Real-Time Disease Forecasting using Climatic Factors: Supervised Analytical Methodology

Abstract

Weather being an uncontrollable factor, often has direct effects on human mortality rates, physical health, mental injury and other health outcomes. Extreme climate incidences and gradual changes in weather are making us more vulnerable to disease outbreaks. In general, there are three ways by which variation in climate affects such diseases: by affecting the virus, the vector or host and spread of a disease. According to 1996 World health organization (WHO) report, 30 new diseases have come into existence in the past 20 years. Additionally, there has been a re-emergence and redistribution of various arthropod-borne diseases such as dengue, malaria etc. on a global scale. Events like rainfall, humidity, temperature etc. have well-defined role in the transference cycle. Any changes in these events can lead to increase in incidence of these diseases.The worldwide pandemic about abroviral diseases demands the need for developing early warning system (EWS) for infectious diseases by considering climate change. Past studies incorporates only the historical weather statistics into account. However because of increasing uncertainty and climate variability, the traditional systems in this context are getting outstripped. In this paper, we’ll propose our methodology for predicting number of dengue cases that are likely to occur in real time on the basis of five-day weather forecast. Our analysis is applicable globally and enables comprehensive scenarios of daily disease outbreaks to be explored using real-time weather API, preparing society against any health related risks arising due to variability in climate.