Geography 335 - GIS I: May 2015

Introduction:

The purpose of this lab was to propose my own spatial question, and then utilize the knowledge and skills that I have developed throughout the semester to answer my spatial question. More specifically, using geoprocessing skills (vector). The question that I proposed was, where are the best locations to camp in Minnesota? Attached to this spatial question is a certain set of criteria that I have selected. The criteria is as follows:

· Must be further than 25 miles from an interstate.

· Must be within a half mile of the Mississippi River.

· Must be a recreation area.

· Must be contained within a Minnesota State Park.

Following the steps in the geographic inquiry process and utilizing an assortment of geoprocessing tools will allow me to arrive at the best camping locations in the state of Minnesota. Hopefully, my analysis will provide avid campers and nature oriented people with the best camping locations in Minnesota based off my selected criteria, which is listed above.

Data Source:

Based off the parameters of my question, I did not have to do much searching to find the information that satisfied my criteria. I was able to access all the necessary feature classes within the ESRI database, which is a service provided in ArcGIS. More specifically I had to tap into the database connection, where I used the SQL server as the database platform. From this point I used the geogsql.uwec.edu as the instance and used the Esri2013 as the database. After accessing all these files I was able to go into the Esri2013.DBO.USAData and find a majority the feature classes that pertained to my spatial question. My main concern with this data is that every file is from 2013, which means that some of the information could potentially be two years old. For the case of the state parks, interstates, and rivers feature class the 2013 data should be sufficient. However, in terms of the recreation areas feature class this could be outdated. Within the past two years other recreation areas could have been built, which would mean that it would not be included in my analysis. Thus, it is important to address this issue as a potential limitation to my analysis.

Methods:

To begin my lab I had to gather all the necessary data that was relevant to answering my spatial question. Most of my data came from the Esri2013.DBO.USAData, which I mentioned above. The states and counties feature classes I accessed through the mgisdata folder, then used the usa folder, and accessed the usdata.gdb to locate the states and counties. After collecting all my data I created a file geodatabase that would hold all the data for my analysis. By creating a personal file geodatabase it would allow me to remain organized and give me easy access to all my created feature classes.

With all my data being displayed on ArcMap it was time to prep the data and clip certain features to the state of Minnesota. First, I conducted a definition query on the states and counties feature class, in order to display only the counties within Minnesota. After the definition query, I changed the coordinate system to NAD 1983 HARN Stateplane MN_Central_FIPS2202 (Meters). Changing the coordinate system more accurately displays Minnesota and minimizes distortion. From here I began to clip the features classes from the Esri2013.DBO.USAData to the state of Minnesota. This was necessary because the interstates, rivers, recreation areas, and parks feature classes were for the whole United States. Thus, it was pertinent that I clip these feature classes to Minnesota, which was my area of interest. Figure 1 below, which is a display of my workflow, depicts all the clips that I conducted.

After clipping all the USAData to my area of interest, I wanted to select only recreation areas in Minnesota. Under the recreation area feature class it included amusement parks, park attractions, and beaches. I was only concerned about the locations that were titled ‘parks and recreation areas’. In order to disaggregate this information I had to conduct a query, the SQL for this operation was FEATTYPE = 'Park and Recreation Area'. From here I created a new feature class based on this query, I kept the name for this feature class as the default of MN_Rec_Selection. Then I clipped the MN_Rec_Selection to the parks_MN feature class. This step can be seen in Figure 1.

Next I had to focus on the Rivers_Clip feature class and establish a buffer for the Mississippi River. Based off my criteria illustrated above, I needed to establish a half mile buffer on the Mississippi River. After this task I created my new feature class as Miss_Riv_Buff.

Following the buffer on the Mississippi River I also had to use the buffer tool on the MN_Interstates feature class. I created a 25 mile buffer on MN_Interstates. This buffer would allow me to see which camping locations would need to be excluded, because remember that one of my criteria elements stipulated that the camping location needed to be further than 25 miles away from an interstate.

Illustrated in Figure 1 below you can see that my next step was to perform an intersect between the Miss_Riv_Buff and Rec_Parks feature classes. This step resulted in the creation of Rec_Miss_Riv_Int feature class. From here I took that feature class and erased it with the Int_25Buff to arrive at my final answer, which I titled as Ideal_camp. All the steps I mentioned within my methodology can be followed more closely in Figure 1 below.

Figure 1 - This is my workflow which depicts the steps I took to arrive at the answers to my spatial question.

Results:

Throughout my analysis I was trying to determine where the best camping locations are in Minnesota that matched my criteria. Depicted in Model 1 below you can see that my analysis provided me with two camping locations, Crow Wing State Park and Schoolcraft State Park. These two state parks in Minnesota are the best camping locations based off my parameters. Crow Wing State Park and Schoolcraft State Park would provide campers with the luxury of camping near the Mississippi River and offering great opportunities for fishing. Additionally, the further than 25 miles buffer from an interstate allows campers to have a more natural camping experience without the influence of high amounts of human traffic.

Model 1 - These maps show the ideal camping locations in the state of Minnesota based off my criteria.

Evaluation:

Performing this final lab was a great way to display the knowledge and skills that have learned throughout this course. Furthermore, this lab exposed me to the geographic inquiry process, where I was able to develop my own spatial question based off factors that are interesting to me. This lab was a fun project, but it came with a fair share of challenges. I originally wanted to do my analysis on crime data connected to commercial burglaries. I started to develop my question and found data oriented towards commercial burglaries. However, after struggling I decided to scrap that idea and focus on my interest in camping. Subsequently, I had difficulty managing my data and coming with the appropriate tools that would satisfy my criteria. After getting through these hoops and getting my data flow model (Figure 1) reviewed the rest of the lab was smooth sailing. If had an opportunity to redo this lab I would go back to focusing on my original interest in performing an analysis on commercial burglaries. I would need to develop an easier spatial question and have extra available time to complete that project to the fullest.

Sources:

ESRI Software: Esri2013.DBO.USAData and usdata.gdb

Goals

The goal of this lab was to utilize numerous geoprocessing tools for vector analysis, in order to determine suitable bear habitats in Marquette County, Michigan.

Background

Each point on the map represents a bear location, which was collected using a global positioning system (GPS). Based off these points the Michigan DNR would like to know which areas are suitable for them to study bear habitats.

Methods

To begin the lab I had gather all the necessary data. However, all the data had already been uploaded to our class folder, which made this process fairly simple. The landcover information came from the USGS NLCD and the DNR management units can be found here. After familiarizing myself with the data and information it can apparent to me that the bear locations were in an X Y coordinate format that was not from a spatial database. In order to correct this issue I had add the X Y coordinates as an ‘event theme’. An ‘event theme’ is a temporary display of X Y data in ArcMap. Once a created the ‘event theme’ the bear locations appeared as points in ArcMap. Since an ‘event theme’ is only temporary it was necessary to then export the points as a feature class.

After establishing the bear locations as a feature class, I then went ahead and added all the rest of the feature classes that would further assist me in the lab. Next I performed a spatial join between the bear_locations and landcover feature classes. As a result, I was able to determine which land cover type most of the bears were located in when the GPS points were collected. Based off this information I was able to conclude that the top three bear habitat types are within the landcover categories, Mixed Forest Land, Forested Wetlands, and Evergreen Forest Land.

Numerous streams are located within the study area, and based off biologist’s findings, it has been assumed that bears might be found within close proximity to streams. To test their assumption I performed a spatial query, using bear_cover as the target layer and streams as the source layer. When I examined the output table from the spatial query I found that approximately 72 percent of the bears were located within 500 meters of a stream. Since a majority of the bears are located within 500 meters of a stream, it confirms the biologist’s assumption about the relationship between bears and streams. Thus, the distance to streams is an important habitat characteristic.

Next, I wanted to find which areas within Marquette County, Michigan, are the most suitable areas for bear habitats. In order to accomplish this task I had to begin by exploring the various analysis tools within ArcToolbox. I determined that I would be a good idea to start by using the buffer tool from the overlay analysis tool category. I performed a 500 meter buffer on streams feature class, titling the output feature class as streams_buff500m. Then I executed an Intersect between suitable land cover and streams_buff500m. Following the intersect I decided to dissolve my newly created intersect feature class, which removed the internal boundaries.

After prepping my data from the previous steps, I added the dnr_mgmt feature class for Marquette County, MI. At this point I decided to use the Clip tool from ArcToolbox and clip the study_area and dnr_mgmt feature classes together. Next I intersected my newly clipped feature class (Clip_DNR) with my Suit_Buff_Int_Dis feature class, which established Suit_DNR_Area feature class. Following that operation I decided it would be a good idea to dissolve the Suit_DNR_Area, which would eliminate the internal boundaries.

In order to provide the Michigan DNR with a more suitable habitat model for bears in Marquette County, I needed to provide them with bear management areas that are at least five kilometers from an Urban or Built up lands. I began this objective by taking the landcover feature class and performing a query on Urban or Built up lands. From this query I created a new feature class from the selected areas and titled the output feature class as Urban_Built. From this point I buffered the Urban_Built feature class by five kilometers. Following the buffer I then utilized the erase tool from the ArcToolbox, naming my output feature class DNR_Suit. As result of all these steps I finally arrived at my answer which can be further observed in Figure 1.

After arriving at my answer, which provides the Michigan DNR with a suitable habitat for bears within Marquette County, I went ahead and created a visually appealing map. I tried my best to create an acceptable Legend which would allow the intended audience to easily understand the data that is being depicted in my map. Furthermore, I provided a second map within my data frame of the state of Michigan with Marquette County being highlighted in yellow. Highlighting Marquette County within the state of Michigan will provide the intended audience with a perspective into the specific geographical location where the suitable bear habitats are positioned.

To become more familiar with Python, I was requested to write a few simple commands to perform a couple of the geoprocessing operations that I conducted in the lab. I began loading ArcGIS python window and docking it to the bottom of screen. From here I ran a buffer on the streams feature class, but this time I used a one kilometer buffer instead of the 500 meter buffer that I used in the lab. After the buffer, I wrote a code that would perform an intersect operation between the results from my new buffer and the suitable land use. Finally, I wrote a command within the python window which would erase the buffer of the urban areas that were utilized in my lab. The results from my commands in python can be seen below in figure 4.

Results:

Figure 1 depicts the Marquette County Suitable Bear Habitats. Figures 2 and 3 shows my data flow model which I used in objectives two through six. Figure 4 is the command codes that I wrote to perform the steps in the last objective of the lab.