Research Activities

IDEAS-Watershed is organized around six Research Activities

This illustrates the relationship of these activities with the three SBR cornerstones. The Reaction Network activity is centered in the Reaction Network cornerstone, the three synergistic Partnership Activities are within the Watershed Hydrobiogeochemistry Cornerstone, and the CONUS Activity is in the Continental Hydrology cornerstone. The Shared Infrastructure Activity works common needs with shared solutions across these activities, and helps move information and capability across the scales associated with each cornerstone.

SFA Partnership Activities

The three Partnership Activities, each undertaken jointly with one of SBR’s interdisciplinary Science Focus Areas (SFAs) (LBNL, ORNL, PNNL), addresses biogeochemical cycling in streams across a wide range of stream orders in disparate climates and land-use conditions, from

  • [LBNL] Watershed Function SFA: East River Use Case – a pristine mountainous headwater catchment
  • [ORNL] Critical Interfaces SFA: East Fork Poplar Creek Use Case – a fourth-order mercury-contaminated stream in deciduous broadleaf forest in a semitropical humid climate
  • [PNNL] River Corridor SFA: Columbia River Use Case – to a heavily managed ninth-order river.

The three regional SFAs share a long-term objective of elucidating how surface water/groundwater exchanges across a range of temporal and spatial scales control watershed-scale biogeochemical cycling.

Common near-term modeling activities for the interdisciplinary SFAs focus on

  • surface water and groundwater exchange in watershed headwaters and stream/river corridors,
  • how biogeochemical processes in spatially limited metabolically active zones interact with steady and unsteady hyporheic exchange flows to control reach-to-watershed-scale export of nutrients and inorganic contaminants, and
  • developing multiscale strategies that allow coupled hydrologic and biogeochemical processes to be represented at their native scales but expressed at watershed scale, rather than the current status of hydrologic modeling at large scale and biogeochemical modeling at smaller scales.

Watershed Function SFA: East River Use Case (LBNL)

Digital elevation map of the East River Watershed at 0.5 m resolution zooming into a subsystem comprising hillslopes and the floodplain, and further into a 2-meander subsection within the floodplain.

The Watershed Function SFA seeks to determine how perturbations to mountainous watersheds (e.g., floods, drought, early snowmelt) impact the downstream delivery of water, nutrients, carbon, and metals. The study area for this use case will be the East River Watershed (Hubbard et al., 2018), a representative mountainous watershed in the Upper Colorado River Basin. The domain contains pronounced gradients in hydrology, geomorphology, or type of biome that challenge conventional watershed hydrobiogeochemical models. Currently, no single model can capture all relevant processes across this domain at fine resolution.

Aggregation of the system behavior across subsystems and scales is a central need for the Watershed Function SFA. For example, it seeks to evaluate what the contribution of hillslope or floodplain processes is to net exports of the East River.

Variable-resolution unstructured mesh for the Copper Creek sub-catchment, north of region b, with mesh refinement in and around the streams

IDEAS-Watersheds Partnership with the Watershed Function SFA at LBNL aims to develop a multiscale modeling framework that will allow us to consider processes at different resolutions within the watershed, including both the software tools and workflows required to enable this framework.

Development will be primarily based on the Advanced Terrestrial Simulator (ATS) with a specific emphasis in the newly available reactive transport capabilities. This involves the Alquimia geochemical interface and the geochemical engines PFLOTRAN and CrunchFlow. A central aspect of the work will also be the meshing tools that will support the generation of unstructured meshes with variable resolutions. We intend to develop a simulation capability that relies on multiple resolution unstructured meshes to dynamically adjust the process resolution and efficiently perform the simulation over large spatial extents.

 

Critical Interfaces SFA: East Fork Poplar Creek Use Case (ORNL)

Hg transformation to MeHg in East Fork Poplar Creek, Tennessee, is the use case for the Partnership Activity with ORNL’s Critical Interfaces SFA. The transformation of Hg to MeHg is biologically mediated, coupled to in-stream nutrient cycling, and occurs only in anaerobic conditions in spatially limited regions. IDEAS-Watersheds will work with the Critical Interfaces SFA to address those multi-physics and multiscale modeling challenges. (Image provided by Oak Ridge National Laboratory.)

Transient storage zones (TSZs) surrounding stream channels are locations where the downstream movement of water is delayed in comparison to the main channel flow. Metabolically active transient storage zones (MATSZs) are important subsets of TSZs that are responsible for a significant portion of carbon, nutrient, and trace metal processing, thus affecting stream biogeochemistry and, ultimately, downstream water quality.

The Critical Interfaces SFA at ORNL is elucidating and quantifying the role that coupled hydro-biogeochemical processes associated with MATSZs have on trace element fate and transformation in low-order freshwater stream systems, using the transport of Hg and its microbially mediated transformation to the neurotoxin methylmercury (MeHg) in East Fork Poplar Creek (EFPC), Tennessee as its use case shown in the Figure.

IDEAS-Watersheds Partnership with the Critical Interfaces SFA at ORNL will develop a process-rich modeling framework that enables laboratory-scale experiments to be linked with reach-scale field observations, contributing to the over-arching strategy of a new multiscale modeling methodology (Painter 2018) that makes it possible to tractably represent the effects of redox zonation and other fine-scale geochemical phenomena in reach-to-watershed–scale models. The approach extends highly successful residence-time frameworks to accommodate nonlinear multicomponent reactions and transient flows. We will complete the implementation and testing of the new hyporheic zone model as a subgrid model in integrated surface/subsurface flow models represented in Amanzi-ATS and apply it to two low-order streams.

River Corridor SFA: Columbia River Use Case (PNNL)

Illustration of proposed modeling framework of PNNL River Corridor SFA to predict the cumulative effects of smaller-scale river corridor hydrobiogeochem¬ical processes on watershed system functions.

The River Corridor SFA at PNNL explicitly studies interactions among variable river surface elevation (“stage”), hydromorphic setting, and hydrogeologic heterogeneity to determine how those interactions influence river corridor hydrobiogeochemical function. The river corridor science is conducted in the context of larger watershed processes that define boundary fluxes and exert other controls on hydrologic exchange. Therefore, our model framework must simulate land-surface and groundwater processes over domains much larger than the river corridor itself.

IDEAS-Watersheds Partnership with the River Corridor SFA at PNNL aims to achieve the SFA’s goal in advancing fundamental understanding of the hydrobiogeochemical function of dynamic river corridor ecosystems and translating the fundamental process understanding into predictive, interoperable models across watersheds, shown in the Figure.

Managing the workflows for multiscale modeling of this type is a significant computational challenge, which includes multiscale integration of fundamental process understanding (genome to watershed scales) as well as integration with distributed data (data assimilation and uncertainty quantification) and coupling multiple modeling platforms (PFLOTRAN, CLM, SWAT, OPENFOAM, NEXSS). Additionally, it is important that the river corridor module developed by PNNL can be used as an interoperable component in a watershed modeling system so it can contribute to other projects. This will require flux couplers with standardized inputs and outputs to allow interaction with other codes such as PFLOTRAN-CLM, ATS, and WRF-Hydro. Partnership with IDEAS-Watershed will allow the PNNL SFA to leverage Alquimia for river corridor biogeochemistry, the xSDK for scalability and unstructured grid refinement along river corridor cells, and new workflow tools in data management and analyses.


CONUS Activity

Map of the expanded CONUS 2.0 domain developed with IDEAS-Watersheds. The box outlines the original CONUS 1.0 domain. The new domain extends to the coastlines and covers all tributary areas to the CONUS. This extent is consistent with the domain of the national water model and additionally the model grid has been setup to align with the National Water Model simulation grid to facilitate direct coupling and comparisons.

Simulating integrated flow over continental scales at so-called hyper-resolution is an identified grand challenge in computational hydrology (Wood, Roundy et al. 2011, Bierkens, Bell et al. 2015). To adequately capture feedback between deeper subsurface flow, the land energy budget, and the lower atmosphere, explicit connections need to be made between these systems in large-scale models. Explicitly simulating these interactions is computationally expensive and therefore often simplified (either in process or resolution) in continental and global modeling approaches.

The IDEAS-Watersheds CONUS Activity will address this gap by supporting the development of an integrated hydrologic modeling platform of CONUS using ParFlow-CLM. The CONUS model, shown in the map of CONUS 2.0, bridges across our study areas and provides a scaling framework from the reach scale up to watershed and regional systems.

The CONUS Activity works to advance a continental-scale simulation platform that can bridge between individual sites and provide larger-scale context across the SFAs to provide an integrated modeling framework that spans the United States at a relatively high spatial resolution.

Roadmap of CONUS model development. IDEAS-Watersheds CONUS Activity will expand the current five layer subsurface model with improved spatial representation of alluvial aquifer thickness, depth to bedrock, and deeper confined aquifer systems. Additionally, improved physical parameterizations of subsurface snow and land-surface/land energy processes will be explored.

The proposed continental modeling will build from the existing CONUS model, applying a phased approach to explore multiscale, multi-physics treatments of terrestrial hydrology modeling from the bedrock into the atmosphere. The continental research proposed here falls under three categories: (1) model development, (2) software development and sustainability, and (3) continental simulations. The continental simulations will be used to evaluate the spatial scales and physical settings in which lateral groundwater flow has the largest impact on overall watershed behavior and sensitivity to stress. Additionally, we will use EcoSLIM to apply particle tracking to the hydrologic simulations and evaluate flow paths and residence times across watershed scales and under different hydrologic and management scenarios.

Current simulations are primarily designed to represent natural (i.e., predevelopment) hydrology with the exception of groundwater extractions, which have been included in several test cases. Moving forward with the new CONUS2.0 domain, additional parameterizations will be added to capture spatially distributed consumptive water use, irrigation, and groundwater pumping. Finally, in conjunction with the NWM team, we will explore the impact of different overland flow and land-surface formulations using the PF-WFR-Hydro coupling that is being developed.

 


Reaction Network Activity

Development of reaction networks to describe biogeochemical transformations is a central component of the SBR research. All SFAs in one form or another develop and use reaction models for aqueous complexation, surface complexation, mineral dissolution-precipitation, and microbially mediated reactions. In particular, the fine-scale SFAs at Stanford Linear Acceleratory Center (SLAC), Lawrence Livermore National Laboratory (LLNL), and Argonne National Laboratory (ANL) focusing on developing mechanistic models of biogeochemical transformations in subsurface environments at fine scales, using molecular, genomic, metabolic, solid-phase, and aqueous chemistry data.

The IDEAS-Watersheds Reaction Network Activity bridges across the SFAs and develops the use of process-explicit reaction models for aqueous complexation, surface complexation, mineral dissolution-precipitation, microbially mediated reactions, microbial dynamics, and similar processes.

The collaboration activities will focus on two major aspects: (1) develop new approaches and workflows to incorporate new data and knowledge into the reaction models, including the derivation of reaction networks from metagenomics data and reaction parameters; and (2) implement reaction models that reflect the new understanding needed in the code CrunchFlow.


Shared Infrastructure Activity

Although the modeling strategies adopted by various activities within SBR differ in important details, they share significant modeling challenges and infrastructure needs. Shared computational infrastructure needs include workflow tools for transferring information across scales, customized meshing, and model-data integration, and interfaces to couple multiple process-based codes and support code interoperability.

The IDEAS-Watersheds Shared Infrastructure Activity will coordinate the development of the shared infrastructure by creating and advancing existing multiscale workflow tools and software interfaces to support interoperability.

Example relationship of modeling domains across scales: the multiscale workflow activity will enable model output to be used to drive boundary conditions of models at finer scales. Workflows will also be developed to work with model input data in native formats and scales, and map it to the desired computational mesh. Particle-based methods are being advanced to analyze scaling relationships and identify potential strategies for bridging fine-scale mechanistic models and understanding to larger scales.

The Shared Infrastructure can be shared among the SFAs to support those multi-scale models and reduce duplication of effort and ensure long-term sustainability of the capability.

The IDEAS-Watersheds Shared Infrastructure Activity includes:

  • Multiscale Workflow Tools for
    • hydrobiogeochemical simulations and analyses across multiple resolutions and scales
    • multi-resolution unstructured meshes that conform to stream/river geometries and allow refinement in and near river corridors
    • using metagenomics data for reaction network generation
    • etc.
  • Interfaces
    • Continue development of formalized interfaces to support code interoperability, focusing on further development of Alquimia and a land-model interface
  • Sustainable Software Ecosystem

The IDEAS-Watersheds Shared Infrastructure Activity

  • Extends the BER ecosystem of interoperable software tools and components to better represent hydrobiogeochemical processes in dynamic stream/river corridors along the river continuum.
  • Provides software tools and workflows that can be shared among the SFAs as well as the critical (and currently incomplete) experience base on how to parameterize and use the new capabilities.