Property:CSDMS meeting abstract
From CSDMS
This is a property of type Text.
2
Simulation models are explicit descriptions of the components and interactions of a system, made dynamic in software. In Coupled Human-Earth Systems Science, we most often employ simulation to conduct controlled experiments in which key socio-ecological parameters are varied, and changes to system-level dynamics are observed over time. An interesting emergent property of these kinds of experiments is that they produce a range of possible outcomes for any set of initial conditions. Thus, rather than use simulations to explain particular case studies from the past, they are better suited to examine the dynamics of ancient systems in a more general way. Model parameters need to be determined and model output needs to be validated, however. So, our simulations *do* need to be connected to empirical data; a useful model must be capable of producing the same *kinds* of patterns observed in the archaeological record (but not *only* these patterns). It is often difficult, however, to connect model output to real data. In this presentation I draw upon research and modeling techniques being developed by the Mediterranean Landscape Dynamics Project to explore ways of connecting the output of simulation models to the kinds of proxy records that we typically use to learn about the past, such as the stratigraphic record, human artifact densities, and phytolith and charcoal accumulation. +
Slip at the ice-bed interface (basal motion) dominates the flow of many glaciers, and it is uncertain whether this velocity component will increase or slow in a warmer world. Past results from an idealized flowline glacier model show that declining basal motion induces a two-phase response that initially accelerates glacier retreat in response to climate warming on a multidecadal timescale but lessens centennial-scale retreat and mass loss. In the present work, we utilize existing field-collected and remotely-sensed constraints on ice thickness, ice surface velocity, and the change in each of these terms to constrain the current rate of basal motion and its change over the past ~40 years. We focus our analysis on glaciers with well-constrained ice thickness, mass balance, and velocity records. Utilizing these constraints, we employ a simple flowline model to estimate the contribution of varying basal motion to observed changes in surface velocity across the study glaciers. We then estimate these glaciers’ retreat and thinning responses to changing velocity and compare these with the magnitudes expected from atmospheric warming, constrained by published point measurements, mass balance models, and snowline observations. These results will constrain the extent to which evolving ice dynamics have amplified or mitigated the response of global glaciers to climate change over past decades. Further, this knowledge will provide insight into the potential importance of varying basal motion on projections of future glacier change, with implications for global sea level rise as well as local water resource and ecosystem management. +
Slow-moving arctic soils commonly organize into striking large-scale spatial patterns called solifluction terraces and lobes. Though these features impact hillslope stability, carbon storage and release, and landscape response to climate change, no mechanistic explanation exists for their formation. Everyday fluids—such as paint dripping down walls—produce markedly similar fingering patterns resulting from competition between viscous and cohesive forces. Here we use a scaling analysis to show that soil cohesion and hydrostatic effects can lead to similar large-scale patterns in arctic soils. A large new dataset of high-resolution solifluction lobe spacing and morphology across Norway supports theoretical predictions and indicates a newly observed climatic control on solifluction dynamics and patterns. Our findings provide a quantitative explanation of a common pattern on Earth and other planets, illuminating the importance of cohesive forces in landscape dynamics. These patterns operate at length and time scales previously unrecognized, with implications toward understanding fluid-solid dynamics in particulate systems with complex rheology. +
Soil creeps imperceptibly downhill, but also fails catastrophically to create landslides. Despite the importance of these processes as hazards and in sculpting landscapes, there is no agreed upon model that captures the full range of behavior. Here we examine the granular origins of hillslope soil transport by Discrete Element Method simulations, and re-analysis of measurements in natural landscapes. We find creep for slopes below a critical gradient, where average particle velocity (sediment flux) increases exponentially with friction coefficient (gradient). At critical there is a continuous transition to a dense-granular flow rheology. Slow earthflows and landslides thus exhibit glassy dynamics characteristic of a wide range of disordered materials; they are described by a two-phase flux equation that emerges from grain-scale friction alone. This glassy model reproduces topographic profiles of natural hillslopes, showing its promise for predicting hillslope evolution over geologic timescales. +
Soil moisture state has a critical role on subsurface-land surface-atmosphere energy and water balance. Yet, there is still no consensus on how to initialize atmospheric-hydrologic models to improve the representation of soil moisture content. Lack of accurate observational soil moisture data is the root of this issue. Although there has been progress in providing remotely sensed soil moisture data (e.g., Soil Moisture Active Passive (SMAP) data), their resolution is not adequate for high-resolution simulations. As an alternative approach, many atmospheric-hydrological simulations use various spin-up periods prior to the start of their analysis to perturb and improve the low-resolution soil moisture with precipitation. It has been shown that such method can improve soil moisture distribution in some studies in comparison to observational data. However, starting simulations from earlier times can cause divergence from accurate initial atmospheric conditions, which were obtained from observational data when simulation reaches the analysis period of interest. Therefore, there is a tradeoff between starting several days or hours before the analysis period in accurate representation of atmospheric data versus soil moisture input. In this study, we evaluated the sensitivity of a high-resolution (150-m) Weather Research and Forecasting (WRF) model to initialization starting point. We ran five nested domains with 12150-, 4050-, 1350-, 450-, and 150-m resolutions to downscale NCEP North American Regional Reanalysis (NARR) to our domain of interest encompassing Baltimore-Washington metropolitan area. The five domains were run in three scenarios starting 4, 7, and 14 days before the analysis period. Land surface temperature (LST) output was compared to LandSat data to investigate the impact of initialization starting point on model’s LST predictability. Results indicate that while the three scenarios underperformed in prediction of the urban heat island, there was no significant difference among the three scenarios. We determined that one of WRF’s thermal roughness parameterizations, which improves LST simulation over nonurban areas, caused significant errors in LST prediction over urban areas. Further simulations and analysis are underway to improve urban LST prediction. The three case scenarios will be compared against LandSat again when urban LST prediction is improved.
Son alluvial fan system, a megafan situated at the foothills of Vindhyans, is governed by the endogenic and exogenic process operating in the Ganga foreland basin. The megafan is interspersed with a number of structural features in the bedrock overlain by quaternary alluvial cover viz., Munger Saharsa Ridge Fault (MSRF), East Patna fault (EPF), and West Patna fault (WPF), some other reported tectonic features. A number of studies have attempted to decipher the recorded signatures of these underlying bedrock structural features and related tectonoclimatic activities in in the form of geomorphic anomalies and sedimentological evidences. In this study, χ‐transform index and χ‐anomalies, in combination with stream channel sinuosity, channel steepness index (ksn), channel concavity index (θ), geomorphology, and field evidences, have been used to examine if these structural features be highlighted on the low relief megafan surface with bedrock-alluvial mixed to thick alluvial cover (upto 1000 m thick). Drainage basin divide (in)stability measured through across divide χ‐anomaly map which proven to be an important tool for quantification of basin and channel network geometry behaviour, has been found to highlight the areas with active structural activities around the reported bedrock structures in the experimental study. Geomorphology and field evidences corroborate the findings of this study. +
Source-to-sink (S2S) studies seek to explicitly link the denudation of continents with the building of basin stratigraphy in an effort to infer tectonic and climatic drivers of surface change. Quantitative models for S2S systems must incorporate geomorphic processes at both source and sink, yet more effort has been devoted to developing landscape evolution models in source terranes than equivalent models for sedimentation in marine basins. In particular, most marine sedimentation models use local linear diffusion approximations for sediment transport, which have been shown to yield reasonable stratigraphy in shallow marine environments but struggle to reproduce diagnostic features of deep marine deposits. The lack of model predictive power in deep marine environments precludes the full closure of S2S sediment budgets.
We present a model for marine sedimentation with two simple modifications allowing non-local sediment transport: 1) a mechanism for sediment bypass on steep topographic slopes, and 2) a parameter allowing long-distance transport over vanishingly gentle slopes. We use Bayesian inference techniques to constrain four model parameters against the stratigraphy of the Orange Basin in southern Africa. We compare modeled against observed stratigraphy over 130 Ma of margin evolution. Our best-fit simulations capture the broad structure of the observed record, and imply non-negligible roles for both non-local model elements: sediment bypass at steep slopes and long-distance runout over gentle slopes. Residual misfit between our best-fit simulations and the stratigraphic data indicate that additional components of transport dynamics—likely hemipelagic sedimentation, grain size variations, or ocean bottom currents—might be required to achieve the longest transport distances observed in the sedimentary record. Results suggest that full closure of Earth’s sediment mass balance for S2S studies requires moving beyond local diffusion approximations, even at the longest timescales. Relatively simple modifications to modeled transport dynamics can lead to better agreement between modeled and observed stratigraphy, and may enable improved inference of landscape perturbations from the stratigraphic record.
Stream channels that cross strike-slip faults play an essential role in the long-term landscape response. So far, numerical models of strike-slip faults have simulated fluvial erosion assuming purely detachment-limited conditions. The detachment-limited theory assumes that the erosion is controlled by material that is detached from the channel bed and is always transported by the flow. As an alternative, erosion in channels can be represented by the transport-limited theory, which assumes that sediment is always available but may or not be transportable depending on the flow capacity. Extreme environments such as the Atacama Desert in Northern Chile, are evidence of strike-slip faulting with channels covered by alluvial deposits, suggesting that the landscape is best represented by a combination of detachment-limited and transport-limited conditions. Based on the most recent strike-slip fault model we incorporate and couple the effect of the SPACE (stream power with alluvium conservation and entrainment) 1.0 Landlab component in Python. The SPACE component can freely transition between detachment-limited and transport-limited conditions offering a closer representation of what is observed in the natural world. The results of coupling SPACE with strike-slip faults models are contrasted against the models that apply only detachment-limited conditions, to identify the action of a layer of sediment in landscape modification under variable strike-slip fault conditions. The concluding remarks of this work contribute to testing the accuracy of simplifying channel erosion processes to the commonly used stream power equation in strike-slip fault settings. +
Stream discharge is often used to drive sediment transport models across channel networks. Because sediment transport is nonlinear, discharge arising from precipitation resolved at 1-hr resolution may simulate bedload differently than discharge arising from daily total precipitation distributed evenly over 24-hrs. In this study, we quantify the bias introduced into a network-scale bedload transport model due to this simplification in forcing. Specifically, we examine the difference between bedload transport capacity driven by 1- vs 24-hr precipitation derived stream hydrographs at channel network locations varying from lowland pool-riffle channels to upland colluvial channels in a watershed where snow accumulation and melt can affect runoff processes. Bedload transport error is expressed as the ratio of cumulative transport capacity driven by 1-h to the 24-h hyetographs. We find that, depending on channel network location, cumulative error can range from 10-20% to more than two orders of magnitude. Surprisingly, variation in flow rates due to differences in hillslope and channel runoff do not seem to dictate the network locations where the largest errors in predicted bedload transport capacity occur. Rather, spatial variability of the magnitude of the bankfull-excess shear stress and changes in runoff due to snow accumulation and melt exert the greatest influence. As bankfull-excess shear stress decreases in the upstream direction, the largest bedload transport capacity errors occur in upland channels. These findings have implications for flood-hazard and aquatic habitat models that rely on modeled sediment transport driven by coarse-temporal-resolution climate data. +
Subglacial hydraulics significantly affects the ice dynamics in Greenland and Antarctic ice sheets, however, has been poorly understood due to the lack of data. Here we present an OpenFOAM-based one-dimensional subglacial model, conduitFoam, to study the hydraulics and ice dynamics of polar ice sheets. This model solves the coupled mass conservation equations for ice and water, the momentum and energy conservation equations for water, with a lake-conduit or moulin-conduit system as constraint boundaries. The model is validated using the theoretical solution applied in early melting stage and lake melting stage of the Greenland ice sheet and can be used to infer the subglacial conduit properties and the ice sheet dynamics in both seasonal and diurnal melting situations. +
Submarine slope failure is a ubiquitous process and dominant pathway for sediment and organic carbon flux from continental margins to the deep sea. Slope failure occurs over a wide range of temporal and spatial scales, from small (10e4-10e5 m3/event), sub-annual failures on heavily sedimented river deltas to margin-altering and tsunamigenic (10-100 km3/event) open slope failures occurring on glacial-interglacial timescales. Despite their importance to basic (closing the global source-to-sink sediment budget) and applied (submarine geohazards) research, submarine slope failure frequency and magnitude on most continental margins remains poorly constrained. This is primarily due to difficulty in 1) directly observing events, and 2) reconstructing age and size, particularly in the geologic record. The state of knowledge regarding submarine slope failure preconditioning and triggering factors is more qualitative than quantitative; a vague hierarchy of factor importance has been established in most settings but slope failures cannot yet be forecasted or hindcasted from a priori knowledge of these factors.<br><br>A new approach to address the knowledge gaps outlined above is using machine learning to quantitatively identify triggering and preconditioning factors that are most strongly correlated with submarine slope failure occurrence. This occurs in three general steps: 1) compile potential predictors of slope failure occurrence gridded and interpolated at desired resolution, 2) compile predictands (specific values that we wish to predict), and 3) recursively test predictor/predictand correlation with observed data until the strongest correlations are found. Potential predictors can be parsed into categories such as morphology (gradient, curvature, roughness), geology (clay fraction, grain size, sedimentation rate, fault proximity), and triggers (seismicity, significant wave height, river discharge). Predictands (i.e. training data) are various proxies for slope failure occurrence, including depth change between bathymetric surveys and sediment shear strength. The initial test sites are heavily sedimented, societally important river deltas, as they host both frequent slope failures and ample predictor/predictand measurements. Once predictors that strongly correlate with submarine slope failure occurrence are identified, this approach can be applied in more data-poor settings to further our current understanding of global submarine slope failure distribution, frequency, and magnitude.
Subsurface flow dynamics are largely controlled by pressure gradients generated by surface flow and differences in permeability. In most models, surface and subsurface flows are decoupled, with effects on one another only considered over relatively large time scales. However, at smaller time scales, these two flows interact and modify each other's structures and properties. In this study, we developed a fully-coupled free-surface/subsurface Large Eddy Simulation model to investigate the spatiotemporal variations in velocity and pressure, particularly near the bed surface. We validated our model by comparing it to experimental data from a laboratory simulation of open channel flow on a simulated salmon redd bed made of coarse granular sediment, using non-toxic index-matched fluid and stereo Particle Image Velocimetry (PIV). Our model accurately captured subsurface flow lines, velocity magnitude and direction, and superficial velocity profiles throughout the water column. With our validated model, we investigated the effects of subsurface hydraulic conductivity on the whole flow field. +
Surface processes are constantly reworking the landscape of our planet with perhaps the most diverse and beautiful patterns of sediment displacement known to humanity. Capturing this diversity is important for advancing our knowledge of systems, and for sustainable exploitation of natural resources by future generations. From a modeler's perspective, great diversity comes with great uncertainty. Although it is understandably very hard to quantify uncertainty about geological events that happened many years ago, we argue that modeling this uncertainty explicitly is crucial to improve our understanding of subsurface heterogeneity, as stratigraphy is direct function of surface processes. In this modeling work (and code), we aim to build realistic stratigraphic models that are constrained to local data (e.g. from wells, or geophysics) and that are, at the same time, subject to surface processes reflected in flume records. Experiments have improved tremendously in recent years, and the amount of data that they generate is posing new challenges to the surface processes community, who is asking more often the question "How do we make use of all this?" Traditional models based on differential equations and constitutive laws are not flexible enough to digest this information, nor were they created with this purpose. The community faces this limitation where the models cannot be conditioned on experiments, and even after exhaustive manual calibration of unobserved input parameters, these models often show poor predictive power. Our choice of inverse modeling and (geo)statistics (a.k.a. data science) was thus made knowing that these disciplines can provide the community with what we need: the ability to condition models of stratigraphy to measurements taken on a flume tank. +
Suspended sediment concentration, flux, and river discharge are essential indicators of river ecosystem health and reflect watershed-scale processes. Monitoring these variables is labor-intensive, leading to sparse and geographically biased observations and the development of models to fill in the observational gaps. These models generally use either climatological data or satellite images to estimate one of these variables. In this work, we present a novel deep learning model that can leverage multiple data sources with different temporal characteristics to produce continuous daily estimates of suspended sediment concentration (SSC), suspended sediment flux (SSF), and discharge. The model first encodes daily hydrological data from the ERA5-Land reanalysis using a Long Short-Term Memory network and water color data from Landsat satellites using a Multi-Layer Perceptron network, then merge these encoded data sources using a cross-attention decoder. We train and test the model on a large dataset of in- situ observations from 630 river sites over 43 years in the contiguous United States, covering a wide range of watersheds and conditions. We produce SSC, SSF, and discharge predictions with respective relative errors of 54%, 73%, and 28%, and relative bias of -15%, -19%, and -3%. We use our model to create a dataset of continuous daily SSC, SSF, and discharge for all large rivers in the contiguous United States. This new model architecture provides a valuable tool for monitoring river systems, addressing limitations of single-source models and offering a framework applicable to other Earth systems monitoring problems where integrating diverse data streams may be useful. +
Suspended sediment concentrations and fluxes on continental shelves impact light attenuation and primary productivity, as well as geomorphology and incorporation of particles into sea ice. As Arctic permafrost thaws, increasing riverine delivery and shoreline erosion, it is particularly important to understand how sediment sources influence suspended sediment concentrations and transport. To investigate these topics, this study analyzed results from a coupled hydrodynamic - sediment transport numerical model, namely the Regional Ocean Modeling System (ROMS) - Community Sediment Transport Modeling System (CSTMS). The model was implemented for the Alaskan Beaufort Sea Shelf for the 2020 open water (nearly ice-free) season, and accounted for processes such as riverine delivery, winds, larger-scale currents, and sediment erosion, transport and deposition. Building on previous work, we categorized riverine and seabed sediments into 26 distinct classes that allow us to distinguish among material originating from different rivers and sections of the seabed. Analysis focused on the spatial distribution of riverine and nearshore sediments over the course of an open water season, as well as the extent to which each sediment class contributed to high turbidity events. Preliminary results suggested that aggregated mud delivered by rivers during an open water season stayed within water depths of 0 – 10 m for at least a month and a half, while some unaggregated mud was transported to deeper regions during this time. Additionally, high turbidity events mainly occurred due to local resuspension, as opposed to riverine plumes, for both aggregated and unaggregated mud. Ongoing and future work include analysis of uncertainty due, for example, to sediment settling velocity and other properties. Overall, these findings suggest that high turbidity events are driven by sediment delivered to the continental shelf during previous open water seasons or winters, as opposed to new terrestrial/riverine inputs.
A
Sustainability of the anthroposphere is a result of a multitude of decisions made concerning social, economic and environmental questions. Decision makers who would like to ensure sustainable development as an emerging characteristic of humanity are challenged by the complexity of a planetary system re-engineered by an increasingly powerful global species. Examples of such problems are sustainable urban growth and the food-water-energy nexus. Tools to reliably assess the consequences of decisions from local to global level are not readily available.
In particular, current capabilities for assessing the various impacts of climate variability and change, as well as other changes are inadequate. The Group on Earth Observation (GEO) recognized this emergency and promoted several initiatives that can help address this shortcoming. One of them is the GEO Model Web initiative. The goal of this initiative is to develop a dynamic modelling consultative infrastructure of intercommunicating models and datasets to serve researchers, managers, policy makers and the general public. It focuses on enhancing interoperability of existing models and making them and their outputs more accessible. The development of the Model Web holds the promise of more decision support tools becoming available. These tools would allow decision makers to ask “What if” questions prior to the implementation of decisions and support adaptive management and responsive design. The Model Web will also benefit researchers by making it easier to run model experiments and model comparisons or ensembles, as well as help highlight areas needing further development. The Model Web would support a synchronization across different spatial and temporal scales and across the languages of different disciplines, thus making the System of System (SoS) more intelligent. The beauty of having a SoS like this is that it amplifies the signal. An immediate application is the emerging geodesign approach to the design of sustainable built environments.
The Model Web is developed in the framework of the Global Earth Observation System of Systems (GEOSS) implemented by GEO. The observing, modelling and other systems that contribute to GEOSS must be interoperable so that the data and information they generate can be used effectively. The Committee on Earth Observation Satellites (CEOS) is promoting interoperability through the Virtual Constellations concept, the Sensor Web approach, and by facilitating model interoperability and access via the Model Web concept.
The Model Web is a concept for a system of interoperable models and data capacities communicating primarily via web services. It would consist of an open-ended, distributed, multidisciplinary network of independent, interoperating models plus related datasets. Models and datasets would be maintained and operated and served by a dynamic network of participants. In keeping with the SoS approach, the Model Web initiative will explore the interoperability arrangements necessary to integrate multi-disciplinary environmental model resources. The approach of loosely coupled models that interact via web services, and are independently developed, managed, and operated has many advantages over tightly coupled, closed, integrated systems, which require strong central control, lack flexibility, and provide limited access to products.
Developing a long-term perspective, a logical next step would be the Internet of Models (IOM). Comparable to the already developing Internet of Things (IOT), which is predicted to connect by 2020 more than 50 billion “things” talking to each other without human interaction (or even knowledge), the IOM would have models talking to each other when needed without human interaction. If we compare the IOT to the nerve system of a human body, then the IOM would be the brain of the human being. Key to the development of IOT and IOM are standards that allow “things” and “models” to communicate when needed and to exchange information as needed (similar to the role of standards in the success of the WWW). Frameworks for model interactions are already developing (e.g. Object Modelling System, ModCom, the Invisible Modelling Environment, the Open Modelling Interface: OpenMI, the Spatial Modelling Environment: SME, Tarsier, Interactive Component Modelling System: ICMS, Earth System Modeling Framework: ESMF, SEAMLESS-IF , CSDMS, etc.), but they are not sufficient to achieve the Model Web (or the IOM). A major effort to develop the standards for the IOT is under way, and a similar effort to needed for the IOM standards. The combination of IOT and IOM would greatly enhance science capabilities, early warning, assessments of impacts, etc.
2
