Property:CSDMS meeting abstract
From CSDMS
This is a property of type Text.
A
This presentation discusses the implementation of component-based software design in Eco-hydrologic modeling. As a first step, we present development and integration of a radiation component that uses the local topographic variables to compute shortwave and longwave radiation data over a complex terrain for modeling Eco-hydrologic dynamics. This component is integrated to a central element that develops and maintains a grid, which represents the landscape under consideration. This component communicates with various other components such as ‘vegetation component’ and ‘soil moisture component’. This component is adapted from the Channel-Hillslope Integrated Landscape Development (CHILD) Model code and has been enhanced. Preliminary results of this study demonstrate the advantages of adopting component-based software design such as improved flexibility, interchangeability and adaptability. +
This presentation or poster will discuss the latest developments of the CUAHSI Hydrologic Information System including 1) the new open source server components built using PHP and MySQL specifically to support citizen science; and 2) the desktop application HydroDesktop with its extensions for search and discovery of data on the 100 servers of the CUAHSI data network. The presentation or poster will include a discussion of the potential integration of HIS data sources in CSDMS modeling efforts and potential for integration of the CSDMS modeling architecture with the HydroDesktop client application. +
This research aims to understand the evolution of the shoreface of sandy, wave-dominated coasts. Using energetics-based formulations for wave-driven sediment transport, we develop a robust methodology for estimating the morphodynamic evolution of a cross-shore beach profile. We compare how shallow water wave assumptions and linear Airy wave theory affect the estimation of morphodynamic shoreface evolution, in contrast to previous work, which has applied shallow water wave assumptions across the entire shoreface. The derived cross-shore sediment flux formula enables the calculation of a steady state (or dynamic equilibrium) profile based on three components of wave influence on sediment transport: two onshore-directed terms (wave asymmetry and wave drift) and an offshore-directed slope terms. Equilibrium profile geometry depends on wave period and grain size. The profile evolution formulation yields a morphodynamic Péclet number that can be analyzed in terms of perturbations around the steady-state profile. The diffusional, offshore-directed slope term dominates long-term profile evolution. A depth-dependent characteristic timescale of diffusion allows the estimation of an effective morphodynamic depth of closure for a given time envelope. Theoretical modeled computations are compared to four field sites along the Eastern US coastline. For each of these four field sites, we use hindcast wave data to determine a representative wave height and period using a weighted frequency-magnitude approach. Using the characteristic wave quantities for each site, we compute the equilibrium profile and the morphodynamic depth of closure, showing reasonable similarities between the computed equilibrium profiles and the actual profiles. In addition, the estimated morphodynamic depth of closure matches well with the location of the visually estimated depth of closure (based upon slope break) for each site. Overall, the methodology espoused in this paper can be used with relative ease for a variety of sites and with varied sediment transport equations.
2
This study aims to fundamentally assess the impact of sea level rise (SLR) on vegetated, muddy coastlines. This includes an assessment of the resilience of coupled salt marsh-mudflat and mangrove fringe-mudflat coastlines under different sea level rise scenarios.
Traditionally, the design of coastal protection measures revolved around the use of hard structures to ensure a certain level of design safety against flooding of the coastal hinterland. However, with the effects of climate change: sea level rise, increased intensities and frequencies of storms; these solutions appear to be unsustainable. Building-with-Nature strategies have reinforced the value of vegetated foreshores, as being capable of allowing for a flexible and adaptive response to climate change. They attenuate wave energy, stabilize and may heighten the foreshore at a rate that matches that of sea level rise. Important parameters related to the resilience of vegetated foreshores to sea level rise are site specific and include sediment supply, wave climate, tidal range, sea level rise rates, type of vegetation cover, vegetation dynamics and topography.
Process-based numerical modelling tools are critical towards enhancing the understanding of the processes governing the morphological development of vegetated-mudflat systems. Limited studies have quantified the impact of sea level rise on the resilience of these intertidal systems with a key focus on determining the tipping points and the governing processes for bio-geomorphological development. Therefore, we applied an open-source 2D process-based model (Delft3D) that couples intertidal flow, wave-action, sediment transport, morphodynamic development with the vegetation dynamics for temporal and spatial growth and decay of vegetation and bio-accumulation.
The vegetation growth model was developed using MATLAB, which was then coupled with a depth averaged Delft3D model. For the salt marsh species, the growth model was based on that of a population dynamics approach whereas the mangrove growth model was based on a windows of opportunity approach. The model setup was inspired by conditions within the Dutch South Western Scheldt and the Guyana coastline for the salt marshes and mangroves respectively. The numerical model and the coupling approach were validated quantitatively against existing theory, data and laboratory studies; after which the system’s resilience against sea level rise was examined.
Spatial equilibrium of the marsh-mudflat system was attained within 120 years with wave action and sediment dynamics being key triggers. The mangrove fringe-mudflat model however attains equilibrium on longer timescales. The subsequent imposition of a 100 year period of rising sea-level (1.1m) in salt marsh-mudflat systems revealed the biomass accumulation as a critical determinant for the drowning rate. Though, initially highly resilient against the exponential increase of sea level rise, the marsh system starts to drown as channels incise the platform after 50-60 years. This corroborates recent studies which predict a decline in the carbon sequestration potential of salt marshes within the North Sea. Contrastingly, the mangrove fringe-mudflat system proved resilient after a 100 year period of extreme SLR and the increases in drag gained from their extensive mangrove root network and the below ground biomass accumulation proved to be the main drivers. However, after 150 years, there is a shift in the nature of the system as it starts to drown. Results show survival for both systems in sediment rich areas. Overall, the model can be applied to assess the vulnerability and resilience of vegetated coastal areas impacted by sea-level-rise worldwide. Thereby, proving to be a useful tool for developing countries where data is scare. Both the Delft3D software and MATLAB tools used in this study are open source and freely available online: https://oss.deltares.nl/web/delft3d. The running of the model requires the use of MATLAB versions 2013 or higher. This software can be attained through purchase, student version or trial online: https://nl.mathworks.com/products/matlab.html. With regards to the hardware required, a standard PC with minimum 8GB RAM. Additionally, the MATLAB source code will be made available via the Environmental Modelling and Software Journal (ESM) once published.
This study investigates riverine sediment dispersal within a bedrock confined estuary in British Columbia, Canada using the HydroTrend and Sedflux models. The models are evaluated using multibeam and acoustic backscatter surveys, piston cores, and grab samples across the Skeena Estuary and its contiguous marine areas. The data has been compiled to produce an overview of seabed geomorphology, texture, and sedimentation rates in the estuary and marine approaches. The model HydroTrend was used to estimate incoming sediment load from the Skeena River. Model estimates of suspended sediment load are higher than past estimates due to a large contribution of suspended sediment from a portion of the Skeena watershed previously excluded due to a lack of available hydrographic data. Over thirty kilometres from the river mouth, cores recovered mud sequences in the deeper proximal bedrock confined channels that indicated sedimentation rates of up to 2.83 cm yr-1. These deeper estuarine passages are seaward of the sandy deposits that make up the delta platform. In comparison, sedimentation rates in the further offshore marine approaches to the Skeena Estuary are as low as 0.004 cm/yr. Sedimentation rates within the estuary agree with the SedFlux model outputs using the HydroTrend sediment load results. More specifically, a sedimentation rate of 2.9 cm/yr was predicted using the SedFlux model at the same distance from the river mouth as the mud sequence radiocarbon dated cores. A relatively high sedimentation rate and seaward fining trend in grain size are interpreted as indicators of high riverine input to the seabed regionally. This initial evaluation of model performance encourages further examination of sedimentation conditions in the Skeena Estuary, including those of importance to eelgrass beds and major port development areas. +
This work explores how feedbacks between erosion and sediment production in landscapes with layered stratigraphy influence channel evolution. In layered rocks, contrasts in erodibility cause erosion rates to vary through space and time, complicating landscape response to external forcing from climate and tectonics. Recent studies have used the detachment-limited stream power incision model to explore the complex variations in erosion rates that arise from channel incision through layered rocks. However, these studies do not capture the effect of sediment cover on channel evolution. This work uses the recently developed Stream Power with Alluvium Conservation and Entrainment (SPACE) model (Shobe et al. 2017) to explore how sediment cover influences landscape evolution and modulates the topographic expression of erodibility contrasts in mixed bedrock-alluvial rivers incising through horizontally layered rocks.
The SPACE model allows for the simultaneous treatment of bedrock, fully alluvial, and mixed bedrock-alluvial channels and transitions smoothly between detachment- and transport-limited behaviors. Here, we use the SPACE model to explore how sediment load influences effective erodibility in layered strata, motivated by topographic and lithologic variability found in the Guadalupe Mountains of Texas and New Mexico. We use the Landlab Toolkit to simulate fluvial incision through alternating horizontal layers of hard and soft rock using the SPACE model. While the SPACE model does not treat individual grains, the relative influence of grain size is modeled by systematically varying particle settling velocity and the erodibility of the alluvial across model runs. We find that sediment cover strongly modulates landscape response to uplift, and model runs with “finer” sediment (lower particle settling velocity and more erodible alluvium) reach a steady average elevation more quickly than model runs with coarser sediment. As particle settling velocity is increased, normalized channel steepness increases in soft rock layers and decreases in hard rock layers. We also explore how sediment flux at the watershed outlet varies as soft and hard layers are exposed in different proportions. Finally, we compare how erosion rates vary through space and time as relative sediment size increases. This work illustrates the importance of feedbacks between erosion and sediment production for landscape evolution, particularly in layered rocks where erosion rates vary in space and time.
A
Though it enhances the exchange of porewater and solids with the overlying water, the role that sediment resuspension and redeposition play in biogeochemistry of coastal systems is debated. Numerical models of geochemical processes and diagenesis have traditionally parameterized relatively long timescales, and rarely attempted to include resuspension. Meanwhile, numerical models developed to represent sediment transport have largely ignored geochemistry. Here, we couple the Community Sediment Transport Modeling System (CSTMS) to a biogeochemical model within the Regional Ocean Modeling System (ROMS). The multi-layered sediment bed model accounts for erosion, deposition, and biodiffusion. It has recently been modified to include dissolved porewater constituents, particulate organic matter, and geochemical reactions.
For this talk, we explore the role that resuspension and redeposition play in biogeochemical cycles within the seabed and in benthic boundary layer by running idealized, one-dimensional test cases designed to represent a 20-m deep site on the Louisiana Shelf. Results from this are contrasted to calculations from an implementation similar to a standard diagenesis model. Comparing these, the results indicate that resuspension acts to enhance sediment bed oxygen consumption. +
Through funding provided by the US Integrated Ocean Observing System, five open source 3-D hydrodynamic models for Chesapeake Bay have been compared to each other and to EPA monitoring data for hindcasts of the years 2004 and 2005. The aim of this project is to provide NOAA, EPA, other government agencies, and the larger modeling community meaningful guidance on the relative accuracy, efficiency, complexity and likely utility for federal operational and scenario modeling of a suite of community models available for simulating hydrodynamics and oxygen dynamics in Chesapeake Bay. The focus of the present paper is on the hydrodynamic comparison of:
# the ChesROMS model (http://ches.communitymodeling.org/models/ChesROMS/index.php)
# the CBOFS2 model (http://cedb.asce.org/cgi/WWWdisplay.cgi?265616)
# the CH3D model (http://www.chesapeakebay.net/publication.aspx?publicationid=55318)
# the EFDC model
# the UMCES ROMS model
These models represent a range of resolutions (from ~5,000 to ~50,000 wetted cells). The models do similarly well in reproducing 3-D, time-dependent temperature fields. Bottom salinity is significantly improved with increases in horizontal resolution that better capture the structure of narrow, deep channels. Seasonal variation in density stratification is surprisingly difficult for all the models to capture well, and density stratification is not found to be especially sensitive to horizontal or vertical resolution within the range of resolutions considered. The hydrodynamics in general are not particularly sensitive to refinements in offshore climatological forcing, nor to refinements in riverine input, nor to refinements in spatial resolution of wind forcing. Lateral and longitudinal advection is sensitive, however, to seasonal changes in wind velocity and direction, suggesting that typical seasonal changes in wind forcing may be more important than seasonal changes in local stratification in controlling transfer of oxygen to deep channels susceptible to hypoxia.
2
Throughout the world’s oceans, there are hundreds of coral atolls. These structures provide valuable habitats to plants and wildlife, and the islands on them are home to hundreds of thousands globally. Atoll islands, known as motu, are geologically quite young; they form relatively quickly, over hundreds or thousands of years. While there has been significant work in studying the effect of ocean waves on coastal morphologies, there is little such work specifically focusing on coral atolls and the motu upon them. In this study, we examine the morphologies and morphometrics of over 80 atolls in the Indian and Pacific oceans and quantify the morphometrics of the reef flats and motu with direction-based binning. We gathered recent ocean wave data from over 600 locations around these atolls using NOAA’s Wave Watch III simulations and quantified the local wave climate for each of our atolls. After analyzing these two separate datasets, we compared the effect of local wave climate on atoll morphology. In our analysis, we have found regional trends in wave climate, specifically we have observed directional differences in wave size prevalence between the wet and dry seasons. We have also noted possible relationships between reef width and wave height and between motu width and the 50 year wave height. As we continue to investigate these data, these findings will allow us to better understand the processes driving motu evolution and may be used to infer potential morphological changes in the presence of changing wave climates. +
Tidal inlets are crucial land-forms that control the exchange of water and sediment between the open sea and the back-barrier basin. Despite the well accepted relationship between tidal prism and inlet cross section area, some questions about the geometry of tidal inlets remain open. What processes set their width to depth ratio (or aspect ratio)? What control the presence of a single-threaded versus a compound channel within the same inlet? Do these relationships change with human activities? These questions were investigated by creating and analyzing a database of inlet geometry and numerical modeling. We found that inlet’s aspect ratio has a weak dependence on tidal range, wave height. Despite the scatter, we developed relationships between inlet width and aspect ratio for three separate types of inlets, i.e. engineered inlet, natural inlet with single channel and natural inlet with multiple channels. From the observation, we found that multiple channels tend to form when the width of natural inlet exceeds 1 km. To investigate the long-term morphological evolution of inlets, we developed an idealized barrier-inlet system in Delft3D (a 2D hydromorphodynamic model). We found that two parameters affect the aspect ratio the most: the transverse bed gradient factor for sediment transport (αbn) and the global/maximum dry cell erosion factor (θsd), which controls erosion of dry cells adjacent to a wet cell. Including dry cell erosion is necessary to widen an initially narrow inlet; removing the dry cell erosion effectively “armors” the side of the inlets and thus produces inlets with small aspect ratio. From model sensitivity analysis, we found that αbn = 10 and θsd = 0.8 provides inlet configurations that best match with observations. When inlets were able to widen, we found that aspect ratio has a weak dependency on the initial width. Despite the dry cell erosion and the transverse parameters being a simplified representation of bank erosion processes, their calibration allows to reproduce realistic inlet geometries.
Tidal marshes store blue carbon because biomass production by vegetation exceeds organic matter decomposition. When methanogenic microorganisms drive decomposition, organic biomass decomposes into methane, a greenhouse gas with a higher warming potential than carbon dioxide. As sulfate availability increases sulfate-reducers outcompete methanogens, and methane production decreases. Such a shift from methanogenesis to sulfate reduction as the predominant decompositional pathway can occur within tidal marshes experiencing sea level rise (SLR), as marsh inundation by saline water increases. Additionally, SLR can lead to changes in marsh morphology and extent. To address this interplay, we adapt a cross-shore numerical model for the evolution of a marsh-lagoon system to predict methane emissions over decadal time scales and under different SLR scenarios, via the addition of a novel biogeochemical module. We compute total methane emissions by integrating the methane flux at each location over the width of the marsh platform, which is controlled by the rate of SLR, the wave energy in the lagoon, and the rate of marsh upland migration. We calculate the methane flux at a given location as a function of its distance from the edge of the marsh/lagoon boundary and the labile carbon available for decomposition. We test the morphodynamic component of the model on marshes along the Great Bay near the outlet of the Mullica River in southern New Jersey. The model can reproduce the magnitude of morphological change seen in the historical data from 1986-2020. In particular, the model captures that the marsh is eroding faster at the marsh/lagoon boundary than it is being gained by landward migration of the marsh/mainland boundary. Preliminary results of the coupled biogeochemical and morphodynamic model show that generally methane emissions increase with higher rates of SLR, however certain environmental conditions allow for scenarios in which higher rates of SLR lead to lower methane emissions.
A
Tidal systems are biogeomorphic systems of great relevance, providing important ecosystem services and coastline protection against storms. The dynamics of these systems, currently threatened by the acceleration in the rate of global sea level rise (SLR) and the decrease in sediment supply, are governed by complex interactions between hydrological, ecological, and geomorphological processes. How do salt-marsh ecosystems respond to changes in the environmental forcings? What is the role physical and biological processes and of their interactions through eco-geomorphic feedbacks in controlling salt-marsh dynamic response to these changes and the existence of possible equilibrium states? To address these important issues and improve our understanding of the chief eco-geomorphic processes controlling salt-marsh response to current changes, we have developed a suite of eco-morphodynamic models accounting for complex two-way interactions between ecological and geomorphological processes. We find that vegetation crucially affects the equilibrium marsh elevation, marsh resilience to accelerations in SLR rates, and the morphological features of salt marsh channels. As soon as the platform is colonized by vegetation, plants crucially affect the local hydrodynamic circulation, favor channel incision, enhance particle settling by a reduction of turbulence levels within the canopy, promotes trapping sediment, and provides organic material. Model results suggest that highly productive and sediment-rich marshes will approach new equilibrium states in response to changes in the rate of SLR faster than sediment-poor or less productive marshes. Moreover, marshes exposed to large tidal ranges are more stable, and therefore more resilient to changes in the rate of SLR, than their microtidal counterparts. We also find that marshes are more resilient to a decrease rather than to an increase in the rate of SLR, and they are more resilient to a decrease rather than to an increase in sediment availability. Our modeling approaches emphasize that biological and physical interactions are crucial in determining the observed spatial patterns in the biological and in the geomorphic domains. The existence of feedbacks between physical and biological processes affects the evolutionary trajectories of saltmarsh ecosystems, and the reversibility of such trajectories, thus highlighting the importance of accounting for biogeomorphic feedbacks to obtain realistic representations of the system dynamics in response to climatic changes.
2
To answer geomorphological questions at unprecedented spatial and temporal scales, we need to (a) parse terabyte-scale datasets (DEMs), (b) perform millions of model realizations to pinpoint the parameters which govern landscape evolution, and (c) do so with statistical rigor, which may require thousands of additional realizations.
A core set of operations underpin many geomorphic models. These include determination of terrain attributes such as slope and curvature; flow routing; depression flooding and breaching; flat resolution; and flow accumulation.
Here, I present new, best-in-class algorithms which perform the foregoing. I explain how they are implemented in a high-performance, open source C++ library called RichDEM which is accessible to general practitioners via Python. This design is novel among terrain analysis software and I argue that it is necessary for moving the field forward in a way which allows for rapid scientific development and practitioner adoption. +
To better understand large-scale delta-network responses to fluctuating discharge, we focused on the evolution of a single channel-island node within a delta network. Using the Surface Transport and Earth-surface Processes (STEP) basin, we were able to construct and observe the evolution of mouth-bar systems and subsequent flow bifurcation around an individual island in transport-limited, turbulent conditions. Overhead time-lapse images, laser-altimetry scans, and a low-cost particle tracking velocimetry system allow us to characterize the flow and depositional evolution of our experimental islands. Two alternating discharges that model flood and interflood transport (6 l/s, 0.355 l/s) with uniform sediment (170 microns) were used to create two characteristic sediment advective lengths. Floods transport sediment in full suspension (P_flood at inlet = 0.16), while interfloods transport sediment as bedload (P_interflood at inlet = 2.7). The consequent deposits are distal steep deposits from floods raining sediment out of suspension, and proximal low-angle, leveed deposits from interfloods laterally advecting sediment and floods remobilizing sediment down-system. By varying the frequency of floods (one every 20s-20 mins) while keeping sediment and water mass constant across experiments, we are able to control the time and spatial organization of these two deposit types and examine the effect on bifurcation length and bifurcation incidence time. While the deposits are initially spatially segregated, as the interflood deposit and flood deposit accumulate sediment over time, the interflood deposit encroaches onto the flood deposit. Flow routes from the interflood deposit to the flood deposit and bifurcates because of a preferential slope gradient around the distal deposit. Rather than a single hydrodynamic condition dictating the location of bifurcation, the length to a bifurcation can be described by the intersection of multiple distributions of topographies from the variable flow of solids.
Topography, material properties, and gravitationally driven groundwater flow together act to control hillslope stability. Although it is well known that material strength and hydraulic conductivity differences can alter slope stability via feedbacks with groundwater, comparatively little is understood about the role of stratigraphic sequencing in governing how hillslopes fail. In northwest Washington State, the recent occurrence of the large-volume, high-mobility SR-530 landslide brought focus to hazards associated with large terraces of glacial sediment that inundate the valleys of the western Cascades. However, observations from high-resolution LiDAR topographic data show significant differences between terraces in adjacent valleys, and both geologic and subsurface data show that each site has a unique stratigraphic configuration. Here we hypothesize that variations in the bed thickness and sequencing of glacial sediment packages within ice-marginal terraces control landslide volume and failure style. Using a three-dimensional limit-equilibrium model, Scoops3D, we show that the variable distribution of silts, clays, sands, and tills have a first-order control on both the volume and location of failures along a terrace. Predicted landslide volumes vary by over an order of magnitude between different stratigraphic configurations. Variably saturated groundwater flow simulations show that hydraulic conductivity contrasts between glacial units lead to perched water tables with localized zones of high pore fluid pressure, and in most cases (but not all) the failure pattern set by stratigraphy is amplified by the presence of groundwater flow. Model results from a range of synthetic stratigraphic configurations show that a twofold increase in the thickness of glaciolacustrine clays produces a tenfold increase in predicted landslide volume, consistent with topographic observations. Knowledge of subsurface stratigraphy may therefore help toward quantitative assessment of deep-seated landslide potential in sedimentary landforms.
Traditional deltaic landscape evolution models often simplify sea level rise (SLR) as a smooth curve to represent the impact of long-term increase in sea level. However, annual variability in SLR can significantly impact these dynamic ecosystems. This study employs a computationally efficient biophysical model to investigate the effects of varying SLR scenarios on Barataria Basin, a complex estuarine environment. Simulations were conducted using three smoothed SLR estimates (2m, 1.1m, and 0.6m by 2100) and four additional scenarios incorporating annual variability around each of the smoothed curves. Our findings suggest that Barataria Basin could experience substantial land loss, potentially ranging from 50,000 to 70,000 acres between 2030 and 2070 due to annual variability in SLR. These results underscore the critical need to consider SLR variability in coastal planning and management strategies to mitigate the impacts of rising sea levels. +
Traditional land management practices in Hawai‘i emphasize interconnected relationality and reciprocity between environmental systems, epitomized through the ‘mauka to makai’ or ‘mountain to sea’ philosophy. This management approach is, at times, at odds with typical modeling techniques which might consider individual rather than holistic ecosystem impacts. Here, we present a modeling framework developed in partnership with local stewardship organizations to support culturally relevant ‘mauka to makai’ land management and adaptation in evolving natural-human systems.
This work is centered on the ahupua‘a (traditional land division) of Puʻuwaʻawaʻa, Hawai‘i, which includes the Puʻuwaʻawaʻa forest reserve, Kīholo Bay state park reserve, and the adjoining ocean. This ahupua‘a supports a rich cultural legacy (traditional fishponds), vulnerable ecosystems (e.g., dryland forests, groundwater-fed anchialine pools, a fringing coral reef), and active recreation and cultural practice opportunities. Quantifying changing environmental conditions across the diverse systems in the Puʻuwaʻawaʻa ahupua‘a requires the integration of models capable of representing atmospheric, oceanic, coastal, and terrestrial systems over management timescales. Here, we develop a modeling framework that couples a stochastic climate emulator (Anderson et al., 2019) that probabilistically simulates future hydro-met-ocean conditions (e.g., waves, water levels, precipitation, temperature) with a nearshore wave propagation model, and a groundwater model. The outputs of the model framework are used in simple hazard exposure analyses to track critical hazard proxies (e.g., number of times fish pond infrastructure is overtopped, number of consecutive storm events, etc.) under a range of climate change scenarios and support decision-making with the entire ahupua‘a system in mind.
The modeling framework and results analysis enhance our understanding of future interconnected ecosystem impacts. They were co-developed with local stewardship organizations through an iterative process (including bi-weekly meetings, in-person gatherings, site visits, etc.) to ensure modeling decisions and hazard proxies directly address community concerns.
* Anderson, D., Rueda, A., Cagigal, L., Antolinez, J. A. A., Mendez, F. J., & Ruggiero, P. (2019). Time-Varying Emulator for Short and Long-Term Analysis of Coastal Flood Hazard Potential. Journal of Geophysical Research: Oceans, 124(12), 9209–9234. https://doi.org/10.1029/2019JC015312.
Traditional sediment transport models are constrained by their spatial and/or temporal resolution, requiring high computational power and consuming extensive processing time. This limitation poses a substantial challenge in modeling two-dimensional landscape evolution over long-term scales and in predicting the impacts of system dynamics, such as climate change. To address this shortcoming, we introduce a novel Deep Learning (DL) framework that integrates Convolutional Neural Networks (CNN) with Long Short-Term Memory (LSTM) to capture the spatiotemporal morphodynamics of rivers in response to flooding events. This framework combines three models to simulate essential hydrodynamic and morphodynamic features: water depth, flow velocity, and bed change, which operate in a continuous loop to ensure dynamic updates to bed topography. The proposed framework was trained and evaluated using a dataset generated by HEC-RAS, a physics-based model, for a 22 km segment of Ninnescah River in Kansas. The hydrodynamic results demonstrate proficiency in capturing the flood dynamics, where the average Root Mean Square Error (RMSE) across the testing dataset is 0.19 m and 0.04 m/s for water depth and flow velocity, respectively. These hydrodynamic features are essential for the bed-change model, which exhibits high accuracy with a normalized RMSE and an R2 of 27% and 0.93, respectively, at the end of the testing dataset. Furthermore, the trained framework can generate predictions 4700 times faster than HEC-RAS. This work signifies a paradigm shift in the long-term simulation of river evolution and sets the stage for exploring new frontiers in fluvial morphodynamic modeling. +
Transport-limited gravel-bed rivers are ubiquitous across Earth's upland environments. Sediment transport processes, while notoriously difficult, are better-understood than bedrock erosion, meaning that solutions to transport-limited river long profiles can help us gain a physics-based toehold into landscape evolution. Here we demonstrate how the coupling of equations for gravel transport, channel morphodynamics, and simple flow hydraulics that produce steady-state river profiles and show how they respond to changes in climate and tectonics. This coupled set of equations is analytically solvable for special cases, and we have also developed efficient semi-implicit numerical solutions that can solve millions of years of landscape evolution in seconds. Gravel-bed rivers become steeper as the sediment-to-water supply ratio increases, and become less concave as uplift rates (relative to input sediment supply and valley dimensions) increase. These distinctive responses allow us to use transport-limited gravel-bed rivers as recorders of climatic versus tectonic influence on river systems. +
Tropical Cyclones (TCs) are an extreme meteorological event that occurs in many locations globally. These events cause high levels of precipitation and flooding as it makes landfall. Extreme flooding events are known to cause increased suspended sediment flux and discharge in nearby rivers, thus altering sediment dynamics in short time periods. This relationship between TC related precipitation and sediment dynamics of impacted rivers has yet to be studied at a continental scale.
This project will be using the WBMsed model to simulate sediment and streamflow dynamics at a continental scale to find the total influence of TC influenced precipitation from 1990-2019. The products will be analyzed in ArcGIS to find spatial and temporal trends and hotspots of influence. The following research questions will also be answered: (1) What is the influence of TC related precipitation on sediment and streamflow dynamics? (2) How does modern anthropogenic conditions affect sediment and streamflow dynamics influenced by TC precipitation? (3) What will be the future impacts of TC related precipitation on sediment and streamflow dynamics in the United States? +
