The Final Project

Introduction:

For lab four, the final project of this class I was required to ask a simple question and answer it with the use of data that was readily available. My question is: Where should people retire in the area around the city of Eau Claire Wisconsin? To answer this question I started with three simple parameters:

1)      Due to wanting to live long enough to fully enjoy one’s retirement the area needs to be within two miles of a hospital.

2)      Because retirees love golf and have a relatively high chance of commuting via golf cart anyways, the area needs to be within 1 ½ miles of a golf course.

3)      Due to retirees wanting to have family visit often but are not keen on hearing the sound of cars driving by at all hours the area needs to be between ½ and 2 miles from a major roadway.

One might wonder why this question needs to be answered. I believe that it is relevant due to the high number of people who are projected to retire and will probably like to move to one of the nicer cities in the United States. Based on my knowledge of what retirees enjoy I believe this map would work well for those people in particular.

Data Sources:

Figure 1. Esri has classified a dog obedience school as a college or university. 

U.S. Institutions, 2013, Digital map. ArcGIS content team. Esri. Redlands, California.

                This is a point dataset that at first seemed accurate. But when inspecting the attribute table for this data a remarkable entry was observed (Figure 1). Apparently, Esri thinks that a dog training business should be in the same category as universities and colleges.

U.S. Large Area Landmarks, 2013, Digital map. ArcGIS content team. Esri. Redlands, California.

                This data appears to have locational accuracy but also seems to have left out smaller features that should have been included. It is quite generalized.

Major_Roads, mgisdata

                This is a very limited data set that only shows interstates or state highways. The lines are not accurate and often overlap other feature classes which, in reality, are bordered by the road. Seeing as this data is not going to be used at small of a scale, these inaccuracies should not subtract from the overall results of the project. 

U.S. Water Bodies, 2013, Digital map. ArcGIS content team. Esri. Redlands, California.

                This water bodies dataset from ESRI appears far more accurate than the aforementioned U.S. Institutions dataset listed above. When overlaid on a satellite imagery base map of the Eau Claire area the polygon features from this water bodies dataset matched up with both the lake and river shores far more accurately than expected. 

 Methods:

The first task involved formatting the above data for this project. By using the project tool all data sets were reprojected from their original WGS84 projections to the NAD83 central Wisconsin state plane system. An 11-mile wide circle, centered on downtown Eau Claire, was then created to define the area of interest (Figure 2) and to remove unneeded data from the three datasets via the clip tool. 

Figure 2. A circular area of interest defines the data used for the project. 

 Once the data is formatted, the perimeters set for the project are ready to be met. The first thing I did was set out to determine the proper distance of the area from a major road. I created two buffers on the major road feature set to determine both the minimum and maximum distance. Then I used the erase tool to subtract the near distance from the far distance to define the area (Figure 3).

Figure 3. Blue polygon represents an area close to major roads but far enough away so as not to hear the traffic. 

Next I used the buffer tool to determine the proper distances from the hospitals and golf courses. Then, along with the road distances layer I made earlier, I combined all of them with the intersect tool. The resulting polygons were fairly large so I had to think of another place that retired people would like to avoid. Seeing as dealing with traffic day to day is difficult (especially with a golf cart) and that young people are often rude I decided that any proper retiree would like to avoid any malls in town. I created another feature class for the two malls in the area and then used the buffer tool to create a 1 ½ mile radius around them. I then used the mall buffer polygon to erase features from the combination of positive factors (distance from roads, hospitals, and golf courses).

Originally I thought this would be as far as I would need to go so I started to map the data. Unfortunately I encountered a problem that only became apparent when I used satellite imagery as a base map. Some of the polygons representing where people should retire were located in the middle of the lakes and rivers that Eau Claire was built on (Figure 4). Seeing as Eau Claire doesn’t have the accommodations for house boats I cannot recommend retirees living in the middle of a lake.

Figure 4. Retirees should not live in the middle of lakes. 

Figure 5. Flow model of preparing water features data for this project: re-projecting, and clipping to the area of interest.  

 To fix this issue I imported a water feature file, re-projected it to the proper coordinate system and clipped it to my area of interest (Figure 5). From here it was easy to integrate it into the flow model for the entire lab buy simply using the union tool to combine a 500 foot buffer of water features to the 1 ½ mile mall distance buffer. From here it was a simple process of using the dissolve tool on the suitable retiree areas to give the data that polished look. The image below is the data flow model I used to meet all the questions criteria (Figure 6).

Figure 6. Data flow model used to meet the requirements of the question. 

Results:

Figure 7. Finished map showing retirement areas. 

The above image is the map I made with the retirement data (Figure 7). I decided to keep it simple and avoid cluttering it with too much data. I have displayed the retirement data and the three feature types that were important in its creation.

Evaluation:

I really liked this project, it was an opportunity to challenge myself with my own question and move forward with the project with very little outside help or guidance.  If I were to do this project differently I would have liked to use more accurate data and to have used more feature types to determine the suitable areas for retirement. If I did then the map above probably wouldn’t be endorsing living on an airport runway. The map could have also been improved by generalizing the hospital point data seeing as they are overlapping in several different areas. I am glad that I used Arcmap’s model maker program to for every step of this project. There were at least two times that I needed to do major edits to my workflow that would have been very difficult to do without the model maker. 

Lab 3: Black Bear Habitats

Goal:

Lab three was designed to introduce students to working on complex geographical questions. The main premise of it was to study bears in Marquette County Michigan. With the use of GPS data and other data from the Michigan Center for Geographic Information (Supplied by Dr. Hupy) I was able to plot the bears’ locations. From this I was able to learn what their preferred habitat was, and find what parts of this habitat fell within the Michigan DNR’s territory.

This lab was divided out into 8 sections:

1.       Mapping a GPS flat file of black bear locations.

2.       Assess what type of land cover these bears prefer to spend their time.

3.       Judge whether or not streams are important when determining habitat.

4.       Map suitable bear habitat according to the above parameters.

5.       Find areas that area managed by the DNR.

6.       Exclude areas in proximity to urban areas.

7.       Build a flow model.

8.       Learn the absolute basics of python coding in ArcMap.

Objective 1:

For the first objective of the lab I was asked to take an excel file that contained locations gathered by GPS of areas which bears frequented. Seeing as it was an excel file with no object ID I was limited in how I could manipulate the data. In order to overcome the limitations of a flat file I had to define the coordinate system and import the data into ArcMap. After it was imported I compared the locations of the excel file point data with a shape file showing forest cover in my area of interest. I did this to make sure I used the correct coordinate system when importing the excel file location data. All the points fell within the study area so I was confident enough to export the excel point data into the marquette_bear_study.gdb. Exporting the data only creates a copy of the file with an object ID meaning that the attribute table can now be edited and individual points can be adjusted.

Objective 2:

Objective two is more interested in the spatial relationships between the locations of the bears and the type of land cover of the locations. Seeing as the location data for the bears only have

an ID number and coordinate data I needed to perform a simple spatial join to get land cover data included in the bear location attribute table. Figure 1 shows the results of the join. Notice the matching ID numbers for both tables and how land cover information has been added to the new bear_cover table.

Figure 1. Results of spatial table join, before (Right) and after (Left).

Now that the bears locations have been joined with the land cover type polygons I can summarize the forest type field to find out what kind of terrain bears like. Below is table that I was able to make from the combined land cover data and bear locations (Figure 2).

Figure 2. Summary table of Bears per land cover type.

From the table it is easy to tell that bears like to hang out in mixed forest land, forested wetlands, and evergreen forest land. I will need these three cover types for use later.

Objective 3:

Objective three requires me to find the percentage of bears that are within 500 meters of a stream. In order to do this I selected the bear locations according to their distance from the nearest stream. According to the lab if the percentage is above 30 then biologists consider it important for calculating bear habitat areas. According to the data 49 of 68 bears (~81.6%) were within 500 meters of a stream when their GPS locations were recorded meaning that stream locations are an important consideration when studying bear habitat. For this lab I was required to symbolize the 500 meter distance from streams. To do this I used a buffer tool to create a new dissolved polygon feature class (Figure 3) around streams. 

Figure 3. 500 meter buffer area around streams. Cyan points represent bears within 500M of a stream.

I don’t like how the stream buffer layer extends outside of the land cover polygon. Seeing as this data will probably be shown on a map and because I wanted to use a new tool, I clipped the stream buffer polygon to get it to look a little better (Figure 4). I found out later that I did not need to do this to improve the appearance of the final product but I believe the exercise was still relevant to this lab.

Figure 4. Top: map showing 500 meter buffer polygon around streams. Bottom: 500 meter buffer polygon clipped by underlying forest cover polygon.  

Objective 4:

For Objective four I was to make a Bear Habitat feature class. From the earlier objectives I learned that Bears don’t like to be too far from streams and that they love mixed forest land, forested wetlands, and evergreen forest land. These two different features are what I will use to determine the ideal bear habitat. Seeing as they are both polygons with attributes I want I used the intersect tool. See the results in the image below (Figure 5).

Figure 5. Top: intersection between a 500 meter stream buffer and forest types bears like to create suitable habitat areas. Bottom: Suitable habitat areas generalized with the buffer tool.

               

Once a suitable habitat feature was made I generalized it using a buffer tool (Figure 5). While there was some loss of information the data’s appearance was improved and still works perfectly fine for my purposes.

Objective 5:

The fifth objective had me bring in a feature class representing DNR coverage. The Michigan DNR wanted to know what areas of suitable bear habitat was within their areas of influence. In order to figure this out I used the intersect tool to preserve all areas which are within both the DNR coverage and bear habitat (Figure 6).

Figure 6. DNR coverage (green) within the bear habitat.

Objective 6:

The DRN are not interested in fostering an inviting habitat for bears close to developed areas in the county. So for Objective six I was required to remove the portions of the DNR bear management areas that are close to urban or built up land. So in the same fashion as I used to determine the areas bears like to visit I determined the areas they shouldn’t get to close to. I used the buffer tool to create a 5 Kilometer areas around urban areas. Once this polygon was created I could use it as an erase feature on the DNR coverage areas (Figure 7).

Figure 7. Viable areas for DNR management in red.

Objective 7:

For the seventh objective of this lab I was required to make a cartographically pleasing map and a workflow model for the entire lab. Here is what I was able to come up with:

Figure 8. Completed map of Bear Habitat study area in Marquette Michigan. 

Figure 9. Workflow Model for Figure 8.

Objective 8:

Objective eight required me to use many of the same tools as which I used earlier in the lab. There was one key change in the methods though: I was supposed to code them into ArcMap with the use of python. I have no coding experience but after a few minutes of messing around with the interface it actually turns out to be moderately easy. I was able to make a one kilometer buffer around the rivers feature class, intersect that output with the bear friendly land cover feature set, and erase any bear habitat from the display that was within five kilometers of urban areas. Below I have included a screen shot of the code and the resulting feature classes (Figure 9).

Figure 9. Python code used to create Feature Classes. 

Lab 2

The objective of this lab was to familiarize students with some techniques of data collection and publishing maps in a professional format. I was tasked with downloading Census data for the state of Wisconsin which I then manipulated so it would be compatible with ArcGIS programs. I would then map this data and publish it to an ESRI online service.

Objective 1: Downloading Census Data

For the first objective of this lab I found information from the United States Census Bureau American FactFinder website. The first dataset which I downloaded was the total population of Wisconsin counties. The data set was located by entering parameters within the advanced search function of the website. I ended up with a zip file containing simple spreadsheets with the raw population data. None of the data contained spatial representation. This led me to downloading the Wisconsin counties shapefile from the FactFinder website.

Objective 2: Joining data sets

Once both the data sets which I had downloaded were imported into the ESRI ArcMap program I had to combine them using the “join” feature from ArcMap. I was able to use GEO_ID, an attribute shared by both tables, as the key field for my join. After completing the operation the attribute table of the county shapefile included the population data from the other downloaded file. But when attempting to symbolize the population data I found out that I had to further manipulate the data. The state population was a sting field type (typically used to map nominal data). To map quantities on ArcMap the data type must be set as a quantitative data type. I was able to add an attribute field containing the same information as the original but with a double field. This is much better suited to expressing numbers and ArcMap recognizes that the data can be mapped quantitatively. Below I have included an image of the symbolized population data and the joined attribute table (Figure 1).

Figure 1. Symbolized population data with labeled attribute table types, note the identical numerical data in the original and new fields.

Objective 3: Displaying Housing Data

The third objective for this lab was quite similar to the previous parts of this lab. I downloaded another data set, total Housing in Wisconsin Counties, from the FactFinder website. I joined the information with the Wisconsin shapefile attribute table using the same processes and operations as earlier. The only difference being that I defined the new attribute file data type as long integer instead of double field type. I will concede that this field type requires more memory than double floating but I wanted to see how ArcMap would interact with this slight difference in how a field is defined. I noticed no change in how the data were symbolized.

Objective 4: Build a Layout

My next objective is to make a cartographically sound layout for them. I designed a landscape document with two data frames, each of which displayed a map of Wisconsin. I set the scale of both data frames to 1:4,550,000 and used a Nad 1983 HARN projection designed for Wisconsin (WKID: 3071). I added basic cartographic elements such as scale bars, north arrows, and a legend for both data frames. I choose not to include a base map of any type because I feel that it would overcomplicate a map of this scale where spatial relationships are not pertinent to precise location. I have attached the completed map below (Figure 2).

Figure 2. Completed map of Wisconsin population and housing.

Objective 5: Publishing to ArcGIS Online

For this objective I was tasked to publishing the map document I created to ArcMap online and editing the data display using the services which ESRI offers on their website. After filling out a basic summary for the map I had to remove both table joins to publish the map. Seeing as I had already made new fields and transferred the data to them this was not a problem for the display of data. From this point publishing the data was not a problem.

The final step of this lab was to share my completed webmap to the UWEC Geography and Anthropology organization (One needs to be in this group to see this map). Here is the web link to view the map: http://uwec.maps.arcgis.com/home/webmap/viewer.html?webmap=88b46650ec2b49fa89045e2f402efc8e. 

Lab 1

  For this lab I was required to symbolize data for the downtown/central business district area of Eau Claire Wisconsin. The project was designed to show different types of data which may have an influence on the decision making for the Confluence Project of Eau Claire; a mixed use assortment of buildings that combines contemporary ground level businesses with campus housing for fine arts students in the above floors. The location of the confluence project, as its name may suggest, is on the southern bank of where the Eau Claire River flows into the Chippewa River. It covers several adjacent lots and is being funded by a combination of funds from the University of Wisconsin Eau Claire, business investments, and donations from community members. At the time of this writing the Confluence Project is currently under construction.   

  As this is the first lab for this class it was designed to familiarize students with some of the basic functions of ArcMap. Through the completion of this project I was able to view attribute tables for data sets, create a shapefile to delineate the area of the Confluence Project, and use basic cartographic techniques to display the data. One can see my results below in Figure 1. 

Figure 1. Assortment of maps displaying pertinent information for the Eau Claire Confluence Project.