Performance assessment of interpolation techniques for investigation Covid-19 spread in Türkiye

Authors : Duygu Arıcan, Nursu Tunalıoğlu
Pages : 42-57
Doi:10.9733/JGG.2025R0004.E
View : 75 | Download : 71
Publication Date : 2025-04-30
Article Type : Research Paper
Abstract :Throughout history, viruses have posed significant threats to human life and health. In the context of the historical pandemics, Covid-19, rapidly spread across continents and was declared a pandemic by the World Health Organization on 11 March 2020. The first case in Türkiye was detected on the same date. Understanding the spatial distribution of the Covid-19 is crucial for effective public health planning and intervention. Geographic Information Systems (GIS) technology can be leveraged as a visualization aid to map the geographical distribution of the disease, the potential risk factors, and the resources available for treatment and prevention. To effectively map and analyze the spatial distributions, and local/global dynamics of the Covid-19 virus, various GIS-based interpolation methods were employed. To understand these dynamics, this study presents a detailed spatial analysis using interpolation methods to evaluate spatiotemporal changes on seasonal levels in the Covid-19 pandemic in Türkiye. Seasons investigated in a 1-year period were determined as follows: Spring, from 20 March 2021 to 18 June 2021; Summer, from 19 June 2021 to 17 September 2021; Autumn, from 18 September 2021 to 17 December 2021; and Winter, 18 December 2021 to 18 March 2022. Seasonal case distribution maps produced from city-level and district-level seasonal case data utilizing Inverse Distance Weighting (IDW), Radial Basis Function, Spline interpolation, and Empirical Bayesian Kriging (EBK) interpolation methods. Finally, the spread of Covid-19 in Türkiye was investigated on the seasonal scale, and interpolation results were assessed by standard deviation, mean absolute error, and root mean square error. The results of this study demonstrated that the period of highest incidence of cases of Covid-19 in Türkiye was winter. Overall, when considering error metrics, EBK and IDW generally proved to be the most reliable methods across different scales and conditions. In contrast, Spline interpolation’s tendency to overfit the data made it less suitable for these datasets.
Keywords : Covid, Pandemi, Veri görselleştirme, Mekânsal enterpolasyon, Performans değerlendirmesi

ORIGINAL ARTICLE URL