Meteorological association for prevalence dynamics of Acute Encephalitis Syndrome: a modeling study

• Published in: Modeling earth systems and environment 2022-06, Vol.8 (2), p.2249-2259
• Main Authors: Kumar, Praveen, Parth Sarthi, Pradhan, Bhakuni, Bharat
• Format: Article
• Language: English
• Subjects: Additives; Aetiology; Agricultural practices; Atmospheric models; Bayesian analysis; Chemistry and Earth Sciences; Computer Science; Earth and Environmental Science; Earth Sciences; Earth System Sciences; Ecosystems; Encephalitis; Environment; Environmental factors; Humidity; Land cover; Land use; Math. Appl. in Environmental Science; Mathematical Applications in the Physical Sciences; Mathematical models; Original Article; Physics; Probability theory; Rain; Rainfall; Relative humidity; Skin temperature; Statistics for Engineering; Temperature
• ISSN: 2363-6203; 2363-6211
• DOI: 10.1007/s40808-021-01225-1

Acute Encephalitis Syndrome (AES) has been treated as a ‘mysterious disease’ because of the unknown aetiology in most of the reported cases. AES is assumed to have a linkage with meteorological variables and other environmental parameters in and around the Gorakhpur district of Uttar Pradesh, India. Hence, the current research is aimed to frame a model for AES association with meteorological variables. The Land Use Land Cover (LULC) changes are also considered in the analysis since 78% (approximately) of the land of Gorakhpur is under agriculture practices. The data for AES cases and meteorological variables rainfall (RF, mm), relative humidity (RH, %), and temperature (°C) are considered from Jan 2012 to Dec 2017. The Bayesian Generalized Additive Model (GAM) builds an association between AES and the meteorological variables and predicts AES cases over the study area. Various GAM models for different possible combinations of meteorological covariates are analysed. The least Deviance Information Criterion (DIC) (83.90) is found for the model with covariates-Relative Humidity (RH, %), Skin Temperature (SKT, °C), Two-meter Temperature (T2, °C), the Month of the Year (MoY), and lag-1 of AES cases. The impacts of covariates on AES cases are found non-linear and inconsistent. Results show the positive effect of the covariates in the following ranges; RH in the range of 55–63%, SKT greater than 30 °C, and 2-m temperature greater than 25 °C. Similarly, a positive impact of MoY is observed for July to September. The predictive ability of the model is assessed using an appropriate cross-validation technique. The absolute percentage error is found in the range of 0.5–47.6% monthly. The root mean square errors are also observed within the range of 0.05–2.0.