Solving a Spatial Problem of Water
Quality
By: Joseph Mandelko
The tourist town of Algoma Wisconsin
is plagued by poor water quality. The town of 3,200 people is in Kewaunee
County on the Door Peninsula. This particular area of Wisconsin has a bedrock
called the Niagara Escarpment, because of the formation of this escarpment the
water table is in some cases a few feet below the surface and the escarpment is
uniquely porous. These special situations result in a water table that is
easily contaminated, much easier than most other areas in the United States.
Compounding the issue is the fact that Kewaunee County has become the home of
over 17 large dairy farms, or CAFOs. These CAFOs use liquid manure to spread on
their field and if they over water, which usually happens, the liquid manure
runs into the water table and as a result appears in citizens wells.
The goal of this spatial inquiry is
to find a place on public park land, an already established area of meeting for
people in Wisconsin, where citizens can obtain filtered drinking water. I
thought it best to have the potential location be within two miles of the city
center and in an area where most of the population lived. I also wanted the
potential areas to be at least 100 meters away from rivers and roads to prevent
additional runoff and flood water so the area providing filtered water was not
compromised. My intended audience would be the people of Kewaunee County and
the upper Northeast corner of Wisconsin. The water quality issue is not all
that well known outside of Northern Kewaunee and this map shows, for people not
familiar with Algoma, where a filtered water source could be. The city of
Algoma could also use this if they ever decided they wanted to place a filtered
water tank somewhere in its vicinity for citizens to obtain water until the
pollution problem is solved.
In order to answer this question I
had to pull data from a few sources. After initial background information was
gathered I needed to find spatial data. I used USA Census 2010 data for the
roads and population area I wanted to include. I used DNR data from Wisconsin
to show where city parks were located. I could have gotten this data from the
Census website and the DNR data from the parks information on the DNR website.
However since I had access to data already in useable formats in my University
of Wisconsin Eau Claire folder I used the data I had available to me there.
While I feel my end result is accurate for its intended purpose I do have some
concerns with the data. First of all the parks data. The data does not say what
the parks are currently used for. For example, one of the parks is on Lake
Michigan and while most of it is away from the water’s edge it would not be
wise to place a drinking water source in the sand on the shores of a Great
Lake. This is a concern prior knowledge brings up, not a concern that is easily
addressed by the dataset. I also have the concern of the population centers,
the data is five years old and no doubt there has been some population movement
recently since water quality has become a growing concern.
The methods I used to answer my
question are most easily explained through looking at my data flow model
(figure 1) I used to organize my question. However it will make more sense supplemented
by a description as the data in the dataflow model has already been clipped. First I had to assemble my data that I wanted included in the
final project. Since using all of the “detailed” roads, rivers, counties, and
population tracts takes a large amount of processing power I clipped those
classes. After creating a feature class of Kewaunee County by making a layer
from my selected county I was able to clip the other classes by that feature
class. When I had clipped what I needed I had the roads, population tract,
rivers, and parks of Kewaunee County. At that point I was ready to narrow down
my information to the Algoma population tract using the summarize tool.
Eventually I had a feature class intersected into all of my positive values.
From there I could subtract what I didn’t want, those would be the areas within
100 meters of a river or road. I created each of those buffers in separate
feature classes then intersected them together and erased them from the
positive feature class I had made. The result was essentially the final map
(figure 2). I had to use a smaller amount of tools than expected though I ended
up relying heavily on clip, intersect, erase, and buffer tools.
In the end I came up with my answer
which appears in green on the map (figure 2). The green is the park land that
would be acceptable for a public water source to be placed in a park, in
Algoma, and within reach of its residents. The tools I used narrowed down the
initial results. I used the erase too to subtract areas I didn’t want and what
was deemed an acceptable area by my methods was left on the top. There is a
place for filtered public water sources to be placed in Algoma.
Overall I believe the project, in
its simplicity, was a success. There is a lot more data that could be helpful,
some of the issues were explained earlier. I think if I were to do this again
and was able to work with raster data it would be ideal to place a well, not a
tank, as the water source for Algoma. This would require a lot more information
about the areas geology and the layout of the town in order to ensure the new
public well would not be contaminated. If this information was known it could
create a very real solution to a very real problem, not just a quick fix. One
of the issues I also had with this is that there are so many options of things
to do and data to look at that it can be overwhelming so choosing a question to
answer was difficult. I also had trouble using the data flow model builder so
while I had an initial data flow model for my approach to the question I wasn’t
able to run it through the process and ended up using the tools in a capacity
outside of the model builder. It wasn’t an issue and I still answered my
original question but I know the process would have been smoother run through model
builder. There are improvements I would make in my own process if I were to do
it again but it was a good project to do. It is important how to ask and answer
your own questions using GIS, my understanding of data, tools, and methodology
were tested.
(Figure 1)
(Figure 2)
