GIS Aided Characterization of Groundwater Arsenic Contamination in The United States
Term Project, GIS in Water Resources, Fall 2000
Prepared by: Hazim Tugun
Project Outline:
1.5 Health effects and exposure to arsenic
1.6 Regulatory status of arsenic
3. Mapping of data, findings and evaluations using ArcView GIS
3.1 Acquiring of geospatial groundwater As data and creating a map of groundwater As concentrations
3.2 Mapping the groundwater arsenic concentrations as an interpolated grid surface
3.4: Identification of public use groundwater sources that exceed the current EPA limit of 50 ppb
Arsenic has been the mysterious and inevitable killer of many people in many detective stories and movies. Indeed, arsenic has been a very deadly poison at high concentrations. It is also mysterious in the sense that people are exposed to it most of the time, fortunately at much lower concentrations than the deadly levels, and usually without knowing. As Karl Vogel said in 1928: "It is uncanny thought that this lurking poison (arsenic) is everywhere about us, ready to gain unsuspected entrance to our bodies from the food we eat, the water we drink and the air we breathe." So, what is this intriguing poison?
Arsenic, As, is a gray, brittle, semi-metal element that is odorless and tasteless. It has four oxidation states, -3, 0, 3 and 5. The oxidized states, As(III) and As(V), are known as Arsenite and Arsenate, respectively. Arsenic naturally exists throughout the earth's crust, often as Arsenic Sulfide and Metal Arsenates, or Arsenites. This form of arsenic is inorganic. When arsenic combines with carbon and hydrogen, as is the case in plants and animals, it becomes organic, which is usually much less harmful than the inorganic arsenic (2).
Arsenic has been used in a wide variety of disciplines, ranging from medicine to manufacturing. It has been used as a cure for diseases such as syphilis or leukemia (1). On the other hand, it has been used as an alloying agent in the manufacturing of semiconductors, lasers, transistors, and in the processing of wood preservatives, ammunition, paper, pigments and metal adhesives (2).
The main sources of arsenic in the environment are either natural or man-made. The natural sources include naturally existing minerals/ores (such as pyrites), soils which usually are sinks for the weathered form of arsenic compounds, and mineral-rich geothermal waters. The man-made sources are usually the industrial effluents, which may be direct discharge of arsenic compounds into soil, water, air (which results in atmospheric deposition of arsenic). Such industries include copper smelters (See smelter on Ruston Point, Tacoma, Superfund Site), pesticide and wood preservative manufacturers, among many others.
Geochemistry of arsenic determines the fate and transport of arsenic in the environment. Arsenic as an element is insoluble in water. The oxidized forms, or compounded forms, are usually more soluble in water (2). Two types of processes mainly control the fate and transport of arsenic in the environment: (i) Adsorption and desorption; (ii) Solid-phase precipitation and dissolution (3).
In groundwater, arsenate and arsenite are the two common forms of arsenic. Arsenate is present under oxidizing conditions whereas arsenite is present under more reducing conditions. In the natural pH range of groundwater, arsenate exists with a negative charge whereas arsenite is neutral. Therefore, the adsorption reactions between a variety of aquifer contents, such as iron oxides, are stronger for arsenate than arsenite. However, as the pH of the water increases (i.e. water becomes more alkaline), desorption of arsenate as well as arsenite from materials like iron oxides increase. Since adsorption/desorption is usually pH dependent, changes in groundwater pH can result in changes in groundwater arsenic concentrations due to increase/decrease in adsorption/desorption. An observed phenomenon, which fits well to the theory above, is as follows: Solid-water reactions in the ground usually consume H+ ions in the water, which tends to increase the pH of the water with increasing residence times along the water pathways. Since the prevalent iron-oxide surfaces in the ground can hold a large amount of adsorbed arsenate, the increased pH of the water slowly induces the desorption of arsenate from iron-oxide surfaces, thus easily increasing the arsenic concentration in groundwater above regulatory limits (3). In addition to variations in pH, redox reactions play an important role in the sorption processes. Since arsenate is more strongly adsorbed than arsenite, reduction of arsenate to arsenite may mean an increase in groundwater arsenic concentration. A reverse-process example may be the oxidation of arsenite to arsenate (less soluble form) where the groundwater is excessively withdrawn causing the subsoil to be aerated and approach oxidizing conditions with the introduction of oxygen from the atmosphere.
On the other hand, precipitation and dissolution processes play their role in groundwater arsenic contamination. As a solid-phase compound containing arsenic dissolves into or precipitates out of groundwater, arsenic also dissolves into or precipitates out of the groundwater together with the containing phase. Examples of such solid phases are iron oxides and sulfide minerals. Iron oxide dissolves under reducing conditions and precipitates under oxidizing conditions. The case is reversed for iron sulfide although it is less stable than iron oxide (3).
1.5 Health effects and exposure to arsenic
There are various pathways for the exposure of a person to arsenic. The main exposure pathway is usually through the ingestion of food containing arsenic at very low levels. Such food items include fish or vegetables, or any other member of the food chain feeding on arsenic containing pray. In other words, this pathway can be called the ingestion of bioaccumulated arsenic. Other pathways are breathing sawdust or smoke from burning wood containing arsenic, and ingesting/inhaling/contacting contaminated water, soil or air at waste sites or near areas naturally high in arsenic content (4).
The health effects from exposure to inorganic arsenic vary with the conditions of exposure, such as the duration and dose of exposure. The health effects can be categorized as acute (short-term) or chronic (long-term). In case of acute health effects, arsenic can be deadly at high concentrations. Ingestion of ~60 ppm arsenic in water or food can be fatal. Lower levels of exposure can lead to nausea, vomiting, diarrhea, abnormal heart rhythm, blood vessel damage and a "tingling" sensation in hands and feet.
Long term exposure to low levels may or may not be harmful yet may as well be painful and deadly after prolonged exposures. The level of arsenic exposure from daily intake of arsenic containing food (at miniscule levels) usually does not harm, and in some cases have been argued to be beneficial. However, prolonged exposure to higher levels (at least greater than 50-150 ppb) may cause skin pigmentations, keratoses and skin cancers. Other cancers, such as lung and bladder cancers, have been observed among people who drank from arsenic contaminated wells in Taiwan in 1966. In United Kingdom, where medicinal arsenic (aka Fowler's solution) was given to patients, it has been shown that prolonged exposure to low levels has caused bladder cancer among the patients treated with Fowler's solution. To be more specific, in terms of duration of exposure, after several years of low level arsenic exposure, skin lesions start to appear. Hyper and hypopigmentations and keratoses of the hands and feet follow. After 10 years or so, skin cancers are expected. Other cancers such as lung, kidney, liver, bladder cancers, are expected after exposure to 500 ppb of arsenic (1).
Arsenic contamination is a concern in many areas of the globe. The biggest and worst example of groundwater arsenic contamination is in Bangladesh. The victims of arsenic poisoning from groundwater contaminated with arsenic can be seen here.
1.6 Regulatory status of arsenic
In response to the ubiquitous presence and dangerous health effects from arsenic exposure, regulatory agencies mandated maximum arsenic limits in environmental media, specifically water. At first, these limits were set to protect people from acute effects. The current arsenic limits mandated by the United States, United Kingdom, and Bangladesh in drinking water is 50 ppb. World Health Organization used to recommend the same limit, yet lowered it to 10 ppb after discoveries of the long-term health effects as a result of prolonged exposure to low levels of arsenic. For the same reason, the USEPA has proposed to lower the current drinking water standard of 50 ppb to 5 ppb (1).
The main objective of this project is to characterize groundwater contamination in the United States using the capabilities of ArcView, GIS. More detailed, specific objectives are:
To create a map that will show the geospatial distribution of groundwater arsenic concentrations in the United States.
To use this distribution map as a base to characterize, if possible:
The extent of contamination relative to the current EPA limit of 50 ppb throughout the US and at regions where groundwater is harvested for public use.
The possible sources of contamination.
The distribution of observed health effects relative to the distribution of arsenic contamination.
3. Mapping of data/findings and evaluations using ArcView GIS:
3.1 Acquiring of geospatial groundwater As data and creating a map of groundwater As concentrations
Groundwater arsenic concentration data was downloaded from USGS NAWQA (National Water Quality Assessment) Program database. This original data file consisted of concentrations of arsenic in 18,850 samples from potable groundwater throughout the United States, covering a time period from 1973 to 1997. The file also had geographic locations of each of the groundwater sampling points in lat/long (degrees, minutes, and seconds), the water use associated with the sampled groundwater source, the well depth and the sampling date/time. The downloaded data file was in TAB-delimited ASCII format (Fig.1).
Fig.1: Original arsenic data in TAB-delimited ASCII format
To be able to display it in ArcView, the file was first opened in Excel and edited for data format conversion errors (e.g. mixed up columns). Then, the data associated with Alaska, Hawaii and Puerto Rico were deleted to form the study area as the United States except Alaska, Hawaii and Puerto Rico. The next main process in Excel was to convert the lat/long (degrees, minutes and seconds) into decimal degrees (Fig.2). Before opening the file in ArcView, all the data columns were edited for proper format to be accurately read by ArcView and the file was saved as a ".dbf" file.
Fig.2: Arsenic data edited and saved in Excel as a ".dbf" file.
Fig.3: Arsenic data converted to an attribute table in ArcView.
In ArcView, a new project was created and the prepared arsenic data was added as a new table into the project. To display the geospatial arsenic concentrations in a View, the table was added into the view as an event theme, using the View Menu/Add Event Theme command. X and Y fields were chosen as the longitude and latitude data, respectively, from the arsenic table (Fig.3). Thus, the groundwater arsenic concentrations were properly displayed in ArcView. To finish the process of creating the base map of the US with groundwater arsenic concentrations, a map of geographic US (retrieved from previous class exercises) was added. The resulting map was projected by projecting the View using Albers, Equal Area Projection of the United States (Fig.4).
Fig.4: Resulting map of US with groundwater arsenic concentration point theme in Albers Equal Area-Conic projection.
As can be seen from the map in Fig. 4, most of the groundwater samples exceeding the EPA limit of 50 ppb are in the western, north-western part of the US, with a few others scattered in the north and north-east.
3.2 Mapping the groundwater arsenic concentrations as an interpolated grid surface
The groundwater concentrations were also mapped as an interpolated grid surface to create a visual representation of the concentration gradients throughout the country. The interpolation was performed on the groundwater arsenic concentration point theme. In ArcView, Surface/Interpolate Grid command was used while the View window and the groundwater arsenic concentration theme ware active. The cell size and the extent of the grids to be formed were specified in the command window that followed (Fig.5). The default cell size given by ArcView was chosen. For the Z Value Field, which would be used to determine cell value, concentration was chosen.
Fig.5: Process showing the steps involved in creating an interpolated grid theme after adding in the GeoProcessing Wizard in ArcView.
The resulting grid theme covered the extent of the whole United States in a rectangle form. Thus, the theme extended beyond the boundaries of the US at points further in from the outer most boundaries such as that of Florida and Texas. For this reason, clipping of the grid theme based on the United States theme was tried using the GeoProcessing Wizard of ArcView. However, the process was not possible since ArcView did not accept clipping of a grid theme. Therefore, the theme was screen-captured and pasted into paint where it was edited to exclude the areas of the grid theme extending beyond United States boundaries.
The interpolated grid map was adjusted to represent hot spots above the current 50 ppb limit (Fig.6) and the proposed 5 ppb limit (Fig.7) for arsenic. In both pictures, the hot spots are shown in the darkest color.
Fig.6: Interpolated grid surface, showing hot spots with arsenic concentrations greater than the current EPA limit of 50 ppb.
Fig.7: Interpolated grid surface, showing hot spots with arsenic concentrations greater than the proposed EPA limit of 5 ppb.
The map in Fig.6 again shows the hot spots to be mostly in the western half of the United States and a few scattered throughout the eastern and northern parts. The region with the largest hot spot is the western part of Nevada towards its California border. This region was expected to have such a bigger hot spot, since the point theme map in Fig.4 shows a densely populated number of samples exceeding the 50 ppb limit. The hot spots in Fig.7 are much more populated especially in the western half again. Almost all of Nevada and Arizona seem to exceed the 5 ppb limit. While Nebraska, Illinois and Indiana were below the 50 ppb limit in Fig.6, it can seen that a large fraction of these states exceed the proposed 5 ppb limit in Fig.7. There are also more regions in the northern and eastern parts that exceed the proposed 5 ppb limit.
In this step, locations of groundwater sources used for domestic, public supply and recreational uses were identified. See the original arsenic data file for detailed descriptions of these three water uses. These locations were important to identify since people are more prone to be exposed to arsenic via the listed three water uses in these areas than in other locations via other water uses (e.g. cooling). To identify these locations, a query was built for the arsenic concentration attribute table, asking ArcView to select those locations where the water use was either of the three water uses (Fig.8).
Fig.8: Query built to identify groundwater sources used for domestic, public supply and recreational purposes.
The selected points from the built query were converted to a shapefile using the Theme/Convert to Shapefile command in ArcView. Then, the newly created shapefile was displayed on the United States map as a point theme (Fig.9). The black points on the map in Fig.9 are the sampled groundwater sources which are used for domestic, public supply and recreational uses (from this point on, these groundwater sources will be called public use groundwater sources). Using Field/Statistics command in ArcView, the number of groundwater samples or the total number of records for the original arsenic concentration theme was found (Total number of records = 18,571). Likewise, the number was found for the newly created point theme displaying groundwater sources with the three listed water uses (Total number of records = 7152). The ratio of the latter number to the former shows that about 40% of the sampled groundwater sources are for public use.
Fig.9: A map identifying the public use groundwater sources.
3.4: Identification of public use groundwater sources that exceed the current EPA limit of 50 ppb
After identifying the groundwater sources that may subject people to arsenic exposure more than the others (Fig.9), it would be interesting and informational to also identify among these groundwater sources the ones that exceed the current 50 ppb limit. To do this step, a query was built on the point theme seen in Fig.9 which asked ArcView to select the groundwater sources that have arsenic concentrations greater or equal to 50 ppb. The selected points were converted into a shapefile for easier access and use for subsequent sections of the project. The results are shown in the following map:
Fig.10: Public use groundwater sources that exceed 50 ppb.
Fig.10 indicates the public use groundwater sources that exceed 50 ppb and also rank them in three categories with increasing concentration. The public use groundwater sources that exceed 50 ppb follow a similar geographic pattern as those of groundwater sources exceeding 50 ppb in previous maps in Fig.4 and Fig.6. Using Field/Statistics as before, 98 out of the 7152 (1.4%) public use groundwater sources exceeded the EPA limit of 50 ppb. Although this percentage may seem low, the number of states with public use groundwater sources exceeding 50 ppb is 14 out of the 59 states. This may be an indication of how widespread the groundwater arsenic contamination is throughout the United States.
As can be seen in Fig.10, some of the detected arsenic concentrations even exceeded 500 ppb. These regions may have such high arsenic concentrations probably due to extreme favorable natural conditions for arsenic or waste sites contaminated with high levels of industrial arsenic. A closer look at the locations of the maximum arsenic concentrations recorded can indicate which states and counties lead the nation in terms of groundwater arsenic contamination. To do this, a county polygon theme from National Atlas of the United States was added to the existing View, and Theme/Select by Theme command was used to intersect the counties with the public use groundwater sources exceeding 50 ppb (Fig.11). The selected counties were again converted into a shapefile for easier use and access. Using the new shapefile, the first five counties having the highest individual arsenic concentration were identified manually (Fig.12).
Fig.11: Counties with public use groundwater sources having arsenic concentrations greater than 50 ppb.
Fig.12: A chart showing the first five counties with highest measured public use groundwater arsenic concentration.
The chart in Fig.12 shows that Churchill County in Nevada has the highest public use groundwater arsenic concentration with 1400 ppb, about 30 times greater than the EPA limit of 50 ppb. The fifth highest arsenic concentration, about 200 ppb is at San Bernardino in California. It is also noteworthy that 4 out of the 5 shown counties are from the western part of the United States, maybe indicating that western half of the United States beats the other half not only in magnitude of arsenic concentration but also in the frequency of occurence of high arsenic concentrations (i.e. in the number of states/counties).
As explained in the Background, the health effects of exposure to arsenic are known and document in some cases (e.g. in Bangladesh or Taiwan). A common and big category of health effects reported for long-term exposure to arsenic is cancer. These cancers usually start out with skin cancer and continue with internal cancers, such as bladder, lung and kidney cancers upon extended exposure. Since it is known that there is an increase in risk of contracting cancer from exposure to arsenic, it might be possible to see a correlation between the distribution of reported cancer cases and groundwater arsenic concentrations in the United States. To determine whether there is a possible correlation, cancer mortality rate dataset from the National Atlas of the United States was downloaded. This dataset consisted of a polygon theme, detailing various cancer mortality rates (#/100,000 people) for each county in the United States from 1970 to 1994. The reported cancers ranged from breast cancer to oral cavity cancer. The dataset was added to the View, creating a new cancer mortality rate theme.
To start with, cancer mortality rates for all cancers was looked at. Two maps showing the different ranges of cancer mortality rates for all cancers in males and females were created and displayed with public use groundwater sources having arsenic concentrations greater than 50 ppb (Fig.13 and Fig.14).
Fig.13: All cancers male mortality rates with public groundwater supplies greater than 50 ppb.
Fig.14: All cancers female mortality rates with public groundwater supplies greater than 50 ppb.
The map in Fig.13 seem to clearly show that all cancer male mortality rates are higher for the eastern parts of the United States than for the western parts. This general pattern does not fit well with that of public use groundwater sources contaminated with high levels of arsenic. As was shown before, higher concentrations of arsenic in groundwater tend to be larger in the western part of the United States, as opposed to the increased cancer mortality rate in the eastern part of the United States. On the other hand, the map in Fig.14 does not reveal a very particular pattern for the all cancers female mortality rates. A slight pattern of increased death rates on the west coast and in the northeastern regions can be observed. A simple select by theme and query analysis shows that 30 out of the 35 counties having public use groundwater sources containing arsenic concentrations above 50 ppb are among the counties with highest death rates (137 - 278 / 100,000 people). It is hard to derive any conclusions from this observation because the increased cancer rate distribution do not match well with that of the increased arsenic concentrations. Thus, it might be a coincidence that those counties happen to have the highest range of all cancers female mortality rates.
The possible correlation between the distributions of arsenic concentrations and male/female bladder cancer rates was also investigated. Again, two maps showing the male and female bladder cancer rates were created using the cancer database (Fig.15 and 16).
Fig.15: Bladder cancer male mortality rates with public groundwater supplies greater than 50 ppb.
Fig.16: Bladder cancer female mortality rates with public groundwater supplies greater than 50 ppb.
Both maps in Fig. 15 and Fig.16 show that there is not an observable correlation between bladder cancer mortality rates and increased arsenic concentrations in public use groundwaters. There is a wide scatter of bladder cancer mortality rates throughout the United States. A possible explanation for not seeing a pattern may be the lack of specific cancer data pertaining to specific causes such as ingestion of contaminants (in this case, arsenic) in drinking water. The cancer cases presented in these maps may have been caused by other causes than arsenic in groundwater, such as smoking or inhalation of high levels of volatile organic chemicals in the workplace. In other words, the effects of exposure to arsenic, such as contracting cancer, can easily be wiped out or shadowed by other environmental or personal factors (e.g. What would be the effect of long term exposure to low levels of arsenic for a business man who smokes, commutes in heavy traffic for hours and gets his clothes dry-cleaned everyday? He may experience adverse health effects due to arsenic, but most probably those effects will be miniscule compared to those experienced due to his smoking, inhalation of exhaust gases during traffic and tetrachloroethylene from his clothes everyday! It should be noted that smoking is the number one lung cancer cause in the United States, and tetrachlorethylene is a known carcinogen). Therefore, there can be thousands of factors in a country like United States that will cause much more emphasized adverse health effects before those from exposure to low levels of arsenic can be distinctly identified. For this reason, detailed and controlled epidemiological studies of arsenic exposure may be performed to obtain data which may be useful in assessing how different levels of arsenic exposure affect people's health adversely, especially in the United States (See Arsenic Health Effects Research Program at University of California, Berkeley).
To look at natural sources of arsenic, US Geological Data was downloaded from USGS mineral resources spatial database. The database consisted of bedrock data for the conterminous US. However, it did not contain mineral information. The downloaded data was displayed in ArcView together with public use groundwater sources exceeding 50 ppb. Then, Theme/select by theme command was used to select the regions of the data (i.e. polygons) that correspond to arsenic concentrations greater than 50 ppb. 99 different types of bedrock were selected. A map was created to show the first six bedrock types with the highest frequency to have arsenic concentrations greater than 50 ppb (Fig.17).
Fig.17: A map showing different bedrock types with occurrences of arsenic concentrations greater than 50 ppb.
A chart showing the frequency ranking of the six bedrock types was also made (Fig.18).
Fig.18: A chart showing the frequency of different bedrock types with groundwater arsenic concentrations greater than 50 ppb.
From the map in Fig.17 and chart in Fig.18, it can be seen that the quaternary rocks seem to dominate the regions with arsenic concentrations greater than 50 ppb than the other 5 bedrock types do. The second in ranking was Pliocene continental bedrock type.
The groundwater arsenic concentrations had a large distribution ranging from 0 to 1400 ppb throughout the US. The mappings of the groundwater arsenic concentrations in the US showed that groundwater arsenic concentrations tend to be higher in the western part of the US than the other parts. Similarly and expectedly, there was a significantly higher number of groundwater samples exceeding the current EPA limit of 50 ppb in the western half of the country than the other parts. Specifically, Nevada appeared to be the leading state in terms of the extent of contamination. There were a few other regions in the north and east which exceeded the 50 ppb limit. In the case the proposed 5 ppb limit would be in effect, a much larger fraction of the country's groundwater sources exceeded the limit. While almost all of Nevada and Arizona exceeded the limit, Illinois, Indiana and Nebraska, which did not exceed the 50 ppb limit, also exceeded the proposed 5 ppb limit.
A little less than half of all the analyzed groundwater sources were used for domestic, public supply or recreational purposed. These water uses were looked at since they are the water uses from which people would more likely be exposed to arsenic in the groundwater. Of these public supply groundwater sources, about 1.5% exceeded the current 50 ppb limit. The significance of this number cannot be easily judged in terms of considering possible health concerns associated with the use of that particular class of groundwater. The reason for this may be that information related to the level of treatment of the contaminated groundwater at those regions and other environmental factors that may expose people to arsenic is not available. However, it gives an idea of what percent of the country's groundwater supplies would need to be evaluated for further use because of arsenic contamination and subsequently treated for public use. Certainly, this percentage will increase with the proposed 5 ppb limit in effect. Despite of the low percentage obtained, the number of states with public use groundwater sources exceeding the current limit was 14 out of 50. This would also indicate the wide geographic distribution of groundwater sources exceeding 50 ppb in a large country like US. In a sense, these GIS aided mappings and evaluations of the extent of arsenic contamination may help the regulators have a rough estimate of the extent of water treatment to be planned and related costs to be considered in the case of the 5 ppb limit. As was done here, more specific regions having public use groundwater sources with the highest arsenic contamination could be found. It was found that Churchill county in Nevada lead the nation with a maximum measured arsenic concentration of 1400 ppb, followed by Washington county in Idaho.
The obtained health data was not useful in characterizing the distribution of observed adverse health effects relative to that of groundwater arsenic concentration. Neither the all cancer mortality rates nor the bladder cancer mortality rates showed a pattern that was close to that of the distribution of groundwater arsenic concentrations. This is because the health effects that would be observed due to exposure to arsenic in groundwater only are probably wiped out by the effects of other environmental or personal factors such as smoking. In other words, the adverse health effects of long term exposure to very low levels of arsenic may not be as significant as other adverse health effects due to other factors, and thus may not be easily distinguished from those other health effects. In order to be able to distinguish adverse health effects of arsenic from those of other factors, very controlled studies should be conducted and thus very cause-specific adverse health effect information should be collected, as it is now being done in some universities in the US. GIS could have been much more helpful in the health effect aspect of this project if such detailed, cause-specific data were available.
Finally, possible sources of arsenic were investigated. The mapping of the different bedrock types in the US showed that quaternary bedrock type dominated the regions with groundwater arsenic concentrations greater than 50 ppb. Mineralogical data related to these bedrock types or underlying rocks in these areas could help verify which minerals were the largest natural contributors of arsenic to groundwater in the US. Therefore, the sources section of this project can be developed further by incorporating mineralogy data. In addition, data related to industries discharging arsenic into the environmental media could help identify which regions of the US were more severely affected from the man-made sources of arsenic than the natural sources. Such data could be collected and compiled into tables from sources such as EPA's Toxic Release Inventory System, and be incorporated into the mappings of groundwater arsenic contamination.
1. The Arsenic Website Project.
4. Arsenic: Agency for Toxic Substances and Disease Registry
