Lab Five: Mini-Final Project

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

Lab 4: Vector Analysis with ArcGIS

Figure 4

Goal:
The goal of this lab was to make a map of suitable bear habitats in Marquette County, Michigan, by using several geoprocessing tools.

Background:
Marquette County, in the upper peninsula of Michigan, is one of the few places in the midwestern region of the United States of America to have a remaining black bear population. Having a bear population can lead to some problems, especially when bears interact with humans. Bears can get into garbage, and once accustomed to human food will often forgo their natural diets for one entirely comprised of human food. This leads to bears breaking into houses and becoming nuisances. Once this happens, it is impossible to save the bear, and it must be killed.

In order to preserve the native bear population, the Michigan DNR wants to determine what determines the bear's ideal habitat, and how much of this habitat lies within DNR managed land. They must ensure the bears stay away from human interaction at all costs, so the bear management areas are to be at least 5 kilometers away from any urban areas, for mutual benefit.

Methods:
In order to create this map, I first learned how to map X Y coordinates from an excel spreadsheet. I added the coordinates as an "event theme", then exported them as a feature class once they were mapped. Then I spatially joined the bear locations and the and cover datasets in order to determine in which land cover types the bears were found. I then summarized the resulting dataset in order to determine which land cover types bears were preferred.

I learned how to use a buffer tool, and used it to create a feature class showing the areas within 500 meters of streams. I then used the dissolve tool to simplify this streams feature class by removing the internal boundaries.

After making a feature class of the three land cover types bears preferred, I used the intersect tool to show just the parts of these land cover types that were also within 500 meters of a stream. I then used the dissolve feature to remove the internal boundaries of the newly formed feature class.


I then intersected the new feature class with one showing DNR management zones in Marquette county, so they can determine which parts of the management zones are also good bear habitat. Next, I dissolved the internal boundaries to simplify the appearance of the dataset.

In order to account for the urban areas, I made a feature class of the land cover areas that were labeled 'Urban', then created a 5km buffer around them. After dissolving the internal boundaries of the buffer, I used the erase tool to eliminate the parts of the DNR management zones that fell within 5km of urban areas.

I then made an cartographically pleasing map of my results, showing both the DNR managed rural bear habitat and all the areas of suitable bear habitat on one map, and the land cover types of the region on the other. I also included a locator map, so those viewing the maps can determine the location of the study areas within Marquette county and the state of Michigan. (Figure 4) I also made a data flow model to show the processes I used in order to obtain my final datasets. (Figure 5)

Results:
The maps show that bears were commonly found within 500 meters of streams, but were also found near lakes, reservoirs and wetland areas. If I were to remake the map, I would add a 500 meter buffer to those land cover types, rather than just one around the streams in the area. The maps also show that although there is a large expanse of ideal bear habitat within the study area, only a small quantity of it falls within DNR managed areas far from urban areas.

Figure 5

Sources: USGS NLCD, Michigan DNR, Michigan Center for Geographic Information

Lab 3: Introduction to GPS

Introduction: The goal of this lab is to learn how to collect features with GPS and learn how to import the collected features into ArcMap in order to map them. I was to create a database, prepare it, then load it onto a Trimble Juno. After learning how to use the Trimble, I was to use it to collect point, line, and polygon features around campus. Finally, I was to import my data into ArcMap, and make a map showing my data.

Methods: First, I made a Geodatabase, which I subsequently named Lab 3. In this Geodatabase, I made six feature classes: point, line, polygon, as well as separate copies of these feature classes to use for practice. All of these feature classes were made using the NAD_1983_HARN_Wisconsin_TM meters coordinate system. I then imported a shape file of the campus buildings to my geo-database and a raster image of campus, so that I could use them as a point of reference when collecting data. I imported them using the import feature class single, and import rasterdataset tools respectively. I then changed the symbology of my feature classes so I would be able to quickly tell the difference between my two series of data when in the field.

Next, I learned how to prepare the geo-database to be added to the Trimble Juno. In order to do this I enabled the ArcPad data manager extension, with which I deployed all of the layers I made into a folder. Then I cut the folder and pasted it on the storage card of the Juno.

 After ensuring that the data was deployed properly I the Juno out to the campus commons and practiced using the Juno by capturing points, polygons, and lines into the practice feature classes. After I felt familiar with the program, I began collecting data points, polygons, and lines for the real feature classes. I used point averaging in order to collect points for light poles and trees. I also used the point averaging feature in order to create polygons of the geometrically designed green spaces, as their straight lines would allow for fewer data points. In order to craft a polygon of a decorative pavement circle in the sidewalk I used the point-streaming feature so that I could more accurately capture the shape of the feature. I then captured the footbridge between the sidewalk circle and Davies center by using point averaging to capture two vertices to make a line.

When I felt content with the data I collected, I returned to the lab to check it back into the computer. In order to do this, I used the ArcPad data manager to check the files back into the computer. After they were properly checked in, I used the feature classes to make a map of the campus commons, using an aerial photo as the basemap.

Results: Upon adding the collected data to the map, it became immediately obvious that small errors can occur frequently when collecting data points with a handheld unit. The data tended to become more skewed, as the points I collected got closer to nearby buildings. In order to compensate for these cartographic errors, I slightly shifted the positions of the affected vertices, so they may more accurately depict the desired features.

Figure 3

Sources: GPS data collected by Peter Sawall on 4-16-2014

NAIP 201X

Lab 2: Downloading GIS Data

Introduction:
This lab was designed to teach me how to acquire data from the U.S. census Bureau and how to combine the data to shapefiles in order to make maps.

Methods:
First, I learned how to acquire data from the U.S. Census website. To do this, I visited the American Factfinder website of the U.S. Census Bureau. Then I downloaded an attribute table for census data showing the total populations of the counties of Wisconsin. Then I learned how to download a shapefile showing the counties of Wisconsin from the U.S. Census Bureau’s website. Next, I learned how to link the data together. I opened the annotated metadata set in excel, changed the file to an MS Excel file, then added the excel file and the shape file to ArcMap. I opened the attribute tables of the two files and created a table join based off of the GID attribute, so they would link properly. Then, I mapped the data, adjusting the symbology to a graduated color scale, so the population would be shown in each county.

I then made a map that shows how much of the population of each county lives in urban and rural areas. First, I acquired data from the U.S. Census Bureau’s website. Then, I changed the data to a MS Excel file before inserting it and the shapefile I previously downloaded to new dataframe. Next, I joined the new data to the shapefile based off of the GID attribute. I created a custom map projection based off of the Lambert Conformal Conic to make the state appear slightly more aesthetically pleasing than the Mercator projection it was originally projected in at the cost of minor distortion. I symbolized the data on a graduated color scale, and had it display the number of persons living in rural areas normalized by the total population of the county. I adjusted the colors so the more rural counties would be shown in a dark green with the shade becoming progressively lighter as the counties became less rural.

Results:
The maps show that as the population decreases, the percentage of the population that lives in rural areas increases.

Sources: I acquired the data from the U.S. Census Bureau’s website.

Lab 1: Base Data

Goals and Background:

The city of Eau Claire, Wisconsin invests a lot of time and money into the performing arts, but the city has been struggling with inadequate facilities, as the primary theaters were built in the early-mid 20th century. The Confluence project aims to fill the current theatre needs and provide a performing space large enough to allow performances the city has not yet held, “including major touring Broadway-style productions that currently cannot be effectively staged in any existing venue in Eau Claire” (University of Wisconsin-Eau Claire [UWEC], 2014). The main goal of this work was to prepare several base maps to provide context for the confluence projects location. It was also to familiarize myself with many different types of data relating to government and land use. This included building a layout with the major thematic feature classes, creation of a legal description, and digitization of the site.

Methods:

In order to learn more about the major feature classes, I made a map showcasing each of them individually. First, I used arcmap to digitize the “proposed site” of the confluence project. Then I used this information to make six separate maps showing the proposed location in regards to: Civil Divisions, Census Boundaries, PLSS Features, Parcel Data, Zoning, and Voting Districts.
 

The first map is one showing the different civil divisions of Eau Claire County. In order to make it, I used the “Bing World Imagery” for my basemap. I then added the county boundary and civil divisions to the basemap, made the county boundary hollow, and adjusted the symbology of the civil divisions layer so it would show the municipality types in different colors. Next, I made the “Civil Divisions” layer slightly transparent so the basemap can be viewed for reference. I then pasted the “proposed site” layer.

The second map shows the population density of the region surrounding the proposed site. First, I added “Bing World Imagery” for the basemap. Next, I added the “block groups” and “tracts” to the map. I made both layers slightly transparent, to show the basemap, and made the “tracts” layer hollow, to merely show the boundaries. I adjusted the symbology of the “block groups” to show the differences of population density in the region with graduating shades of pink. The “proposed site” layer was then added to the map.

The third map shows where the proposed site falls within the quarter-quarters of its township according to the Public Land Survey System. In order to show this information, I added the “PLSS quarter quarter” layer to the “Bing World Imagery” basemap. To allow simplicity, I made the “PLSS quarter quarter” layer hollow, and made it a bright shade of green so it would easily stand out against the basemap.

The fourth map shows the parcels and roads in the area surrounding the “proposed site”. In order to make this map, I added the “parcel area”, “centerlines”, and “water” layers to the “Bing World Imagery” basemap. In order to better show the information in reference to the area, I made all of the layers slightly transparent. I made the “parcel area” layer cyan, and the centerlines yellow, so both would be extremely visible. I made the “water” layer a muted shade of blue, so it would be still visible, but not provide too much distraction. I then added the “proposed site” layer to the map.

           

The fifth map shows the zoning of the region surrounding the proposed site. First, I added “Bing World Imagery” for the basemap. Then, I added the “centerlines” and “zoning class” layers to the basemap. I made both layers slightly transparent, to show the basemap, making the “centerlines” layer a bright green color, for greater contrast. Next, I adjusted the symbology of the “zoning class” layer, combining the many specific zoning classes into six more general classes. I then adjusted the symbology to show the different zoning classes in the region with different colored sections. The “proposed site” layer was then added to the map.

           

The sixth map shows where the proposed site falls within the voting districts of the city. In order to show this information, I added the “voting districts” class to the “Bing World Imagery” basemap. Then, I made the “voting districts” layer into a light orange color, and made it slightly transparent so the districts could be seen in reference to the basemap. Finally, I added the “proposed site” layer to the map.

Results:

The maps help paint the picture of the proposed site of the confluence project. The Census Boundaries map shows that the proposed site is in a rather densely populated part of downtown Eau Claire. Zoning map shows that the proposed site will be linked by a footbridge to large residential neighborhoods, just across the river. The zoning map also shows that there are several swaths of public land in the area surrounding the proposed site, which will only add to the beauty of the performing venue. From what I’ve learned from these maps, the confluence project will make a wonderful addition to the Eau Claire’s rich history of music in a very beautiful way.

Figure 1

References

University of Wisconsin-Eau Claire (2013, October 13). Frequently asked questions: The Confluence Project. Retrieved from http://www.uwec.edu/News/more/confluenceprojectFAQs.htm