GIS 3015 Lab 7: Choropleth Mapping

The seventh lab of Cartographic Skills further explored the role one's choice of classification method plays in presenting data.  Additionally, we explored the strengths and weaknesses of choropleth maps for communicating data.  In the simplest of terms, a choropleth map features units shaded according to a value (e.g., population, median income).  This seems simple enough, yet choropleth maps can easily misrepresent data if one isn't careful.  To ensure data are represented accurately, they must be standardized.  As an example:  Rather than symbolizing a simple population count of a State, the population value could be divided by the state's area to get population per square mile (or whatever unit of measure one chooses).  Another potential weakness of choropleth maps is the assumption that the represented values are homogeneous across each unit (e.g., assuming crime rates are constant throughout each state).  Values may not be as uniform as represented.

The maps above were created using both ArcMap and Adobe Illustrator.  After setting up the map in ArcMap (turning on State labels for the color map, choosing the natural breaks classification method, creating the legend, etc.), Illustrator was used to finalize the end product (adjusting the size and position of Alaska and Hawaii along with the State labels of New England, for example).  The greyscale map was created mostly in Illustrator by converting the color map to grey and deciding the best way to classify each regional division based on calculations completed in Microsoft Excel.  This map focuses on regional population trends as opposed to the more refined State level focus of the color map, although the latter also indicates similar regional trends.

GIS 4043 Lab 6: Projections Part II

Our lab project for week six proved to be the most difficult yet rewarding thus far in the course.  We were required to learn and make use of multiple distinct GIS skills to produce the required map.  First, we needed to choose two adjacent Quads in Escambia County, Florida and download the eight associated digital orthophoto quarter quads from labins.org.  Making note of the coordinate system used, we then defined the coordinate system of the downloaded files to the Florida Stateplane system.  We added vector files representing Florida's major roads, Quad index, and county boundaries that were previously downloaded from FGDL.org after using the Project tool to reproject the vector files' coordinate system to match the DOQQs.  We also used Microsoft Excel to calculate the decimal coordinates of a dataset provided in degrees, minutes and seconds; the spreadsheet data was added to the map with a coordinate system of WGS 1984 defined, then exported as a shapefile, and finally reprojected to the Florida Stateplane system.  Having accomplished the above steps, we were required to produce a map document clearly communicating this information.

Although this lab was certainly a challenge, that we were required to accomplish the above tasks with minimal specific instruction made me feel much more confident in my abilities in GIS.  As noted in the lab instructions, the ability to find and incorporate available online GIS data is a valuable skill, and this lab was an excellent introduction.

GIS 3015 Lab 6: Data Classification

The primary objective of our lab for week six was to experience the differences between classification methods in how our data are presented.  The first map above presents the same data filtered through four different classification methods.  None of them lie, but all tell only versions of the truth.  Being aware of how one's chosen classification method affects data presentation and audience perception is crucial in understanding how to communicate your analysis.

As can be seen above, the four classification methods focused on for this lab were:  Natural breaks, equal interval, quartile and standard deviation.  The equal interval classification scheme is determined by taking the data range (from the highest to the lowest sample value) and dividing it by how many classes one has chosen (e.g., five in the example above).  This produces a classification that is not  responsive to the data; some categories may have the bulk of samples while others may have few or even none.  The quartile method is, in a sense, the opposite of equal interval by creating classes such that each have the same number of samples.  Standard deviation, as the title suggests, creates categories based upon statistical standard deviations.  This method, while certainly useful for some people, is more difficult to interpret than quartile and equal interval.

The second map above focuses on the natural breaks classification method.  This method was designed to create classes based upon how the data are grouped; ideally, each class represents a data group and class breaks do not divide groups.  As noted in our lab instructions, this is ArcMap's default classification method and, in my opinion, most clearly communicates the given data.

GIS 4043 Lab 5: Projections Part I

The lab for week five saw us investigate the ramifications of using different coordinate systems for the same geographic data.  Among other activities, we were instructed to create one map with three dataframes illustrating the differences between three coordinate systems.  Switching between the dataframes in data view clearly demonstrated the differences as the map of Florida shifted.  Placing the dataframes side by side in layout view for our exported map made the differences more difficult to see.  Thus, we created a new field for the shapefiles' attributes that calculated the area for each county.  Four counties were then selected to display their area in each dataframe as a more precise demonstration of the differences between coordinate systems.  Doing this reinforced the importance of knowing the coordinate systems of one's datasets and how to harmonize them.  For example, layers with different coordinate systems may be displayed correctly together yet prevent use of analysis tools as well as consume more computing resources as they are drawn. 

For the design of the exported map above, I decided to keep it relatively simple.  As suggested in the lab instructions, I kept the symbology consistent across the dataframes.  My creative input was limited to applying a light blue background to the dataframes and a light tan overall background.  I was momentarily baffled as to how I could ensure a single scale bar would work for all three dataframes until I remembered to manually input the same scale ratio for each dataframe.

GIS 3015 Lab 5: Spatial Statistics

Lab five of Cartographic Skills forced us to dive into the dreaded field of statistics.  The major portion of this module involved the completion of the ESRI online course "Exploring Spatial Patterns in Your Data Using ArcGIS."  The above map was taken from the first exercise of that course with the finalizing addition of the map essentials (legend, scale bar, north arrow, title, author, date, and data source).

I am not very knowledgeable in the field of statistics, although I am aware of its utility in detecting trends and significant "hidden" characteristics of data collections.  I was pleased, therefore, to see the ESRI course explain the various statistical tools in ArcMap in easy to understand terms (for the most part).  For the map above, we utilized the Mean Center, Median Center, and Directional Distribution tools from the Spatial Statistics toolbox.  The Mean Center tool averages the locations of all input data (weather stations in this case) and creates a layer symbolizing the average location.  The Median Center tool does the same for the median location of input data.  Although not identical, the similar locations for the mean and median centers indicate the data distribution is close to normal.  Finally, the Directional Distribution tool creates a layer indicating the orientation of the data distribution.  In this case, the orientation of the weather stations is primarily east-west with a slight northeast-southwest tilt.

As with many of the powerful tools built in to ArcMap, merely scratching the surface of the Spatial Statistics toolbox reveals the many options one has to analyze whatever types of data are available.  

GIS 3015 Lab 4: Typography

Our fourth Cartographic Skills lab continued our odyssey into using Adobe Illustrator to create maps beyond what could be done with ArcMap alone.  As in the previous lab, this week we worked solely with Illustrator to design our map document.  Our focus this time, however, was typography and proper labeling techniques.  We were given a basic, unlabeled map of the Marathon region of the Florida Keys and were required to label the keys, cities, bodies of water and additional selected features.  We were given no further limitations on how to complete the task other than to follow basic labeling and typographic guidelines.  On the contrary, we were encouraged to be creative with our map as long as these guidelines were followed.

I must say, however, that this map proved to be frustrating to label.  The geography of the keys make consistent, clear labeling difficult.  For example, the narrow keys allow for few labels to be placed within the island outlines, yet I did not want to place all the labels over water with multiple indicator lines radiating in every direction.  In hindsight, the latter may have been the better, more consistent option.  Instead, whenever possible I placed the Key labels within the landmass outlines while the remainder were placed over water with indicator lines.  I decided not to use a symbol to denote Key location as I believe the geography makes their extent clear.

Regarding other design choices, I made use of the 'pen' tool to create paths without fill or stroke then used the 'type on a path' tool to create the curved water body labels as well as the labels for Deer Key and Grassy Key.  I used the 'smudge stick' effect on the blue background and a 'canvas' effect over the artboard to make the map more interesting (hopefully without detracting too much from the map itself).  I used the title to experiment with a variety of different effects, settling finally on a combination of feather, outer glow and drop shadow.

Although Illustrator was challenging at first, I am beginning to learn more and enjoy experimenting with different options and effects.  This has become a challenge for me, as I probably spent too much time playing with different effects without much improvement in the final map.  As I learn more, I hope to become more efficient with the program.

GIS 4043 Lab 4: ArcGIS Online and Map Packages

This week's lab gave us an opportunity to interact with ESRI's online functionality and training.  I already had an account for training purposes for an earlier course but had never registered with ArcGIS online or accessed ArcGIS online through ArcMap.  Initially, I was unable to register and was forced to reset my password two times before I successfully connected with ArcGIS online.  I had no problems connecting thereafter, although ArcMap did kick me off at one point and required a second login.

The major portion of this lab involved working through sections of the ESRI online course "Creating and Sharing Map Packages for ArcGIS 10.1."  This course served as a good introduction to utilizing ArcGIS online functions via ArcMap and web browser.  The ease with which maps can be created, edited and shared by and for multiple parties is impressive.  I look forward to experimenting further with ArcGIS online and learning more of its potential.

The screenshots below show the item description of the maps uploaded as part of the ESRI online course.  This provides another example of the importance of good metadata; I would not enjoy wading through search results that lacked useful descriptions and source data information.