Coastal flooding poses increasing threats to communities worldwide, necessitating accurate and hyper-local inundation forecasting for effective emergency response. However, real-world deployment of forecasting systems is often constrained by sparse sensor networks, where only a limited subset of locations may have sensors due to budget constraints. To approach this challenge, we present Diff-Sparse, a masked conditional diffusion model designed for probabilistic coastal inundation forecasting from sparse sensor observations. Diff-Sparse primarily utilizes the inundation history of a location and its neighboring locations from a context time window as spatiotemporal context. The fundamental challenge of spatiotemporal prediction based on sparse observations in the context window is addressed by introducing a novel masking strategy during training. Digital elevation data and temporal co-variates are utilized as additional spatial and temporal contexts, respectively. A convolutional neural network and a conditional UNet architecture with cross-attention mechanism are employed to capture the spatiotemporal dynamics in the data. We trained and tested Diff-Sparse on coastal inundation data from the Eastern Shore of Virginia and systematically assessed the performance of Diff-Sparse across different sparsity levels (0%, 50%, 95% missing observations). Our experiment results show that Diff-Sparse achieves upto 62% improvement in terms of two forecasting performance metrics compared to existing methods, at 95% sparsity level. Moreover, our ablation studies reveal that digital elevation data becomes more useful at high sparsity levels compared to temporal co-variates.

Coastal flooding poses significant risks to communities, necessitating fast and accurate forecasting methods to mitigate potential damage. To approach this problem, we present Diff-Flood, a probabilistic spatiotemporal forecasting method designed based on denoising diffusion models. Diff-Flood predicts inundation level at a location by taking spatiotemporal context into account. It utilizes inundation levels at neighboring locations and digital elevation data as spatial context. Inundation history from a context time window, together with additional co-variates are used as temporal context. Convolutional neural networks and cross-attention mechanism are then employed to capture the spatiotemporal dynamics in the data. We trained and tested Diff-Flood on coastal inundation data from the Eastern Shore of Virginia, a region highly impacted by coastal flooding. Our results show that, Diff-Flood outperforms existing forecasting methods in terms of prediction performance (6% to 64% improvement in terms of two performance metrics) and scalability.

We study exascale parallel algorithms for the selection of intervention or monitoring strategies in massive realistic socio-technical networks through scalable Influence Maximization (InfMax) algorithms.We employ novel techniques to enable efficient scaling on up to 8k nodes of OLCF Frontier, with 65k AMD GPUs and 458k AMD CPU cores.Current state-of-the-art InfMax tools are limited to networks with only a few million actors (vertices) and a few hundred million interactions (edges).By overcoming these limitations, ICS '25, June 08-11, 2025, Salt Lake City, UT, USA Minutoli et al.we show that our approach is capable of processing a realistic social contact network of the United States with 285 million nodes and about 8 billion edges.This two ordersof-magnitude improvement over the previous state-of-theart is obtained by leveraging algorithmic advancements for the InfMax problem and designing several problem-specific approaches to overlap communication with computation, improve GPU efficiency, and lower the application's memory requirements.We evaluate strong scaling for computing 10k most influential seeds using up to 8k nodes of an exascale system, and weak scaling from 128 to 8k system nodes for seed sets ranging from 625 to 40k seeds.We achieve the fastest-known runtime of 25 minutes while performing 48 million diffusion simulations totaling 2.31 petabytes to identify 40k influential seeds using 8k nodes, and take 5.75 minutes to identify 10k seeds while using 4k nodes.

The Russian invasion of Ukraine in February 2022 has led to the largest forced migration crisis in Europe since World War II, with millions displaced both internally and internationally. Among the displaced, approximately 4.2 million individuals have returned, highlighting the significance of return migration as a critical phase in the migration continuum. Existing studies on return migration are limited in scope, relying on survey-based approaches that suffer from demographic bias, lack of validation against ground truth, and inability to account for uncertainty. We propose a novel computational framework for modeling the return of conflict-induced migrants, using agent-based models (ABMs) and their surrogates. These models are grounded in hazard functions and account for sociopolitical contexts. Our proposed ABMs outperform baseline methods in estimating return migration from Poland to Ukraine by at least 42% and by as much as 57% in terms of normalized root mean squared error (NRMSE). Further, to illustrate the utility of such models for policymakers, we conduct two case studies that estimate the duration of displacement and characterize the demographic breakdown among the returnees.

Preventing the spread of infectious diseases requires implementing interventions at various levels of government and evaluating the potential impact and efficacy of those preemptive measures. Agent-based modeling can be used for detailed studies of the spread of such diseases in the presence of possible interventions. The computational cost of modeling epidemic diffusion through large social contact networks necessitates the use of parallel algorithms and resources in order to achieve quick turnaround times. In this work, we present Loimos, a scalable parallel framework for simulating epidemic diffusion. Loimos uses a hybrid of time-stepping and discrete event simulation to model disease spread, and is implemented on top of Charm++, an asynchronous, many-task runtime that enables over-decomposition and adaptive overlap of computation and communication. We demonstrate that Loimos is able to achieve significant speedups while scaling to large core counts. In particular, Loimos is able to simulate 200 days of a COVID19 outbreak on a digital twin of California in about 42 seconds, for an average of 4.6 billion traversed edges per second (TEPS), using 4096 cores on Perlmutter at NERSC.

Large-scale population displacements arising from conflict-induced forced migration generate uncertainty and introduce several policy challenges. Addressing these concerns requires an interdisciplinary approach that integrates knowledge from both computational modeling and social sciences. We propose a generalized computational agent-based modeling framework grounded by Theory of Planned Behavior to model conflict-induced migration outflows within Ukraine during the start of that conflict in 2022. Existing migration modeling frameworks that attempt to address policy implications primarily focus on destination while leaving absent a generalized computational framework grounded by social theory focused on the conflict-induced region. We propose an agent-based framework utilizing a spatiotemporal gravity model and a Bi-threshold model over a Graph Dynamical System to update migration status of agents in conflict-induced regions at fine temporal and spatial granularity. This approach significantly outperforms previous work when examining the case of Russian invasion in Ukraine. Policy implications of the proposed framework are demonstrated by modeling the migration behavior of Ukrainian civilians attempting to flee from regions encircled by Russian forces. We also showcase the generalizability of the model by simulating a past conflict in Burundi, an alternative conflict setting. Results demonstrate the utility of the framework for assessing conflict-induced migration in varied settings as well as identifying vulnerable civilian populations.

Evacuation planning is an essential part of disaster management where the goal is to relocate people under imminent danger to safety. However, finding jointly optimal evacuation routes and a schedule that minimizes the average evacuation time or evacuation completion time, is a computationally hard problem. As a result, large-scale evacuation routing and scheduling continues to be a challenge. In this paper, we present a game-theoretic approach to tackle this problem. We start by formulating a strategic routing and scheduling game, named the Evacuation Game: Routing and Scheduling (EGRES), where players choose their route and time of departure. We show that: (i) every instance of EGRES has at least one pure strategy Nash equilibrium, and (ii) an optimal outcome in an instance will always be an equilibrium in that instance. We then provide bounds on how bad an equilibrium can be compared to an optimal outcome. Additionally, we present a polynomial-time algorithm, the Sequential Action Algorithm (SAA), for finding equilibria in a given instance under a special condition. We use Virginia Beach City in Virginia, and Harris County in Houston, Texas as study areas and construct two EGRES instances. Our results show that, by utilizing SAA, we can efficiently find equilibria in these instances that have social objective close to the optimal value.

Evacuation planning is a crucial part of disaster management. However, joint optimization of its two essential components, routing and scheduling, with objectives such as minimizing average evacuation time or evacuation completion time, is a computationally hard problem. To approach it, we present MIP-LNS, a scalable optimization method that utilizes heuristic search with mathematical optimization and can optimize a variety of objective functions. We also present the method MIP-LNS-SIM, where we combine agent-based simulation with MIP-LNS to estimate delays due to congestion, as well as, find optimized plans considering such delays. We use Harris County in Houston, Texas, as our study area. We show that, within a given time limit, MIP-LNS finds better solutions than existing methods in terms of three different metrics. However, when congestion dependent delay is considered, MIP-LNS-SIM outperforms MIP-LNS in multiple performance metrics. In addition, MIP-LNS-SIM has a significantly lower percent error in estimated evacuation completion time compared to MIP-LNS.

Synthetic data is gaining rapid importance in many application domains due to privacy issues, bias, lack, or simply unavailability of real data. It is important that the synthetic data be a good representation of real data for successfully completing the task at hand. Thus, devising characteristic validation metrics is crucial and remains an open problem in many domains (e.g., image generation). Good validation metrics must be able to disentangle the differences between the quality and the variability coverage of the synthetic data. We propose to use a 3-dimensional metric (precision α, recall β, coverage γ) to describe the fidelity and diversity of the synthetic data. In this paper, we improve on existing definitions of precision, recall, and coverage to extend to large scale time series data. Traditional nearest neighbor manifolds from the literature are replaced by unsupervised learning techniques such as clustering to deal with large scale fine resolution time series while computing the validation metrics. The proposed metrics are employed to validate synthetic data in the domain of residential energy demand. In addition, we extend these definitions to datasets that have a natural hierarchical structure. We propose a hierarchical data-tree model in which precision, recall, and coverage can be computed at multiple inherent (and/or custom) levels of groupings of the data.

Evacuation planning is an essential part of disaster management where the goal is to relocate people in a safe and orderly manner. Existing research has shown that such problems are hard to approximate and current methods are difficult to scale to real-life applications. We introduce a notion of fairness and two related objectives while studying evacuation planning, namely: minimizing maximum inconvenience and minimizing average inconvenience. We show that both problems are not just NP-hard to solve exactly, but in fact are NP-hard to approximate. On the positive side, we present a heuristic optimization method MIP-LNS, based on the well-known Large Neighborhood Search framework, that can find good approximate solutions in reasonable amount of time. We also consider a multi-objective problem where the goal is to minimize both objectives and solve it using MIP-LNS. We use real-world road network and population data from Harris County in Houston, Texas (a region that needed large-scale evacuations in the past), and apply MIP-LNS to calculate evacuation plans for the area. We compare the quality of the plans in terms of evacuation efficiency and fairness. We find that the solutions to the multi-objective problem are superior in both of these aspects. We also perform statistical tests to show that the solutions are significantly different.

Residential consumers have become active participants in the power distribution network after being equipped with residential EV charging provisions. This creates a challenge for the network operator tasked with dispatching electric power to the residential consumers through the existing distribution network infrastructure in a reliable manner. In this paper, we address the problem of scheduling residential EV charging for multiple consumers while maintaining network reliability. An additional challenge is the restricted exchange of information: where the consumers do not have access to network information and the network operator does not have access to consumer load parameters. We propose a distributed framework which generates an optimal EV charging schedule for individual residential consumers based on their preferences and iteratively updates it until the network reliability constraints set by the operator are satisfied. We validate the proposed approach for different EV adoption levels in a synthetically created digital twin of an actual power distribution network. The results demonstrate that the new approach can achieve a higher level of network reliability compared to the case where residential consumers charge EVs based solely on their individual preferences, thus providing a solution for the existing grid to keep up with increased adoption rates without significant investments in increasing grid capacity.

We study allocation of COVID-19 vaccines to individuals based on the structural properties of their underlying social contact network. Using a realistic representation of a social contact network for the Commonwealth of Virginia, we study how a limited number of vaccine doses can be strategically distributed to individuals to reduce the overall burden of the pandemic. We show that allocation of vaccines based on individuals' degree (number of social contacts) and total social proximity time is significantly more effective than the usually used age-based allocation strategy in reducing the number of infections, hospitalizations and deaths. The overall strategy is robust even: (i) if the social contacts are not estimated correctly; (ii) if the vaccine efficacy is lower than expected or only a single dose is given; (iii) if there is a delay in vaccine production and deployment; and (iv) whether or not non-pharmaceutical interventions continue as vaccines are deployed. For reasons of implementability, we have used degree, which is a simple structural measure and can be easily estimated using several methods, including the digital technology available today. These results are significant, especially for resource-poor countries, where vaccines are less available, have lower efficacy, and are more slowly distributed.

We study the role of vaccine acceptance in controlling the spread of COVID-19 in the US using AI-driven agent-based models. Our study uses a 288 million node social contact network spanning all 50 US states plus Washington DC, comprised of 3300 counties, with 12.59 billion daily interactions. The highly-resolved agent-based models use realistic information about disease progression, vaccine uptake, production schedules, acceptance trends, prevalence, and social distancing guidelines. Developing a national model at this resolution that is driven by realistic data requires a complex scalable workflow, model calibration, simulation, and analytics components. Our workflow optimizes the total execution time and helps in improving overall human productivity.This work develops a pipeline that can execute US-scale models and associated workflows that typically present significant big data challenges. Our results show that, when compared to faster and accelerating vaccinations, slower vaccination rates due to vaccine hesitancy cause averted infections to drop from 6.7M to 4.5M, and averted total deaths to drop from 39.4K to 28.2K nationwide. This occurs despite the fact that the final vaccine coverage is the same in both scenarios. Improving vaccine acceptance by 10% in all states increases averted infections from 4.5M to 4.7M (a 4.4% improvement) and total deaths from 28.2K to 29.9K (a 6% increase) nationwide. The analysis also reveals interesting spatio-temporal differences in COVID-19 dynamics as a result of vaccine acceptance. To our knowledge, this is the first national-scale analysis of the effect of vaccine acceptance on the spread of COVID-19, using detailed and realistic agent-based models.

The COVID-19 global outbreak represents the most significant epidemic event since the 1918 influenza pandemic. Simulations have played a crucial role in supporting COVID-19 planning and response efforts. Developing scalable workflows to provide policymakers quick responses to important questions pertaining to logistics, resource allocation, epidemic forecasts and intervention analysis remains a challenging computational problem. In this work, we present scalable high performance computing-enabled workflows for COVID-19 pandemic planning and response. The scalability of our methodology allows us to run fine-grained simulations daily, and to generate county-level forecasts and other counterfactual analysis for each of the 50 states (and DC), 3140 counties across the USA. Our workflows use a hybrid cloud/cluster system utilizing a combination of local and remote cluster computing facilities, and using over 20,000 CPU cores running for 6-9 hours every day to meet this objective. Our state (Virginia), state hospital network, our university, the DOD and the CDC use our models to guide their COVID-19 planning and response efforts. We began executing these pipelines March 25, 2020, and have delivered and briefed weekly updates to these stakeholders for over 30 weeks without interruption.

Evacuation planning methods aim to design routes and schedules to relocate people to safety in the event of natural or man-made disasters. The primary goal is to minimize casualties which often requires the evacuation process to be completed as soon as possible. In this paper, we present QueST, an agent-based discrete event queuing network simulation system, and STEERS, an iterative routing algorithm that uses QueST for designing and evaluating large scale evacuation plans in terms of total egress time and congestion/bottlenecks occurring during evacuation. We use the Houston Metropolitan Area, which consists of nine US counties and spans an area of 9,444 square miles as a case study, and compare the performance of STEERS with two existing route planning methods. We find that STEERS is either better or comparable to these methods in terms of total evacuation time and congestion faced by the evacuees. We also analyze the large volume of data generated by the simulation process to gain insights about the scenarios arising from following the evacuation routes prescribed by these methods.

It is well known that physical interdependencies exist between networked civil infrastructures such as transportation and power system networks. In order to analyze complex nonlinear correlations between such networks, datasets pertaining to such real infrastructures are required. However, such data are not readily available due to their proprietary nature. This work proposes a methodology to generate realistic synthetic power distribution networks for a given geographical region. A network generated in this manner is not the actual distribution system, but its functionality is very similar to the real distribution network. The synthetic network connects high voltage substations to individual residential consumers through primary and secondary distribution networks. Here, the distribution network is generated by solving an optimization problem which minimizes the overall length of the network subject to structural and power flow constraints. This work also incorporates identification of long high voltage feeders originating from substations and connecting remotely situated customers in rural geographic locations while maintaining voltage regulation within acceptable limits. The proposed methodology is applied to the state of Virginia and creates synthetic distribution networks which are validated by comparing them to actual power distribution networks at the same location.

Prophylactic interventions such as vaccine allocation are some of the most effective public health policy planning tools. The supply of vaccines, however, is limited and an important challenge is to optimally allocate the vaccines to minimize epidemic impact. This resource allocation question (which we refer to as VaccIntDesign) has multiple dimensions: when, where, to whom, etc. Most of the existing literature in this topic deals with the latter (to whom), proposing policies that prioritize individuals by age and disease risk. However, since seasonal influenza spread has a typical spatial trend, and due to the temporal constraints enforced by the availability schedule, the when and where problems become equally, if not more, relevant. In this paper, we study the VaccIntDesign problem in the context of seasonal influenza spread in the United States. We develop a national scale metapopulation model for influenza that integrates both short and long distance human mobility, along with realistic data on vaccine uptake. We also design GreedyAlloc, a greedy algorithm for allocating the vaccine supply at the state level under temporal constraints and show that such a strategy improves over the current baseline of pro-rata allocation, and the improvement is more pronounced for higher vaccine efficacy and moderate flu season intensity. Further, the resulting strategy resembles a ring vaccination applied spatiallyacross the US.

Network-based computer simulation models are powerful tools for analyzing and guiding policy formation related to the actual systems being modeled. However, the inherent data and computationally intensive nature of this model class gives rise to fundamental challenges when it comes to executing typical experimental designs. In particular this applies to model validation. Manual management of the complex simulation work-flows along with the associated data will often require a broad combination of skills and expertise. Examples of skills include domain expertise, mathematical modeling, programming, high-performance computing, statistical designs, data management as well as the tracking all assets and instances involved. This is a complex and error-prone process for the best of practices, and even small slips may compromise model validation and reduce human productivity in significant ways.

Analysis of structural properties and dynamics of networks is currently a central topic in many disciplines including Social Sciences, Biology and Business. CINET, a cyber infrastructure for such studies, introduced the concept of supporting network analysis as a service. The basic idea is to allow experts in various disciplines to focus on obtaining domain-specific insights from the results of network analyses instead of worrying about programming details and allocation of computational resources needed to carry out the analyses. A basic version of CINET was released in May 2012. This paper discusses CINET 2.0, a significantly enhanced version that supports complex network analyses through a web portal. CINET 2.0 has already been used for teaching courses related to Network Science at several US universities. In this paper, we discuss how CINET 2.0 significantly extends CINET 1.0 through enhancements to some components and the addition of new components.

Simple diffusion processes on networks have been used to model, analyze and predict diverse phenomena such as spread of diseases, information and memes. More often than not, the underlying network data is noisy and sampled. This prompts the following natural question: how sensitive are the diffusion dynamics and subsequent conclusions to uncertainty in the network structure? In this paper, we consider two popular diffusion models: Independent cascade (IC) model and Linear threshold (LT) model. We study how the expected number of vertices that are influenced/infected, for particular initial conditions, are affected by network perturbations. Through rigorous analysis under the assumption of a reasonable perturbation model we establish the following main results. (1) For the IC model, we characterize the sensitivity to network perturbation in terms of the critical probability for phase transition of the network. We find that the expected number of infections is quite stable, unless the transmission probability is close to the critical probability. (2) We show that the standard LT model with uniform edge weights is relatively stable under network perturbations. (3) We study these sensitivity questions using extensive simulations on diverse real world networks and find that our theoretical predictions for both models match the observations quite closely. (4) Experimentally, the transient behavior, i.e., the time series of the number of infections, in both models appears to be more sensitive to network perturbations.

Simple diffusion processes on networks have been used to model, analyze and predict diverse phenomena such as spread of diseases, information and memes. More often than not, the underlying network data is noisy and sampled. This prompts the following natural question: how sensitive are the diffusion dynamics and subsequent conclusions to uncertainty in the network structure? In this paper, we consider two popular diffusion models: Independent cascades (IC) model and Linear threshold (LT) model. We study how the expected number of vertices that are influenced/infected, given some initial conditions, are affected by network perturbation. By rigorous analysis under the assumption of a reasonable perturbation model we establish the following main results. (1) For the IC model, we characterize the susceptibility to network perturbation in terms of the critical probability for phase transition of the network. We find the expected number of infections is quite stable, unless the the transmission probability is close to the critical probability. (2) We show that the standard LT model with uniform edge weights is relatively stable under network perturbations. (3) Empirically, the transient behavior, i.e., the time series of the number of infections, in both models appears to be more sensitive to network perturbations. We also study these questions using extensive simulations on diverse real world networks, and find that our theoretical predictions for both models match the empirical observations quite closely.

Networks are an effective abstraction for representing real systems. Consequently, network science is increasingly used in academia and industry to solve problems in many fields. Computations that determine structure properties and dynamical behaviors of networks are useful because they give insights into the characteristics of real systems. We introduce a newly built and deployed cyberinfrastructure for network science (CINET) that performs such computations, with the following features: (i) it offers realistic networks from the literature and various random and deterministic network generators; (ii) it provides many algorithmic modules and measures to study and characterize networks; (iii) it is designed for efficient execution of complex algorithms on distributed high performance computers so that they scale to large networks; and (iv) it is hosted with web interfaces so that those without direct access to high performance computing resources and those who are not computing experts can still reap the system benefits. It is a combination of application design and cyberinfrastructure that makes these features possible. To our knowledge, these capabilities collectively make CINET novel. We describe the system and illustrative use cases, with a focus on the CINET user.

This paper develops a formal model of process migration that describes programs, processes, and the migration of those processes within a migration realm. A migration realm is a group of processors modeled as finite state machines. The model is motivated by a migration application between software and Field Programmable Gate Array (FPGA) hardware, and the theorems of the model guide the use of FPGA resources while guaranteeing complete and correct execution of a process. By defining different types of migration realms this paper also develops a migration realm taxonomy.

This work demonstrates that road traffic simulation of entire metropolitan areas is possible with reconfigurable supercomputing that combines 64-bit microprocessors and FPGAs in a high bandwidth, low latency interconnect. Previously, traffic simulation on FPGAs was limited to very-short road segments or required a very large number of FPGAs. Our data streaming approach overcomes scaling issues associated with direct implementations and still allows for high-level parallelism by dividing the data sets between hardware and software across the reconfigurable supercomputer. Using one FPGA on the Cray XD1 supercomputer, we are able to achieve a 34.4/spl times/ speed up over the AMD microprocessor. System integration issues must be optimized to exploit this speedup in the overall simulation.

Often reconfigurable systems are reported to have 10× to 100× speedup over that of a software system. However, the reconfigurable hardware must usually be combined with software to form an entire system. This system integration presents a hardware/software co-design problem with many system engineering issues. Here, we present traffic acceleration on the Cray XD1 supercomputer and describe the costs involved in different hardware/software trade-offs.