Discrete Reaction-Diffusion Model Resolves Periodic Epidemic Thresholds and Movement Dynamics

A mathematical investigation published in Advances in Continuous and Discrete Models establishes threshold dynamics and asymptotic behaviors for a discrete-time reaction-diffusion susceptible-infected-susceptible epidemic system operating in a time-periodic and spatially heterogeneous environment. Researchers Ruiqiang Zhuo, Zhiming Guo, and Ruibin Jiang formulated the mathematical framework to capture infectious disease transmission driven by periodic environmental fluctuations and discrete-time observational structures. While classical continuous-time models rely on differential systems and continuous operator semigroups, infectious disease surveillance, intervention deployments, and human contact patterns frequently occur across discrete temporal units such as days or weeks. By framing seasonal epidemic transmission through discrete time steps across a bounded spatial habitat, the study rigorously defines the basic reproduction number and determines how spatial diffusion influences long-term disease persistence and extinction.

The underlying theoretical motivation addresses the seasonal variability commonly observed in respiratory and vector-borne diseases, where transmission rates, host vulnerability, and vector proliferation fluctuate over regular cycles. In natural transmission systems, seasonal factors alter both contact intensity and biological recovery rates, leading to periodic changes in disease pressure across geographical areas. The mathematical formulation considers a population distributed across a one-dimensional habitat bounded by spatial coordinates zero and L, subject to homogeneous Neumann boundary conditions ensuring no population flux crosses the external boundary. Susceptible and infected population densities advance over positive integer time steps, governed by spatial movement, natural survival, demographic recruitment, recovery transitions, and nonlinear disease incidence evaluated through a characteristic contact transmission function.

To rigorously analyze the spatial and temporal interactions of the system, the authors converted the discrete-time reaction-diffusion formulation into an equivalent system of nonautonomous integrodifference equations utilizing explicit Green functions. In this representation, spatial movement across the habitat is governed by specific diffusion kernels that incorporate hyperbolic cosine and sine functions weighted by the square root of the individual diffusion coefficients for susceptible and infected individuals. The model integrates recruitment into the susceptible class alongside local survival rates, recovery proportions, and nonlinear saturation mechanics that reflect transmission prevalence. Through this integral formulation, the periodic epidemic process can be examined as a composite operator mapping continuous spatial population densities across each discrete time cycle.

Prior to addressing disease invasion, the researchers evaluated the demographic baseline of the host population in the absence of infection. They proved that the uninfected system admits a unique positive periodic state that is globally asymptotically stable within the disease-free space. Utilizing the Banach contraction mapping principle, the investigation established that the composite Poincaré operator governing the uninfected population acts as a strict contraction on the positive cone of the underlying continuous function space. Linearization around this periodic trajectory demonstrated that local perturbations decay uniformly over successive transmission cycles, guaranteeing that the host population stabilizes toward a predictable, globally attractive spatiotemporal background upon which potential epidemic introductions can be analytically assessed.

To determine the threshold criteria governing disease outbreaks, the authors defined the basic reproduction number by decoupling the linearized infection subsystem at the disease-free state. Unlike continuous-time periodic systems that derive the reproduction number from differential operators and exponential decay semigroups, the discrete formulation constructs a next generation operator from two distinct components within a complete temporal period: an infection operator representing newly generated infections and an evolution operator accounting for the movement, survival, and non-recovery transitions of infected hosts. The spectral radius of this composite next generation operator was formally designated as the basic reproduction number, capturing the total expected secondary cases generated by an infected individual over the lifespan of an infection across seasonal variations and spatial movement.

The study further linked the basic reproduction number to an associated weighted spectral eigenvalue problem, proving that the reproduction number is strictly reciprocal to the unique positive principal eigenvalue of the system. Through an adjoint operator formulation and Krein-Rutman spectral theory, the authors demonstrated that the basic reproduction number shares an exact threshold correspondence with the principal Floquet multiplier governing linearized perturbation growth. Specifically, the basic reproduction number is strictly less than one if and only if the principal Floquet eigenvalue is less than one, equals one when the eigenvalue equals one, and strictly exceeds unity when the eigenvalue is greater than unity. This spectral linkage ensures analytical coherence between operator-theoretic invasion metrics and direct stability evaluations.

A central theoretical finding of the investigation concerns the asymptotic behavior of the basic reproduction number and the principal eigenvalue under extreme dispersal regimes of infected individuals. When the diffusion rate of infected hosts increases toward infinity, the spatial dispersal kernel converges uniformly to a flat spatial distribution across the domain length L. Under this infinite diffusion limit, spatial structure averages out, causing the basic reproduction number to approach an explicit constant composed of the integrated spatiotemporal averages of infection and survival-evolution operators over a full periodic cycle. Spatial mixing under rapid dispersal effectively homogenizes the epidemic risk across the entire domain into an aggregate global balance.

Conversely, when the diffusion rate of infected individuals tends toward zero, the spatial mobility of infected hosts ceases, confining disease transmission to localized sites across the domain. Under this vanishing diffusion limit, the researchers demonstrated that the overall basic reproduction number approaches the spatial maximum of the localized reproduction number across all individual points in the habitat. Because the global reproduction threshold converges to the highest local transmission risk in the absence of mobility, restricting the diffusion of infected hosts in environments containing high-risk hotspots does not prevent disease persistence, as isolated pockets of intense transmission can independently sustain and amplify the pathogen.

The authors established complete threshold dynamics for the discrete-time reaction-diffusion system based entirely on the magnitude of the basic reproduction number. By constructing a specialized discrete-time Lyapunov functional paired with the positive adjoint eigenfunction of the linearized system, the study proved that whenever the basic reproduction number is less than or equal to one, the disease-free periodic solution is globally asymptotically stable throughout the entire state space. Under this condition, infected population densities decay to zero regardless of their initial spatial distribution, demonstrating that disease eradication is guaranteed if the aggregate reproduction threshold does not exceed unity.

When the basic reproduction number strictly exceeds unity, the disease-free periodic state becomes unstable, and the system exhibits uniform persistence across the habitat. Applying abstract dynamical systems persistence theory and Morse decompositions to the discrete Poincaré map, the authors proved that the infected and susceptible populations remain bounded away from zero by a strictly positive constant for all sufficiently large time steps. Furthermore, uniform persistence guarantees the existence of at least one positive endemic periodic state. Notably, the authors highlighted that these discrete-time threshold and asymptotic outcomes maintain dynamic consistency with existing continuous-time reaction-diffusion models, bridging discrete surveillance modeling with classical continuous epidemic theory while outlining open questions regarding differing dispersal kernels and discrete bifurcation phenomena.

The structural formulation of the system relies on specific analytical hypotheses regarding demographic parameters and contact transmission. Across the discrete periodic framework, the individual survival rate and the non-recovery rate of infected hosts are constrained strictly between zero and one, sharing an identical temporal period with the positive recruitment function of susceptible individuals. Disease transmission incorporates a prevalence parameter and a continuously differentiable contact function that equals one at zero, possesses a strictly negative first derivative, and exhibits a non-negative second derivative. This contact function identifies the proportion of susceptible hosts who remain uninfected following contact with infectious individuals, while its complement captures the proportion transitioning into the infected class. Furthermore, the authors proved that all solutions to the system remain uniformly bounded for all large time steps by a finite constant determined by domain length, recruitment maxima, and survival ceilings, ensuring that population densities cannot expand indefinitely irrespective of initial conditions.

These periodic mathematical relationships capture concrete biological patterns observed across distinct pathogen families. In respiratory infections such as influenza and COVID-19, disease transmission rates typically surge during winter and spring when colder temperatures and indoor crowding facilitate contagion, while recovery rates may decline due to strained medical resources. In contrast, higher temperatures and increased outdoor activities during summer and autumn lower transmission intensity while allowing recovery rates to rebound. Similarly, vector-borne pathogens including malaria and dengue fever exhibit regular seasonal cycles driven by mosquito reproduction, which expands rapidly during rainy periods and contracts during dry intervals.

To isolate the specific effect of spatial variation, the authors analyzed a spatially homogeneous benchmark where survival and non-recovery rates vary over time but remain independent of spatial position. In this setting, the discrete-time reaction-diffusion model yields a basic reproduction number identical to that of purely temporal periodic difference equations, because spatial dispersal operators integrate across the habitat to unity and the next generation operator becomes self-adjoint. The study also highlighted key analytical differences from autonomous systems. While autonomous models with time-invariant coefficients allow variational characterizations to directly establish how the principal eigenvalue shifts monotonically with the diffusion rate of infected individuals, time-periodic nonautonomous equations preclude variational representations, requiring the authors to develop custom operator approximations and spectral bounding techniques.

Subject of Research: Discrete-time reaction-diffusion epidemic modeling in periodic environments

Article Title: Dynamics of a discrete-time reaction-diffusion SIS epidemic model in a time-periodic environment

Article References: Zhuo, R., Guo, Z., & Jiang, R. (2026). Dynamics of a discrete-time reaction-diffusion SIS epidemic model in a time-periodic environment. Advances in Continuous and Discrete Models, 2026(1), Article 64. https://doi.org/10.1186/s13662-026-04100-4

Image Credits: AI Generated

DOI: 10.1186/s13662-026-04100-4

Keywords: reaction-diffusion model, discrete-time epidemic model, basic reproduction number, time-periodic environment, spatial heterogeneity, SIS model, integrodifference equations, spectral radius, next generation operator, threshold dynamics, uniform persistence, Lyapunov functional, disease transmission, dispersal kernel, mathematical biology, infection dynamics