Translating Big Data into Smart Data for Veterinary Epidemiology

VanderWaal, Kimberly; Morrison, Robert B.; Neuhauser, Claudia; Vilalta, Carles; Perez, Andres M.

Substantial spatial heterogeneities exist in the occurrence of infectious diseases, and management and analysis of large spatial datasets represents another facet of voluminous data (34) . Numerous spatially explicitly datasets exist for environmental and climatic factors [e.g., Ref. (35) ], land cover and use [e.g., Ref. (36, 37) ], distributions of at-risk, reservoir, and vector populations [e.g., Ref. (37) (38) (39) ], and satellite imagery and remote sensing products (40). In addition, the increasing use of GPS tracking devices creates a rich source of data on the movement of people, vehicles, and animals that can be used to dynamically represent exposure and transmission dynamics (41) . These data can be combined with geo-referenced disease data [e.g., Ref. (42) ] to identify environmental correlates of disease through ecological niche modeling, thus contributing to our ability to understand and map a pathogen's geographic distribution (9, (43) (44) (45) . By utilizing near real-time updates in environmental data and locations of new cases, risk maps can become evolving rather than static representations of risk (43, 46) . Remote sensing, in particular, could be re-framed as a high velocity source of data, as many satellite-based data are updated at regular intervals (43) . Ultimately, the ability to predict the occurrence of pathogens through space and time will allow for more effective targeted surveillance and control.