Herd Immunity with Spatial Adaptation Based on Global Prevalence Information
Akhil Panicker, V. Sasidevan · 2025
Read Herd Immunity with Spatial Adaptation Based on Global Prevalence Information by Akhil Panicker, V. Sasidevan (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2507-20747.txt.
Introduction
Akhil Panicker and Sasidevan study how people changing their behavior can blunt an epidemic, in a 2025 paper. During an outbreak, many pull back: they travel less and meet fewer people, guided by news of how widespread the disease is. The pair model this response in space, where a share of the population shrinks its range of contact as case counts rise.
They test three ways the responding share can grow with case counts: steady, as a power curve, and as an S-shaped curve. In a well-mixed population they compute the response level needed to stop an outbreak, and in a fixed spatial layout they bound it. A simple proportional response adds nothing over a steady share, they find, while the S-shaped response can make case counts swing up and down.
Catalog record
| CREATOR | Akhil Panicker, V. Sasidevan |
|---|---|
| YEAR | 2025 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 52 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2507-20747
Browse the full library · About the Tower of Babel
Full text
The opening of the work, from the library's hosted text:
Herd Immunity with Spatial Adaptation Based on Global Prevalence Information Akhil Panicker, V. Sasidevan Abstract During an epidemic outbreak, individuals often modify their behavior in response to global prevalence cues, using spatially mediated adaptations such as reduced mobility or transmission range. In this work, we investigate the impact of distance-based adaptive behaviors on epidemic dynamics, where a fraction of the population adjusts its transmission range and susceptibility to infection based on global prevalence. We consider three adaptation scenarios: a constant adaptive fraction, a power-law dependence and a sigmoidal dependence of adaptive fraction on global prevalence. In the spatially well-mixed regime, we analytically obtain critical adaptation thresholds necessary for epidemic mitigation and in the spatially static regime, we establish bounds for the thresholds using continuum percolation results. Our results indicate that a linear adaptive response to prevalence provides no additional advantage over a constant adaptive fraction in controlling outbreaks, and a highly superlinear response is required to suppress epidemic spread. For a sigmoidal adaptation, we identify conditions under which oscillations in prevalence can emerge, with peak prevalence exhibiting a non-monotonic dependence on the width of the sigmoidal function, suggesting an optimal parameter range that minimizes epidemic severity. We obtain prevalence, final epidemic size, and peak prevalence as functions of adaptation parameters in all adaptation scenarios considered, providing a comprehensive characterization of the effects of spatial adaptation based on global prevalence information in shaping adaptive epidemic dynamics. I Introduction Mathematical modeling of epidemics plays an important role in understanding and predicting the dynamics of infectious diseases Ottar2018 ; WangStati ; Keeling2008 ; Anderson1985 ; Rothman2008 ; schn23 ; Ashby2021 ; Fred2001 . When combined with knowledge from diverse disciplines and real-world data, epidemic models help us to make informed decisions to manage pandemics effectively. By integrating various pharmaceutical and non-pharmaceutical interventions, epidemic models provide quantitative framework for suggesting better control measures and mitigating the impact of outbreaks wong2020 ; Markosocial ; Funk2009 ; Buscarino2008 ; Xinyu2025 . Compartment models and their several variants have long served as the foundation of epidemic modeling Kermack27 ; Vinc1978 ; Peng2019 ; Song2009 ; Khan2018 . These models divide the population into various compartments based on the state of infection of an individual, the typical compartments being susceptible (S), infected (I), recovered (R), exposed (E), etc. The conventional compartmental models assume a well-mixed population and can be easily extended to consider populations with specific contact structures using network-based approaches Turner2020 ; Pastor2015 . A key assumption in conventional compartmental models is that the contact rate between susceptible and infected individuals remains constant over time Kermack27 . However, in most situations, as the prevalence of infection increases within a population, the interaction between susceptible (S) and infected (I) individuals may decrease, reflecting a behavioral adaptation effect. An earlier example of an epidemic model accounting such behavioral changes was by Vincenzo et al Vinc1978 . The effect of complex social responses on epidemic dynamics has since been explored using models incorporating saturating or non-monotonic interaction functions between the susceptible and the infected population Dongmei2007 ; Ruan2003 ; Kar2019 ; Khan2018 . The dynamical behavior of epidemic models with such nonlinear incidence rates has been extensively studied in the past providing a deeper understanding of how human behavior may influence epidemic dynamics (for a review of such models See Funk2010 ). A key driver of adaptive behavior is risk perception, which influences precautionary actions such as social distancing, reduced mobility, and mask usage Moinet2018 ; Hehuang2023 ; Rongcheng2023 ; DanYang2024 ; Kar2019 . Risk perception may be based on local information, such as the infection status of immediate neighbors in a contact network, or on global information, such as population-wide prevalence reported by media or public health agencies. Several studies have modeled risk perception as a probabilistic response function influencing individual behavior Cui2008 ; Liu2007 ; chang2021 ; Paulo2022 ; Arthur2021 ; Vrugt2020 ; Sharma2019 ; Maier2020 ; Nowak2011 ; Caley2008 ; Mgosac21 ; Glaubitz2020 ; Xinyu2025 . Studies show that a sufficiently strong nonlinear increase in risk perception can lead to epidemic extinction in specific network structures Bagnoli2007 ; Herrera2019 , highlighting the importance of adaptive behavior by individuals in shaping epidemic outcomes. In addition to behavioral adaptations, the spatial nature of disease transmission has received increasing attention in recent years Barthelemy2011 ; Akhil24 ; Huang2016 ; loring2020 . Accounting for this spatial effects is particularly relevant for airborne diseases such as COVID-19, where physical distance between individuals and the spatial contact structure (as opposed to the topological contact structures typically assumed in conventional models) plays a critical role in transmission dynamics Gross2020 ; Peng2019 . In the spatial models of epidemic, agents are assumed to occupy positions in a plane and two agents form a potential disease transmitting link between them only when their physical separation is closer than a particular value. Just as in the conventional framework, one may consider a well-mixed spatial population or a static spatial contact network as representing two opposite extremes of agent mobility. The latter typically make use of the language of Random Geometric Graphs and has garnered increased attention in the recent past Saha2023 ; Spatial2020 ; loring2020 . Such graphs have a close association with the continuum percolation models in statistical physics and in the context of epidemic spreading, the percolation threshold can act as a determining factor in deciding whether a disease undergoes large-scale spreading in the population or not Akhil24 ; Mello2021 . In this work, we integrate the two key aspects discussed above-prevalence-based behavioral adaptation and spatial constraints-to analyze epidemic dynamics in a spatial setting. Specifically, we consider a spatial Susceptible-Infected-Recovered (SIR) model with global incidence-based adaptation by a fraction of agents in the population. The assumed spatial nature of the process naturally lends itself to incorporate various non-pharmaceutical interventions, such as reduction in mobility, adherence to social distancing, and the use of protective measures like masks, all involving a spatial scale chang2021 ; Paulo2022 ; Arthur2021 ; Vrugt2020 ; Sharma2019 ; Maier2020 ; Nowak2011 ; Caley2008 ; Mgosac21 ; Glaubitz2020 ; Xinyu2025 . There are a few recent studies which look into various adaptive mechanisms implemented at both the global and individual levels, examining scenarios where adaptation is informed by prevalence information or by local interactions. Oscillatory dynamics in prevalence have been observed in models where adaptation depends on prevalence and feature a well-defined adaptation threshold for containing the epidemic from spreading Jianping2021 ; Glaubitz2020 ; Akhil24 . In this paper, employing a spatial setting, we analyse the impact of three different global-prevalence-based adaption mechanisms employed by the agents on the dynamics of an epidemic. In all cases, we assume that a fraction of the population takes adaptive action based on the global prevalence of the disease. The three scenarios differ in the functional dependence of the adaptive fraction on the global prevalence and cover a wide spectrum of potential behavioral responses. Specifically, the three scenarios correspond to a constant adaptive fraction independent of prevalence, a power-law form and a sigmodial form. In the spatial model, it is assumed that an infected agent has a characteristic transmission radius which is the distance up to which it can potentially pass on the disease. Alternatively, a susceptible