Geog483

Final Project

John Middendorf

 

For our final project in Geog 483, we are determining priority conservation areas within Centre County based on the following criteria:

 

 

We are given the necessary data to compute suitable areas, namely:

 

All data uses an Albers Conical Equal Area Projection, with standard parallels of 40 and 42 degrees, and a central meridian longitude of  78 degrees west.  The datum is NAD 27 using the Clarke 1866 ellipsoid.  Units are in meters.

 

INTRODUCTION AND PROCESS

The general plan in this project is to convert the vector shapefiles into raster data which are then reclassified into yes/no divisions (yes = 1) based on the criteria stated above. Once all the qualifications are in raster data, all rasters are then multiplied together using Spatial Analyst to calculate a final raster map of suitable areas.

 

The first step involved finding study sites with the appropriate biodiversity (e.g. bird and mammal species greater than 75 total).  This involved joining the Studysite shapefile with the Species richness database using a common attribute.  Then it was simply a matter of adding a field of total species, and selecting the proper records, and creating a new shapefile.

 

This step eliminated all the study sites which did not meet the biodiversity criteria, simplifying future steps.  Next the “Select by Location” in the main menu was used to find roads that intersected with the suitable study sites:

 

Above: After intersecting roads with study sites that have greater than 75 species.

 

Next up was to buffer the roads appropriately (roads had a 20 meter buffer, highways had a 50 meter buffer, and interstates had a 100 meter buffer), using the Arc 8.x Buffer Wizard after creating a new field with the appropriate buffer distance.  Here, we did NOT use dissolve, because it was simpler  to use the subsequent steps to calculate road areas.

 

Above: Buffered roads that intersect with suitable study sites (un-dissolved buffers).

 

Once we had the buffers, a Union operation was performed in ArcCatalog using the buffered roads and the Studysites that met the biodiversity criteria:

 

Above: After Union operation giving a single layer of adjacent shapefiles.

 

Next was a process to find only the portions of the studysites with the buffered roads eliminated.  This was easily done by examining the Attribute data and selecting records (which represent shapes in the layer) that did not have a buffer distance after the Union operation.  Once the non-buffered records were selected, a new vector file was created:

 

Above: Suitable Study Sites with buffered roads eliminated.

 

To calculate the area that the buffered roads occupied, it was a matter of knowing the original size of the study sites (given within attribute data), and subtracting the new area.  Once the road area was calculated, the percent road area was found.  All this was done in attribute data of the new suitable study site shapefile (sans buffered roads), by creating three new fields and using the Vbasic code to calculate area (the code is easily found in the help box associated with the Calculate Values command).

 

Above:  Result of calculating percentage of roads within each study site, with suitable sites selected.

 

Above: the study site map of the above attribute data selection (percent roads less than 10 percent).

 

Finally, we created a new shapefile (Export data) of new study sites with less than 10 percent buffered road area.  This was our last Vector operation, all future operations were done creating Raster data.

 

Above: Suitable study sites after road and biodiversity criteria applied.

 

RASTER OPERATIONS

Initially, I chose a 100 meter grid:

 

Above: 100 meter grid specification for Spatial Analyst.

 

I then creating a grid from the processed suitable study sites without buffered roads completed above:

 

Above: Raster grid of suitable study sites without buffered roads.

 

Above: close up of raster grid of suitable study sites without buffered roads. Grid size equals 100 meters.

 

Next was a series of raster grids.  These were simply a matter of creating a raster grid from the given shapefile, then reclassifying based on requirements and making permanent.

 

Above: Reclassified Habitat grid.

 

Above: Reclassified Landuse grid.

 

Above: Reclassified Ownership grid.

 

Above: Reclassified Slope grid.  Grid size = 100 meters.

 

FINAL CALCULATIONS

Using Spatial Analyst, with the above 5 raster grids processed with suitable areas given a value of 1 and unsuitable areas with a value of 0, it was a simple matter to multiply all raster grids together to get the final suitable area map:

 

Above: Suitable areas using 100 meter grid

 

This gave us the number of 100meter by 100 meter cells that met all the suitable criteria:

 

.

Above: count of suitable area cells of raster calculation based on 100 meter grid.

 

Using the number of cells (7179), it was a simple matter to calculate square miles of suitable area, giving us 27.72 square miles, using the 100 meter grid.

 

REWORKING THE DATA

After looking in detail with the results, I found the 100 meter grid too coarse, so picking up from the beginning of the raster manipulations; I looked at a 50 meter grid and a 20 meter grid:

Studysite without buffered roads using 50 meter grid.

 

Above: Studysites without roads using a 20 meter grid.

 

The 20 meter grid, on the same scale as the buffered roads, appeared to be the correct level of detail.  So next, I processed the same steps as above, but with a 20 meter default grid size specified in Spatial Analyst.

 

Above: Suitable areas using a 20 meter cell size.

 

With a 20 meter grid, there were 151631 suitable cells, giving 23.42 square miles of suitable area.  Unfortunately, the finer grid also seemed inadequate, because the slope detail using a 20 meter cell broke up the contiguous areas excessively:

 

Above: using a 20 meter cell size, the slope requirement fragments  contiguous areas.

 

As a compromise, for my final result, I used a 20 meter grid for all grids EXCEPT for the slope grid, which used a 100 meter grid prior to the final Spatial Analyst calculation for the final suitable areas.  My logic here was that the slope requirement applied more appropriately to larger areas than the other requirements.  Using the combined grid sizes, the calculation of the final suitable region came out at 27.49 square miles.

 

Above: Suitable areas using a 20 meter cell size for roads, forests, habitat and ownership, and a 100 meter cell size for slope. Compare with the image above which uses a 20 meter cell size throughout.

 

FINAL PRESENTATION

 

Below are the final presentation maps:

 

Above: Final presentation showing suitable sites, with three regions shown in more detail.

 

Above: A layout showing five criteria used in selecting the final sites, to accompany the main layout map, above.

 

DATA SOURCES: Penn State Online course data provided.

 

COURSE CONCLUSION (to instructors)

This course has been a great way to get to know some of the advanced features of ArcGIS, which should put me in good stead as I progress with a position with the county’s assessors office.  The biggest challenge will be converting all parcel data from an existing MIMS database into ArcGIS and ArcEditor.  Subsequently, we plan to create interactive web pages sourcing the data so citizens will be able to access parcel data online.  Leads to anyone with experience and knowledge of MIMS conversions would be appreciated.

 

Also, I will not be continuing the program this coming semester, but I will be picking it up again this summer for the subsequent semester.  Please advise if I need to make arrangements for a break (my wife and I are off for our honeymoon!).

 

Thank you.