Lab 3- Vector Analysis with ArcGIS

Goal: 

The goal of this lab was to use ArcGIS geoprocessing tools to understand vector analysis and apply this skill to a real world example: determining suitable bear habitat in a section of Marquette County, Michigan

Background:

 The purpose of this exercise was to gain familiarity with ArcGIS geoprocessing tools such as intersect, buffer, and dissolve for vector analysis.  The combination of these tools is a powerful skill that create maps for a variety of purposes.  In this lab, the question asked was where suitable bear habitat is located within DNR management zones.  Using land cover, the study area, and bear locations derived from GPS trackers, this question was fully answered and displayed in map. 

Methods:

Objective 1:

For this objective an excel file of bear locations in a study area in Marquette County, Michigan were exported into ArcGIS as an event theme.  Because an event theme is only a temporary display of X,Y coordinates, the event theme was exported as a feature class that would then give it an ID number allowing spatial analysis.  This feature class was saved in my geodatabase for this lab.  

Objective 2:

In this objective, the goal was to figure out what type of land cover the bears were in when their position was recorded with the GPS.  To do this, The bear locations and land cover feature classes were intersected putting the land cover data for each bear point within that attribute table.  Afterwards, the 'minor type' field was summarized to show the top three habitats bears were found in: Mixed Forest Land, Forested Wetlands, and Evergreen Forest Land.  

Objective 3:

To find out if bears are mostly found around streams, a 500 meter buffer was created around the streams in the study area.  In order to get an accurate count, the buffer zones around the streams were dissolved into one polygon.  Then this polygon was intersected with the bear locations to find how many bears were withing 500 meters of a stream.  Doing some simple statistics, approximately 72% of the bears recorded were found within the buffer around streams.  

Objective 4:

Based on the previously found information, suitable areas for bear habitat were calculated.  In the land cover feature class, a query was done for Mixed Forest Land, Forested Wetlands, and Evergreen Forest Land, the three major land covers bears were found in.  This selection was then intersected with the 500 meter buffer zone around the streams.  The resulting feature class showed favored bear habitat within 500 meters of a stream.  To make this all one area, the dissolve tool was once again used.  

Objective 5:

In order to determine what suitable bear habitat is found within DNR management lands, the management lands were intersected with the suitable bear habitat found in objective four.  First, the management lands were clipped to only include those found in the study area and those lands were dissolved into one polygon.  The final product was a feature class of the suitable bear habitats that could be managed on DNR lands.  

Objective 6:

The final task was to find bear management areas that were at least 5 kilometers away from urban or built-up land.  In the land cover feature class, a query was done for 'urban and built-up land' which was then exported as its own feature class.  The dissolve tool was used to make this one polygon.  Then the buffer tool was used to make a 5 kilometer buffer around these areas.  Finally, the erase tool removed all bear habitat that fell within the urban and built-up land buffer.  

Objective 7:

For the last objective, basic coding in python was used to perform some of the earlier geoprocessing that was done in ArcGIS.  A buffer was created around streams of 1 kilometer, an intersection of the suitable bear habitat and the stream buffer was completed, and the suitable lands near the urban area buffer was removed using the erase function.  All these were simple tasks, shown in Figure 1, that introduced the concept of using python.

Figure 1. Python script performing a buffer, intersection, and erasure.

 

Results:

The map in Figure 2 shows all suitable bear habitat based on land cover, proximity to streams, within DNR management zones, and at least 5 kilometers away from urban developments.  Most of the bear habitat falls to the northwest of the study area within Marquette County besides one section the the northeast.  These habitats are represented as many small sections across the state.  From a management perspective, these areas would have to be connected by corridors to allow the animals to move from place to place, especially in the case of a bear.  Providing natural corridors for the bears to use would also decrease the chance for human-bear interactions.  These results are only based on 68 tracked bear locations, which can only tell so much about where the bears prefer to be.  The bear habitat may change from season to season and year to year depending on where their food supplies migrate to, which should be considered before designating land for bear habitat.  The data flow used to create this project was recorded using Microsoft Visio software and depicted in Figure 3.

Figure 2. Map of suitable bear habitat in Marquette County, Michigan created using ArcGIS.

Figure 3. Data flow model showing the steps used to create the map of suitable bear habitat within the study zone of Marquette County, Michigan.

Sources:

State of Michigan Open GIS Data, State of Michigan, 2017, http://gis-michigan.opendata.arcgis.com/. Accessed 4 May 2017.
 Michigan 1992 NLCD Shapefile by County, State of Michigan, 2002, http://www.mcgi.state.mi.us/mgdl/nlcd/metadata/nlcdshp.html. Accessed 4 May 2017.
 Wildlife_mgmt_units, Michigan Department of Natural Resources (MDNR), 2001, http://www.dnr.state.mi.us/spatialdatalibrary/metadata/wildlife_mgmt_units.htm. Accessed 4 May 2017.
Michigan Geographic Framework: Marquette County, Center for Shared Solutions and Technology Partnerships, 2014, http://www.mcgi.state.mi.us/mgdl/framework/metadata/Marquette.html.  Accessed 4 May 2017.