Rich dynamics of a partially degenerate reaction-advection-diffusion mosquito-borne disease model with multiple transmission routes, latent periods and human behavior change

The spatiotemporal dynamics of mosquito-borne diseases are driven by host movement, vector dispersal, and environmental heterogeneity, while human behavioral responses often considerably modulate intervention effectiveness. To quantitatively capture these factors, we propose a degenerated reaction-advection-diffusion model that unifies multi-route host infections, host-vector transmission, and individual movement, with host psychological factors embedded as an endogenous component of control, yielding a framework more reflective of realistic epidemic settings. A generalized incidence describes direct contact, enhancing model adaptability. By refined estimates and inductive method, we establish the well-posedness of solutions and the existence of a global attractor. Following next-generation operator, the definition of basic reproduction number\, (BRN) R₀ is given, along with its explicit expression. Threshold analysis shows that for R₀< 1, the disease-free steady state\, (DFSS)\, is globally asymptotically stable\, (GAS); for R₀> 1, the disease becomes endemic and the model has at least one positive steady state\, (PSS). In particular, in a spatially homogeneous environment without advection, this PSS is globally attractive. Furthermore, with no advection, we derive the GAS of DFSS when R₀ = 1, as well as the asymptotic profiles of PSS as susceptible diffusion rates tend to 0 or infinity. Numerical simulations indicate that a higher advection coefficient greatly reduces the early-stage epidemic size, while human behavior effectively lowers the outbreak peak, especially in high-risk areas during the early phase. Finally, using the Markov chain Monte Carlo\, (MCMC) method, the model successfully fits Zika virus data from Brazil, demonstrating its potential for practical application.