Projections and Data Collection

The lab this week consisted of two parts: projections and data collection.  Both of these things important for being able to create useful and meaningful maps.

Projections 

For the projections I downloaded a map of Florida's counties from FGDL.org, a resource for managing and distributing GIS data for the state of Florida.  The initial map I downloaded used the Albers projection.  Projections transform GIS data so that it can be mapped on a two dimensional surface.  The projection used determines how things like the shape or area, so it is important to use the correct projection when designing a map.

To explore how different projections looked, the task was to reproject our Florida map to the UTM 16 and Florida State Plane N projections.  This was done using the geoprocessing tool "project" tool and inputting our original map and choosing the new coordinate system to create the reprojection.

A map with three projections of Florida highlighting the differences in area between the projections.

The different projections looked fairly similar, but we further investigated the differences by considered the area in square miles of four different counties: Alachua, Escambia, Miami-Dade, and Polk.  Between the three projections the area of the counties was not always calculated to be the same.  This is especially apparent when looking at the Albers versus the UTM 16 projection.  The UTM 16 projection is best for the western part of Florida, so eastern counties like Miami-Dade are more distorted and thus the measurements are less accurate.  The Albers projection is designed to consider the entirety of the state of Florida so no one part is considerably more distorted than the rest.  Florida State Plane North has a similar problem to UTM 16, though it is less exaggerated.

We also learned how to handle raster projections and input their correct coordinate system to have them properly display on the map so the image doesn't show up in the middle of the ocean.

Data Collection

The other half of this lab involved data collection.  To begin, I downloaded the ArcGIS collector app onto my smartphone.  The assignment was to collect location data on public safety features and describe their condition (good, fair, and poor).  I chose to gather data on the fire hydrants located in the Oakland neighborhood of Pittsburgh, PA.

Before going out into the field I had to set up my map so that it was ready to collect data.  To do this, I created a condition domain and which had the three different choices with a description of what the choices meant.  I then created a new feature class for the fire hydrants I would be documenting.  This consisted of the condition domain, a raster field for photos, and a text field for any notes I may have.  I also set the symbology for the symbols, with blue being the best condition, green being fair, and red being poor.  I enabled editing on the file so I could collect data using my phone and I was all set to begin my field work.

I choose fire hydrants because they are plentiful and by their nature have to be accessible so I didn't have to worry about trespassing.  I collected just over 40 data points and only got a couple of funny looks when I took pictures!

A collection of fire hydrant locations and conditions in Oakland, Pittsburgh.

Did you know that the colors on fire hydrants have meanings?  There are no national regulations on fire hydrant colors, but in Pittsburgh the color of the body of the fire hydrant relates to the size of the main that the hydrants connection to (source).  Red connects to the smallest, yellow the medium, and green the largest pipes.  Dark blue also exists in Pittsburgh, but only for waterworks purposes which probably explains why I did not see any.  The caps relate to how much pressure the hydrant provides with orange being the lowest, followed by white, and then blue.  So having hydrants with unambiguous coloring is important to the ability to fight fires.  In cases were a white cap was badly discolored by rust to the point of it being orange I would mark these as poor quality as in an emergency ambiguity is the last thing you would want to deal with.  If I was doing this as a formal survey I could add additional domains for the body color and cap color so that someone could easily query the different types.

We were instructed to try out creating map packages via ArcGIS Pro and also importing our data into Google Earth via .kml files.  Of the three, ArcGIS Online was my favorite to use because it is straightforward and intuitive.