Engineering Papers⌕ Search

Engineering topics

Bennett, Andrew

Publications and source records attributed to Bennett, Andrew.

Simulation-based inference for parameter estimation of complex watershed simulators

High-resolution, spatially distributed process-based (PB) simulators are widely employed in the study of complex catchment processes and their responses to a changing climate. However, calibrating these PB simulators using observed data remains a significant challenge due to several persistent issues, including the following: (1) intractability stemming from the computational demands and complex responses of simulators, which renders infeasible calculation of the conditional probability of parameters and data, and (2) uncertainty stemming from the choice of simplified representations of complex natural hydrologic processes. Here, we demonstrate how simulation-based inference (SBI) can help address both of these challenges with respect to parameter estimation. SBI uses a learned mapping between the parameter space and observed data to estimate parameters for the generation of calibrated simulations. To demonstrate the potential of SBI in hydrologic modeling, we conduct a set of synthetic experiments to infer two common physical parameters – Manning's coefficient and hydraulic conductivity – using a representation of a snowmelt-dominated catchment in Colorado, USA. We introduce novel deep-learning (DL) components to the SBI approach, including an “emulator” as a surrogate for the PB simulator to rapidly explore parameter responses. We also employ a density-based neural network to represent the joint probability of parameters and data without strong assumptions about its functional form. While addressing intractability, we also show that, if the simulator does not represent the system under study well enough, SBI can yield unreliable parameter estimates. Approaches to adopting the SBI framework for cases in which multiple simulator(s) may be adequate are introduced using a performance-weighting approach. The synthetic experiments presented here test the performance of SBI, using the relationship between the surrogate and PB simulators as a proxy for the real case.

54 ENVIRONMENTAL SCIENCES↗

Spatio–Temporal Machine Learning for Regional to Continental Scale Terrestrial Hydrology

Integrated hydrologic models can simulate coupled surface and subsurface processes but are computationally expensive to run at high resolutions over large domains. Here we develop a novel deep learning model to emulate subsurface flows simulated by the integrated ParFlow–CLM model across the contiguous US. We compare convolutional neural networks like ResNet and UNet run autoregressively against our novel architecture called the Forced SpatioTemporal RNN (FSTR). The FSTR model incorporates separate encoding of initial conditions, static parameters, and meteorological forcings, which are fused in a recurrent loop to produce spatiotemporal predictions of groundwater. We evaluate the model architectures on their ability to reproduce 4D pressure heads, water table depths, and surface soil moisture over the contiguous US at 1 km resolution and daily time steps over the course of a full water year. The FSTR model shows superior performance to the baseline models, producing stable simulations that capture both seasonal and event–scale dynamics across a wide array of hydroclimatic regimes. The emulators provide over 1,000× speedup compared to the original physical model, which will enable new capabilities like uncertainty quantification and data assimilation for integrated hydrologic modeling that were not previously possible. Our results demonstrate the promise of using specialized deep learning architectures like FSTR for emulating complex process–based models without sacrificing fidelity.

54 ENVIRONMENTAL SCIENCES↗

A science paradigm shift is needed for Earth and Environmental Systems Sciences (EESS) to integrate Knowledge-Guided Artificial Intelligence (KGAI) and lead new EESS-KGAI theories

The focal area of this white paper is learning from complex data through the use of AI techniques and AI-derived model components. Specifically, we advocate for research programs to develop knowledge-guided AI (KGAI) in the Earth and Environmental Systems sciences (EESS) as a basic research paradigm that is separate from (but supports) any specific Earth system model, modeling components, and modeling workflows, and even separates from specific hypothesis-driven questions about individual Earth system processes.

54 ENVIRONMENTAL SCIENCES↗

Mouthpart conduit sizes of fluid-feeding insects determine the ability to feed from pores

Fluid-feeding insects, such as butterflies, moths, and flies (20% of all animal species), are faced with the common selection pressure of having to remove and feed on trace amounts of fluids from porous surfaces. Insects able to acquire fluids that are confined to pores during drought conditions would have an adaptive advantage and increased fitness over other individuals. Here we performed feeding trials using solutions with magnetic nanoparticles to show that butterflies and flies have mouthparts adapted to pull liquids from porous surfaces using capillary action as the governing principle. In addition, the ability to feed on the liquids collected from pores depends on a relationship between the diameter of the mouthpart conduits and substrate pore size diameter; insects with mouthpart conduit diameters larger than the pores cannot successfully feed, thus there is a limiting substrate pore size from which each species can acquire liquids for fluid uptake. In conclusion, given that natural selection independently favored mouthpart architectures that support these methods of fluid uptake (Diptera and Lepidoptera share a common ancestor 280 mya that had chewing mouthparts), we suggest that the convergence of this mechanism advocates this as an optimal strategy for pulling trace amounts of fluids from porous surfaces.

59 BASIC BIOLOGICAL SCIENCES↗

The Ingestion of Fluorescent, Magnetic Nanoparticles for Determining Fluid-uptake Abilities in Insects

Here, fluid-feeding insects ingest a variety of liquids, which are present in the environment as pools, films, or confined to small pores. Studies of liquid acquisition require assessing mouthpart structure and function relationships; however, fluid uptake mechanisms are historically inferred from observations of structural architecture, sometimes unaccompanied with experimental evidence. Here, we report a novel method for assessing fluid-uptake abilities with butterflies (Lepidoptera) and flies (Diptera) using small amounts of liquids. Insects are fed with a 20% sucrose solution mixed with fluorescent, magnetic nanoparticles from filter papers of specific pore sizes. The crop (internal structure used for storing fluids) is removed from the insect and placed on a confocal microscope. A magnet is waved by the crop to determine the presence of nanoparticles, which indicate if the insects are able to ingest fluids. This methodology is used to reveal a widespread feeding mechanism (capillary action and liquid bridge formation) that is potentially shared among Lepidoptera and Diptera when feeding from porous surfaces. In addition, this method can be used for studies of feeding mechanisms among a variety of fluid-feeding insects, including those important in disease transmission and biomimetics, and potentially other studies that involve nano- or micro-sized conduits where liquid transport requires verification.

36 MATERIALS SCIENCE↗

An Update on Improvements to NiCE Support for PROTEUS

The Department of Energy Office of Nuclear Energy's Nuclear Energy Advanced Modeling and Simulation (NEAMS) program has supported the development of the NEAMS Integrated Computational Environment (NiCE), a modeling and simulation workflow environment that provides services and plugins to facilitate tasks such as code execution, model input construction, visualization, and data analysis. This report details the development of workflows for the reactor core neutronics application, PROTEUS. This advanced neutronics application (primarily developed at Argonne National Laboratory) aims to improve nuclear reactor design and analysis by providing an extensible and massively parallel, finite-element solver for current and advanced reactor fuel neutronics modeling. The integration of PROTEUS-specific tools into NiCE is intended to make the advanced capabilities that PROTEUS provides more accessible to the nuclear energy research and development community. This report will detail the work done to improve existing PROTEUS workflow support in NiCE. We will demonstrate and discuss these improvements, including the development of flexible IO services, an improved interface for input generation, and the addition of advanced Fortran development tools natively in the platform.

22 GENERAL STUDIES OF NUCLEAR REACTORS↗

NASA Inverse Methods/Data Assimilation

An overview of NASA's Third International Summer School on Inverse Methods and Data Assimilation which was conducted at Oregon State University from July 22 to August 2, 2002, is presented. Items listed include: a roster of attendees, a description of course content and talks given.

Bennett, Andrew↗