Final Project: Finding Potential Locations for a New Biological Field Station

Introduction: Pigeon Lake—a biology field station—used to be an integral part of the University of Wisconsin-Eau Claire’s foundations of biology course as well as an educational retreat for other schools within the Eau Claire and Chippewa County. However, it recently closed. The purpose of this project is to find suitable land for a new biological field station in which 5 criteria are ideally met: 1) within forested area, 2) an appropriate distance from major roads, 3) nearby inland still waters (lakes, swamps, ponds, etc…), and 4) nearby moving waters (rivers and streams). These criteria are used in hopes they will provide the potential area for a new biology field station with diverse ecosystems to study and a greatly diverse resident organism population.

Data Sources: I originally gathered all necessary data from the UWEC database connection to Wisconsin DNR 2014 data and ESRI US Census data. This proved to have problems fitting together once projected. However, once I found feature classes within the Wisconsin DNR 2014 database to replace the ESRI data, the feature classes were able to line up properly. The earliest feature class I obtained from here was updated in 2011, so all feature classes are fairly recent. All metadata describes feature class attributes as the proper criteria needed for this project. I have no more remaining concerns surrounding the data.

Methods: Using model builder, I first projected all data layers to the NAD 1983 (2011) Zone 15N projected coordinate system (Figure 1). I then clipped all data layers by the study area layer (Eau Claire and Chippewa counties). I created a 1km buffer surrounding major roads and erased county forests within 1km of major roads. I then made a 3km buffer surrounding streams and a 5km buffer surrounding water bodies. All buffers were set to dissolve all internal boundaries. Finally, I intersected the county forest areas away from major roads, the streams buffer, and the water bodies buffer to create the data layer that contains suitable lands for a new biological field station.

Figure 1. The final model builder displaying the overall process of the project. This model builder was used to create the final map.

Results: The results yielded appropriate locations for a new biological field station if interest and budget allows for action (Figure 2).The locations fit all 4 criteria that helps to support biodiversity and diversity of habitats for educational purposes. 

Figure 2. Final map displaying possible locations for a new biological field station in Chippewa and Eau Claire County. 

Evaluation: This project contained very easy concepts and tools while also giving valuable information in the output. If I needed to do another project like this, I would not hesitate to agree to it. However, I did face some challenges during the process. Like I said previously, there was a problem with the original data from two separate sources that would not fit together once projected. There was also a technical issue of the model builder freezing up on the last intersect tool. However, that was mitigated by dissolving boundaries between feature classes created in previous steps. Eventually, the process ran correctly and the result was obtained. 

Sources: Wisconsin DNR 2014 Database

Lab 3: Determining Bear Habitat with Geoprocessing Tools and Python

Goal: The goal of this lab is to find suitable black bear habitat within a study area of Marquette County, Michigan using geoprocessing tools.

Background: The habitat is based on 4 criteria and geared towards DNR management purposes and thus takes several steps and tools to determine.

I first used a typical Data Flow Model Method to which there were several objectives:

1: Mapping XY Coordinates from an Excel File

2: Determining Forest Type in Which Bears were Found

3: Determining Legitimacy of Stream Proximity as a Suitable Habitat Criteria and Creating Feature Classes for the Criteria

4: Finding Suitable Bear Habitat Based on Two Criteria: Proximity to Streams and Forest Type

5: Finding Suitable Habitat Within the DNR Management Districts

6: Finding Suitable Bear Habitat Away From Urban Areas or Built Up Lands

Python was then used to find suitable bear habitat in a similar fashion to gain exposure to coding.

Objective 1: Mapping XY Coordinates from an Excel File

After previewing the bear locations data excel file, I noted that the data consisted of x, y coordinates and must be added to the map as an event theme. However, since the data does not have an ObjectID field, after the XY data was added to ArcMap, I needed to export the data and save it as a feature class on my personal lab 3 geodatabase.

Objective 2: Determining Forest Type in Which Bears were Found

To find the bear habitat, I generated a new feature class that has the ID number of the bear and the forest type in which it was found. To do this, I selected by location land cover types by bear locations. This layer contained only land cover areas that bears were recorded being within. I then joined the data from the land cover selection with the bear locations based on spatial location to get the bear and land cover feature classes on the same attribute table.

I found the top 3 habitat types (3 habitat types with the most bears found within the land cover type) by selecting by attributes for each of the 6 habitat types in the minor type field.

 The amount of selected attributes equals the amount of individual bears in each habitat type.

Objective 3: Determining Legitimacy of Stream Proximity as a Suitable Habitat Criteria

 Since it was indicated that bears may most likely be found close to streams, I tested this statement by determining if bears were, in fact, found close to the streams.

First, I selected by location bear locations within 500 meters of a stream with bear_locations (XY coordinates of individuals) as the target layer and streams as the source layer. 

I then created a feature class that represents distance from streams. I used the clip tool to clip streams at the boundary of bear_cover. I then used the Buffer tool to create a feature class that contained a 500 meter radius surrounding all streams within the bear_cover feature class. I named this feature class within500m_stream_area.

Objective 4: Finding Suitable Bear Habitat Based on Two Criteria: Proximity to Streams and Forest Type

To find the suitable areas of habitat, I used both land cover types in which bears have been located (bear_cover) as well as areas 500 meters from streams (within500m_stream_area). Since land cover types in bear_cover had multiple polygons within their boundaries, I used the dissolve tool to remove the internal boundaries. I then intersected the two criterial feature classes to find areas in common. The resulting feature class is possible suitable habitat for bears.

Objective 5: Finding Suitable Habitat Within the DNR Management Districts

If the DNR is interested in taking steps to maintain this habitat, they must be able to know the areas which areas of suitable habitat lies within their management land. To find this, I first used the dissolve tool to eliminate the internal boundaries within DNR management lands. I then used the clip tool to find only the areas that DNR land management regions and suitable habitat have in common.   

Objective 6: Finding Suitable Bear Habitat Away From Urban Areas or Built Up Lands

It may be of interest to the DNR to only include habitat at an appropriate distance from urban areas or built-up lands in their criteria for suitable bear habitat which they can manage. To do this, I first created a new layer by selecting by attributes “urban or built up land” from the land cover layer and used dissolve to remove the internal boundaries. I then used buffer to extend the boundary 5km from the actual urban or built up land cover. Finally, I used the erase tool to find the final bear management lands from the suitable bear habitat within DNR lands that were not included in the boundary of 5km from urban or built-up lands.

Using Python

I ran 3 functions in python to find suitable bear habitat. The first function created a buffer of 1 kilometer from streams. The second intersected the stream buffer and suitable habitat found previously in the lab. The final function performed an erase analysis on the suitable habitat and the buffer of 5 kilometers radius around urban areas found previously.

Results:

The results suggest that most bears prefer forest wetlands, mixed forests, and evergreen forests over all other forest types as these forest types had the highest amount of individuals within their boundaries. 

Forested Wetlands: 9 bears

Mixed Forest Land: 9 bears

Evergreen Forest Land: 5 bears

It was also found that streams play an important role in suitable black bear habitat as 72% of all bears in the sample were found within close proximity of streams. Distributions of bears was observed to be mostly within  the northern regions of Marquette County, away from urban areas which indicates this may be another important criteria of bear habitat and should be considered when determining suitable habitat within the county. Therefore, the final map of suitable black bear habitat in Marquette County, MI includes 4 criteria: 1) land cover (forest) types in which bears have been located, 2) areas within 500 meters of streams, 3) regions at least 5 kilometers away from urban areas or built up lands, and 4) areas within DNR management lands.

Figures:

Figure 1: Suitable Black Bear Habitat found using geoprocessing tools.

Figure 2: Data Flow Model of the steps in finding suitable black bear habitat.

Figure 3: Python code used to find suitable bear habitat.

Sources: Data Collected from Michigan Center for Geographic Information, USGS NLCD, and DNR (May, 2015).

http://www.mcgi.state.mi.us/mgdl/nlcd/metadata/nlcdshp.html

http://www.dnr.state.mi.us/spatialdatalibrary/metadata/wildlife_mgmt_units.htm

http://www.mcgi.state.mi.us/mgdl/framework/metadata/Marquette.html

Lab 2: Downloading GIS Data

Introduction: The goal of this lab was to become familiar on downloading and mapping data from an online source. This involved several objectives that included:

1. Downloading Census data and shapefiles from the US Census Bureau (for both Total Population in Wisconsin and a variable of my choosing)
2. Combining the shapefiles with the data in ArcGIS
3. Mapping the information and building a layout suitable for both maps

Methods:

Downloading
First, I downloaded the data and shapefiles I needed by visiting the American FactFinder database at http://factfinder2.census.gov/faces/nav/jsf/pages/searchresults.xhtml?refresh=t. I used the advanced search tools to find data that contained the total population census information and occupancy status information from 2010. I used the SFI because I only needed basic standard Census data. I then downloaded a shapefile map of Wisconsin that included all counties.

The data files were downloaded and then saved as an excel workbook for uploading to ArcGIS.

Combining Shapefiles and Data

After bringing in the shapefiles and data tables onto ArcGIS, I wanted to combine the shapefile with the variable I was mapping (Total Population and Occupancy Status seperatly). Before joining the attribute tables of the two files, I wanted to preview what this would look like to make sure no problems existed with the files and if they shared common attributes I could use to join them.
 I opened both attribute tables separately, then under Table Options in one of the table windows, chose “Arrange Tables” à New Vertical Tab Group.

The attribute table for Occupancy status was listing the titles of the attribute data as a feature ID. To solve this, I opened the document in excel and deleted the problem column.

 To permanently join the tables together, I then right-clicked on the shapefile in Table of Contents and under “Joins and Relates,” I chose “Join.” I chose the field to join the shapefile with the Excel Workbook Attribute Tables by GEO_ID. Because table joins link tables together based on common attributes, I joined the shapefile and the excel workbook of decennial census information by GEO_ID because they share the same geographic id field (GEO#id). 

Mapping 

 I mapped the total population data by quantities with a graduated color scheme. I mapped the total population per county. However, the total population attribute did not map at first because it was imported as a string field type. To fix this problem, I added a new field as a "double" field type. I right clicked this newly added field and brought up the field calculator and chose the field that contained the total population. I was then able to map total population of individuals in each county,

For the shapefile joined with occupational status data, I mapped number of vacant houses per county and normalized it by the total number of houses per county to get a vacancy rate. This was also mapped quantitatively with a graduated color scheme.

Results:

Wisconsin US Cencus Data by County 2010: Percent Vacant Housing and Total Population. End product maps.

Source: U.S. Department of Commerce, United States Census Bureau (2010). American FactFinder - Search. Retrieved March 17, 2015 from http://factfinder2.census.gov/faces/nav/jsf/pages/searchresults.xhtml?refresh=t.

Lab 1: The Confluence Project

Background: In 2012, Clear Vision Eau Claire--an initiative group working to develop a collaborative vision for Eau Claire announced  intentions to construct "The Confluence Project." This project will is supported in partnership by UW-Eau Claire and the Eau Claire Regional Arts Center. It will consist of a new community arts center and student housing and commercial retail as well as three performance spaces, galleries, studios, and more. This complex is to be located in downtown Eau Claire at the river's confluence across from Phoenix Park and is to start construction in 2014.


In this lab, I prepare a basic report that contains base maps with all relevant information surrounding the Confluence Project. In order to do this, I had to:

1. digitize a proposed site for the Confluence Project
2. write up a legal description of the site parcels
3. and build several maps containing this information.


Digitizing a Proposed Site:
I first created a new geodatabase and feature class to store the proposed site as a new feature. The attributes were collected in the feature "Proposed Site" by adding parcel area data to the map and editing the proposed site using a polygon and without the use of End Snapping and Vertex snapping to insure the vertices and nodes of the proposed site match. The two polygons traced the outline of the parcel area and were saved in the "Propsed Site" feature.

Proposed area feature for The Confluence Project(shown in orange)

Writing a Legal Description of the Proposed Site:
I used the identify tool to find the parcel information for the parcels the proposed site resided in. I read documents explaining how townships and towns differ, what civil divisions are defined as, and how sections are described.                  

Parcel Information using Identify 

Legal Descriptions:

Parcel 1:

Parcel Number : 02365  

PIN: 1822122703200042068

Street Number : 128  

Street Name : Graham Ave. 

Township: 27

               Range: 9

               Section #: 20

               PLSS-DESC: NENW2027N09W

Parcel 2: 

Parcel Number : 020357  

PIN: 182212270920004263

Street Number : 202  

Street Name : Graham Ave. 

Township: 27

               Range: 9

               Section #: 20

               PLSS-DESC: NENW2027N09W

Building Maps:

Civil Divisions: displays the civil divisions between the city of Eau Claire and the surrounding area within Eau Claire County. I added the county boundary and civil division classes as well as the proposed site feature classes to a world imagery basemap.

Census Boundaries: shows population densities around the proposed site for the Confluence Project with the proposed site, tracts, and population 2007 feature classes (population 2007 was normalized by square mile) added to a world imagery basemap.

PLSS Data (Public Land Survey System): shows quarter quarter parcel boundaries with the proposed site, PLSS quarter quarter feature classes added to a world imagery basemap.

City Parcel Data: shows the parcels the proposed site occupies and the parcels surrounding it. The proposed site, centerlines, parcel area, and water feature classes were added to a world imagery basemap.
Zoning: displays the zoning districts surrounding the confluence project. The proposed site, centerlines, and zoning classes (grouped into commercial, industrial, public park, residential, and transportation) were added to a world imagery basemap.
Voting Districts: displays the voting districts in the area surrounding the proposed site. Proposed site and voting wards 2011 feature classes were added to a world imagery map. All voting districts are labeled on the map.

All six maps were laid out on a 11x17 landscape page format complete with legends (where necessary), scale bars, north arrows, and separate titles. The source (City of Eau Claire and Eau Claire County 2013) is written on the bottom right of the document. The file was exported as a .jpeg with 300 dpi.

Results: