Development of two-dimension epidemic prediction model

Abstract

Epidemic prediction is a crucial foundation of disease control policy-making. Owing to the high population connectivity of current epidemics, it is essential to capture the spatial transmission of infectious diseases. However, most models currently used in epidemic prediction are single-point models, and they can only capture the time-dynamic increase of cases in limited areas. In this study, we develop a two-dimension epidemic prediction model by introducing diffusion processes, which take spatial transmission epidemics into account. We utilize mathematical theorems to prove a well-posed solution of the model. In addition, we also consider various influencing factors that affect the spread of epidemics, and introduce multiple parameterization schemes. Results suggest that this two-dimension model provides more precise predict the spatial and temporal distribution of confirmed cases. The regional average prediction score of COVID-19 in July 2022 in Lanzhou is 76.5 % and COVID-19 from May 1st to May 31st, 2023 in China is 70.7 %,respectively. Our results offer a scientific foundation for further study on the prediction of spatial epidemics, which contributes to an in-depth understanding of epidemic dynamics and provides valuable reference for the formulation of public health strategies and policies.