The XY(Z)s of Geosupport coordinates
Everyone wants to know where they are and where they need to go. Most people use addresses to solve this problem. Give them a street number and name and they can find their way to that spot. For mapping or spatial analysis, some people use XY coordinates. The Department of City Planning’s Geosupport application is designed to provide various facets of geographic information for its user. XY coordinates are one of them. Geosupport offers three sets of XY coordinates. Most people unfamiliar to the program have no idea what the three sets of coordinates mean or why there are three different sets in the first place. The goal of this post is to explain why there are three different sets of coordinates, what the difference between the three are, and when to use them.
Why?
As a newcomer, if you went to GOAT, City Planning’s web interface to Geosupport, you might look at the interface and think that it’s relatively straightforward. But as you play around with the application, you may realize that you don’t fully understand what the output fields from each function capture, and why there are so many of them. You may ask yourself, for example:
“Why am I getting three different XY coordinates for the same address? Shouldn’t there only be one?”
Each XY coordinate is the output from a different address function. Each address function takes a house number, street name, and borough as inputs. They differ, however, in what kinds of attributes and location data they return. This is because behind the scenes, each function maps the input address to a different set of lookup tables. Function 1A outputs property information for the address or place name like tax block and tax lot information. Function 1E outputs information on the street and city services for the address or place name, such as whether the street has a bike lane or what is the election district. Function AP outputs information such as building identification number for the address point, which is tied to a specific structure. Because of the different natures of the address functions, each function outputs different XY coordinates related to the nature of that address function.
What?
The figure on the left illustrates three different coordinates for the same address. City Planning’s headquarters at 120 Broadway in Manhattan is used in this example. For each function, you would input a house number, street name, and the borough of this address. Knowing that Function 1A provides property information, Function 1E provides street and city services, and Function AP provides information about the specific building, helps understand why there are different coordinates for the same address, but now the question is what location are the coordinates actually representing.
To start, let’s address the coordinates that seems to be placed in front of the building but are not located in the building. These coordinates are outputted from the function 1E, which tells us that these coordinates have to do with the street of the address. Function 1E interpolates where the coordinates of the address are on the street, given the range of high and low address numbers assigned to that street segment. It is difficult to tell in the figure, but the coordinates is slightly offset from the centerline (an imaginary line running through the middle of the street) to the side of the street where the address is located.
The point labeled “1A” is on the coordinates outputted from function 1A, which tells us that these coordinates have to do with the property level information of the address. Function 1A retrieves the XY coordinates that are derived from a tax lot centroid using your input as the key. Tax lots are outlined in red on the map. For 120 Broadway, the entire parcel is within view of the map so you can see how 1A is the centroid of that tax lot.
The point labeled “AP” is located at the coordinates outputted by function AP, which tells us that these coordinates has to do with the address point of the address. Function AP retrieves the XY coordinates of the address point for the address. The address point usually reflects the posted address of the buildings and is often the location of the entrance to the building. Address points are located inside the building footprint, on the side of the building adjacent to the street name included in the address. This point is perhaps closest to what you would imagine you’re getting when you ask for the XY coordinates of a location.
When?
Now that we’ve covered what these coordinates mean, we can finally move onto when to use these coordinates. You may have noticed that point 1A is somewhat removed from the other two points in Figure 2. Tax lots are parcels of land that are of indeterminate size, and can range from 50ft by 50ft to 1000ft by 1000ft. The 1A coordinates reflect that. For smaller tax lots, the 1A coordinates can be within close proximity of the other two sets of coordinates, but for larger tax lots, the coordinates can be many feet away. If your map includes an entire borough or the whole city, it probably doesn’t matter which set of XY coordinates is used. At that scale, the differences are not great. At the neighborhood scale, 1A’s XY coordinates work well for areas of the city that have one building per lot. If the building is centered on the lot, the XY coordinates will fall in the center of the building. However, if you are mapping multiple addresses for a lot that has several buildings, 1A’s XY coordinates for these addresses will be identical and only one point will appear on the map. In this case, AP coordinates would be better, showing the building assigned to each address. 1E coordinates could also be used, but these may be less precise than the AP coordinates. The location of the 1E coordinates is based on interpolation along the street where the address is located. In cases where addresses aren’t distributed evenly along the street segment, the 1E coordinates can end up some distance from the desired location. An example of how different these coordinates are is shown in Figure 2.
The address illustrated in Figure 2 is 291 Roxbury Blvd in Queens. Initially, it is difficult to tell from this map where the actual address is located, but if you recall the previous section, AP coordinates show the most accurate location of an address. In this specific example, if you were looking for the building located at 291 Roxbury Blvd, you would want to use the address function AP. Function 1A in this example would not be useful because this address is located on a tax lot containing hundreds of buildings. The tax lot is so large that its entirety cannot be seen in Figure 2. Thus, the centroid is too far from the address to be used for a closeup map. In Figure 3, you can see clearly the tax lot in its entirety. The red lines indicates the boundaries of the tax lot.
Function 1E in this example is also not useful because of how the building is placed in relation to the street. As you can see in Figure 2, the interpolation algorithm approximates the coordinates in the middle of the street segment when 291 Roxbury is actually closer to the end of the street segment. This address function retrieves accurate coordinates in areas where buildings are arranged in rectangular blocks. Figure 4 shows the line segment of the Roxbury Blvd segment and why Geosupport approximates the 1E coordinates where it did.
Geosupport can provide us with a lot of information, but can be difficult to figure out how to pick which information is relevant for a given use. Understanding what the different address functions do and what their XY coordinates mean is a good first step in working with Geosupport.
