Tuesday, May 2, 2017

Final Project

It's hard to believe it, but it is already time for the final project. As with any final, we were required to utilize all the skills we've covered throughout the semester in the completion of the project. The project entailed creating a map displaying two thematic data sets over one geographic area along with creating a map with visual hierarchy and proper cartographic design principles in use. There were two options to choose from for the final project, the first option was to map provided data on mean SAT scores and participation percentages in the United States for the year 2014, and the second option was to find our own data and topic of mapping. I chose to use the provided data and create a map depicting the SAT data compiled by The College Board.

In order to display the two data sets, mean composite score and participation percentages, together, two different thematic methods were used. Participation percentages were mapped using choropleth techniques, and the score data mapped using graduated symbols. In order to import the score and participation data into an ArcMap document, Excel spreadsheets were created and then the “join and relate” tool was used to sync the state-by-state data to the corresponding state. A base map of the United States was obtained from the U.S. Census website and utilized in the process of joining the score and participation data.

The participation data was classed into five different classes using the Natural Breaks/Jenks method. Five classes were deemed appropriate upon examining the data and determining that five classes most efficiently separated the data into logical groups without placing too many states into one group or another. The Natural Breaks/Jenks method of classification was chosen because the method uses an algorithm to class similar data values together and emphasizes the differences between classes. This method created a visualization of the data that allowed for easy identification of states with the highest, lowest, and similar participation rates.

The mean composite score data was symbolized using graduated symbols. While the score data could have also been symbolized using the choropleth technique, using graduated symbols and a vibrant color for those symbols created a visual hierarchy that placed emphasis on the scores for each state rather than participation percentages. The maximum and minimum size for the symbols was chosen so that the smallest value was easily visible and that the maximum value symbol would fit within each state and still leave enough room for labeling each state. Abbreviations were not used for the states to prevent users unfamiliar with state abbreviations needing to lookup abbreviations for states of interest. The score data was classed in a similar fashion to the choropleth data, five classes using the Natural Breaks/Jenks method of classification. The same procedure used for participation data was used for the score data, a visual inspection of the data followed by experimentation with class number and method, before determining that five classes using the Natural Breaks/Jenks method was appropriate for the data.

The projection chosen for the project was the Albers North American Equal Area projection. This decision was based on the desired map area containing all of the United States and the inclusion of data suitable for choropleth mapping which is measured by enumeration unit (state), necessitating preservation of area. Because a projection that preserves area was chosen, moving Hawaii and Alaska from their exact geographic locations was necessary. Alaska is a large state and it was necessary to create an inset map to place Alaska along with Hawaii below the continental United States in order to ensure the continental portions were large enough to show the detail needed to display the data sets properly. Drop shadows were used to bring the states to the foreground and a bright yellow color used for the graduated symbols depicting the composite scores for each state. The choropleth data is displayed in shades of green to represent land mass and also the changing participation percentages between the states. The color change is gradual and a single hue of color is in use.

I thoroughly enjoyed the Cartographic Skills course and now that it is coming to an end, I feel that my GIS and cartographic skills have been greatly improved. Each assignment during the semester required me to utilize new skills and also those learned previously in Intro2GIS. I have to say that learning to use a new program, Adobe Illustrator, was my favorite aspect of the course. Prior to this semester I was restricted to creating and adding finishing touches to my maps in ArcMap with limited design choices, but now I feel confident that I can create visually appealing maps using Adobe. 

Friday, April 14, 2017

Module 12: The Future of Cartography & Google Earth

This week we focused on where GIS is headed in the future by looking at Volunteered Geographic Information (VGI), Public Participation GIS, Geocollaboration, Geotargeting and Cloud Based GIS and computing. All of the above topics relate to the movement of GIS practices to the web. Through crowd sourcing and various methods mentioned above, more and more people are contributing to map making than ever. While there are many perks to having more people contributing data to maps, data integrity becomes an issue. Through validation of maps and peer review (or methods deemed appropriate by individual organizations), the integrity of data can be maintained and detailed user friendly maps can be created.  

Public Participation GIS is used to encourage citizens to become involved with public decisions by contributing information or ideas in the form of maps, or feedback on supplied maps. This method is effective on getting feedback from communities on proposed building projects. Geocollaboration is a method of collaboration based around GIS that allows users to work together remotely and is frequently involved in emergency services. Geotargeting is used frequently by advertisers in an attempt to show consumers products specific for their region. The location of individuals can be obtained from the user, or by tracing IP addresses. Lastly, Cloud Based GIS and Cloud Based computing is a service that allows for large amounts of data to be stored remotely and accessed from anywhere. This is useful for projects where many people are working with the same data but in different physical locations.

VGI was the main topic of lecture this week and is the leading sources of user generated geographic information. In VGI individuals map areas independent of an overseeing organization and then distribute the information via the web, and this is what can lead to data integrity issues. OpenStreetMap and Wikimapia are two examples of map resources that have been compiled by users rather than by traditional data gathering means. Maps created on OpenStreetMap and Wikimapia tend to be user oriented, mapping biking routes, restaurants, points of interest, etc. But there are also humanitarian efforts that use VGI to map areas of the world that have yet to be mapped, or areas affected by a natural disaster. When individuals volunteer geographic information from sites of natural disasters such as which roads are blocked/destroyed, or where the most damaged occurred, this can help emergency responders decide where resources should go and how to get them there.

The lab assignment this week required us to take data from a previous module, Module 10: Dot Density, and create a map that could be accessed in Google Earth. The intent of the lab was to make us more familiar with ways in which we could share geographic information with people that might not have a strong understanding of GIS. In order to make the data viewable in Google Earth, the data files had to be converted into KML files that could be opened in Google Earth. We completed steps in ArcMap to prepare the data to display appropriately in Google Earth, such as changing the background to hollow since Google Earth shows satellite imagery. Once the data was converted to a KML file and open in Google Earth, the second part of the lab required us to create places of interest and create a tour file. For the tour we were required to begin with a full view of the data like the one show above, and then "visit" a few major cities in South Florida. Creating the tour was fun and frustrating at the same time. I had to attempt the task several times before I had a finished product I was satisfied with. Turning the layers on and off at appropriate times so that the images were clear proved to be the biggest challenge. 

Sunday, April 9, 2017

Module 11: 3D Mapping

This week we focused on creating 3D images in ArcScene and learning the different ways in which mapping information in 3D can be useful. In order to become more familiar with 3D mapping we completed the 3D Visualization Techniques Using ArcGiS training module on ESRI’s training portal. The training required use of ArcScene and techniques used to depict 2D data in 3D. Some of the techniques learned included adjusting base heights, vertical exaggeration, illumination, draping and extrusion.

It was interesting to know that any 2D data could be presented in 3D using some of the techniques listed above. The image below depicts one of my favorite applications of 3D mapping that we learned this week. We were first required to use extrusion to show how far above a landscape buildings stood and then, in the exercise shown below, we used extrusion to adjust the height of parcels based on their monetary value rather than depicting a physical measurement. I like this technique because it allows anyone using the map to quickly determine which parcels are worth the most or the least. This technique can be used to depict population data or a myriad of other types of data traditionally depicted in 2D.  

While mapping in 3D can create visually stimulating maps, there are also some drawbacks to consider. Sometimes 3D maps can be difficult to navigate if the map user is unfamiliar with them. 3D maps allow the user to “fly” around the scene and it is very easy to get lost in the map or disoriented. Another drawback to 3D mapping is that sometimes it becomes easier to unwittingly lie with your map. Taking the parcel map above for example, the map user can adjust their view of the map and parcels that are somewhat similar to each other in value will look exactly the same or, depending on the view, the tallest parcels will look to be shorter than others. Despite the risks of depicting the data incorrectly, 3D maps can be used to create maps that people want to use and look at and that also contain more data than 2D maps by adding additional information in clickable icons. 

Saturday, April 1, 2017

Module 10: Dot Mapping

Dots!! This week we focused on using dots to map phenomena. Dot maps have been in use for a long time, dating back to Dr. John Snow's Cholera/Water Pump map of 1854, but the use of dot mapping has been in decline and the technique is mostly applied to mapping agricultural phenomena.  

Dot mapping is a mapping method that uses dots to display the occurrence of a phenomena in geographic locations. This type of mapping is appropriate for conceptual data and can be used to display overall spatial distribution, and is also useful for identifying spatial patterns. Each dot on the map represents a value deemed appropriate by the cartographer. When creating dot maps it is important to take into consideration the placement of the dots on the map to ensure map users are not misled by the information presented. For example, if dots are placed uniformly by default into enumeration units without regard for other phenomena (phenomena that would prevent the phenomena of interest from occurring), the map could be misleading.

The lab assignment this week required us to create a dot map depicting the population of counties in South Florida. Creating the map this week was quite challenge because of the processing power needed to place the dots correctly on the map. Because there is a lot of wetlands and uninhabitable/protected areas in South Florida, a random placement of dots in each county would not accurately depict populated areas. In order to place the dots correctly and take into account other geographic phenomena that affects where people live, we used a masking technique that placed the dots in urban areas only. As shown in the map, dots are not placed on any areas considered to be a lake or pond, wetland, or stream.

Another important aspect of creating the map for this week was determining an appropriate dot size and value for the dots. I ended up with a dot size of 2 and value of 15,000. While not my favorite choice because the dots are very small on the map, the masking function in ArcMap seemed to handle these values the best and didn't crash when I ran the mask. Also, while working on the map I felt that the mapped area was too large for this type of map and that maybe a county by county approach would have been better. I say this because of the highly populated area around Miami in Miami-Dade County. There is a high concentration of dots in this area and selecting a larger dot size for the map as a whole would cause the dots to be too dense in this area and make the map harder to interpret.

ArcMap crashed on me several times when I attempted to run the mask with all the layers needed for the final map. It took a while but I figured out what order I needed to add each layer and map element to prevent the program from getting bogged down and crashing. It was definitely a lab exercise that required a lot of trial and error with a healthy dose of patience this week. 

Saturday, March 25, 2017

Module 9: Flowline Mapping

The focus for this week was creating effective Flowline Maps. Flowline maps are used to visualize the movement of various phenomena between geographic locations. Qualitative or quantitative data can be used when creating flowline maps, but specific types of flowline maps are considered best for each kind of data set.

There are three types of flowline maps, radial, network, and distributive. Radial flowline maps are appropriate for use with qualitative data and are used for flow of phenomena that is unidirectional. The second type of flowline map is the network map. This type of map is used with quantitative or qualitative data and can be used to show bidirectional flow of phenomena. Lastly, there's distributive flowline maps. Distributive maps are used for quantitative data and show the flow of phenomena from multiple sources to selected destinations. Creating a distributive flowline map was required for this week's lab assignment and was most appropriate because we were mapping immigration to the United States for the year 2007 (flow from multiple destinations into the United States only).

Below is my map. We were required to work exclusively in Adobe Illustrator this week, no ArcMap. Other requirements included displaying a choropleth map of immigration-per-state (%) as an inset (depending on design choice), and the map must include flowlines showing immigration to the U.S. from various regions around the world. For the flowlines, they had to be curved and sized proportionally to the number of immigrants coming from the region of origin. As seen in the map, Asia leads all regions in the number of immigrants coming to the U.S. and so the flowline leaving that region is the largest. In order to determine the size for each flowline we were provided with an Excel Spreadsheet with the data on immigrantion per region. Using Excel it was possible to review the data, determine which regions had the most and least immigrants, and also develop an equation for determining the size of flowlines proportional to the immigration figures.

Once the equation for determining the proportional size was input into Excel, it was then time to experiment with the size of the flowlines to determine how large the biggest line could be without cluttering the map, but also considering how visible the smallest flowline would be. I experimented with different sizes before settling on a line width of 22 for Asia (largest) giving me a line width of 1 for the Unknown values (smallest). This size allowed for emphasis on the flowlines visually without cluttering the map too much or requiring me to reduce the size of the regions of interest.

I  decided to add drop shadows to the flowlines and also changed their opacity to 80%. Visually I liked this effect because the flowlines seem to be moving over the background of the map and they also look distinct from their regions of origin though they are color matched. Other options included making the flowlines 3D or adding an inner glow, but I felt that a drop shadow was enough to make the flowlines stand out from the rest of the map. I also changed the color of the provided choropleth map to a blue scheme so that it would blend in better with my overall color choices. Prior to changing the color the choropleth map was various shades of orange and stuck out like a sore thumb. I consulted colorbrewer2.org to find a blue scheme appropriate for continuous data like that in the inset choropleth map. I also added drop shadows to the labels for each region to visually tie them to the flowlines.

Overall this was a fun and somewhat difficult lab to complete. I feel that I learned a lot about Adobe Illustrator by being forced to complete the entire lab using only that program. As a bonus I feel that in the future I can add finishing touches to my map using Adobe more quickly as a result of completing this lab. I now have a better understanding of where to find things in Adobe and the best way to implement changes.

Monday, March 6, 2017

Module: 8 Isarithmic Mapping

This week in Cartographic Skills we learned about isarithmic mapping. Isarithmic mapping is used to depict smooth, continuous data such as rainfall, elevation and barometric pressure. Two types of data are used in isarithmic mapping, true point and conceptual point. True point contains data values that have been physically measured at point locations and conceptual point is data that has been collected over an area and then an assumption is made that these values occur at point locations.

Maps depicting true point data are called isometric maps and maps using conceptual point data are called isopleth maps. It is important to note that when mapping conceptual point data the data must be standardized for area. A challenge facing map makers when using either type of data is determining values that fall between known control point locations. To remedy this an appropriate method of interpolation (method of determining intermediate values) must be chosen. Some interpolation methods used for isarithmic mapping include triangulation, inverse-distance and kriging.

As seen in my map above, there is also another method of interpolation known as PRISM. PRISM (Parameter-elevation Regressions on Independent Slopes Model) was developed at Oregon State University and this model takes into account the impact of elevation on precipitation levels to fill in unknown measurements between point measurement stations. This is done by using DEM elevation data and assigning values to grid cells. PRISM also factors in other attributes such as location (proximity to coastlines), complex terrain features, etc. Basically, PRISM is a program that can do what it used to take climatologist long periods of time to do, factor in various geographic influences on precipitation to determine precipitation levels in various regions. 

The most common form of isarithmic mapping is a contour map that depicts changes in elevation and uses a series of lines to depict locations with the same elevation. Contour lines can also be used to symbolize physical features of  landscapes with specific symbology for depressions, hills, streams, valleys and saddles. When contour lines are close together it indicates a steep change in elevation whereas contour lines farther apart show areas with a more gradual change in the elevation.

Along with contour lines that depict elevation there are other methods of symbology associated with isarithmic mapping. Hypsometric tints can be used to shade the area between contour lines to assist in creating a 3-D image depicting change in elevation. This is the method of symbology used in my map above (contour lines with hypsometric tinting) and is an effective method for depicting precipitation because elevation has a strong impact on precipitation levels. In addition to hypsometric tints, continuous-tone can be used to depict change occurring across a landscape. The difference between hypsometric tint and continuous-tone is the step wise depiction of change in hypsometric versus the gradual change of continuous-tone. While continuous-tone displays a more realistic portrayal of a physical landscape, it is difficult to match tones on the map with tones/values in the legend because the change is so gradual.

When creating the map for this week we were required to take a look at the data represented in continuous-tone symbology and hypsometric symbology. This allowed for a better understanding of how symbology choice can represent a data set differently. Hypsometric tint appears to be the better choice for this data set based on the fact that PRISM was used and elevation was considered when the values were determined. Shading between the contour lines reminds the user of the effects of elevation on the precipitation levels shown in the map. While compiling the map it was interesting to learn that different methods would need to be used to create legends for hypsometric or continuous-tone symbology. It wasn't a simple insert>legend> and follow along. We learned how to change the orientation of the legend and how to make the values line up with the color schemes of the legends appropriately by converting the legend to graphics, grouping, and un-grouping.

Thursday, March 2, 2017

Module 7: Choropleth & Proportional Symbol Mapping

This week we focused on creating a map using two different mapping techniques, choropleth mapping and proportional symbols. The assignment for this week also required using what we learned last week about data classification methods. We were required to choose a data classification method that we felt best represented the data for both the choropleth and the proportional symbols elements. The choropleth portion represents population densities for various areas in Europe while the proportional symbols represent wine consumption per capita.

My map is above and just like others in the class, I had a lot of trouble with ArcMap and Adobe Illustrator being slow and/or crashing on me. After working on my map a bit in Adobe I decided I'd like to make the proportional symbols for the wine consumption larger but, when I went back to ArcMap, I had so many issues that I decided to stick with the size I'd chosen and successfully exported to Adobe Illustrator. Overall I'm happy with the map I created and feel that I made correct choices for data classification. As seen in the legends there is a logical break between classes for the choropleth and the proportional symbols. After running through the classification methods I chose to use natural breaks for both the population density portion (choropleth) and wine consumption (graduated symbols).

There was also a design decision to be made regarding graduated vs proportional symbols to represent wine consumption per capita. After taking a look at both I went with the graduated symbols. When I used proportional, the smallest values were barely visible on the map and the largest would have overcrowded the map making labeling the countries difficult. Color choice was also important this week and I consulted colorbrewer2.org to find a suitable color scheme to represent the population data (choropleth).

This week was a real test of patience and persistence when completing the map for lab due to the issues encountered in ArcMap and Adobe Illustrator. It was also a challenge finding the right layout, text size, and scale to best represent all the data required. It was definitely a cartographic challenge this week.