Sunday, February 26, 2017

Modeul 6: Data Classification

During this lab I applied four different classification methods (Equal Interval, Quantile, Standard Deviation, and Natural Breaks) to U.S. Census data for Miami Dade County, Florida. I created two maps displaying senior population by tract in Miami Dade, each with different presentation methods, using four data frames for the four classification methods listed above. The first presentation method was mapping percentage of people 65 and older compared to the total populations of each tract, and the second method was displaying the number of people 65 and older normalized by square mile in each tract.

After creating these maps, I compared and contrasted the classification methods to determine which was best suited to represent the spatial data for a specified audience as well as the presentation methods to determine which was best suited to present the distribution of data. I determined that the natural breaks classification method would be the most useful for an audience targeting senior citizens because it revealed useful patterns in the data such as potential extremes and differences between values while also maximizing variance of inter-class values and minimizing variance of in-class values. Although this method does not consider how values are distributed on a number line, I think the natural breaks still did a decent job of maintaining the distribution of seniors as displayed by the other methods. I determined that the normalized presentation method (shown below) does a better job at representing the distribution of seniors in the county because the percentage-based method causes some tracts to be placed in the highest class, while they actually only have a high relative proportion of seniors compared to the total population in that region. Although I think knowing the distribution of percentage of people 65 and up compared to the total population could be useful in conjunction with the other presentation method, it alone does not help the audience gain an understanding about how many seniors actually live in the area.

The map above is the map I created using the four methods of classification and the normalized presentation method. It shows 2010 U.S. Census data by tract normalized by square mile using four different classification methods. In addition to the natural breaks method described above, I used the equal interval method, where the range of all classes are the same. The quantile method places an equal number of values into each class, and the standard deviation method contains classes that include values within a certain standard deviation of the mean and is best for displaying data with normal distributions. This lab helped me gain a thorough understanding of the various classification methods, when each one is appropriate, and how important it is to choose the best one so that viewers have a useful map they can rely on to make important decisions.

Friday, February 24, 2017

Week 6: Projections Part 2

The goal of this lab was to learn how to find spatial data online, define spatial reference for it, and finally apply a single projection for all data layers of a map. This crude map shows where petroleum storage tank contamination monitoring (STCM) sites are located throughout a particular region (defined by a quadrangle) within Escambia County. The map is unfinished because the goal was to show how all data layers aligned because they are all projected in the same coordinate system.

To make the map, I first dowloaded a quad index shapefile and a county boundaries shapefile from FGDL and use them to locate a quadrangle within Escambia County. Then I found the four DOQQ files for this quadrangle in Labins and added these to the map. I also found a major roads layer in FGDL and dowloaded this and reprojected both this one and the county boundaries layer from their original projection of Albers Conical Equal area to the one used by the DOQQs. Although I did not include a DRG file in this map, we did practice defining projections because the particular topographic quadrangle we used from Labins did not contain a spatial reference. I used a list of STCM sites from the Florida Department of Environmental Protection website and converted the coordinates to decimal degrees, which allowed me to import it as XY data into ArcGIS. Finally, I exported this as a shapefile and reprojected it as the same State Plane coordinate system I used on the rest.

This lab taught me how important it is to understand how to find the spatial reference for data and change it if it needs to be in a different projection. I learned how tedious yet important this is in order to end up with a useful map and also how helpful it is to keep track of your data as you go along including file paths and metadata information.

Sunday, February 19, 2017

Module 5: Spatial Statistics

This week I completed a course from ESRI's My Virtual Campus Training called "Exploring Spatial Patterns in Your Data Using ArcGIS." This course focused on how to first explore the spatial distribution of your data and its distribution of values as well as the relationship between these two  using spatial statistics tools and Geostatistical Analyst tools in ArcGIS. This helps you better understand your data so you can choose appropriate analysis methods and edit your data as needed beforehand to get reliable results.

I created this map during the first portion of the ESRI training that covered analyzing the spatial distribution of data. First, I used the Mean Center and Median Center tools to create feature classes with symbols displaying the mean and median centers of weather monitoring stations in Western Europe. The median center is near the actual center of the weather stations, and the median and mean centers are pretty close to each other. Both of these aspects of the spatial exploration of the data support that the stations could be normally distributed. The clusters of stations in Austria and Switzerland pulled the median center slightly south and east. Then I used the Directional Distribution tool to create a standard deviational ellipse. The ellipse has a rotation value 79.5°, which means most of the spatial variation is in the east-west direction. I completed the map in ArcGIS by adding essential map elements and a description to explain what the map is showing.

This ESRI training course was very helpful because it explained not only how to do certain things in ArcGIS, but also why they are important. I plan to utilize additional training courses in the future!

Thursday, February 16, 2017

Week 5: Projections

In this week's lab, we created a map displaying three data frames of Florida counties projected in three different projected coordinate systems. I created a table using data from the attribute tables of each different projection containing the areas of four counties spread throughout Florida. This shows how projected coordinate systems are not perfect and distort some aspect of the region being mapped and how important it is to choose the best projection based on the needs of the map you are creating. In addition to utilizing the Project and Project Raster tools to reproject data, we also learned how projection information for GIS data files is stored as .prj files, worked with three data frames within ArcMap at one time and created a map displaying them, and finally learned how to create new fields in attribute tables and calculate geometric calculations, in this case area of polygon shape features (counties). Although I understood the concepts of geographic and projected coordinate systems prior to this lab, actually working with multiple projections of the same data and comparing their areas helped me understand how much they can differ between one another and therefore how important it is to understand their importance and how to alter them within ArcGIS. The most important thing I learned in this lab was how to define a projection for raster layers and how they typically do not contain spatial reference data with the file.

Monday, February 13, 2017

Module 4: Cartographic Design

The purpose of this lab was to learn how to apply Gestalt's Principles of perceptual organization while creating a map of public schools in Ward 7 of Washington, DC. These four principles include visual hierarchy, contrast, figure-ground, and balance. Three additional objectives of this lab were to learn how to create a map according to the needs of the end user, to symbolize layers by category, and to create a map insert and establish an extent indicator.

The map I created shows public schools within Ward 7 with ordinal data symbolized by type of school using thematic symbols. I included an inset to show where Ward 7 is located relative to Washington, DC as a whole as well as a list of school names. I utilized all of Gestalt's Design Principles including implementing visual hierarchy by making school symbols a bright red color and the top layer, which presents them as the most important element according the the intellectual hierarchy I established prior to making this map. I kept base information such as the roads and boundaries different shades of grey because they are not as important as other elements. I achieved adequate contrast with the colors as well by making the Ward 7 layer the lightest, and the Washington, DC layer and background darker. This also aided in establishing figure-ground, because it accentuates Ward 7, making it appear closer to the viewer than other layers. Finally, I incorporated balance by first making the main map area as large as I could within the data frame, and then assessing the empty space and placing the list of schools, legend, north arrow, scale bar, inset, and author/date/source information within it.

I used ArcGIS to create the majority of this map except for the list of schools, which I created in Excel based on information in the Ward 7 data layer. I was initially going to import it to AI to do additional edits regarding labels and colors; however, after trying several troubleshooting solutions could not figure out how to get the school symbols to load. This forced me to use ArcGIS to edit labels and symbols, which was a good learning experience because I got practice editing labels as annotations, inserting a table from Excel, and using the draw tool to label the river.

Tuesday, February 7, 2017

Week 4: Sharing GIS Maps and Data

In this lab, we learned three different techniques for sharing GIS maps and data including how to create an ArcGIS Online map, how to create and share a map package, and how to create and share a KML Google Earth map via the web. We first made a data table in Excel containing our "top 10" locations based on criteria of our choosing. I chose to do the top 10 US cities for street art because I love art and based my list off an article I found on the TimeOut website titled "The best graffiti across all of America". After creating a table with location information for these 10 cities in Excel, I saved it as a text file and uploaded it into ArcGIS Online to be geocoded and created an online map, which can be accessed by clicking on the image below.

I then opened this data in ArcMap Desktop and created a shapefile for the top 10 cities that were geocoded. Next, I prepared the data to be turned into both a map package, which I shared publicly on ArcGIS Online, and KML file that can be viewed in Google Earth. This lab was very informative because after all, the purpose of analyzing spatial data and creating maps is usually in order to share knowledge with others. Now I know several different ways I can share maps I create with other people, and I can see how all of the methods will come in handy in the future in different situations. 

Sunday, February 5, 2017

Module 3: Typography

The purpose of this lab was to gain experience in typography by applying basic guidelines to a map of Marathon, Florida. The learning objectives were to include essential map elements in the final map, label it based on typographic guidelines, use proper type placement for all features, and demonstrate competency in Ai. 

This image is my final map of Marathon, Florida and includes major towns, names of key islands, points of interest, and signifiant hydrographic features. I also included an inset so viewers can easily see where this key is oriented in relation to the southern tip of Florida. Its intended audience are people interested in visiting the area and who'd like a general understanding of its layout. 

I first created the basic map including the shape features making up the keys, the inset, and the legend in ArcGIS and then exported the file to Ai in order to add labels, legend, title, north arrow, source information, author, date, symbols, and additional customizations to complete the map including the following three. First, I included a drop shadow to the island shape features because it added depth and made them more defined. Second, for the hydrographic features, I labeled them all with the same serif font type, italic style, and cyan color to make it easy for the viewer to interpret them all as water features; however, I used different type sizes to denote relative size differences between bodies of water (e.g. largest font size for the Atlantic Ocean and smallest for the harbors) and also used upper case for larger bodies not relevant only to Marathon Key and sentence case for bodies of water only relevant to the keys being mapped. Finally, I used uppercase to label the areal key features and sentence case for the cities and parks/city features because without some obvious differentiation, it would be difficult to understand that some of the labels were referring to cities and others to islands. 

After completing this lab, I've come to appreciate the typographic guidelines that have been created for cartography because without them deciding which style, size, color, type, etc. of font seems overwhelming due to endless possibilities and labeling features of all types seems much more methodological and therefore do-able for me. Also, because we did all of the labeling manually in Ai, I think the algorithms used to auto-label maps in ArcGIS will become a huge help, but also I realized that being able to edit them in Ai will enhance the map even more.