All Faculty Deposits

Permanent URI for this collection

The "All Faculty Deposits" collection contains works deposited by faculty and appointed delegates from the Elements (EFARs) system. For help with Elements, see Frequently Asked Questions on the Provost's website. In general, items can only be deposited if the item is a scholarly article that is covered by Virginia Tech's open access policy, or the item is openly licensed or in the public domain, or the item is permitted to be posted online under the journal/publisher policy, or the depositor owns the copyright. See Right to Deposit on the VTechWorks Help page. If you have questions email us at vtechworks@vt.edu.

Browse

Recent Submissions

Now showing 1 - 20 of 5277
  • Towards Semantically-Rich Spatial Network Representation Learning via Automated Feature Topic Pairing
    Wang, Dongjie; Liu, Kunpeng; Mohaisen, David; Wang, Pengyang; Lu, Chang-Tien; Fu, Yanjie (Frontiers, 2021-10-20)
    Automated characterization of spatial data is a kind of critical geographical intelligence. As an emerging technique for characterization, spatial Representation Learning (SRL) uses deep neural networks (DNNs) to learn non-linear embedded features of spatial data for characterization. However, SRL extracts features by internal layers of DNNs, and thus suffers from lacking semantic labels. Texts of spatial entities, on the other hand, provide semantic understanding of latent feature labels, but is insensible to deep SRL models. How can we teach a SRL model to discover appropriate topic labels in texts and pair learned features with the labels? This paper formulates a new problem: feature-topic pairing, and proposes a novel Particle Swarm Optimization (PSO) based deep learning framework. Specifically, we formulate the feature-topic pairing problem into an automated alignment task between 1) a latent embedding feature space and 2) a textual semantic topic space. We decompose the alignment of the two spaces into: 1) point-wise alignment, denoting the correlation between a topic distribution and an embedding vector; 2) pair-wise alignment, denoting the consistency between a feature-feature similarity matrix and a topic-topic similarity matrix. We design a PSO based solver to simultaneously select an optimal set of topics and learn corresponding features based on the selected topics. We develop a closed loop algorithm to iterate between 1) minimizing losses of representation reconstruction and feature-topic alignment and 2) searching the best topics. Finally, we present extensive experiments to demonstrate the enhanced performance of our method.
  • Fast and adaptive dynamics-on-graphs to dynamics-of-graphs translation
    Zhang, Lei; Chen, Zhiqian; Lu, Chang-Tien; Zhao, Liang (Frontiers, 2023-11-17)
    Numerous networks in the real world change with time, producing dynamic graphs such as human mobility networks and brain networks. Typically, the “dynamics on graphs” (e.g., changing node attribute values) are visible, and they may be connected to and suggestive of the “dynamics of graphs” (e.g., evolution of the graph topology). Due to two fundamental obstacles, modeling and mapping between them have not been thoroughly explored: (1) the difficulty of developing a highly adaptable model without solid hypotheses and (2) the ineffectiveness and slowness of processing data with varying granularity. To solve these issues, we offer a novel scalable deep echo-state graph dynamics encoder for networks with significant temporal duration and dimensions. A novel neural architecture search (NAS) technique is then proposed and tailored for the deep echo-state encoder to ensure strong learnability. Extensive experiments on synthetic and actual application data illustrate the proposed method's exceptional effectiveness and efficiency.
  • Reservoir based spiking models for univariate Time Series Classification
    Gaurav, Ramashish; Stewart, Terrence C.; Yi, Yang (Frontiers, 2023-06-08)
    A variety of advanced machine learning and deep learning algorithms achieve state-of-the-art performance on various temporal processing tasks. However, these methods are heavily energy inefficient—they run mainly on the power hungry CPUs and GPUs. Computing with Spiking Networks, on the other hand, has shown to be energy efficient on specialized neuromorphic hardware, e.g., Loihi, TrueNorth, SpiNNaker, etc. In this work, we present two architectures of spiking models, inspired from the theory of Reservoir Computing and Legendre Memory Units, for the Time Series Classification (TSC) task. Our first spiking architecture is closer to the general Reservoir Computing architecture and we successfully deploy it on Loihi; the second spiking architecture differs from the first by the inclusion of non-linearity in the readout layer. Our second model (trained with Surrogate Gradient Descent method) shows that non-linear decoding of the linearly extracted temporal features through spiking neurons not only achieves promising results, but also offers low computation-overhead by significantly reducing the number of neurons compared to the popular LSM based models—more than 40x reduction with respect to the recent spiking model we compare with. We experiment on five TSC datasets and achieve new SoTA spiking results (—as much as 28.607% accuracy improvement on one of the datasets), thereby showing the potential of our models to address the TSC tasks in a green energy-efficient manner. In addition, we also do energy profiling and comparison on Loihi and CPU to support our claims.
  • Resilient s-ACD for Asynchronous Collaborative Solutions of Systems of Linear Equations
    Erlandson, Lucas; Atkins, Zachary; Fox, Alyson; Vogl, Christopher; Miedlar, Agnieszka; Ponce, Colin (IEEE, 2023-09-26)
    Solving systems of linear equations is a critical component of nearly all scientific computing methods. Traditional algorithms that rely on synchronization become prohibitively expensive in computing paradigms where communication is costly, such as heterogeneous hardware, edge computing, and unreliable environments. In this paper, we introduce an s-step Approximate Conjugate Directions (s-ACD) method and develop resiliency measures that can address a variety of different data error scenarios. This method leverages a Conjugate Gradient (CG) approach locally while using Conjugate Directions (CD) globally to achieve asynchronicity. We demonstrate with numerical experiments that s-ACD admits scaling with respect to the condition number that is comparable with CG on the tested 2D Poisson problem. Furthermore, through the addition of resiliency measures, our method is able to cope with data errors, allowing it to be used effectively in unreliable environments.
  • Human-AI Collaborative Innovation in Design
    Song, Binyang; Zhu, Qihao; Luo, Jianxi (2024)
    Human-AI collaboration (HAIC) is a promising strategy to transform engineering design and innovation, yet how to design artificial intelligence (AI) to boost HAIC remains unclear. Accordingly, this paper provides a new, unified, and comprehensive scheme for classifying AI roles. On this basis, we develop an AI design framework that outlines expected AI capabilities, interactive attributes, and trust enablers across various HAIC scenarios, offering guidance for integrating AI into human teams effectively. We also discuss current advancements, challenges, and prospects for future research.
  • Data-driven Car Drag Coefficient Prediction with Depth and Normal Renderings
    Song, Binyang; Yuan, Chenyang; Permenter, Frank; Arechiga, Nikos; Ahmed, Faez (American Society of Mechanical Engineers, 2024)
    Generative AI models have made significant progress in automating the creation of 3D shapes, which has the potential to transform car design. In engineering design and optimization, evaluating engineering metrics is crucial. To make generative models performance-aware and enable them to create high-performing designs, surrogate modeling of these metrics is necessary. However, the currently used representations of 3D shapes either require extensive computational resources to learn or suffer from significant information loss, which impairs their effectiveness in surrogate modeling. To address this issue, we propose a new 2D representation of 3D shapes. We develop a surrogate drag model based on this representation to verify its effectiveness in predicting 3D car drag. We construct a diverse dataset of 4,535 high-quality 3D car meshes labeled by drag coefficients computed from computational fluid dynamics simulations to train our model. Our experiments demonstrate that our model can accurately and efficiently evaluate drag coefficients with an R^2 value above 0.84 for various car categories. Our model is implemented using deep neural networks, making it compatible with recent AI image generation tools (such as Stable Diffusion) and a significant step towards the automatic generation of drag-optimized car designs. Moreover, we demonstrate a case study using the proposed surrogate model to guide a diffusion-based deep generative model for drag-optimized car body synthesis. We have made the dataset and code publicly available at https://decode.mit.edu/projects/dragprediction.
  • Generative Design for Manufacturing: Integrating Generation with Optimization Using a Guided Voxel Diffusion Model
    Song, Binyang; Chilukuri, Premith Kumar; Kang, Sungku; Jin, Ran (2024)
    In digital manufacturing, converting advanced designs into quality products is hampered by manufacturers' limited design knowledge, restricting the adoption and enhancement of innovative solutions. This paper addresses this challenge through a novel generative denoising diffusion model (DDM) trained on historical 3D design data, enabling the creation of voxel-based designs that meet manufacturing standards. By integrating a surrogate model for evaluating the manufacturability of generated designs, the proposed DDM is able to optimize manufacturability during the generative process. This paper takes a leap forward from the predominant 2D focus of existing generative models towards 3D generative design, which not only broadens manufacturers' design capabilities but also accelerates the development of practical and optimized products. We demonstrate the efficacy of this approach via a case study on Microbial Fuel Cell (MFC) anode design, illustrating how this method can significantly enhance manufacturing workflows and outcomes. Our research offers a path for manufacturers to deepen their design expertise and foster innovation in digital manufacturing.
  • Drag-guided Diffusion Models for Vehicle Image Generation
    Arechiga, Nikos; Permenter, Frank; Song, Binyang; Yuan, Chenyang (2023-12-15)
    Denoising diffusion models trained at web-scale have revolutionized image generation. The application of these tools to engineering design holds promising potential but is currently limited by their inability to understand and adhere to concrete engineering constraints. In this paper, we take a step toward the goal of incorporating quantitative constraints into diffusion models by proposing physics-based guidance, which enables the optimization of a performance metric (as predicted by a surrogate model) during the generation process. As a proof-of-concept, we add drag guidance to Stable Diffusion, which allows this tool to generate images of novel vehicles while simultaneously minimizing their predicted drag coefficients.
  • Nonparametric Bayesian Functional Clustering with Applications to Racial Disparities in Breast Cancer
    Gao, Wenyu; Kim, Inyoung; Nam, W.; Ren, X.; Zhou, W.; Agah, M. (WILEY, 2024-02)
    As we have easier access to massive data sets, functional analyses have gained more interest. However, such data sets often contain large heterogeneities, noises, and dimensionalities. When generalizing the analyses from vectors to functions, classical methods might not work directly. This paper considers noisy information reduction in functional analyses from two perspectives: functional clustering to group similar observations and thus reduce the sample size and functional variable selection to reduce the dimensionality. The complicated data structures and relations can be easily modeled by a Bayesian hierarchical model due to its flexibility. Hence, this paper proposes a nonparametric Bayesian functional clustering and peak point selection method via weighted Dirichlet process mixture (WDPM) modeling that automatically clusters and provides accurate estimations, together with conditional Laplace prior, which is a conjugate variable selection prior. The proposed method is named WDPM-VS for short, and is able to simultaneously perform the following tasks: (1) Automatic cluster without specifying the number of clusters or cluster centers beforehand; (2) Cluster for heterogeneously behaved functions; (3) Select vibrational peak points; and (4) Reduce noisy information from the two perspectives: sample size and dimensionality. The method will greatly outperform its comparison methods in root mean squared errors. Based on this proposed method, we are able to identify biological factors that can explain the breast cancer racial disparities.
  • Detection of passageways in natural foliage using biomimetic sonar
    Wang, Ruihao; Liu, Yimeng; Müller, Rolf (IOP Publishing, 2022-08-10)
    The ability of certain bat species to navigate in dense vegetation based on trains of short biosonar echoes could provide for an alternative parsimonious approach to obtaining the sensory information that is needed to achieve autonomy in complex natural environments. Although bat biosonar has much lower data rates and spatial (angular) resolution than commonly used human-made sensing systems such as LiDAR or stereo cameras, bat species that live in dense habitats have the ability to reliably detect narrow passageways in foliage. To study the sensory information that the animals may have available to accomplish this, we have used a biomimetic sonar system that was combined with a camera to record echoes and synchronized images from 10 different field sites that featured narrow passageways in foliage. The synchronized camera and sonar data allowed us to create a large data set (130 000 samples) of labeled echoes using a teacher-student approach that used class labels derived from the images to provide training data for echo-based classifiers. The performance achieved in detecting passageways based on the field data closely matched previous results obtained for gaps in an artificial foliage setup in the laboratory. With a deep feature extraction neural network (VGG16) a foliage-versus-passageway classification accuracy of 96.64% was obtained. A transparent artificial intelligence approach (class-activation mapping) indicated that the classifier network relied heavily on the initial rising flank of the echoes. This finding could be exploited with a neuromorphic echo representation that consisted of times where the echo envelope crossed a certain amplitude threshold in a given frequency channel. Whereas a single amplitude threshold was sufficient for this in the previous laboratory study, multiple thresholds were needed to achieve an accuracy of 92.23%. These findings indicate that despite many sources of variability that shape clutter echoes from natural environments, these signals contain sufficient sensory information to enable the detection of passageways in foliage.
  • Une mauvaise publicité pour l’hydroélectricité québécoise
    Calder, Ryan S. D. (La Presse, 2019-10-11)
  • 2023 Fall edition of the Clover News
    Morgan, Erin (2023-09-25)
    The fall edition of the Northampton 4-H newsletter
  • High-Resolution Finite Fault Slip Inversion of the 2019 Ridgecrest Earthquake Using 3D Finite Element Modeling
    Barba-Sevilla, Magali; Glasscoe, Margaret T.; Parker, Jay; Lyzenga, Gregory A.; Willis, Michael J.; Tiampo, Kristy F. (AMER GEOPHYSICAL UNION, 2022-09)
    The 2019 Ridgecrest earthquake sequence manifested as one of the most complex fault surface ruptures observed in California in modern times. The M6.4 foreshock and M7.1 mainshock occurred on an intricate network of orthogonal and sub-parallel faults resulting in observable surface displacement and surface rupture captured by geodetic data. Here we present the application of a high-resolution 3D finite element model (FEM) approach to invert for the detailed fault slip of the entire sequence using complex rheology and fused coseismic Global Navigation Satellite System (GNSS) data with Sentinel-1 differential interferometric synthetic aperture radar and pixel offset data. The heterogeneous FEM and the fused geodetic data set of pixel offsets, interferograms, and GNSS data results in our optimal inversion solution. This preferred solution is a complex, high-resolution non-planar slip model of both the M6.4 and M7.1 events that features three main regions of large slip (6.9+ m), with depths ranging from 2 to 10 km. The regions of slip are bounded by the mainshock hypocenter and the mainshock aftershocks and appear to be related to spatially varying rheological properties. We successfully reproduce a localized region of observed subsidence in the northern portion of the primary fault through the inclusion of a curved fault strand with a significant dip-slip component. The curved fault strand is the site of our maximum slip of 7.4 m at a depth of 4.2 km. The results demonstrate a robust fit from a more complete, detailed model for the entire seismogenic zone with reasonable computational cost, providing new insights into the governing rheologic and structural processes.
  • Characterization of large tsunamigenic landslides and their effects using digital surface models: A case study from Taan Fiord, Alaska
    Corsa, Brianna D.; Jacquemart, Mylene; Willis, Michael J.; Tiampo, Kristy F. (ELSEVIER SCIENCE INC, 2022-03-01)
    On 17 October 2015, a large landslide entered the marine waters of Taan Fiord, Alaska, and generated a displacement wave with a 193 m runup. The wave scoured the surrounding hillslopes of soil and vegetation and deposited significant volumes of material into the fjord, onto hillslopes on the opposite side of the fjord, and on top of Tyndall Glacier. For this study, we generated six, 2-m posting Digital Surface Models (DSMs) using DigitalGlobe/Maxar satellite imagery acquired near-annually between 2012 and 2019, and the Surface Extraction with TIN-based Search-space Minimization (SETSM) high-performance computing algorithm. We aligned the DSMs to exposed bedrock in the 01 March 2014 DSM acquisition, and then used them to characterize topographic and volumetric changes from before and after the 2015 Taan Fiord rock avalanche. We find that the landslide mobilized roughly 77. 0 ± 0.9 Mm3 of material, of which approximately 56.3 Mm3 were deposited in the fjord waters. Furthermore, we quantified an additional 27.2 ± 3.8 Mm3 of material scoured from fjord-adjacent hillslopes and deposited in the fjord waters, providing new constraints on the subaqueous deposition. This is the first time that DSMs have been used to estimate the volume of scour caused by a tsunami and the subsequent changes in extent and volume with time. Our results for the landslide and runout are consistent with field measurements published previously. This study offers improved estimates of both subaerial and subaqueous deposition for the 2015 Taan Fiord landslide and describes additional regional environmental conditions. We identify precursory motion prior to the 2015 landslide, characterize several smaller-scale landslides in the larger Taan Fiord region, delineate terminus positions and associated ice dynamics of the Tyndall Glacier, and detail seasonal changes in vegetation growth and snow melt/accumulation. This work provides important new insights into the geomorphic features and dynamics of this landslide and subsequent tsunami. The interdisciplinary applications associated with DSMs and the accuracy of the measurements presented here demonstrate that these methods are an effective tool to improve our understanding of the pre- and post-landslide processes, for monitoring areas at risk for landslides and other natural hazards, and for rapid response to catastrophic events.
  • Atypical landslide induces speedup, advance, and long-term slowdown of a tidewater glacier
    de Vries, Maximillian Van Wyk; Wickert, Andrew D.; MacGregor, Kelly R.; Rada, Camilo; Willis, Michael J. (GEOLOGICAL SOC AMER, INC, 2022-07-01)
    Atmospheric and oceanic warming over the past century have driven rapid glacier thinning and retreat, destabilizing hillslopes and increasing the frequency of landslides. The impact of these landslides on glacier dynamics and resultant secondary landslide hazards are not fully understood. We investigated how a 262 ± 77 × 106 m3 landslide affected the flow of Amalia Glacier, Chilean Patagonia. Despite being one of the largest recorded landslides in a glaciated region, it emplaced little debris onto the glacier surface. Instead, it left a series of landslideperpendicular ridges, landslide-parallel fractures, and an apron of ice debris—with blocks as much as 25 m across. Our observations suggest that a deep-seated failure of the mountainside impacted the glacier flank, propagating brittle deformation through the ice and emplacing the bulk of the rock mass below the glacier. The landslide triggered a brief downglacier acceleration of Amalia Glacier followed by a slowdown of as much as 60% of the pre-landslide speed and increased suspended-sediment concentrations in the fjord. These results highlight that landslides may induce widespread and long-lasting disruptions to glacier dynamics.
  • Transient ice loss in the Patagonia Icefields during the 2015-2016 El Nino event
    Gomez, Demian D.; Bevis, Michael G.; Smalley, Robert; Durand, Michael; Willis, Michael J.; Caccamise, Dana J.; Kendrick, Eric; Skvarca, Pedro; Sobrero, Franco S.; Parra, Hector; Casassa, Gino (NATURE PORTFOLIO, 2022-06-10)
    The Patagonia Icefields (PIF) are the largest non-polar ice mass in the southern hemisphere. The icefields cover an area of approximately 16,500 km2 and are divided into the northern and southern icefields, which are ~ 4000 km2 and ~ 12,500 km2, respectively. While both icefields have been losing mass rapidly, their responsiveness to various climate drivers, such as the El Niño-Southern Oscillation, is not well understood. Using the elastic response of the earth to loading changes and continuous GPS data we separated and estimated ice mass changes observed during the strong El Niño that started in 2015 from the complex hydrological interactions occurring around the PIF. During this single event, our mass balance estimates show that the northern icefield lost ~ 28 Gt of mass while the southern icefield lost ~ 12 Gt. This is the largest ice loss event in the PIF observed to date using geodetic data.
  • Accuracy, Efficiency, and Transferability of a Deep Learning Model for Mapping Retrogressive Thaw Slumps across the Canadian Arctic
    Huang, Lingcao; Lantz, Trevor C.; Fraser, Robert H.; Tiampo, Kristy F.; Willis, Michael J.; Schaefer, Kevin (MDPI, 2022-06)
    Deep learning has been used for mapping retrogressive thaw slumps and other periglacial landforms but its application is still limited to local study areas. To understand the accuracy, efficiency, and transferability of a deep learning model (i.e., DeepLabv3+) when applied to large areas or multiple regions, we conducted several experiments using training data from three different regions across the Canadian Arctic. To overcome the main challenge of transferability, we used a generative adversarial network (GAN) called CycleGAN to produce new training data in an attempt to improve transferability. The results show that (1) data augmentation can improve the accuracy of the deep learning model but does not guarantee transferability, (2) it is necessary to choose a good combination of hyper-parameters (e.g., backbones and learning rate) to achieve an optimal trade-off between accuracy and efficiency, and (3) a GAN can significantly improve the transferability if the variation between source and target is dominated by color or general texture. Our results suggest that future mapping of retrogressive thaw slumps should prioritize the collection of training data from regions where a GAN cannot improve the transferability.
  • A Machine Learning Approach to Flood Depth and Extent Detection Using Sentinel 1A/B Synthetic Aperture Radar
    Tiampo, K.; Woods, C.; Huang, L.; Sharma, P.; Chen, Z.; Kar, B.; Bausch, D.; Simmons, C.; Estrada, R.; Willis, Michael J.; Glasscoe, M. (IEEE, 2021-01-01)
    The rising number of flooding events combined with increased urbanization is contributing to significant economic losses due to damages to structures and infrastructures. Here we present a method for producing all weather maps of flood inundation using a combination of synthetic aperture radar (SAR) remote sensing data and machine learning methods that can be used to provide information on the evolution of flood hazards to DisasterAware©, a global alerting system, that is used to disseminate flood risk information to stakeholders across the globe. While these efforts are still in development, a case study is presented for the major flood event associated with Hurricane Harvey and associated floods that impacted Houston, TX in August of 2017.
  • 3D models of the leader valley using satellite & UAV imagery following the 2016 Kaikoura earthquake
    Zekkos, D.; Clark, M.; Willis, M.; Athanasopoulos-Zekkos, A.; Manousakis, J.; Knoper, L.; Stahl, T.; Massey, C.; Archibald, G.; Greenwood, W.; Medwedeff, W. (2018-01-01)
    The ability to quickly, efficiently and reliably characterize changes in the landscape following an earthquake has remained a challenge for the earthquake engineering profession. The 2016 Mw7.8 Kaikoura earthquake provided a unique opportunity to document changes in topography following an earthquake on a regional scale using satellite derived high-resolution digital models. Along-track stereo satellite imagery had been collected for the pre-event topography. Satellites were tasked and collected stereo-mode post-event imagery. Both sets of images were used to create digital surface models (DSMs) of the affected area before and after the event. The procedure followed and indicative results for the Leader valley are presented with emphasis on the challenges associated with the implementation of the technique for the first time in this environment. The valley is of interest because of the variety of features it includes, i.e., the large Leader landslide, smaller landslides, stable sloping and flat ground as well as fault rupture lineaments. The open-source SETSM software is used to provide multiple DSMs. Our workflow is described and results are compared against the DSM created using Structure-from-Motion with imagery collected by Unmanned Aerial Vehicles (UAV) and aerial LIDAR. Overall, the sub-meter agreement between the DSM created using satellites and the DSM created using UAV and LIDAR datasets demonstrates viability for use in seismic studies, but features smaller than about 0.5 m are more difficult to discern.
  • Understanding of Contemporary Regional Sea-Level Change and the Implications for the Future
    Hamlington, Benjamin D.; Gardner, Alex S.; Ivins, Erik; Lenaerts, Jan T. M.; Reager, J. T.; Trossman, David S.; Zaron, Edward D.; Adhikari, Surendra; Arendt, Anthony; Aschwanden, Andy; Beckley, Brian D.; Bekaert, David PS S.; Blewitt, Geoffrey; Caron, Lambert; Chambers, Don P.; Chandanpurkar, Hrishikesh A.; Christianson, Knut; Csatho, Beata; Cullather, Richard; DeConto, Robert M.; Fasullo, John T.; Frederikse, Thomas; Freymueller, Jeffrey T.; Gilford, Daniel M.; Girotto, Manuela; Hammond, William C.; Hock, Regine; Holschuh, Nicholas; Kopp, Robert E.; Landerer, Felix; Larour, Eric; Menemenlis, Dimitris; Merrifield, Mark; Mitrovica, Jerry X.; Nerem, R. Steven; Nias, Isabel J.; Nieves, Veronica; Nowicki, Sophie; Pangaluru, Kishore; Piecuch, Christopher G.; Ray, Richard D.; Rounce, David R.; Schlegel, Nicole-Jeanne; Seroussi, Helene; Shirzaei, Manoochehr; Sweet, William; Velicogna, Isabella; Vinogradova, Nadya; Wahl, Thomas; Wiese, David N.; Willis, Michael J. (American Geophysical Union, 2020-07-20)
    Global sea level provides an important indicator of the state of the warming climate, but changes in regional sea level are most relevant for coastal communities around the world. With improvements to the sea-level observing system, the knowledge of regional sea-level change has advanced dramatically in recent years. Satellite measurements coupled with in situ observations have allowed for comprehensive study and improved understanding of the diverse set of drivers that lead to variations in sea level in space and time. Despite the advances, gaps in the understanding of contemporary sea-level change remain and inhibit the ability to predict how the relevant processes may lead to future change. These gaps arise in part due to the complexity of the linkages between the drivers of sea-level change. Here we review the individual processes which lead to sea-level change and then describe how they combine and vary regionally. The intent of the paper is to provide an overview of the current state of understanding of the processes that cause regional sea-level change and to identify and discuss limitations and uncertainty in our understanding of these processes. Areas where the lack of understanding or gaps in knowledge inhibit the ability to provide the needed information for comprehensive planning efforts are of particular focus. Finally, a goal of this paper is to highlight the role of the expanded sea-level observation network—particularly as related to satellite observations—in the improved scientific understanding of the contributors to regional sea-level change.