Abstract: The study took downtown Guangzhou as the study area to observe the spatio-temporal spreading features of Dengue Fever (DF) at micro-scale. We used survey analysis, kernel density analysis and standard deviation ellipse method to explore the spatio-temporal diffusion features. Then, we tried to combine the cross correlation analysis method with geographical detector to analyze the effect of temperature, humidity, atmospheric pressure, different land-use types on dengue fever spread. The result showed that: Firstly, fever transmission had stage character significantly. The infection cases occurred predominantly in the spaces with poor environment and high population density, and spread rapidly. Secondly, the importation cases played a major role in the early stages of DF development. Thirdly, the old town, where there were relatively poor environment, large quantity of elderly population, complicated personnel, and convenient traffic, was a high-risk area of DF outbreak. Fourthly, the influence of temperature, humidity and atmospheric pressure on dengue fever spread was lagging. Temperature and humidity had positive correlationship with the spread of dengue fever, while atmospheric pressure had negative correlationship with the spread. Finally, the distribution of local residents had the closest association with dengue fever spread, and the interactive places among pond, farmland, grassland and public green space could increase the dengue fever infection risk.

Key words: Dengue Fevers, spatio-temporal diffusion characteristics, meteorological factors, land-use type, cross-correlation analysis, geographical detector, Guangzhou