Modeling Arsenic Occurrence in Western United States Using Ordinal Logistic Regression
Location
Space Dynamics Laboratory
Event Website
http://water.usu.edu/
Start Date
3-26-2004 11:45 AM
End Date
3-26-2004 12:00 PM
Description
Arsenic is historically known be toxic to human health. Drinking water contaminated with unsafe levels of arsenic may cause cancer of skin, bladder, lung (and possibly several other internal organs, including kidney, liver and prostate), and non-cancer effects, including manifestations that are distinctive and characteristic of chronic arsenic poisoning. The toxicity of arsenic is suggested by the MCLG of zero μg/L and a very low MCL OF 10 μg/L, that has been a subject of constant scrutiny. US Environmental Protection Agency (US EPA), based on the recommendations of National Academy of Sciences has revised the MCL from 1974 value of 50 μg/L to 10 μg/L. The decision was based on a National-level analysis of arsenic concentration data collected by National Analysis of Water Quality Assessment (NAWQA). Another factor that makes arsenic in drinking water a major issue is the widespread occurrence and variety of sources. Arsenic occurs naturally in mineral deposits and is also contributed through anthropogenic sources. Ground water with moderate (10 to 50 micrograms per liter) to high (greater than 50 micrograms per liter) concentrations of arsenic tend to occur throughout much of the Western United States. These high concentrations of arsenic are generally observed in the following geochemical conditions- basin-fill deposits of alluvial-lacustrine origin, particularly in semiarid areas, volcanic deposits, geothermal systems, and uranium and gold-mining areas. Arsenic occurrence in ground water is heavily dependent on various hydrogeochemical variables and land use. The occurrence of arsenic in high concentrations is common in Western United States, especially in the states of Utah, Montana and California and Nevada. In the current work, the dependencies of arsenic occurrence in ground water, on the various influencing factors like land use, aquifer types etc., is analyzed using the technique of ordinal logistic regression. Ordinal Logistic Regression predicts the probabilities for arsenic concentrations to exceed threshold concentration levels like Maximum Contaminant Level (MCL) and detection limits, given a particular geochemical environment and human activity. The model results were extended for estimating the individual risks and costs posed by the arsenic occurrence in the ground water of the western United States.
Modeling Arsenic Occurrence in Western United States Using Ordinal Logistic Regression
Space Dynamics Laboratory
Arsenic is historically known be toxic to human health. Drinking water contaminated with unsafe levels of arsenic may cause cancer of skin, bladder, lung (and possibly several other internal organs, including kidney, liver and prostate), and non-cancer effects, including manifestations that are distinctive and characteristic of chronic arsenic poisoning. The toxicity of arsenic is suggested by the MCLG of zero μg/L and a very low MCL OF 10 μg/L, that has been a subject of constant scrutiny. US Environmental Protection Agency (US EPA), based on the recommendations of National Academy of Sciences has revised the MCL from 1974 value of 50 μg/L to 10 μg/L. The decision was based on a National-level analysis of arsenic concentration data collected by National Analysis of Water Quality Assessment (NAWQA). Another factor that makes arsenic in drinking water a major issue is the widespread occurrence and variety of sources. Arsenic occurs naturally in mineral deposits and is also contributed through anthropogenic sources. Ground water with moderate (10 to 50 micrograms per liter) to high (greater than 50 micrograms per liter) concentrations of arsenic tend to occur throughout much of the Western United States. These high concentrations of arsenic are generally observed in the following geochemical conditions- basin-fill deposits of alluvial-lacustrine origin, particularly in semiarid areas, volcanic deposits, geothermal systems, and uranium and gold-mining areas. Arsenic occurrence in ground water is heavily dependent on various hydrogeochemical variables and land use. The occurrence of arsenic in high concentrations is common in Western United States, especially in the states of Utah, Montana and California and Nevada. In the current work, the dependencies of arsenic occurrence in ground water, on the various influencing factors like land use, aquifer types etc., is analyzed using the technique of ordinal logistic regression. Ordinal Logistic Regression predicts the probabilities for arsenic concentrations to exceed threshold concentration levels like Maximum Contaminant Level (MCL) and detection limits, given a particular geochemical environment and human activity. The model results were extended for estimating the individual risks and costs posed by the arsenic occurrence in the ground water of the western United States.
https://digitalcommons.usu.edu/runoff/2004/AllAbstracts/21