Perhaps one of the biggest and most dangerous diseases for urban life in the current century is the Covid-19 pandemic. Spatial pattern analysis and risk analysis is a suitable tool in the diagnosis and weather of epidemics and can be understood and help public health diseases. Spatial analysis methods of Geographic Information System (GIS) were used to examine the relationship between the initial statistics of the Covid-19 virus data and the analysis and analysis of spatial patterns in 62 neighborhoods of Khoy city. Therefore, using spatial statistics analysis tools, the spatial distribution pattern was examined and analyzed. The research method is descriptive-analytical; the aim of this study is to use spatial statistics methods to analyze and analyze the use of the Covid-19 spatial virus from the beginning of 2010 to the end of 2023. In order to analyze and analyze spatial patterns, spatial autocorrelation statistics, Moran cluster analysis, global local cluster analysis and identification of hot spots were used. The spatiotemporal pattern showed that the global Moran autocorrelation is highly clustered and less than 1% is likely to be a random cluster pattern. Local Moran autocorrelation statistics for hot and cold spots showed that the northwestern neighborhoods of Khoy city have a high-high clustering (aggregation) feature and the neighborhoods of the western part of Khoy city are hot spots with a 99% confidence level and the southern neighborhoods in the Valiasr township are cold spots for COVID-19 with a 99% confidence level. Comparison of multiscale geographic modeling with other models showed that apart from the performance index with a negative effect; the social and economic, environmental, communication network and physical indicators have a positive effect. Results The degree of dependence index (DOD) of the research data improved with 90% spatial dependence and the R^2 coefficient from 58% to 64% in the multiscale regression model.