Spatio-temporal changes pattern in the hotspot's footprint: a case study of confirmed, recovered and deceased cases of Covid-19 in India

Hotspot detection and the analysis for the hotspot's footprint recently gained more attention due to the pandemic caused by the coronavirus. Different countries face the effect of the virus differently. In India, very little research has been done to find the virus transmission. The paper's main objective is to find changing pattern of the footprint of the hotspot. The confirmed, recovered, and deceased cases of the Covid-19 from April 2020 to Jan 2021 is chosen for the analysis. The study found a sudden change in the hotspot district and a similar change in the footprint from August. Change pattern of the hotspot's footprint will show that October is the most dangerous month for the first wave of the Corona. This type of study is helpful for the health department to understand the behavior of the virus during the pandemic. To find the presence of the clustering pattern in the dataset, we use Global Moran’s I. A value of Global Moran’s I greater than zero shows the clustering in the data set. Dataset is temporal, and for each type of case, the value Global Moran’s I > 0, shows the presence of clustering. Local Moran’s I find the location of cluster i.e., the hotspot. The dataset is granulated at the district level. A district with a high Local Moran’s I surrounded by a high Local Moran’s I value is considered the hotspot. Monte Carlo simulation with 999 simulations is taken to find the statistical significance. So, for the 99% significance level, the p-value is taken as 0.001. A hotspot that satisfies the p-value threshold is considered the statistically significant hotspot. The footprint of the hotspot is found from the coverage of the hotspot. Finally, a change vector is defined that finds the pattern of change in the time series of the hotspot’s footprint. SUPPLEMENTARY INFORMATION: The online version contains supplementary material available at 10.1007/s41324-022-00443-8.