Lab 4: Mini-Final Project

Introduction

At the start of this lab I had the freedom to choose my own research question. I decided to determine the best location to build a cabin in Sawyer County, Wisconsin. This question appealed to me because after a long hectic week of finals, a nice secluded cabin would be ideal. With that being said, this research can help hardworking families or individuals create a home-away-from-home that provides a quiet environment to relax and enjoy the outdoors. Criteria for this cabin includes…

            Location on a lake

            Near a forest or national park

            An isolated location

                        At least 5 miles from nearest city

                        About 2 miles away from any major roads

            Located near a hospital in case of an emergency

These parameters would create the best area for a nice-get away cabin.

Data Sources

Luckily, the data needed to complete this project was found relatively easily in the Wisconsin and ESRI geodatabase. The data necessary to complete this project included Wisconsin county boundaries, lakes, county forests, U.S. parks, hospitals, cities, and major roads. Sources of the data are as follows

            County Boundaries: Wisconsin DNR 2014 Data

Lakes: ESRI 2013 USA Data

County Forests: Wisconsin Forest Inventory & Reporting System (WisFIRS)

U.S. Parks: ESRI 2013 USA Data

            Hospitals: Geographic Names Information System (GNIS) Hospitals

            Cities: Wisconsin DNR 2014 Data

Major Roads: Census 2000 TIGER/Line files

The data I retrieved concerned me in a few ways. For starters, I am concerned about the cities data. I am curious as to what the data determines a city to be. Is it based on area or population? I feel as though Sawyer County may have more cities than portrayed which may alter the findings throughout the project. Initially, I was concerned about the water bodies shown in the map. Originally, the layer had several rivers/streams, along with some rather small bodies of water. After looking through the attribute table, I realized the data had all types of water bodies. After conducting a simple query, I was able to select only the lakes to create a “lakes” layer. Because most of my data is rather static, I was not concerned with the accuracy or chance that the data has changed since collected.

Methods

Before starting the project, I first made a database connection to both the ESRI 2013 geodatabase and the Wisconsin DNR geodatabase. This allowed the process of selecting and adding the necessary layers to the map much easier. The process of my project is shown in Figure 1 as a data flow model.

To begin, in ArcMap, I created a blank document and added the Wisconsin county boundaries. From here, I selected Sawyer County and exported it to create a new layer in the map. Once Sawyer County was a new layer, I then added the needed data layers from the two geodatabases to the map. These data layers included the lakes, county forests, U.S. parks, hospitals, cities, and major roads layers. I then clipped the each data layer with the Sawyer County layer to have only the data within Sawyer was shown. This allows the software to run much faster because it narrows the data being geoprocessed to those only within Sawyer County’s boundaries.

To begin the narrowing process, I put a union between the county forests and U.S. parks layers. I then put a 1 mile buffer around them because I want the cabin to be within that distance from a forest. I then dissolved the buffer to generalize the data. I then put a buffer of 100 feet around all the lakes within Sawyer County. I did this to ensure a lake front property. I then intersected the lake buffer layer and U.S. Parks/county forests layer. This allowed me to narrow down the locations to lakes near a forest. To ensure a somewhat close proximity to a hospital, I put a 20 miles buffer on the hospital layer. I chose 20 miles because that is a reasonable driving distance in the case of an emergency. If the buffer had been smaller it would have put the location closer to cities or major roads. By taking the hospital buffer layer and the lakes and pakrs/forest intersect layer and applying another intersect, I could see all of the desirable locations for a new cabin thus far.

Next, I started to take out the areas I did not want. I put a 2 mile buffer around the major roads layer to ensure a quiet location without noise pollution from traffic but still close enough for easy accessibility. I then determined the distance from a city would be 5 miles. I assumed that any closer would bring more noise and chaos to the area. I then but a 5 mile buffer on the city layer. I then took the two layers containing the areas I do not want and put a union between them.

Now that I know the areas near roads and cities, along with the desirable traits of a cabin location, it is time to take out what I do not want. To do this, I performed an erase between the desirable layer and the undesirable layer. This left me with the lakes near forests and a hospital but away from major roads and cities.

However, after looking at the final product I realized my data and answer did not make any sense. If I had left it the way it was, it would have shown that a desirable location for a new cabin would be in the middle of a lake. To solve this problem I erased the original lakes layer from the output. This left me with just the areas along the lake that were near forests and hospitals and a desired distance from any cities or major roads.

Figure 1: Data Flow Model

 

Results

As a final product, I decided to create two separate maps. The first map shows the desired areas in Sawyer County that meet the specified parameters (Figure 2). The second map, though very similar, shows the buffers created in the process to show the reasoning behind the highlighted desirable areas in both maps. Also, in both maps, I provided a small locator map of Wisconsin along with a large-scale map of the top locations. Large-scale Map #1 shows the locations on Nelson Lake in the Northwest corner of Sawyer County. Nelson Lake has a lot of possible locations for a cabin which makes it a marketable location. Large-scale Map#2 shows the locations on Spider-Clear Lake in the Northern region of Sawyer County. This lake is smaller compared to Nelson Lake which can come with both benefits and disadvantages. This lakes also has a substantial less amount of land to choose to from when deciding where to build a cabin.  

Figure 2: Top Cabin Locations in Sawyer County, Wisconsin

 

Figure 3: Top Cabin Locations in Sawyer County, Wisconsin showing criteria

 

 

Evaluation

 I felt this project did a good job in showing the new skills we have learned throughout the semester in GIS I. If I had to repeat the project, I would add some criteria to it. I am interested in finding out whether or not the areas found desirable are DNR managed land or already owned by others. I could do this by adding the DNR managed land layer in found in the WiDNR geodatabase. Some challenges I faced involved the buffers throughout the project. Some of the preferred distances would have landed me with no locations to choose from. I had to change the buffer distances to ensure a desirable location would result. Something I would like to do in the future is expand the area of interest to the entire state of Wisconsin.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Lab 3: Vector Analysis with ArcGIS

Background

The purpose of this lab was to determine the most suitable land to be used as a bear habitat within the study area in Marquette County, Michigan. While keeping in mind previous locations of bears, presence of streams, proximity to urban areas, and most suitable land type I was able to use data provided by the state of Michigan to select and remove land based on the given criteria.

·         Within at least 500 meters of a stream

·         Favorable land cover type

·         Areas of DNR managed land

·         At least 5 kilometers from an Urban or Built up land cover type

Goal

The goal for this lab is to become familiar with the different types of geoprocessing tools and be able to determine which ones to use to figure out what land is best suitable for a new bear habitat.

Methodology

Throughout this lab’s objectives, I used several different methods to narrow down the suitable land of a new bear habitat. You can see a simplified model of my methods in Figure 1.

Objective one:

During objective one, I was able to explore the data and the file types that were to be used in this lab.

                marquette_bear_study

                landcover

                bear_management

Within these data files, was an excel file specifying the XY coordinates of previous bear locations. Because the excel file is a non-spatial database, I first had to add the coordinates as an “event theme”. An “event theme” allows you to plot these XY coordinates spatially within ArcMap. Once added, I exported the locations to my lab 3 geodatabase.

Objective two:

After adding all of the data from the bear_management_area dataset, I created a unique value map of the land cover by “Minor Type”. I then wanted to determine which land cover type had the most bear locations. By intersecting the land cover and bear_location feature classes, I was able to generate a table that had both the id of the bears and the various types of land cover. For this objective, I focused on the “minor type” field and summarized the field to determine the count of bears in each type. I concluded that the top 3 habitat types were…

                  1. Mixed Forest Land (964)

                  2. Forest Wetland (644)

                  3. Evergreen Forest Land (576)

I then created a separate layer for these 3 land types and named it suitable_land_cover

Objective three:

After receiving information from biologists, I wanted to determine how man of the bears were found in close proximity of streams. To do so, I conducted a query by location and determined that nearly 72% of the bears were found within 500 meters of a stream. Because of this large percentage, I believed it to be important criteria to keep in mind when determining suitable bear habitat locations.  This importance led me to create a 500 meter buffer around all streams within the study area (stream_buff). I then dissolved the buffer to generalize the feature class and clean it up.

Objective four:

Based on the findings up to this point, I decided to intersect the stream buffer and the suitable land cover because of the important role they play in the location of bears. I dissolved the result of the intersection to combine the internal boundaries of the layer to simplify the data. 

Objective five:

Because the bear habitat must be on DNR managed land, I then chose to intersect the suitable land near streams (objective 4 outcome) and the DNR managed land. I also dissolved this output to create a more appealing layer. 

Objective six:

For this task, I was asked to manipulate the data further and select areas that are not near any urban or built up land. To do this, I used a query on the land cover layer and created a new layer of the urban and built up land. I then put a 5 kilometer buffer on the layer (then dissolved it again). With this 5 kilometer buffer, I was able to erase the areas that were within the designated area from the suitable land managed by the DNR leaving us with only areas away from urban land. 

Figure 1: Data Flow Model to find suitable land for bear habitat

Practice with Python:

To gain some practice using python coding, I did some of the previous geoprocessing tools by typing commands within the python window (Figure 2). I proceeded to do a buffer analysis, an intersect analysis, and an erase.

Figure 2: Python Coding Practice

Results

Figure 3: Final Results

 The final results from this lab are shown above (figure 3).This map shows the study area within the Marquette County boundary. It also includes a location map on the right showing where Marquette county is located in Michigan. The map also shows the bear locations, the locations of streams, the suitable land types near streams (objective 4), and the suitable land near streams on DNR managed land (objective 6). Most of the urban and built up areas we wanted to avoid in this selection are found in the southern region of the study area. Overall, all areas shown in the pink/salmon color would be perfect habitat for bears following the criteria. In my opinion, I would select the area in the north east of the study area because of the large area allowing the bears to roam and its numerous streams for the bears. 

Sources

"Michigan 1992 NLCD Shapefile by County." Michigan 1992 NLCD Shapefile by County. Accessed December 6, 2015. http://www.mcgi.state.mi.us/mgdl/nlcd/metadata/nlcdshp.html.

"Michigan Geographic Framework: Marquette County." Michigan Geographic Framework: Marquette County. Accessed December 6, 2015. http://www.mcgi.state.mi.us/mgdl/framework/metadata/Marquette.html.

"Wildlife_mgmt_units." Wildlife_mgmt_units. Accessed December 6, 2015. http://www.dnr.state.mi.us/spatialdatalibrary/metadata/wildlife_mgmt_units.htm.