Keywords
TMDL, Natural Language Processing, TRS tool, Water quality modeling
Start Date
27-6-2018 2:00 PM
End Date
27-6-2018 3:20 PM
Abstract
In the United States, once a waterbody fails to meet its designated-use it is listed as impaired in the so-called 303(d)-list established by the Clean Water Act. A total maximum daily load (TMDL) is developed to remediate the water body and restore the designated use. Development of TMDL, a daily allocation of allowable pollution diet, typically involves water quality modeling to assess loads from various sources for the impairment (e.g., Nutirents, Sediments, Dissolved Oxygen). For each TMDL developed, a report is sent the United States Environmental Protection Agency (USEPA) for approval. We have developed a method based on natural language processing to analyze over 27,000 TMDL reports available form the USEPA representing over 79,000 unique TMDL developed. The analysis of the reports, from 1986 to 2017, show how water quality modeling applied for the TMDL development has evolved, the implication of policy on TMDL development, and adoption of modern techniques (e.g., remote sensing) in TMDL development. An interactive diagram that shows users the relationships between modeling techniques and impairments to assist in choosing examples reports—the TMDL Report Selection tool (TRS tool)—is available at the URL https://occviz.com/tmdl. In this presentation, we will discuss the method used to develop this tool and some analysis results.
TRS Tool—Using data mining and natural language processing to assess the state of TMDL development
In the United States, once a waterbody fails to meet its designated-use it is listed as impaired in the so-called 303(d)-list established by the Clean Water Act. A total maximum daily load (TMDL) is developed to remediate the water body and restore the designated use. Development of TMDL, a daily allocation of allowable pollution diet, typically involves water quality modeling to assess loads from various sources for the impairment (e.g., Nutirents, Sediments, Dissolved Oxygen). For each TMDL developed, a report is sent the United States Environmental Protection Agency (USEPA) for approval. We have developed a method based on natural language processing to analyze over 27,000 TMDL reports available form the USEPA representing over 79,000 unique TMDL developed. The analysis of the reports, from 1986 to 2017, show how water quality modeling applied for the TMDL development has evolved, the implication of policy on TMDL development, and adoption of modern techniques (e.g., remote sensing) in TMDL development. An interactive diagram that shows users the relationships between modeling techniques and impairments to assist in choosing examples reports—the TMDL Report Selection tool (TRS tool)—is available at the URL https://occviz.com/tmdl. In this presentation, we will discuss the method used to develop this tool and some analysis results.
Stream and Session
B3: Sixth Session on Data Mining as a Tool for Environmental Scientists