Introduction:
I
want to know the ideal location to build a house in Grand County,
Colorado. I want the location to be
within the jurisdiction of local fire and police departments, and to be between
three to ten miles from the nearest town. I don’t want to be in a subdivision,
because I would like my home to be a place of solitude and serenity. I can’t
build on protected land, so I can’t build in the national forest, or any other
protected areas. Ideally, the location would be within 10 miles of Winter Park,
because I love skiing at the resort.
Last
year, I attended a school in Grand County, and fell in love with the area’s
beauty. It is my belief that my fellow classmates also would also like to live
in the area, and may wish this information to help them decide where to live. I
could also potentially use this information in the future to convince a spouse
to move to the area.
Data
Sources:
In
order to answer this question I needed: a dataset showing the parcels, data
showing police and fire department jurisdictions, data showing the locations
and limits of towns, all data regarding protected areas, as well as data
showing subdivision locations. I obtained all necessary data from the Grand
County GIS department’s website. The data is all scaled for the county, and has
been in a consistent state of being updated. All of the data is in a projected
coordinate system that is appropriate for the mapped area. I’m concerned that I
couldn’t find dates regarding when the information was last updated.
Methods:
 |
| The original data-flow model Figure 6 |
Initially,
I executed a spatial join to link the ‘parcels‘ feature class with ‘zoning’ so
each parcel would show how it was zoned (Figure 6). Next, I intersected the ‘parcels_spatialjoin_1’
with the feature class showing the fire and police departments’ jurisdiction; so
only the parcels within areas they cover are shown. I wanted the parcels to be
within 10 miles of a town, so I made a ten-mile buffer of the ‘towns’ feature
class, dissolved it, and intersected it with my ‘intersect_parcels_FD’ feature
class. I didn’t want to be too close to a town, so I made a buffer with a
radius of three miles of the ‘towns’ feature class. After dissolving it, I
executed an erase, eliminating all remnants of the prior intersects within the
three-mile buffer.
Next, I executed a union of five different feature classes:
one showing Federal and State owned land, Rocky Mountain National Park, the
Arapahoe National Recreation Area, US Forestry Service land, and all Wilderness
areas. I named the resulting feature class ‘protected_areas’. After the union
was completed, I erased all of the land within ‘protected_areas’ from the
‘parcels_within10_notwithin3mi’ feature class, and I named the resulting
feature class ‘parcels_notprotectedland’. I don’t necessarily want to live in a
subdivision, so I erased all of the parcels that fell within the ‘subdivisions’
from ‘parcels_notprotectedland’ and named the resulting feature class
‘notprotected_nosubdivision’. Next, I performed a buffer of the railroads
feature class, dissolved the buffer, then erased all of the land within three
miles of a railroad from the ‘notprotectedland_nosubdivision’, and made
‘ideal_awayfromrailroad’.
I thought I wanted to live near Fraser, since that’s
where the school was, so selected just the polygons in the ‘towns’ feature class
whose names equaled ‘Fraser’ and performed a buffer on those polygons to show all land within 7 miles of Fraser. After dissolving the buffer, I
intersected the resulting feature class with ‘ideal_awayfromrailroad’, giving
me the final feature class, which I subsequently named ‘Optimal_locations’.
Unfortunately,
as an unexpected side affect of using the intersect tool, all of the resulting parcels
were severed into smaller chunks. I couldn’t accurately see how many parcels
truly fit all of the ideal characteristics, so I created a new model to
preserve the parcels’ dimensions.
 |
| The revised data-flow model Figure 7 |
For
the second model (Figure 7), I again started with a spatial join between ‘zoning’ and
‘parcels’ which gave me the feature class ‘parzone_parzone_join’. I still
wanted to see just the parcels within the police and fire departments’
jurisdiction, however, this time I used a select by location to make a
temporary selection of just the parcels in ‘parcels_zoning_join (2)’ that are
within ‘FirePD’, rather than intersecting the two. I also wanted to again
select only parcels within ten miles of a town, but not within three miles, so
I executed and dissolved two buffers: one with a ten mile radius from ‘towns’
and one with a three mile radius from ‘towns’. When both dissolves were executed
and dissolved, I erased the three-mile buffer from the ten-mile buffer zone,
giving the feature class ‘Town_Buffer_ring’. I then used this new feature class
and the Select Layer by Location tool to select only parcels from the current
selection that are completely within ‘Town_Buffer_ring’. Next, I again executed
a union of the feature classes showing federally managed land, however, this
time I included ‘ConservationAreas’, one I accidentally left out in my prior
model. I named the resulting feature class ‘protected_areas’. After the union
was completed, I used the Select Layer by Location tool, first to remove any
parcels that fell within ‘protected_areas’, then to remove any that fell within
‘Subdivisions’. After the selections were finalized, I used the Make Feature
Layer tool to make the resulting parcels into a feature class, named
‘parcels_zoning_join_Layer1’. After
reviewing my prior model, I realized that being within ten miles of Winter Park
was significantly more important than being within seven miles of Fraser, as
they share many of the same views except for the resort. So, I made and
dissolved a ten-mile buffer of Winter Park and named it ‘Winter_Park_Buffer’.
After adding the layer to the map and realizing that none of the parcels were
near the border of the new buffer, I used an intersect tool to show only the
parcels within both ‘parcels_zoning_join_Layer1’ and ‘Winter_Park_Buffer’. This
intersect gave me my final feature class, ‘Ideal_parcels’.
Results:
After
running my parcel finding model, I was able to narrow the number of potential
locations to sixteen out of the thirty-two thousand parcels in the county. I was
unable to make any actual decision since I can’t view the parcels personally,
but I am relatively certain that all of the parcels are beautiful, as they are
all surrounded by protected land and in a mountainous part of the county. In order to display my findings, I used four different maps. The map on the top left shows three of the different qualifications the parcels had to meet, and I used primary colors and slightly transparent layers to show the locations where both the Fire & Police Department (Red) and Ideal Distance from Towns (Yellow) qualifications were met as orange. I then added the Protected Land layer in white, so that only the areas that aren't on protected land would be visible. The map on the top right shows the parcels that met all of the qualifications of the previous map, then displayed the final qualification and the parcels that fell within the qualifying area. I used the lower two maps to put the parcels in context with both the lay of the land and their location in the county and their locations in Colorado.
 |
| The resulting map Figure 8 |
Evaluation:
I
thoroughly enjoyed this project, for it’s open expectations allowed me to
really look into something that interested me, rather than restricting me to a
specific part of the world. If asked to repeat the project, I would also use
both the DEM and a geology layer to create different feature classes showing
just areas with very steep slopes and rock types that are easily tunneled, as
my dream house would be hewn from a mountainside. I would also use the NLCD to
create a layer showing only areas of Aspen forest, as they are the only
pristine forests left after the county’s pine forests were ravaged by an
invasive species of beetle three years ago. I would also join the parcel layer
to the parcel metadata, so I could show the property values of the parcels and
possibly use that as another, more practical way to eliminate potential land
parcels. The main problems I faced with this project stemmed from the opulence
of data I obtained. I had so many possible ways to select parcels that it took
much deliberation in order to determine in which ways to make the selections.
This project truly helped me bring together all of the different concepts and
tools I learned in this class.
Sources:
Grand County GIS Department. (2014). Digital Data Sets [Data file]. Available from http://co.grand.co.us/170/Digital-Data-Sets
ESRI
