Spatial effect of anti-COVID measures on land surface temperature (LST) in urban areas: A case study of a medium-sized city

Abstract

This case study investigates the magnitude and nature of the spatial effect generated by the anti-COVID measures on land surface temperature (LST) in the city of Târgu Mureș (Marosvásárhely), Romania. The measures were taken by the Romanian government during the state of emergency (March 16 – May 14, 2020) due to the SARS-CoV-2 coronavirus pandemic. The study shows that – contrary to previous studies carried out on cities in China and India – in most of the urban areas of Marosvásárhely LST has increased in the period of health emergency in 2020 concerning the large average of the years 2000–2019. Remote sensing data from the MODIS and the Landsat satellites show, that MODIS data, having a moderate spatial (approximately 1 km) but good temporal resolution (daily measurements), show a temperature increase of +0.78 °C, while Landsat data, having better spatial (30 m) but lower temporal resolution, show an even greater increase, +2.36 °C in the built-up areas. The difference in temperature increase is mainly due to the spatial resolution difference between the two TIR band sensors. The LST anomaly analysis performed with MODIS data also shows a positive anomaly increase of 1 °C. However, despite this increase, with the help of the hotspot-coldspot analysis of the Getis-Ord Gi* statistic we were able to identify 46 significant coldspots that showed a 1–2 °C decrease of LST in April 2020 compared to the average of the previous years in April. Most of these coldspots correspond to factory areas, public transport epicenters, shopping centers, industrial polygons, and non-residential areas. This shows that anti-COVID measures in the medium-sized city of Marosvásárhely had many effects on LST in particular areas that have links to the local economy, trade, and transport. Paired sample t-test for areas identified with LST decrease shows that there is a statistically significant difference in the average LST observed before and after anti-COVID measures were applied. MODIS-based LST is satisfactory for recognizing patterns and trends at large or moderate geographical scales. However, for a hotspot-coldspot analysis of the urban heat islands, it is more suitable to use Landsat data.