Thursday, March 30, 2017

GIS Week 12: Geocoding and Network Analysis

This week in Intro to GIS, we looked at geocoding (locating addresses) and network analysis. We used a TIGER/Line road network file from the US Census Bureau to create an address locator that was used to geocode the addresses from a table of EMS Station locations in Lake County, FL. Most of them matched automatically, but a few had to be located by hand due to discrepancies between the road data and the address table. Then, we chose three of those stations and set up and ran a network analysis to find the fastest route between them. Below is the map that resulted, showing the locations of all the EMS stations along with the best route to get to three selected ones:


Monday, March 27, 2017

GIS Week 10: Vector Analysis

For this lab assignment, we used buffer and overlay tools in ArcMap to produce a map of possible campsites in a section of DeSoto National Forest. Roads and water features were buffered at various distances, and the two resulting sets of buffers were overlaid to find areas that were within the appropriate distance of both roads and water. Finally, areas that fell within conservation areas were eliminated, leaving only those areas that are close enough to roads and water to be suitable campsites, but outside conservation areas.


Sunday, March 26, 2017

Cartographic Skills Module 9: Flow Line Mapping

This week's lab adventure was creating a flow map, or a map that depicts movement (of people, goods, traffic, etc.) from one place to another. In this case, the dataset was numbers of immigrants to the United States from various parts of the world, and the map has two parts: flowlines showing total numbers of immigrants from each region, and a choropleth map of the U.S. showing the percentage of those immigrants that ended up in each state.

This map was more fun to make than I thought it was going to be when I first realized the whole thing needed to be done by hand in Adobe Illustrator (a personal nightmare). There are a lot of stylistic things you can do with this kind of map, starting with drawing the flow lines themselves and having complete control over what they look like and how to place them. For this assignment, we drew the biggest line first, tweaking it until it looked good, and then used Excel to calculate proportional widths for the remaining lines based on the width of the first one. That way the relative size of the lines reflects the numbers of people they represent. 

Once I had the lines ready to go, I made them partly transparent so as not to hide country or continent boundaries, and added drop shadows to make them pop out to the reader. (I also used a drop shadow to make the U.S. stand out, to draw attention to the choropleth data there and to the fact that it's the endpoint for the rest of the data.) I also used a little bit of a gradient pattern for the stroke of the flow lines, and finally, I used AI's text on a path tool to make the flow lines' labels match their curves.

One other thing I did with the lines, which wasn't really part of the assignment, was separate out Canada. The data we were using was totaled by continent, but that meant North America was going to either need two separate arrows or two converging arrows in order to not be misleading, and either way I wanted to get the proportions right. Fortunately, the Excel spreadsheet also broke down the immigration totals by country, even though we didn't need to use that part, so I was able to subtract Canada from the rest of North America and satisfy my compulsion for accuracy.


Friday, March 10, 2017

Cartographic Skills Module 8: Isarithmic Mapping

On the surface, this was a pretty easy week in Cartographic Skills, but I'm continuing to struggle with some of the finer points of using Adobe Illustrator, and this week's map is not my best. So it goes.

The topic of the week is isarithmic mapping, Isarithmic maps are used to map continuous data, such as temperature or topography. The data can be symbolized using graduated color, contour lines, or both. In the lab exercise for this week, we looked at a dataset for annual precipitation amounts across Washington state. Although the data was first mapped using continuous tone symbology, in which the data is unclassed and stretched over the full range of a color ramp, for the final map, pictured below, the data is mapped using hypsometric tints. It is divided into classes (we set the ranges manually) with a color assigned to each class. In this case, contour lines were also created that fall at the divisions between classes, to make it easier to distinguish them when reading the map. 

This type of map, which illustrates the yearly average precipitation over a 30-year time span, is typically used to study long-term climatic trends. This particular dataset was created using the PRISM system, which relate data collected at weather monitoring stations to elevation and other variables in order to interpolate the data with respect to known weather patterns as well as location. The blurb on the map explains PRISM in a bit more detail.


Tuesday, March 7, 2017

GIS Weeks 7-8: Data Search

For the midterm lab assignment for Intro to GIS, we were asked to acquire a list of datasets for a county in Florida, project all of them into the same coordinate system, "clip" state-level or other larger datasets to the shape of the county in question using the tools available in ArcMap, and create 1-3 useful maps.

I was assigned Flagler County, on the state's Atlantic coast. My first set of maps, below, includes one illustrating some key natural and man-made features in the county, including waterbodies, cities and towns, roads, and parks, and one showing Strategic Habitat Conservation Areas symbolized according to their priority ranking (with cities and roads marked to help orient the reader). For both of these maps, the data is overlaid on a Digital Elevation Model, allowing the reader to see how all the features represented interact with the county's terrain.



I had one other dataset that I felt was too complex to be combined with any others, so my third map just shows land cover categories for the county. However, one of the required datasets was a DOQQ, or digital orthophoto quarter-quandrangle, so I also included an inset zoomed in on part of the city of Palm Coast, where the land cover dataset is overlaid on top of the aerial imagery.




Sunday, March 5, 2017

Cartographic Skills Module 7: Choropleth and Proportional Symbol Mapping

In this assignment for Cartographic Skills, we tackled two types of thematic maps, choropleth and proportional symbol, by creating a single map of Europe showing population density (as a choropleth map) and per capita wine consumption (as a proportional symbol overlay).

The map was created in ArcMap using provided data and finished using Adobe Illustrator (that was a LOT of symbols and labels to adjust manually--yikes). I chose natural breaks classification for both population density and wine consumption, because I thought it visualized the data best, showing the overall variation while also capturing outliers at the high end of each range (for example, the Netherlands for population density, and Vatican City for wine consumption, which ended up being a class by itself because it's so much higher than everything else). ArcMap classifies the data automatically based on an algorithm that looks for existing gaps to separate classes. I also chose to use graduated symbols for the wine consumption layer rather than proportional symbols--same concept, but slightly different presentation. With graduated symbols the data is divided into classes for which the whole range of data is assigned one symbol size, and that size isn't directly related to the data value. I thought that in this case it was easier to distinguish the symbols this way (that's also why I used only four classes--I found if I used five, it was hard to tell the difference between the middle symbols anyway, which sort of defeats the purpose) and easier to understand what the symbols represent.

Anyway, the map:

Some general trends you might notice are that Western Europe tends to have higher levels of wine consumption, while Scandinavia and much of Eastern Europe seem to drink less wine. The colder parts of the continent, predictably, tend to have lower population densities. I don't see any very obvious correlation between population density and wine consumption, though that could be a function of the way the data ended up being represented here, but it's still an interesting map.