Document Sample
Arsenic Powered By Docstoc
					Arsenic in the Water

   by   Claude Davila

          Advisors: Kate Calder
                    Noel Cressie
                    Hong Fei Li
   A team of four SSES (Spatial Statistics and Environmental Sciences )
    researchers from OSU among them Catherine Calder and Noel
    Cressie and three researchers from Battelle Memorial Institute was
    awarded a three-year (2004-2007) contract funded by the American
    Chemistry Council's (ACC) Long-Range Research Initiative for the
    research project titled "From Sources to Biomarkers: A Hierarchical
    Bayesian Approach for Human Exposure Modeling".

   The objective of this research is to characterize multi-pollutant
    (arsenic, lead, cadmium, and chromium) human exposures by
    linking sources to biomarkers using a multi-scale hierarchical
    Bayesian statistical model that describes how multi-media pathways
    contribute to direct routes of exposure (inhalation, ingestion,

   This is under the hypothesis that by incorporating several different
    sources of data that inform about pollutant pathways, the model will
    discern patterns in human exposures and allow more informed
    conclusions to be drawn about current and future distributions of
              Introduction: Arsenic
    Natural Sources
   Arsenic occurs naturally in rocks and soil, water, air, and plants and
    animals. It can be further released into the environment through
    natural activities such as volcanic action, erosion of rocks and forest
    fires, or through human actions.

    Anthropogenic Sources
   High arsenic levels can also come from certain fertilizers and animal
    feeding operations. Industry practices such as copper smelting,
    mining and coal burning also contribute to arsenic in our
   Large number of arsenic-containing chemical compounds are
    present in the terrestrial environment: they include the inorganic
    species arsenate, or As (V), and arsenite, or As (III), in addition to
    organic derivatives.

   The inorganic species are the more toxic ones.

   The arsenic species present in groundwater and surface water are
    largely arsenate and arsenite.
Groundwater and Surface water
                                    When it rains or if water is
   Surface water is the             poured on the ground it will
    water in the oceans,             travel downward because of
    streams, lakes and rivers.       gravity.
                                    It gets filtered by the soil but at
                                     the same time it collects certain
   Groundwater is the               minerals or metals such as
    underground water that           arsenic.
    gets tapped when a well         It may be released as a spring,
    gets dug for drinking            drain into a lake or river, travel
    water.                           to the ocean underground or
                                     get pumped up from a well into
                                     somebody’s kitchen sink.
How does Arsenic get into the ground water,
    surface water and drinking water?

   Drinking water can come from either ground water sources (via
    wells) or surface water sources (such as rivers, lakes, and streams).
    Nationally, most water systems use a ground water source (80%),
    but most people (66%) are served by a water system that uses
    surface water.

   Ground water that is in contact with the rocks with high
    concentration in arsenic can have high levels of arsenic.
    Chemistry behind Arsenic getting
            into the Water
   Arsenic needs to be in a soluble form to end up
    in drinking water.
   If there is a reduction of As (V) to As (III), As
    (III) will be more mobile hence it will end up in
   If a certain form of arsenic is present in rock
    (perhaps as waste rock from a mining site) it
    can get attached to Oxygen and turn into the
    mobile form As (III) and end up in the water.
                 Health Effects
Short term exposure (high concentration in
  short amount of time)
   Vomiting
   Throat and stomach pain
   Bloody diarrhea
Long term exposure (low concentration over
  long period of time)
   Circulatory problems (trouble with blood vessels and
   High blood pressure
   Cancer
                Details of this study
   The object of this REU project is      Data that was provided
    to use geostatistics to analyze
    data from public water systems in
    Arizona.                                -Id: Id number of the public
                                             water system
   The data was obtained from
    Water Quality Division, Arizona         -Sys.source: source of the
    Department of Environmental              drinking water
    Quality.                                -x: longitude
   It originally included the              -y: latitude
    information of the PWS, the             -arsenic: log concentration of
    arsenic concentration of each            arsenic
    PWS and the counties served.
   Since exact locations of each
    PWS has not been provided the
    location of each PWS from the
    county that it serves has been
    randomly generated.
   Histograms
    summarize the
    distribution of data
   Horizontal axis is the
    log concentration of
    arsenic in drinking
   Vertical axis is for the
    frequency of counts in
    each bin
Exploratory Analysis: Map of Arizona

                         This is a map of the
                          data locations.
                         There are circles of
                          different sizes indicating
                          different levels of
                          arsenic concentration.
                         The different colors of
                          the circles indicate four
                          different source types:
                          GWP-groundwater primary
                          SW-surface water
                          SWP-surface water primary
Plot of Sys.Source v Arsenic
           Box Plot
 Exploratory Analysis: Contour Plot
Contour Plot displays a 3 dimensional surface on a 2 dimensional plot.
      Background: Geostatistics
   A collection of statistical methods which
    are used in the geosciences
   Used for estimating/predicting the value of
    continuous spatial process at unobserved
    locations given observations at known
   Kriging is a method used to perform
    spatial prediction (to be explained later
            Variogram Analysis
    Consists of an experimental variogram (eg. Empirical variogram)
     calculated from the data and variogram model fitted to the data.

    Spatial dependence - measurements at points close together are
     more similar than those further apart.

    Variogram tells you whether data exhibit spatial dependence.

    Nugget effect is when a variogram, as distance goes to 0, does
     not approach zero variance. The amount by which the variance
     differs from zero is known as the nugget effect.
    Variogram Analysis: Variogram
   The variogram is defined by:

    2 (h)  E{[ Z (u )  Z (u  h)]} }  2

     - 2 (h) Stands for the variogram at h
     - h stands for the distance= ui-uj
     - u (location vector) stands for all possible locations
     - Z stands for the data
     - E stands for mean
    Variogram Analysis: Semi-Variogram
   The function that is used in kriging is the
    semi-variogram which is the variogram
    equation divided by 2.
          (h)  1 / 2 E{[ Z (u )  Z (u  h)]} }

   Kriging : uses the information from variogram
    to find an optimal set of weights that are used
    in estimating a surface at unsampled
Variogram Analysis: Exploratory Plots
    Empirical Variograms- nonparametric estimators
     of the variogram of a spatial process
     2 (h)  1 / | N (h) | N ( h ) [Z (ui )  Z (u ]}2 }
         |N(h)| stands for the unique pairs of locations in N(h)
         N(h) = {(ui, uj): ui- uj=h; i, j=1,……,n}

     Robust Empirical Variogram is used when outliers
      are present and when data is scattered.
                | N(h) |
                         N (h) | Z (u i )  Z (u j ) |1/ 2 ]4
     2 (h) 
                     0.457  (0.494 / | N (h) |)
      Variogram Analysis: Variogram
            Exponential Model
   Exponential Variogram Model: when combined with
    nugget effect it is used for an experimental variogram
    that levels out but has curve all the way up.
Variogram Analysis: Empirical

        Spatial Prediction: Kriging
   Kriging : uses the information from a variogram to find an
    optimal set of weights that are used in estimating a surface at
    unsampled locations.

   Kriging is named after D.G. Krige, a South African mining
    engineer and a pioneer in the application of statistical
    techniques to mining investigations.
Grid of Data and Prediction Locations

 (Left) Grid of the study area and data locations. (Right) Grid of
 prediction locations.
   Groundwater seems to be a common water
    source because most of the data collected is
    from groundwater because there is not much
    surface water in Arizona.
   Higher levels of arsenic concentration are in the
   According to the kriging estimates, log arsenic
    levels of approx.-5 are quite abundant
    throughout Arizona.
   EPA Guideline is 0.010 ppm (this includes
    both organic and inorganic arsenic)
   Calder A., Catherine and Cressie, Noel (2006) Kriging and Variogram
    Models. Ohio: Ohio State University




Shared By: