Time-series modelling of dengue incidence in the Mekong Delta region of Viet Nam using remote sensing data

Abstract

Objective: This study aims to enhance the capacity of dengue prediction by investigating the relationship of dengue incidence with climate and environmental factors in the Mekong Delta region (MDR) of Viet Nam by using remote sensing data. Methods: To produce monthly data sets for each province, we extracted and aggregated precipitation data from the Global Satellite Mapping of Precipitation project and land surface temperatures and normalized difference vegetation indexes from the Moderate Resolution Imaging Spectroradiometer satellite observations. Monthly data sets from 2000 to 2016 were used to construct autoregressive integrated moving average (ARIMA) models to predict dengue incidence for 12 provinces across the study region. Results: The final models were able to predict dengue incidence from January to December 2016 that concurred with the observation that dengue epidemics occur mostly in rainy seasons. As a result, the obtained model presents a good fit at a regional level with the correlation value of 0.65 between predicted and reported dengue cases; nevertheless, its performance declines at the subregional scale. Conclusion: We demonstrated the use of remote sensing data in time-series to develop a model of dengue incidence in the MDR of Viet Nam. Results indicated that this approach could be an effective method to predict regional dengue incidence and its trends.

Keywords

• dengue
• time series modelling
• remote sensing data
• mekong river delta