Cracking the chemical code: a data-science approach to deciphering the chemical information stored in environmental samples

点击翻译按钮获取中文摘要

Numerous natural and anthropogenic processes release organic chemicals into the environment. The underlying hypothesis of this project is that these processes have distinct chemical markers that can be used to uniquely identify each source. Over time, chemicals are transported through the environment and are often captured in lakes, rivers, and the ocean. These water bodies thus contain a chemical record of all processes occurring upstream. This project will develop tools to test water bodies for these chemical markers, using artificial intelligence (AI) tools to screen for all processes that occur in a watershed. This approach holds great promise to efficiently collect more data than existing methods. Successful development of this AI approach will have wide-ranging applications to detect and identify sources of pollution from local to global scales. Broader impacts to society will result from the training of underrepresented communities and development of STEM curricula for high school students. These will lead to diversifying the STEM workforce and increasing scientific literacy. Lakes and other bodies of water can be considered as systems that store chemical information recorded in the form of tens of thousands of molecules in the water and sediment. The goal of this research project is to use artificial intelligence (AI) algorithms to translate the chemical data stored in environmental samples into knowledge about ecosystem processes. This goal will be achieved through specific objectives to: 1) develop diagnostic chemical fingerprints associated with multiple anthropogenic pollution sources, 2) quantify environmental processes the affect the chemical composition in receiving water bodies, and 3) identify pollution sources through source identification in various water bodies including lakes and groundwater. Although current application of chemical forensics is focused on pollution source tracking, the goal of this research is substantially broader. Hundreds to thousands of ecosystem processes occur across the landscape, and successful completion of the research holds promise to track environmental processes through fingerprint identification within a single water sample.This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.