Great Lakes Air Deposition Program
Great Lakes Information Network
Great Lakes Air Deposition Program Sponsored Projects
Return to Project List
Project Title: Monte Carlo Based Multi-Media Fate Model
for the Great Lakes Ecosystem
Chemicals Studied: The model is designed to be used
with a wide range of persistent organic compounds. Such compounds
well-known high priority chemicals such as polychlorinated biphenyls
PCBs), toxaphene, hexachlorobenzene (HCB), dichlorodiphenyltrichloroethane
(DDT), as well as chemicals that have become of increasing concern
in recent years, such as the polybrominated diphenyl ethers (PBDEs).
Given the Monte Carlo method that is used to handle uncertainty,
this model is well suited for application to organic chemicals
for which there is incomplete knowledge of chemical properties and
Geographic Areas: Lakes Huron, Ontario, Michigan, Erie, and
Project Start Date: April 2006
Project Status: Completed
Synopsis: This project has produced an online, interactive modeling application for predicting the fate of chemicals within the Great Lakes environment. This easy-to-use application allows users to determine several important factors relating to how chemicals behave once they have been released into the Great Lakes environment. These include:
(i) Where in the environment does the chemical tend to accumulate (such as in water, soils, sediments, air or fish)?
(ii) How long is the chemical retained in the Great Lakes environment before it degrades or is transported to another region?
(iii) What concentration in the environment would be produced by a given level of emission?
(iv) What is the potential for a chemical to travel to the Great Lakes region from far away?
By providing this capability through an online interface that is easy to access and use, the model gives users a powerful tool for answering questions about the behavior of a given chemical in the Great Lakes system. The modeling program may be run with as little input as the chemical's Chemical Abstract Service (CAS) identification number (of which 33 million are available) although more detailed characteristics may be used as well. The model's ease of use and availability over the Internet make it a great tool for use in educational settings.
While much information can be generated by the model without specifying chemical quantities, knowing how much of a chemical is emitted into the environment allows the user to obtain even more information, such as predicted concentrations in air, water, soils, sediments and fish. It also allows the user to determine the likelihood that a certain concentration known to be a high-risk level would be exceeded. The users can also enter their own emissions information or the website also includes links to pesticide application data from the CropLife Foundation and air emissions data from the Great Lakes Regional Toxic Air Emissions Inventory, which together provide useful release information on more than 350 toxic substances.
The standard version of the model provides the easiest option , or for less advanced users there is also a version that allows the use of Monte Carlo statistical techniques to assess the uncertainty in the models predictions. This option allows users to enter a range of possible values for important variables, rather than entering a single value. The model then computes the likelihood of possible outcomes given that range of inputs. For example, rather than simply predicting that a given chemical's concentration in water would be 2 grams per liter, the results might say it is 90% likely that the concentration would be between 0.5 and 5 grams per liter, with a most like value of 2. Examples are given on the model's website to assist users in choosing a range of values.
The model is configured to represent each of the five Great Lakes basins (Huron, Ontario, Michigan, Erie, and Superior). The user is able to choose which of these five they would like to use as the basis for the model. The program then automatically loads the proper environmental characteristics for that lake basin (such as area and depth of surface water). By modeling with the same inputs but different lake basins selected, comparisons can be made as to how chemicals might behave differently in the different basins. Although the model allows the selection of individual lake basins, the geography within each basin model is not specific enough to allow release points to be defined. Rather, the air and water of each basin are modeled as a single, well-mixed, box.
Potential applications include investigating the behavior of new or unstudied chemicals; comparing emissions and monitoring data; and calculating expected concentrations of chemicals in the environment or the likelihood that such concentrations will exceed known risk values.
Project results have been presented during the 2010-11 GLAD webinar series.
Mario J. Citra, Ph.D.
Syracuse Research Corporation, Environmental Science Center
The Model Website: http://glad.syrres.com/
Supporting Information: Model
Documentation (0.6MB PDF)