Lab 5: Mini-Final Project

Introduction

After downloading forest fire data from the Wisconsin DNR I noticed that there were a number of fires which occurred outside of the protected regions of the state. Based on this I decided to ask the following question: In what areas outside of the already protected area of the state should there be increased protection from forest fires based on fire occurrences from 2007 and 2008? 

My objectives were to first select only the fire data from 2007 and 2008 and create a buffer of 5 miles surrounding each fire to better determine areas where there needs to be increased fire protection. The results from this project would benefit the Wisconsin DNR to develop more extensive fire protection areas within the state. 

Data Sources

In order to answer my study question I needed a variety of data. This included: fire occurrences, fire protection regions, national forest and county forest areas as well as state outline data. I gathered my data from the Wisconsin DNR website where they have all data files they offer to the public. This website I used was: ftp://dnrftp01.wi.gov/geodata/. This data was collected from a reputable source and therefore I was not concerned with the accuracy of the data. However, there was some concerns about the age of the data because it was difficult to determine the age of fire protection region.

Methods


For this lab, I used a variety of tools including: buffer, intersect, erase, union, dissolve and clip. I first started by intersecting the county and national forest datasets to produce one region of fire protection. I then created a new feature layer which selected just the intensive and extensive protection types of fire protection region. I then used the union tool to combine the intensive and extensive protection types to the county and national forest protection areas. Next, I created a new feature layer which selected the 2007 and 2008 fire occurrences from the total occurrence dataset. Once I had the new feature class, I applied a buffer of 5 miles around each of the new data points. After this was done, I erased the fire occurance data points from 2007 and 2008 which were included in the regions of fire protection. Once this was done I was left with only data values which existed outside of fire protected regions in the state. With these regions being buffered they needed to be dissolved so the points appeared continuous. To finish I clipped the data values to make sure that the new regions were all within the state of Wisconsin.

Results


My results from this lab can be seen in the map above. After I ran the model shown in the methods section I produced figure 3 as shown in my final map. This shows the regions in bright green where there should be an increase in fire protection in the state besides those which are already protected.

Evaluation

Overall I enjoyed this project. It provided me with the challenge of not only creating a project of my own but also applying the skills I have learned throughout this course to complete said project. If I were asked to repeat this project I would change the amount of fire occurance data points used. I would increase them from only 2007 and 2008 to maybe 2000 to 2008 for a more complete result. By increasing the number of points taken into consideration I would be able to produce a more complete map of where there should be an increase in the areas which are protected from fires by the Wisconsin DNR. I faced a number of challenges when creating my model because I originally used a spatial join rather than unions. 

Lab 4: Vector Analysis with ArcGIS

Goals

The goal of this lab was to use various geoprocessing tools in order to utilize to perform vector analysis in ArcGIS. For this particular lab I was put in the scenario that I was to help the DNR develop a map showing suitable habitat for bears in Marquette County, Michigan.

Background

To start this lab I was given a number of GPS data points which showed recent bear locations within the study area of Marquette County, MI. I also was given a number of parameters which I needed to take into consideration when developing a map of suitable habitat. These included: stream proximity, land cover type and distance from urban/built-up areas. Another big issue to consider was to make sure that my proposed area was within the DNR management zones which was data also given to me.

Methods

The first step in this process was to develop the land cover types which most typically contain bears. Using the bear location data and the land cover data I was given, I spatially joined the feature classes. Once they were joined together I summarized the land cover feature in order to determine how many bears (based on the location data) were found in each habitat/ land cover type. From the summary I learned that most bears were found in either mixed forest land, forested wetland or evergreen forest land. I then exported these three particular land cover types as the bear_land_cover feature class. 

Next, I needed to determine how many of the bears were found near streams to see if they typically were or were not located near these features. To do this I used the select by location tool to select the bears from the location data which are within 500 meters of a stream. After I did this I found that 49 of the 68 bear locations  (72.06%) met these parameters. Based on this value of 72%, I was able to conclude that streams are of high importance when determining bear habit since biologists would consider a feature such as this important if the value was greater than 30%.

The other important factor that the DNR requires consideration of when working to produce a potential habitat for these bears is whether the areas are within the DNR's management zone. The DNR only has certain zones in which they can set up management areas so it is important to take this into consideration when I develop a final proposal. In order to get the zones ready to analyze I clipped out any of the zones outside of the study area. 

To produce my final proposed bear habitat regions I had to take into consideration the following characteristics: must be within either mixed forest lands, forested wetlands or evergreen forest land, must be within 500 meters of a stream and must be within the DNR management area. The first step was to buffer the streams at 500 meters and intersect these three features to come up with a potential habitat for the bears. However, another factor I need to take into consideration is the proximity these areas have to urban/built-up areas. It is best to keep the bear management lands at least 5 kilometers from any urban areas. After exporting the urban/built-up land class from the overall land cover data I buffered it by 5 kilometers. Next I used the erase tool to eliminate all of the possible bear habitat within 5 kilometers of any urban areas. Once this was done I was left with a final proposed habitat (or in my data model, final_output).

Results

The final map that I produced as a result of the last model can be seen below. This map clearly shows both the potential bear habitat within this region of Marquette County, MI and the areas where I would propose there should be habitat. The proposed habitat is within 500 meters of a stream, within the DNR management zones, at least 5 kilometers from urban/built-up land and located in either mixed forest lands, forested wetlands or evergreen forest land. While there are still some bears which are located in the north-west region of the study area there is no proposed habitat there because overall there is very little DNR management zones in that region of the study area. It would be good based on the lack of DNR management in a part of the study area where so many bears have been found to possibly expand the management zones to better protect more bear habitat. The pink, "potential habitat" includes the proposed habitat as it is the habitat before taking into consideration the proximity to urban/built up land. Based on the results from this lab I would propose that the regions shaded in green are the best regions where bear management zones should be in this portion of Marquette County.

Sources

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

Michigan DNR:
http://www.dnr.state.mi.us/spatialdatalibrary/metadata/wildlife_mgmt_units.htm
http://www.mcgi.state.mi.us/mgdl/framework/metadata/Marquette.html