a blueprint for others interested in leveraging such abilities to produce water quality predictions.} _{https://water.w3.uvm.edu/scripts/bib/journal_article/query2.php?id=4}

 
Title: Improved Representation of Floodplains and Natural Features for Channel Routing 
 
Principal Investigator: Beverley Wemple 

Research Team: Rebecca Diehl, Kristen Underwood, Julianne Scamardo, Eric Roy, Kenneth Johnson 

Insitution(s): University of Vermont 

^{Abstract: Floodplains play an important role in the attenuation of floods, influencing river forecasts and flood inundation predictions, but they are poorly represented in the National Water Model (NWM). This project aims to improve our understanding and modeled representation of the influence of floodplain-channel connectivity on flood celerity and flood routing processes. Our initial use case is situated in the Northeastern US, comprising over 3000 river reaches in the Vermont portion of the Lake Champlain basin. We use high-resolution topographic data to develop river reach morphological signatures in cross section to characterize floodplain types. A supervised machine-learning algorithm was used to cluster reaches based on their hypothesized influences on flood attenuation. The workflow for topographic signature extraction is publicly available on a GitHub repository, with future improvements planned as we integrate planform complexity into our characterization. Future work in project years 2-3 will involve development and testing of hydrodynamic models to validate hypothesized differences in routing for distinct floodplain classes, and for river reaches in our northeast testbed sites, instrumented with water level and inundation tracking sensors. As an outcome of the hydrodynamic modeling, we aim to identify the reach types for which the simpler representation of flood wave routing (i.e., Muskingum Cunge) is appropriate and which settings may require a more computationally expensive approach to optimize efficiency for enhanced performance of river stage forecasts and inundation extent predictions. Products developed through this project will be helpful for the National Water Center (NWC) and its partners (e.g., USGS) in the development of the Next Generation Water Resources Modeling Framework (NextGen) and geospatial data that support national hydrologic modeling applications.}_{https://water.w3.uvm.edu/scripts/bib/journal_article/query2.php?id=5}

 
Title: Forecasting the Incidence and Duration of Harmful Algal Blooms (HABs) at Daily, Weekly and Seasonal Scales 
 
Principal Investigator: Asim Zia

Research Team: Patrick J. Clemins, Panagiotis Oikonomou, Donna Rizzo, Andrew W. Schroth, Safwan Wshah, Peter Isles,Imad Hanoun, Kareem Hanoun, Scott Turnbull, Noah B. Beckage, Mohammad Adil, Montana Bailey, Hakan Unveren, Saul Blocher, Jarlath O’Neil Dunne, Luis D. Espinosa, George Pinder

Insitution(s): Department of Community Development and Applied Economics, University of Vermont, Department of Computer Science, University of Vermont, Department of Civil and Environmental Engineering, University of Vermont, Department of Geography and Geosciences, University of Vermont, Department of Environmental Conservation, Vermont Agency of Natural Resources, Water Quality Solutions, Inc., Vermont EPSCOR, University of Vermont, Rubsenstein School of Environment and Natural Resources, University of Vermont, Department of Electrical and Biomedical Engineering, University of Vermont

^{Abstract: Despite significant advancements in satellite monitoring of Harmful Algal Blooms (HABs), the “accurate” forecasting of HABs and development of “real-time” HABs Early Warning Systems (HABEWS) at finer spatial (200 m to 500 m) and temporal (daily to seasonal) scales still requires a considerable amount of basic and applied research. In this sub-project, we will advance the National Water Model (NWM’s) predictive intelligence for early warnings of HABs at daily, weekly, and seasonal lead time scales by applying machine learning, process-based modeling, and hybrid frameworks that couple the two. This project will scale and test an approach to forecasting HABs in freshwater lakes and estuaries that leverages hydrological forecasts derived from NWM and existing Earth Observation datasets currently being produced in real time through satellites and in situ monitoring systems and sensors. In year 1, WRF-Hydro derived NWM hydrological hindcasts and forecasts will be embedded in an existing Integrated Assessment Model (IAM) computational workflow to drive an already calibrated and validated process-based lake model (AEM3D). The IAM also uses weather data that can be derived from National Weather Model and/or IBM weather forecast products. This workflow will produce high resolution HAB hindcasts and forecasts in two bays of Lake Champlain (Missisquoi Bay and St. Albans Bay). In year 2, we will develop and test a self-learning AI-HABEWS by identifying a best-fitting deep neural network to update AEM3D by validating HAB forecasts through community science monitors, and in situ & satellite sensors. In year 3, we will conduct a sensitivity analysis of HAB forecast accuracy (as predicted by machine-learning, process based, and hybrid forecast models) to hydrological forecasts based on NWM (both WRF-Hydro and Topmodel/NextGen versions), SWAT, and RHESSys models. This sub-project will advance NOAA’s mission to understand and predict the effects of changing climate, weather, and socio-environmental factors on marine ecosystems. This may, in turn, help conserve marine ecosystems and further advance NOAA’s vision of building resilient and healthy ecosystems.} _{https://water.w3.uvm.edu/scripts/bib/journal_article/query2.php?id=6}

 
Title: Integrate observations and field work in support of expanded evaluation data testbeds in the Northeast 
 
Principal Investigator: Asim Zia

Research Team: Patrick Clemins, Eric Roy, Scott Turnbull

Insitution(s): University of Vermont

<sup>Abstract: This cross-cutting project will build a Northeast National Water Model (NWM) Applications Testbed that defines best practices and implementation frameworks for water resource forecasts that make use of NWM outputs and provides the cyberinfrastructure resources necessary to fuse a diverse array of data sources to force and evaluate the forecasts.  

The goal of the project is to support evaluation and research to operations (R2O) for the following three NWM application forecast use cases: (1) Improving the representation of floodplain effects on hydraulic routing and flood inundation modeling and mapping (flood forecasting); (2) Coupling novel low-cost spatially distributed nutrient sensors and NWM output to forecast nutrient loading and inform state implementation of EPA mandated nutrient reduction targets (nutrient forecasting); and (3) Forecasting the incidence and duration of harmful algal blooms (HABs) at daily, weekly, and seasonal scales (HABs forecasting).  

This work supports research on the development of water quantity and quality forecasts in the NWM with the following objectives: (1) Verify river stage time series and flood inundation dynamics at selected sites representative of different hydrogeomorphic settings to inform improved channel routing algorithms (flood forecasting); and (2) Improve the understanding of the relationships between streamflow and transported constituents in support of developing water quality forecasts (nutrient forecasting) and predicting the occurrence of harmful algal blooms (HABs forecasting).  

The Northeast NWM Application Testbed will expand and maintain a river and floodplain sensor network consisting of (1) High-frequency, low-cost sensor systems installed along select NHDplus reaches to monitor river stage and document floodplain inundation timing, duration, and extent; (2) Drone-based sensors to spatially enhance and contextualize at-a-point measurements of river stage; and (3) High-frequency turbidity sensors to guide the development of discharge-concentration relationships. The Testbed will also establish water resource forecasting best practices and implementation frameworks as well as the complementary cyberinfrastructure necessary to provide data and code for forecast evaluation, utilities and containers to analyze and reformat data, documentation for the various data sources, and operational user outreach and facilitation to help the CIROH forecast teams use the Testbed. The Northeast NWM Applications Testbed will serve as an integration and collaboration nexus for NWM applications researchers both regionally and nationally. Through the testbed services, including compute resources, data storage, a forecast workflow and protocol framework, software tools, and facilitation, the Testbed is catalyzing CIROH research across the three use cases by providing an accessible, easy-to-use, standardized set of evaluation data, models, and tools to researchers. In addition, the Testbed is convening discussions among regional and national CIROH participants around evaluation best practices including next generation evaluation metrics (information theory, event-based, and object-based) and error analysis and disaggregation methods. The Testbed will support establishment of new operational NWM application forecasts across the three initial test cases. In addition, as our Testbed users develop their NWM application forecasts, they can contribute to NextGen development by determining which NWM reach forecasts contribute most to the errors in their own NWM application forecast and inform which hydrology models are most appropriate for certain classes of reaches or in need of further development.</sup>_{https://water.w3.uvm.edu/scripts/bib/journal_article/query2.php?id=7}