Spatiotemporal Propagation Research Articles

Visualization of the Spatiotemporal Propagation of Interictal Spikes in Temporal Lobe Epilepsy: A MEG Pilot Study.

Magnetoencephalography (MEG) is clinically used to localize interictal spikes in discrete brain areas of epilepsy patients through the equivalent current dipole (ECD) method, but does not account for the temporal dynamics of spike activity. Recent studies found that interictal spike propagation beyond the temporal lobe may be associated with worse postsurgical outcomes, but studies using whole-brain data such as in MEG remain limited. In this pilot study, we developed a tool that visualizes the spatiotemporal dynamics of interictal MEG spikes normalized to spike-free sleep activity to assess their onset and propagation patterns in patients with temporal lobe epilepsy (TLE). We extracted interictal source data containing focal epileptiform activity in awake and asleep states from seven patients whose MEG ECD clusters localized to the temporal lobe and normalized the data against spike-free sleep recordings. We calculated the normalized activity over time per cortical label, confirmed maximal activity at onset, and mapped the activity over a 10 ms interval onto each patient's brain using a custom-built Multi-Modal Visualization Tool. The onset of activity in all patients appeared near the clinically determined epileptogenic zone. By 10 ms, four of the patients had propagated source activity restricted to within the temporal lobe, and three had propagated source activity spread to extratemporal regions. Using this tool, we show that noninvasively identifying the onset and propagation of interictal spike activity in MEG can be achieved, which may help provide further insight into epileptic networks and guide surgical planning and interventions in patients with TLE.

Time-aware personalized graph convolutional network for multivariate time series forecasting

The use of graph structures to model variable dependencies in multivariate time series data holds promise. Probabilistic graph models, sampling discrete graphs from neural network-parameterized distributions, have shown efficacy. However, the current approaches ignore that dependencies are dynamic and constrained by heterogeneous information. For instance, interactions among traffic data’s roads are constrained by traffic volume and network topology. Furthermore, the existing spatio-temporal graph propagation mechanism has failed to account for the impact of the evolutionary patterns of the nodes themselves on future outcomes. To address these limitations, this research introduces a novel approach: a time-aware personalized graph convolutional network (TPGCN) consisting of time-aware discrete graph structure estimation (TADG) and dynamic personalized graph convolutional (DPGC) components. The TADG module confronts the challenge of graph learning by framing it as an inference issue within a heterogeneous information landscape. Through dynamic inputs, it deduces graph structures that are influenced by the varying nature of interactions. DPGC, drawing inspiration from personalized PageRank, models variable effects while considering evolutionary patterns and external factors. By dynamically fusing these two types of information at every information aggregation, our approach outperforms existing state-of-the-art methods, as demonstrated on eight benchmark datasets.

Transport causality knowledge-guided GCN for propagated delay prediction in airport delay propagation networks

Flight delays pose a worldwide challenge that significantly affect the safety and efficiency of air transportation systems. However, propagated delay prediction, as well as its causality among airport delay propagation networks, has not considered some crucial issues regarding spatiotemporal dependence and propagation relationships. Thus, this study proposes a transport causality knowledge-guided extended graph convolutional network (GCN) framework to tackle these issues. In particular, a causality knowledge-guided airport delay propagation network (ADPN) is developed using the second modified transfer entropy (SMTE) principle. Furthermore, a causality-embedded adjacency matrix is utilized by an extended GCN for propagated delay prediction. Comprehensive validations and results indicate that the proposed method benefits significantly from the causality knowledge, and increases the prediction performances up to 15.51%. Thus, transport causality is significant and efficient for understanding propagated delay features and airport delay propagation network characteristics.

Temporal and Spatial Propagation Characteristics of the Meteorological, Agricultural and Hydrological Drought System in Different Climatic Conditions within the Framework of the Watershed Water Cycle

Investigating systematic meteorological–agricultural–hydrological drought propagation within an integrated watershed framework is crucial yet challenging for advancing robust early warning systems and targeted resilience strategies. To address this gap, this study employs the standardized precipitation evapotranspiration index (SPEI), standardized soil moisture index (SSMI), and nonlinear joint hydrological drought index (NJHDI) to characterize meteorological, agricultural, and hydrological drought, respectively. Utilizing the strongest correlation method, variations in the propagation time along moisture movement pathways within the system are quantified using the Soil and Water Assessment Tool (SWAT) model. The Yellow River basin (YRB) is selected as the case study area. Key results reveal distinct seasonal patterns in meteorological–agricultural propagation across the basin, with escalated spring and summer timescales. However, weaker correlations and limited seasonality emerge for agricultural–hydrological linkages due to sustaining baseflow buffers. Specifically, the arid and semiarid region exhibited shorter propagation time with 1–5 months, while the semiarid and semihumid region displayed longer timescales with 7–12 months. These findings provide valuable scientific references for enhancing real-time early warning systems tailored to coupled watershed systems. The integrated methodology underscores the importance of unraveling fine-scale spatiotemporal propagation variability for localized drought resilience.

Enhanced brain structure-function tethering in transmodal cortex revealed by high-frequency eigenmodes

While the link between brain structure and function remains an ongoing challenge, the prevailing hypothesis is that the structure-function relationship may itself be gradually decoupling from unimodal to transmodal cortex. However, this hypothesis is constrained by the underlying models which may neglect requisite information. Here we relate structural and functional connectivity derived from diffusion and functional MRI through orthogonal eigenmodes governing frequency-specific diffusion patterns. We find that low-frequency eigenmodes contribute little to functional interactions in transmodal cortex, resulting in divergent structure-function relationships. Conversely, high-frequency eigenmodes predominantly support neuronal coactivation patterns in these areas, inducing structure-function convergence along a unimodal-transmodal hierarchy. High-frequency information, although weak and scattered, could enhance the structure-function tethering, especially in transmodal association cortices. Our findings suggest that the structure-function decoupling may not be an intrinsic property of brain organization, but can be narrowed through multiplexed and regionally specialized spatiotemporal propagation regimes.

Full-field dynamic response of low-velocity impact on an elastic plate using 3D digital image correlation

Measurement of the impact process is challenging due to its short duration and high frequency response. Due to the very small strain in low-velocity impacts, the requirements for the temporal and spatial resolution of test methods are extremely high. Traditionally, strain gauges or lasers are commonly used to measure the dynamic response of a single point, or quasi-static loading is used to achieve full-field response. To finely describe the transient full-field deformation process of low-velocity impacts, a high-speed imaging system combined with 3D digital image correlation (DIC) technique is set up to investigate the instantaneous responses of impact on a polyvinyl chloride (PVC) plate at the ms and με level. The full-field deformation of the contact surface and the whole process of elastic wave propagation are measured. The measurement by DIC closely matches the measurement by strain gauges and agrees well with the numerical simulation results. The measurement reveals that the “multi-micro-impacts” phenomenon may happen during a complete impact process due to the complex strain wave propagation. This work provides a benchmark case for low-velocity impact full-field testing, and the results show that 3D DIC can accurately evaluate dynamic behaviors of low-velocity impact between flexible bodies and can accurately describe the spatiotemporal propagation characteristics of strain waves.

Spatiotemporal propagation of a time‐periodic reaction–diffusion SI epidemic model with treatment

AbstractThis work is concerned with the spatiotemporal propagation phenomena for a time‐periodic reaction‐diffusion susceptible‐infectious (SI) epidemic model with treatment in terms of the asymptotic speed of spread and periodic traveling waves. First, the asymptotic speed of spread is characterized and the spreading properties of the model are analyzed by combining the periodic principal eigenvalue problem, comparison method, and the uniform persistence idea for a dynamical system. Second, by constructing suitable super‐ and subsolutions for truncation problems corresponding to the traveling wave system, the existence of periodic traveling waves is established via the fixed point theorem twice. It turned out that the asymptotic speed of spread coincides with the minimum wave speed of periodic traveling waves. Finally, via numerical simulation, the effects of some important parameters (such as diffusion rate, treatment rate, etc.) on the spreading speed are discussed, and the asymptotic properties of the periodic traveling waves are explored.

The differing effects of sleep on ictal and interictal network dynamics in drug-resistant epilepsy.

Sleep has important influences on focal interictal epileptiform discharges (IEDs), and the rates and spatial extent of IEDs are increased in non-rapid eye movement (NREM) sleep. In contrast, the influence of sleep on seizures is less clear and its effects on seizure topography are poorly documented. We evaluated the influences of NREM sleep on ictal spatiotemporal dynamics and contrasted these with interictal network dynamics. We included patients with drug-resistant focal epilepsy who underwent continuous intracranial EEG with depth electrodes. Patients were selected if they had 1-3 seizures from each vigilance state, wakefulness and NREM sleep, within a 48-hour window and under the same anti-seizure medication. A 10-min epoch of the interictal iEEG was selected per state, and IEDs were detected automatically. Twenty-five patients (13 females; 32.5±7.1 years) were included. The seizure onset pattern, duration, spatiotemporal propagation and latency of ictal high-frequency activity did not differ significantly between wakefulness and NREM sleep (all p>0.05). In contrast, IED rates and spatial distribution were increased in NREM compared to wakefulness (p<0.001, Cliff's d=0.48 and 0.49). The spatial overlap between vigilance states was higher for seizures (57.1±40.1%) than IEDs (41.7±46.2%) (p=0.001, Cliff's d=0.51). In contrast to its effects on IEDs, NREM sleep does not affect ictal spatiotemporal dynamics. This suggests that once the brain surpasses the seizure threshold, it will follow the underlying epileptic network irrespective of the vigilance state. These findings offer valuable insights into neural network dynamics in epilepsy and have important clinical implications for localizing seizure foci. This article is protected by copyright. All rights reserved.

Local development, metropolitan sustainability and the urbanization-suburbanization nexus in the Mediterranean region: A quantitative exercise

The present study investigates the influence of economic growth and socio-demographic change on long-term urbanization and suburbanization processes in a divided region of Southern Europe (Athens, Greece). A two-block Partial Least Squares (tPLS) analysis decomposed metropolitan growth into distinctive processes of local development in ‘central’ and ‘peripheral’ areas considering changes over time (1966–2008) in 14 indicators representative of multifaceted territorial dynamics in the Mediterranean space. By quantifying the bi-directional (spatio-temporal) propagation of socioeconomic impulses, the empirical results of tPLS have delineated substantial differences in the development path of inner cities and suburbs, fueling spatial divides in social conditions and increasingly heterogeneous processes of regional development. While inner cities experienced population growth and settlement densification, peripheral areas underwent rapid suburbanization with spatial concentration of economic functions and a generalized settlement dispersion. Providing further insights in the analysis of the differential mechanisms of metropolitan development in Western Europe, results of this study confirm the specificity of urbanization and suburbanization processes at the base of the increased polarization in central and peripheral areas typical of Mediterranean countries.

Cultural and artistic dissemination of traditional Chinese silk patterns in the 17th-18th centuries in the context of big data

Abstract Silk has evolved a unique silk culture in its long history, and silk patterns have condensed the external aesthetics of silk culture and shown distinctive national characteristics. This paper firstly constructs the cultural communication network of the 17th-18th century Maritime Silk Road through big data technology and uses social network analysis to study the structural characteristics of the cultural and artistic communication network. The MPI multidimensional spatio-temporal communication evolution model is used to delineate the network’s layers further, and exploratory spatial data analysis is introduced to explore the spatio-temporal communication evolution pattern of the network. Finally, in the empirical analysis, 200 years of silk pattern data from 1600-1800 in China were selected to establish four dimensions and 11 indicators data for measurement. A comprehensive MPI multidimensional spatio-temporal propagation evolution model was derived by principal component analysis to derive the contribution of each indicator. The results show that the multidimensional cultural dissemination index of traditional silk pattern culture and art from 1600 to 1800 was 0.7461 and approximated to 0.8, and the dissemination influence was 0.2252, respectively, rising to 97%, indicating that the dissemination of silk pattern culture and art in the 17th-18th centuries was a flourishing trend. The MPI multidimensional spatial and temporal communication evolution model proposed in this paper effectively visualizes and analyzes the communication channels of silk pattern culture and art, which provides a guiding reference for the revival and communication of modern Chinese traditional culture.

TSDTVOS: Target-guided spatiotemporal dual-stream transformers for video object segmentation

Video object segmentation automatically separates the interested objects from the background across a video sequence and was an active research area in recent years. The crucial challenge lies in investigating an effective architecture to fully exploit spatiotemporal correlation in a given video sequence for achieving accurate segmentation results. In this paper, we propose a novel semi-supervised Transformer-based framework called Target-guided Spatiotemporal Dual-stream Transformers (TSDT) with two separate streams to enable effective spatiotemporal context propagation. Technically, the temporal stream is used to aggregate rich temporal cues from past frames, while the spatial stream is trained to encode object location and appearance information stored in the current frame. To compress and integrate temporal features, a target guidance block (TGB) is designed to retrieve target information in the past video flow under the guidance of the current frame. The experimental results on video object segmentation benchmarks demonstrate the feasibility and effectiveness of the proposed framework. Codes and trained models are available at https://github.com/zhouweii234/TSDTVOS.

Does seizure propagate within or across intrinsic brain networks? An intracranial EEG study

BackgroundUnderstanding the spatiotemporal propagation profiles of seizures is crucial for the preoperative assessment of epilepsy patients. The present study aimed to investigate whether seizures exhibit propagation patterns that align with intrinsic networks (INs). MethodsA quantitative analysis was conducted to examine ictal fast activity (IFA). The Epileptogenicity Index (EI) was employed to assess the epileptogenicity, spectral features, and temporal characteristics of IFA. Intra-network and inter-network comparisons were made regarding the IFA-related metrics. Additionally, the metrics were correlated with Euclidean distance. Network connection maps were generated to visualize seizures originating from different INs, allowing for comparisons between distinct groups. ResultsData for 81 seizures in 43 subjects were captured using stereoelectroencephalography implantation. Three metrics were compared: EI, time involvement (TI), and energy ratio index (ERI). Intra-network channels exhibited higher EI, earlier involvement of IFA, and stronger high-frequency energy. These findings were further validated through subgroup analyses stratified by neuropathology, seizure type, and seizure origination lobe. Correlation analyses revealed a negative association between distance and both EI and ERI, while distance exhibited a positive correlation with TI. Seizures originating from different INs exhibited varying propagation characteristics. ConclusionsThe study findings highlight the dominant role of intra-network dynamics over inter-network during seizure propagation. These results contribute to our understanding of seizure dynamics and their relationship with INs.

Spatiotemporal patterns and propagation mechanism of meteorological droughts over Yangtze River Basin and Pearl River Basin based on complex network theory

Investigating spatiotemporal pattern and propagation mechanism of meteorological drought is of great significance to develop early warning system for drought-affected regions and other drought types. In this study, the spatiotemporal characteristics and evolution trend of meteorological drought represented by the standardized precipitation evapotranspiration index (SPEI) during 2016–2021 across the Yangtze River basin (YZRB) and Pearl River basin (PRB) were analyzed. The drought spatial synchronization extent and propagation mechanisms (i.e., propagation source, sink, direction, strength and distance) were also investigated through constructing the undirected and directed complex networks, respectively. The results indicate that during 1961–2021, meteorological droughts in the middle and lower reaches of the YZRB and PRB present an alleviating trend, while droughts over their upper reaches that are relatively far from coastal regions are aggravated. The drought synchronization extents at the spatial scale are classified into seven communities, where the drought events within the same community are more likely to occur synchronously and their characteristics are more similar. Moreover, the drought propagation source regions are mainly distributed over the eastern coastal regions of the YZRB, the upper coastal regions of the PRB and the headwaters of the YZRB, from where droughts start to propagate along the WSW (or WNW) direction, NE direction and ESE direction to interior regions, respectively. The sink regions are mostly distributed in the YZRB and PRB middle reaches and the southwestern parts of the YZRB upper reaches. Meteorological drought events in most regions have a propagation distance of hundreds of kilometers. This study provides new tools and insights on drought propagation and prediction and can help policy-makers formulate advanced preventive measures.

A Quantitative Framework for Propagation Paths of Natech Domino Effects in Chemical Industrial Parks: Part I—Failure Analysis

Along with global climate change and industrialization, domino effects caused by Natech events occurred frequently in chemical industrial parks over the past decades. Previous research has not yet proposed a reliable method to obtain all possible paths of Natech domino effects, and moreover, a risk assessment and mitigation system has not been established. The present work aims to develop a quantitative framework for propagation paths of Natech domino effects, which can effectively safeguard the sustainable development of chemical industrial parks. The presentation of this work is divided into two parts: Part I (current paper) proposes a path probability calculation method that can simultaneously consider multiple primary accident scenarios and multi-level domino effects triggered by natural disasters. The proposed method transforms the propagation paths of domino effects into the paths of directed graph by constructing the equipment failure state transition matrix and the equipment failure state transition probability matrix. The depth-first traversal algorithm is used to obtain all possible propagation paths and their propagation probabilities, providing data support for the quantitative risk assessment and prevention and control measures presented in the accompanying paper (Part II). The case study shows that the probability of equipment failure caused by multi-level domino effects triggered by Natech accidents is higher than that of conventional accidents. However, the present work only considers the spatial propagation of domino effects, while their spatio-temporal propagation remains as a further direction for this area of inquiry.

Propagation of quasi-diffraction-free lommel-airy vortex wavepackets in chiral media

Lommel beams have significant applications in optical communication and optical manipulation due to the changeable symmetry of transverse intensity distribution and constantly varying orbital angular momentum. In this paper, a novel quasi-diffraction-free Lommel-Airy vortex (LAiV) wavepackets is proposed, and spatiotemporal dynamics of the LAiV wavepackets propagating in chiral media are investigated. Results indicate that several variables such as topological charge group velocity asymmetry factor and chiral parameter could be used to regulate the profiles of the 3D spatiotemporal vortex rings. Particularly, the spatiotemporal propagation trajectories and intensity patterns of the LAiV wavepackets could be manipulated by altering the asymmetry factor and the chiral parameter These results would deepen the understanding of the spatiotemporal propagation properties of the LAiV wavepackets and are expected to promote the development of the novel optical communications in chiral media.

The Way Toward Growth: A Time-series Factor Decomposition of Socioeconomic Impulses and Urbanization Trends in a Pre-crisis European Region

The present study investigates long-term urbanization and suburbanization trends in a divided region of Mediterranean Europe (Attica, Greece). A time series (1965–2008) dynamic factor analysis of 14 indicators that reflect different aspects of metropolitan growth, allowed estimating the impact of economic expansion and social change on the local development path characteristic of the study area. To quantify the spatio-temporal propagation of socioeconomic impulses to local development, Attica was partitioned in two spatial domains, the ‘Greater Athens’ area (hereafter the ‘core’ district) and the rest of the metropolitan region (hereafter the ‘ring’ district). By integrating Multi-way Factor Analysis with Continuous Wavelength Transform and rapidity-of-change metrics grounded on complex thinking, the exploratory approach proposed here has delineated latent mechanisms of urban expansion, indicating substantial divergences in the development path of the two districts. While the ‘core’ district experienced population increase and settlement densification, the ‘ring’ district underwent a suburbanization process resulting in a moderate—and relatively slow—concentration of economic activities and social functions. Reflecting a non-linear propagation of growth impulses from urban to rural areas, the economic interplay between ‘core’ and ‘ring’ areas delineates a complex development path accelerating spatial polarization in central and peripheral locations. Our results definitely highlight the importance of ‘system thinking’ in regional studies, applied economics, and social research.

Spatio-temporal propagation and reconstruction for low-light video enhancement

Different from the single low-light image enhancement, the key for video enhancement is to improve temporal stability in spatial and temporal varying manner. However, most methods fail to effectively align images or features of adjacent frames to fuse spatio-temporal information and achieve limited enhancement performance, especially in dynamic videos with scene changes and moving targets. In this paper, we propose a spatio-temporal propagation and reconstruction (STPR) network to enable video representation learning for low-light video enhancement. Specifically, we develop a recursively supervised-based pyramid residual dense structure for extracting expressive context features from video sequences, the unique structural design enlarges the receptive field in multi-scale space with limited training parameters. Moreover, to strengthen the interaction of across-frame features, a feature propagation subnet is present to achieve precise alignment and effective fusion of spatio-temporal features. After obtaining the aggregated features of adjacent frames, the spatio-temporal feature reconstruction subnet exploits temporal dependence to improve the quality of target frames and enforce temporal consistency. Experimental results on several low-light video datasets demonstrate that our network outperforms state-of-the-art methods both in quantitative and visual comparisons.

Identifying spatiotemporal propagation of droughts in the agro-pastoral ecotone of northern China with long-term WRF simulations

Understanding the spatiotemporal evolution of droughts is critical for food security and water allocation. In this study, we propose an approach to construct the linkage of the propagation from meteorological drought to agricultural drought to explore the simultaneous spatiotemporal evolution of droughts based on a 3-dimensional clustering algorithm. We first evaluate the improvement of the downscaled high-resolution outputs from the Weather Research and Forecasting (WRF) model to reanalysis data in detecting wet-dry variations in the agro-pastoral ecotone of northern China (APENC). The WRF simulation results well characterize the evolution of droughts with a high correlation coefficient of 0.82 (p<0.05), which implies that the WRF downscaling model provides reliable hydrometeorological results for assessing and analyzing long-term wet-dry variations. Subsequently, we identify and track the spatiotemporal evolution of each individual drought event in the APENC for the first time. Based on the standardized precipitation index calculated from the WRF outputs, a total of 185 drought clusters were identified during 2000–2017, with 28 drought events lasting at least 3 months duration occurring in the APENC. The characteristics of the different drought events vary considerably. Droughts in the APENC have a larger percentage of propagation to the northeast and northwest and migrated hundreds of kilometers from the source sites. The results of the propagation of meteorological droughts to agricultural droughts indicate that 53.75% of the meteorological drought events progressed further to agricultural droughts. Agricultural droughts exhibit spatial variations consistent with meteorological droughts. Agricultural droughts have 1.69 months onset lag time, 2.12 months termination lag time, and overall longer duration than meteorological droughts in the APENC. Our findings provide valuable information for the mechanism and prediction of drought propagation.

Analysis of spatio-temporal propagation of occipito-frontal spikes in childhood epilepsies by 3D sequential voltage mapping and dipole localization.

We sought to study the spatio-temporal propagation of occipito-frontal spikes in childhood epilepsies by voltage mapping and dipole localization and identify types of occipito-frontal spikes based on onset, propagation, and stability of their dipoles. Sleep EEG data of children, aged 1-14 years, with a minimum 1 h of recording from June 2018 to June 2021, were analyzed to identify occipito-frontal spikes. In total, 150 successive occipito-frontal spikes were manually selected from each EEG and using a source localization software were averaged using automated pattern matching with a threshold of 80%, and sequential 3D voltage maps of averaged spike were analyzed. Stability quotient (SQ) was calculated as the total number of averages/150. Stable dipole was defined as SQ ≥ .8. Dipole analysis was performed with principal component analysis using an age-appropriate template head model. Ten children with occipito-frontal spikes were identified; five with self-limited epilepsy with autonomic seizures (SeLEAS) and five with non-SeLEAS epilepsies. Three types of occipito-frontal spikes were identified: (1) narrow occipito-frontal spikes with stable dipoles seen in all five children with SeLEAS which were "apparently" synchronous and bilateral clone-like with an occipito-frontal interval of 10-30 ms and a homogeneous propagation pattern from a unilateral medial parieto-occipital region to an ipsilateral mesial frontal region; (2) wide occipito-frontal spikes with stable dipoles seen in one child with non-SeLEAS and developmental and/or epileptic encephalopathy with spike-wave activation in sleep (D/EE-SWAS) with an occipito-frontal interval of 45 ms, caused by focal spike propagation from a deeper temporal focus to ipsilateral peri-rolandic cortex; and (3) wide occipito-frontal spikes with unstable dipoles seen in four children with non-SeLEAS lesional epilepsies with an occipito-frontal latency of >50 ms and heterogeneous propagation patterns with poor intra-individual dipole stability. We successfully identified different types of occipito-frontal spikes in childhood epilepsies. Although the term "occipito-frontal" is used to describe these spikes on the 10-20 EEG system, true propagation from occipital to frontal regions is not necessary. It is possible to differentiate idiopathic from symptomatic cases by analyzing the stability quotient and the occipito-frontal interval of occipito-frontal spikes.

Spatial networks and the spread of COVID-19: results and policy implications from Germany

Spatial networks are known to be informative about the spatiotemporal transmission dynamics of COVID-19. Using district-level panel data from Germany that cover the first 22 weeks of 2020, we show that mobility, commuter and social networks all predict the spatiotemporal propagation of the epidemic. The main innovation of our approach is that it incorporates the whole network and updated information on case numbers across districts over time. We find that when disease incidence increases in network neighbouring regions, case numbers in the home district surge one week later. The magnitude of these network transmission effects is comparable to within-district transmission, illustrating the importance of networks as drivers of local disease dynamics. After the introduction of containment policies in mid-March, network transmission intensity drops substantially. Our analysis suggests that this reduction is primarily due to a change in quality—not quantity—of interregional movements. This implies that blanket mobility restrictions are not a prerequisite for containing the interregional spread of COVID-19.