Search

The new SARS-CoV-2 coronavirus has spread rapidly around the world since it was first reported in humans in Wuhan, China, in December 2019 after being contracted from a zoonotic source. This new virus produces the so-called coronavirus 2019 or COVID-19. Although several studies have supported the epidemiological hypothesis that weather patterns may affect the survival and spread of droplet-mediated viral diseases, the most recent have concluded that summer weather may offer partial or no relief of the COVID-19 pandemic to some regions of the world. Some of these studies have considered only meteorological variables, while others have included non-meteorological factors. The statistical and modelling techniques considered in this research line have included correlation analyses, generalized linear models, generalized additive models, differential equations, or spatio-temporal models, among others. In this paper we provide a systematic review of the recent literature on the effects of climate on COVID-19’s global expansion. The review focuses on both the findings and the statistical and modelling techniques used. The disparate findings reported seem to indicate that the estimated impact of hot weather on the transmission risk is not large enough to control the pandemic, although the wide range of statistical and modelling approaches considered may have partly contributed to the inconsistency of the findings. In this regard, we highlight the importance of being aware of the limitations of the different mathematical approaches, the influence of choosing geographical units and the need to analyse COVID-19 data with great caution. The review seems to indicate that governments should remain vigilant and maintain the restrictions in force against the pandemic rather than assume that warm weather and ultraviolet exposure will naturally reduce COVID-19 transmission.

The near-repeat phenomenon usually occurs with any crime. Hence, to implement preventive measures, it is of great interest to figure out at which spatio-temporal scale crimes are more likely to be repeated by offenders. The Knox test is the most used statistical tool for evaluating the presence of the near-repeat phenomenon given a dataset of crimes that are located in space and time. The classic version of this test assumes that crime risk is homogeneous in both space and time, although this assumption rarely holds in reality. Therefore, the main goal of this article is to highlight the necessity of adjusting the standard version of the Knox test, including spatial and temporal effects that allow for the consideration of crime risk heterogeneity. In this regard, a methodology that has already been proposed for addressing this issue is described and adapted. This methodology is then put into practice through a dataset of burglaries recorded in the city of Valencia (Spain) in 2016 and 2017. The results yielded by both versions (standard and adjusted) of the Knox test confirm that the near-repeat phenomenon took place for the burglaries that occurred in Valencia during the period under investigation. However, using the adjusted version of the Knox test leads to a reduction in the number of spatio-temporal intervals that are declared as statistically significant. This fact should be born in mind before making decisions on preventive measures.