maximum amount of exposure information and to develop effective and relevant applications of such information. One important objective should be to describe, reconstruct, and forecast real-world exposures more accurately and more efficiently. To be effective, exposure science needs to adopt a systems-based approach that, to the extent possible, considers exposures from source to dose and from dose to source and considers multiple levels of integration, including time, space, and biologic scales in connection with multiple stressors in human and ecosystem populations.
In the near term, exposure science needs to develop strategies to expand exposure information rapidly to improve understanding of where, when, and how exposures occur and their health significance. Data generated and collected would be used to evaluate and improve models of exposure for use in generating hypotheses and developing policies. New exposure infrastructure (for example, sensor networks, environmental monitoring, activity tracking, and data storage and distribution systems) will help to refine or replace existing measurement and monitoring strategies. This process will help to identify the largest knowledge gaps and reveal where gathering of more exposure information would contribute the most to reducing uncertainty.
In the field of environmental health, substantial investment and progress have been made in recent years to collect and improve access to genomic, toxicology, and health data (for example, Davis et al. 2011; CTD 2012) and to provide information on chemical toxicity and inform and guide research. However, those data have historically lacked the extensive and reliable exposure information required for examining environmental contributions to diseases and assessing health risks. The Environmental Protection Agency (EPA) ExpoCast program, initiated to address that research gap, is intended to advance the characterization of exposures to support toxicity testing (Cohen Hubal et al. 2010a) and in the long term to link exposures to health outcomes. There is still a growing demand to collect more exposure data to populate emerging exposure databases (for example, Gangwal 2011) and to facilitate linkages with toxicity and environmental-fate data and with manufacturers’ production and use data.
An Exposure Infostructure
Exposure data are often scattered among such widely dispersed sources that it is difficult to relate them (Egeghy et al. 2012). Several initiatives in the United States and abroad aim at developing tools to integrate those data sources
1“Data landscape” is a term used in informatic and computational analyses. The term implies stepping back and looking at the data available, identifying data rich and data poor areas, and seeing what the data “landscape” looks like.