A Spatio-Temporal Analysis of COVID-19 Spread in Türkiye

This study investigates the regional and temporal dynamics of Coronavirus Disease 2019 transmission in Türkiye by integrating downscaled environmental covariates with urban footprint analysis. The aim is to explore how spatial and climatic factors influence virus spread, particularly within urban contexts. Multi-temporal satellite imagery and remote sensing techniques were used to derive 15-day interval datasets between April and August 2020. Environmental and climatic variables (e.g., temperature, humidity, wind, carbon monoxide, particulate matter, ammonia, Normalized Difference Vegetation Index) were statistically downscaled from coarse-resolution data (e.g., European Centre for Medium-Range Weather Forecasts (ECMWF) Reanalysis v5, Copernicus Atmosphere Monitoring Service (CAMS)) using ensemble machine learning models, including Random Forest, Generalized Additive Model, and Multivariate Adaptive Regression Splines. Urban footprints were extracted from Landsat 8 imagery via Random Forest classification, achieving 90% accuracy. These downscaled covariates were spatially matched with the urban footprint to represent conditions within densely built environments. The Geographically and Temporally Weighted Regression model was implemented at the (Nomenclature des Unites Territoriales Statistiques, or Nomenclature of Territorial Units for Statistics) NUTS 1 regional level to identify the primary predictors of the spread of the novel coronavirus, SARS-CoV-2, also known as “COVID-19.” A comprehensive analysis revealed that eight variables, namely Normalized Difference Vegetation Index (NDVI), carbon monoxide, wind speed, humidity–CO, humidity–NH3, temperature–O3, particulate matter, and land surface temperature, exert a substantial influence on transmission rates. The Normalized Difference Vegetation Index demonstrated a consistent negative correlation, indicating that urban areas with higher levels of green space may serve as a barrier against the spread of viruses. Regional variations are influenced by physical, ecological, and climatic characteristics, which in turn modulate infection dynamics. The results of the study underscore the significance of adapting public health measures to the urban configurations, green infrastructure, and ventilation potential of each locale. The study underscores the relevance of integrating urban planning, remote sensing, and geospatial modeling in managing future outbreaks. Comprehensive understanding, which has been demonstrated in the various technical approaches examined in this study, is crucial for the development of effective urban resilience to combat future pandemics.

Cite this article as: Görmüş, S., Cengiz, S., & Yüzbaşı, B. (2026) A spatio-temporal analysis of COVID-19 spread in Türkiye. Forestist, 76(1), 0071, doi:10.5152/forestist.2026.25071.