Thursday, December 15, 2016

Lab 4, Finding a Location for a Future Golf Pro Shop

Goals and Background
     The purpose of this lab is the determine ideal locations of where a golf pro shop should be located in the city of Eau Claire. To do this, five different feature classes were used to filter criteria to find the best possible locations. These include zoning_areas, WITracts,  majroads, InfGolfCenter, and ECGolfCourses. The city limits feature class (EC_City_Limits) was also used, but no tools were performed on it The ideal location for a golf pro shop, in this lab, is located in a commercial zoning class, is within one and one half mile of a golf course, is within a quarter mile of a main road, is in a census tract which has an average median household income of at least $45,000, and is at least three quarters of a mile away from the Infinity Golf Center (a golf pro shop already in Eau Claire). This information would be useful for someone who is looking to start up a golf pro shop. Also, anyone would be able to use this information, and could be useful for people looking for locations for a golf repair shop.

Data Sources
     The data sources for each initial feature class varied. The EC_City_Limits, and the zoning_areas feature class came from the City of Eau Claire's geodatabase which our university has some access to. The WITracts, and the median household income table were downloaded off of the Census Fact Finder's website. The majroads feature class came from the US geodatabase which comes with the Textbook Mastering ArcGIS. The InfGolfCenter, and the ECGolfCourses feature classes were created manually by digitizing the golf courses, and store as a polygon. There aren't any major concerns with the data sources, but a couple of minor ones. First, the majroads feature class was not really to the scale of Eau Claire. In some extreme cases, the roads were as much as 50 ft off from the actual road. Most of the time, this wasn't an issue though. Also, when digitizing, its probable that the polygons were not the exact boundaries of the golf course, or Infinity Golf Center shop. Both of these minor concerns aren't really big enough to have a significant impact on the final result,

Methods
Step 1: Query the Zoning Areas to Include only the Commercial Zoning Class

     First, a query was done on the zoning_areas feature class to select only the commercial zoning class features. Then, these features were created into a new feature class named EC_Commercial. The data flow model for this step is shown below in Fig 4.0.

Step one's Data Flow Model
Fig 4.0: Data Flow Model for Step 1


Step 2: Create a Feature Class with Census Tracts Which Have an Average Household Median                 Income Greater Than $45,000

     To do this, both the WITracts shapefile, and the Median Household Income table had to be downloaded from the Census Bureaus Fact Finder website. After being downloaded, they were exported into ArcMap. Then, a join was performed between them using the Geo_id field as the common key. This made the median household income field available in the WITracts attribute table. To get the feature class to only display those census tracts which have an average household median income greater than $45,000, a query was done. To make the feature class more cartographically pleasing, the Dissovle tool was ran on it. The data flow model for this step is shown below in Fig 4.1.
     
Step Two's Data Flow Model
Fig 4.1: Data Flow Model for Step 2

   
Step 3: Buffer Out the Main Roads in Eau Claire

      In order to do this step, a query must be done on the mjroads class to only select the roads which will be considered main roads. These roads include Hwy 53, S Farewell St, Clairemont Ave, and Interstate 94. Then, with these roads selected, a new feature class was made. Next, this feature class was buffered out one quarter mile. Lastly, the Dissolve tool was ran on it to get rid of the internal boundaries. The data flow model for this step is shown below in Fig 4.2.
Step Three's Data Flow Model
Fig 4.2: Data Flow Model for Step 3

Step 4: Create and Buffer Out the Golf Course in Eau Claire
  
     First, there was no shapefile which had the golf courses, so a feature class was created in ArcCatalog. Then, digitizing was done using satellite imagery to create polygon features for the golf courses. Next, these polygons were buffered out one and one half miles to meet the criteria for the location of the golf pro shop. The new feature class was then dissolved to make eliminate the internal boundaries. The data flow model for this step is shown below in Fig 4.3.
Step Four's Data Flow Model
Fig 4.3: Data Flow Model for Step 4

Step 5: Create and Buffer Out Infinity Golf Center

     The Infinity Golf Center also didn't have a shapefile, so a new feature class was created in ArcCatalog. Then, just like the golf courses, the Infinity Golf Center was digitized as a polygon. After this, the polygon was buffered out three quarters of a mile. This is an area which the golf pro shop cannot be located in, as it would be too close to a competitor. The data flow model for this step is shown below in Fig 4.4.
Step Five's Data Flow Model
Fig 4.4: Data Flow Model for Step 5

Step 6: Intersect the Appropriate Feature Classes

   In this step, the criteria for which the golf pro shop is to be located in is intersected. These feature classes include ECGolfBufDis, ECRdsBufDis, Tract45KDis, and EC_Commerical. This new output layer feature class is given the name GoodGolfStoreLoc. The data flow model for this step is shown below in Fig 4.5.
Step Six's Data Flow Model
Fig 4.5: Data Flow Model for Step 6

Step 7: Erase the Infinity Golf Center Buffer From the Good Golf Store Locations 
             Feature Class

     Prior to doing this step, the Infinity Golf Center was already located in the best locations for a golf pro shop. Because there shouldn't be too many golf pro shops located next to each other, the Erase tool was used between GoodGolfStoreLoc and InfGolfCenterBufDis. Doing this will result in the final output layer (PrimeGolfStoreLoc) which will represent the best places to start up a golf pro shop in Eau Claire based upon the criteria stated in the introduction. The data flow model for this step is shown below in Fig 4.6.

Fig 4.6: Data Flow Model for Step 7
Results

Fig 4.7: Data Flow Model
Figure 4.7 displays the data flow model for all of the steps put together. In total, six different tools were used. These tools include Query, Join, Buffer, Dissolve, Intersect, and Erase. In general, the model flows from right to left, top to bottom. The final output feature class is shown in the very bottom right.

Ideal Golf Pro Shop Locations
Fig 4.8: Ideal Golf Pro Shop Locations
     This map shows the ideal golf pro shop locations in bright red. These are the locations which meet all of the criteria listed in the introduction section. Most of the possible locations are along Hwy 53, and Clairemont Ave. Right in downtown would've also been part of the area if it were not for that area's household income being lower than $45,000. Before using the Erase tool to erase away the Infinity Golf Center buffer area, the location of the Infinity Golf Center met all of the previous criteria.This is good because it shows that the criteria is well suited for golf pro shops. One negative thing about the map is that the ideal golf pro shop possible locations is quite a small area. This is mostly due to zoning restrictions.
     Also added to the map were some of the features which the criterion was based on. These include the Eau Claire city limits, the golf courses, and the main roads. An inset map is also present to show where Eau Claire is located in Wisconsin.

Evaluation
     Overall, this project went pretty well. It took a while to develop a spatial question which would work, but other than that there weren't any issues. Because there weren't any shapefiles online which contained the golf courses, or Infinity Golf Center, they had to be created manually. This process didn't take long, as only five golf courses had to be digitized, and there is only one Infinity Golf Center. If the project were to be redone, the only thing that should be changed is the criterion for the household median income. Instead of querying the areas by tracts, it could be done using block groups which are smaller. This would have led to an increase in the number of possible locations for the golf pro shop.

Sources

Data Access and Dissemination Systems (DADS). (n.d.). American FactFinder Retrieved December 14, 2016 from http://factfinder2.census.gov/faces/nav/jsf/pages/searchresults.xhtml?refresh=t  

City of Eau Claire's geodatabase

Friday, December 9, 2016

Lab 3: Vector Analysis with ArcGIS

Goals and Background
     The purpose of this lab is to determine suitable bear habitat within a certain study area of Marquette County, Michigan. Factors which influence bear habitat will be overlaid with each other to determine the best habitat locations. One goal is to show how to properly use a data flow model. The data flow model in this post will show the process which will contain each input needed to create the map. Antoher goal of this lab is to create a map showing bear habitat areas in a study area in Marquette County. The last goal of this lab is to demonstrate a very basic understanding of some simple Python coding. The tools Buffer, Intersect, and Erase will be coded.

Methods
Step 1: Export and map GPS coordinates from an excel file into ArcMap.

     To do this, first, the excel document was added to ArcMap. Then, the data was added through the Add XY Data label under the File tab. The X-Field and Y-Field were changed to "POINT_X" and "POINT_Y". Also, the coordinate system was changed to "NAD_1983_HARN_Michigan_GeoRef_Meters". After hitting the "OK" button, ArcMap made a shapefile by default. Instead of using the shapefile, it was exported into a feature class with the name bear_locations.

Step 2: Determine Bear Habitat

     First, all of the necessary feature classes were added: BearLocations, Landcover, Streams, study_area, and dnr_mgmt. Then, with BearLocations and Landcover, an Intersect was performed creating a new layer called bear_cover. By doing this, the land cover type can be assigned with the bear locations. A summarize was performed on the MINOR_TYPE field of the bear_cover feature class in order to see which land cover types were the most common with bear locations. The three types include Mixed Forest land, Forested Wetlands, and Evergreen Forest Land. With this information a query was performed on the Landcover feature class to select any entry in the MINOR_TYPE field which had any three of the above land cover types. After the selection, a new layer was created with the selected attributes and was given the name PrimeBearLandcover. The data flow model for this step is shown below in Fig 3.0.

Data Flow Model for Step 2
Fig 3.0: Data Flow Model for Step 2

Step 3: Link Bears to Streams and Land Cover

     In this step, it was determined that streams were a very important part of suitable bear habitat. This was determined by creating a buffer of 500 meters from the feature class Streams, and then performing the Dissolve tool on it. The name assigned to it was Stream500mBufferDissolve which was then intersected with PrimeBearLandcover to see how many bears were located within this stream buffer. This outputted the new feature class named PrimeBearHabitatNearStreams. It turns out that the vast majority of the bears were located within this stream buffer zone. For clarity, PrimeBearHabitatNearStreams is dissolved creating a new feature class named PrimeBearHabitatNearStreamsDissolved. This feature class shows the bear habitat which have close proximity to a stream and have the suitable land cover type. The data flow model for this step is shown below in Fig 3.1.
Data Flow Model for Step 3
Fig 3.1: Data Flow Model for Step 3

Step 4:
Locate DNR Management Areas Within the Study Area

     The feature class dnr_mgmt had polygons located inside of polygons. This is because there are groupings within the management areas. However, for this analysis, all that mattered was the DNR management areas, not the groups. To fix this, the Dissolve tool was used on dnr_mgmt. The new output feature class was given the name Dnr_mgmtDissolved. This feature class was then intersected with the feature class PrimeBearHabitatNearStreamsDissolved. The name given to this feature class was BearhabStreamsDnr_mgmt. This new feature class shows the DNR management areas located within 500 meters of a stream and areas that have the three main land cover types. The data flow model for this step is shown below in Fig 3.2.
Data Flow Model for Step 4
Fig 3.2: Data Flow Model for Step 4

Step 5: Find Bear Habitat 5 km Away from Urban and Built Up Areas

     In order to find bear habitat that is 5 km away from urban or built up areas, these areas must be identified. These can be identified by using the MAJOR_TYPE field in the Landcover feature class. First, a query must be performed to select only those entries with urban/built up in them. Then, a new layer and feature class is created using the selected attributes. This new feature class is given the name UrbanBuiltUpLand. This is then buffered out 5 km. The buffer tool is ran on it and the new feature class UrbanBuiltupLand5KmBuffer is created. Because this has many polygons in the layer the Dissolve tool is ran on it as well. This creates a new output feature class titled UrbanLanDissolved5KmBuffer. A clip is performed using the feature class created in step 5: PrimeBearHabitatNearStreamsDissolved and UrbanLanDissolved5KmBuffer. This outputted a feature class given the name BearHabWithin5Km. This has an area which has suitable bear habitat and is within 5 km of urban or built up lands. However, the DNR wants to know suitable bear habitat areas which are farther than 5 km away from urban or built up lands. In order to display this area on the map, the feature class BearHabWithin5Km must be placed above PrimeBearHabitatNearStreamsDissolved in the table of contents. By doing this, PrimeBearHabitatNearStreamsDissolved will display the suitable bear habitat at least 5 km away from urban or built up lands. The DNR management lands must also be displayed on the map to be at least 5 km away from urban or built up lands. To do this, the Erase tool is used between the BearhabStreamsDnr_mgmt and UrbanLandDissolved5KmBuffer feature classes. This feature class is given the name PrimeBearLocations. The data flow model for this step is shown below in Fig 3.3.
Data Flow Model for Step 5
Fig 3.3: Data Flow Model for Step 5

Step 6: Map the Data

     Using the features PrimeBearLocations, BearHabWithin5Km, Streams, BearLocations, and PrimeBearHabitatNearStreams, a cartographically pleasing map was created. An inset map, scale bar, legend, title, base map, and north arrow were all added to the map. Some transparency was added to the polygon feature classes so contrast was was pleasing to the eye. The map shows the prime bear habitat areas within with sections being separated by a 5 km buffer from urban or built up lands. The map is displayed in the results section below in Fig 3.6.

Step 7: Write Some Python Code

     The Buffer, Intersect, and Erase tools were all successfully used by using the ArcGIS Python window. Basic code was written to run these tools. The purpose of doing this was to see that writing code is a much more powerful way to run tools than to do it through the user interface of ArcMap. Below, in Fig 3.4, are the lines of code which have been put into a word processor to make it look neater. Headers have been added which are not part of the code to make it easier to identify the tool being used. Both the entry and results lines are displayed.
Python Script for Buffer, Intersect, and Erase
Fig 3.4: Python Script for Buffer, Intersect, and Erase
Results


Data Flow Model
Fig 3.5: Data Flow Model
     Above is the data flow model for this entire lab. It was created by combining all of the individual data flow models from each step. The blue ovals represent original feature classes which haven't been changed or modified with any tools The yellow rectangles represent the tools used on feature classes. The green ovals represent feature classes which have been created. The model generally runs from right to left, and top to bottom.

Bear Study Area Map
Fig 3.6: Bear Study Areas Within Marquette County, Michigan

     The map above displays the features described in Step 6. Streams are a very important part of bear habitat. As seen in the map, most of the bear locations occur within the buffer zone of 500 meters of a stream. Also important is the 5 km buffer which can be seen by the change in color of the habitat area. Bear locations are much more common outside the buffer zone than within. This seems appropriate because bears tend to stay away from humans. The DNR management areas which overlay with the suitable bear habitat, are shown in a dark purple. This area is small enough where there is no point where a bear was tracked in this area. This has probably more to do with the area being pretty small than anything else. The ideal bear habitat is found in both the dark and light purple areas.

Sources

Michigan Geographic Framework: Marquette County. (2014, June 1). Retrived December 9, 2016 from http://www.mcgi.state.mi.us/mgdl/framework/metadata/Marquette.html

Wildlife_mgmt_units. (2001, August). Retrieved December 9, 2016, from http://www.dnr.state.mi.us/spatialdatalibrary/metadata/wildlife_mgmt_units.htm 

Michigan 1992 NLCD Shapfile by County. (2002, January 11). Retrieved December 9, 2016 from http://www.mcgi.state.mi.us/mgdl/nlcd/metadata/nlcdshp.html

Sunday, November 20, 2016

GIS Extra Credit: Geography Open House

Overview
     The geography open house is an event which took place from 5:30 -7:00, Thursday, November 17 and was located in Phillips Hall. The open house was an informal event where individuals could talk with geography students and professors in order to learn about more about what UW-Eau Claire has to offer in the Geography and Anthropology department. Students presented on their capstone projects and on geography courses. Below are a few of topics which were presented.

Drones
     One student used a drone to collect data about the water content of vegetation at the Eau Claire Golf and Country Club. Using the data collected, he displayed the data on a map which showed areas that were dry, areas that the sprinkler missed, and the difference between water content mainly on grassy surfaces. After analyzing the data, he found that fairways were severely under watered when comparing them to the greens.
     A different student used drones in a different matter. He used one to collect footage of the large flooding event on the Eau Claire river which occurred earlier this fall. Footage was collected and then displayed in a short film. He was using his own drone, and was presenting for geography 390 - Unmanned Areal Systems to show how the topics learned from this course can be applied to real world events.

Ground Penetrating Radar
     Another student worked with a team this past summer to locate a lost Jewish man's body buried since the Nazi's buried him in World War II. Based off of eye witness and written accounts his team was able to identify a possible location for this lost body. They used ground penetrating radar across a large area to collect the data needed to search for the exact location. This student searched through the data to find anomalies in the soil layer to pinpoint the most likely location of where this body could be found. Him and his team then identified this spot and began digging. They found a tombstone and a skeleton. Currently the DNA from the skeleton is being tested to see if it matches family members to see if it indeed the correct skeleton.

Remote Sensing
     One student presented on our department's quantitative methods course (geography 370). He wanted to identify the change in a landscape over time. He used areal photographs taken 30 years apart and then used a computer program to identify the pixels which had changed the most over this time. Most of the change was due to urbanization, although vegetation also played a role.

Conclusion
     All of these presentations were very informative and showed how applicable different skills can be in the job market. The students who presented their findings using data collected from drones kept on reiterating this point. The drones were very cool. They are a fairly new and growing field to go into as well. The time taken to attend was well worth it as the information taken in was very significant.



Saturday, November 19, 2016

Lab 2: Downloading GIS Data

Goals and Background
     The purpose of this lab is to download data from the U.S. Census Bureau’s website and then use this data to produce and share demographic maps with UW-Eau Claire - Geography and Anthropology. Goals include creating Joins between tables, formatting excel files correctly to add them to ArcMap, sharing and publising a map using ArcGIS online, and properly downloading .csv files from the U.S. Census Bureau’s website. Data downloaded was searched and looked at before it was downloaded in order to see which data sets were adequate to produce into maps.
 
Methods
Step 1: Download 2010 U.S. Census data, and the Wisconsin shapefile from the U.S. Census Bureau.

     This included navigating to the U.S. Census Bureau Fact Finder Website found here. Then, by browsing through different datasets and by applying proper filters, two datasets where chosen to be downloaded. These data sets are titled TOTAL POPULATION with the ID P1, and TENURE BY AGE OF HOUSEHOLDER with the ID H17. Then, the Wisconsin shapefile was download from the same website.

Step 2:  Join the downloaded tables to the Wisconsin shapefile.

     First, the .csv files had to be properly formatted, and saved as a .xls in order to add them to ArcMap. Then, a Join was created between the Wisconsin shapefile and the annotation table. This was done using the GEO_ID field as the common key.

Step 3:  Map the data.

     Prior to making the map, a new field for both Population and Number of Rent Units was created which had a number datatype so it could be mapped. The datatype was originally string when downloaded straight from the U.S. Census Bureau’s website. Then, these fields were mapped from the Wisconsin shapefile on two different data frames using a quantities graduated colors map for both. Jenks classification method was used on both maps. The map which displayed population was given 5 natural breaks, and the map which displayed the number of renter occupied housing units was given 6 natural breaks. The number of breaks was chosen to make the map cartographically pleasing. A base map was added to the backdrop of  both maps along with a legend, title, scale-bar, and north arrow.

Step 4: Create and share a web map with the Population by county data.

     This was done by signing in to ArcGIS online through ArcMap using an enterprise login. After logging in, a feature service was created from the ArcMap document. In creating a service, the base maps and Joins had to be deleted because the map couldn’t be published with these. An item description was entered in before publishing to ArcGIS online. Once the map was published to ArcGIS online, the map settings had to be changed in order to show a Pop-up box when a county was clicked on. The Pop-up shows the title of the map, the name of the county, and the population of the county. After this, the map was shared with UW-Eau Claire – Geography and Anthropology.

Results
Population by county in Wisconsin
Fig: 2.0: Population by county in Wisconsin


   
     Above is the map that was published to ArcGIS online and shared with UW-Eau Claire - Geography and Anthropology. The Pop-up window shows the title of the map, the county name, and the corresponding population total. The full map can be found here.
Fig 2.1: Demographic information of Wisconsin by county
Fig 2.1: Left- Number of renter occupied housing units by county
Right- Total population in Wisconsin by county  

     Above are the two maps created in ArcMap. The spatial distribution of population increases as one moves from southeast to northwest. The same is true for the spatial distribution of renter occupied units. The reason why this relationship between the two variables correlates very well is fairly logical as one might expect that there would be more renter occupied housing units where there are more people. It is also important to note that there is an increased number of renter occupied housing units relative to population where college towns are present. These include but are not limited to UW-Eau Claire - Eau Claire county, UW-River Falls -  Pierce county, UW-Superior - Douglas county, UW-Stevens Point - Portage county, UW- La Crosse - La Crosse county, and UW-Plattville - Grant county.

Sources

Data Access and Dissemination Systems (DADS). (n.d.). American FactFinder. Retrieved November 19, 2016 from http://factfinder.census.gov/faces/nav/jsf/pages/index.xhtml?refresh=




Wednesday, October 26, 2016

Lab 1: Eau Claire Confluence Project

Goals and Background
      The purpose of this lab is to look at the spatial distribution of different data sets around the Confluence Project in Eau Claire Wisconsin. The goal of this lab is to be comfortable with using these data sets and map them in six separate maps overlaying the proposed Confluence Project site.
      I began this project by digitizing the proposed Confluence Project area with parcel data, and an imagery base map. Then, I viewed several different data sets accessed through the city and county of Eau Claire's geodatabases and mapped them with the proposed Confluence Project site. I looked at PLSS lines, land parcels, population density, zoning class, land cover, land use, political boundaries, hydrologic  features, and US Census features. Then, I read up a bit on the Public Lands Survey System (PLSS). I learned about how the PLSS divides up the land into sections and how they are important for townships. All of this was prepping me for making the six maps.

Methods
      The next step was to create the maps. First, I added the proposed Confluence Project feature and the imagery base map to all of the data frames. For the Civil Divisions map, I added the Civil Divisions, and County_Boundary feature classes. In the Census Boundary map, I added the BlockGroups, and Tracts feature classes. For the Census Boundaries map, I also used the the data to create a graduated colors maps with six classes using Jenks natural breaks method displaying population per square mile. For the PLSS Lines map, I added the PLSSqq feature class. For the Eau Claire City Parcel Data map, I added the Centerlines, Water, and ParcelArea feature classes. In the Zoning map I added the ZoningClass, and Centerlines feature classes. In the Zoning map, I also created a unique values map by grouping similar zoning classes together. For the Voting Districts map, I added in the VotingWards2011 feature class. I added labels and gave them halos to make them more legible. For all the maps, I changed the transparency on the appropriate layers to make the map neat and visually pleasing, and so you can see the imagery base map. I pulled all of the feature classes from the Eau Claire county or city geodatabases.

Results
Eau Claire Confluence Project Proposed Site
Fig 1.0: Eau Claire Confluence Project Proposed Site
      Looking at the maps we can identify some spatial aspects about the Confluence Project proposed site. In the Civil Divisions map, it is clear the the proposed site is located in a city (Eau Claire). The Census Boundaries map shows us that the proposed site is located in a tract where the population is between 3,801 and 5,037 people. The PLSS Lines map shows us that the proposed site is located in two different quarter quarter sections. The Eau Claire City Parcel Data map shows us the proposed site's proximity to the Chippewa River, and that the site takes up four different land parcels. The Zoning map displays the zoning classes which immediately surround the proposed site. Lastly, The Voting Districts map shows us that the proposed site is located in ward number 31, and shows the which wards surround the site as well.

Sources
City of Eau Claire and County of Eau Claire geodatabases

Lippelt, D. Irene. Underestimating Wisconsin Township, Range, and Section Land Descriptions.
      Wisconsin Geological and Natural Hisoty Survey, Education Series 44, 1-4.