Picture 2: Hydrography Inaccuracies and Road Inaccuracies

Picture 1: Centerline Inaccuracies

Figure 3: Incompleteness of ReferenceUSA

Laura Tolkoff

GIS Assignment 4

March 9, 2010

Project Description

The City of Somerville, under pressure from MASSBIKE, the Livable Streets Coalition and various nonprofit agencies, seeks to add two miles of bike lane to its current active transit network. We hope to increase bicycle ridership to promote active transit, reduce the loss of open space for parking, and improve ambient air quality in Somerville. Thus, this project uses Road Centerlines and Hydrography to better understand the road layout, width of roads, and streets paths that can accommodate bicycle traffic near the Charles River. We look at existing bike trails through the city to see where we can extend our paths. We also use ReferenceUSA to locate Somerville grocery stores and convenience stores so that our bike paths are convenient for errands such as grocery shopping. This project also looks at the location of schools, in hopes that a protected bike path will increase bicycle ridership among children. The City of Somerville requires that this project requires a high level of positional accuracy (+/- 1 foot) for the following reasons. First, adequately sized and positioned bike lanes are necessary to protect bicyclists from traffic and vehicle-related deaths. Second, the city of Somerville has a restricted budget to improve its active transit network and requires a high level of accuracy so as to work within this budget.

Datasets

(bold text indicates optional layers)

Road Centerlines from a state: M:\state\MassGIS\Infrastructure|EOTROADS_ARC.shp

Road Centerlines from Census TIGER dataset: M:\state\ma\MassGUS\Census\CENSUS2000TIGER_ARC

Streetmap USA: M\country\USA\ESRIDataMaps906\streetmap_usa

Orthophotography: ArcGIS World Imagery

Census TIGER hydrography: M:\state\ma\MassGIS\Census\CENSUS2000TIGERHYDRO_POLY.shp

Hydrography from city: M:\City\Boston\BRA\water.shp

Bike Trails: M:\state\ma\MassGIS\Conservation_Recreation\BikeTrails_Arc

Reference USA: Grocery and Convenience Stores in Somerville

Open Space: M:\city\Boston\BRA\openspace.shp

Open Space:M:\state\ma\MassGIS\

Schools: M:\state\ma\MassGIS\Infrastructure\schools_pt.shp

Positional Relationship of Centerline Data and Orthophotography

The Massachusetts Centerlines (EOTROADS) is more positionally accurate than the Census 2000 TIGER files. The TIGER files are very clearly intersecting houses and other parcels shown in the Orthophotography. However, we cannot assume that the Orthophotography is entirely accurate either. Figure 1 pictures the positional discrepancies of the TIGER files, symbolized in blue.

Positional Relationship of Hydrography Data

The Hydrography Data from the BRA are clearly incorrect, as demonstrated by Figure 2. Typically, bodies of water are not jagged and angular. Moreover, this data set shows clear disruptions in the body of water compared to the Orthophotography. The water is also shown to cover a significant portion of land, roads, and parcels.

Quantitative Positional Accuracy, US National Map Accuracy, Table of Scales

The National Map Accuracy standards say that for a map at the scale of 1:1200, the points should be within +/- 3.33 feet. However, these datasets would not be accurate because the distance between the orthophotography streets and other datasets is far off, upwardsof 64 feet in some places. Picture 1 shows a discrepancy of almost fifty feet.

Qualitative Assessment of Positional Accuracy, Appropriateness of these layers for Positional Accuracy

There are logical inaccuracies in the hydrography layers, which display the Charles River and other water bodies as being jagged and angular (see picture 2). For creating bicycle paths near the river, the city might want to consider using an alternative data source that is more accurate. There is also a high level of relative positional inaccuracies, as demonstrated in the picture below. The Tiger Centerlines are not close to the true positional values of the centerlines, so the City of Somerville would most likely want to use the BRA centerline layer for its bicycle path extension project.

Completeness

The Hydrography data from both TIGER and the BRA is incomplete, since a large section of the Charles River is missing when compared with the orthophotography (see Figure 2). Figure three also demonstrates that the ReferenceUSA data is not complete, since the White Hen Convenience Store is missing from the data set.

Currency

The most relevant data layer was the schools_pt layer, which was updated in April 2008. The MassGIS EOT Roads layer was updated in 2005, though the TIGER files from 2000 are clearly out-dated. Because the city of Somerville needs a high rate of precision, we need to use the most current data layers possible.

Attribute Accuracy

The schools_pt layer has a high level of attribute accuracy and completeness, including the grade levels, phone number, and fax machine. However, it also includes principal/headmaster names, which may not be current since the set was last updated in 2008.

The ReferenceUSA (grocery and convenience store) was not entirely accurate, since some points were unmatched. The attribute table is also incomplete because it did not contain all executive names or stores (see figure 3), and may not be up-to-date.

The Bike Trails attribute table is fairly incomplete, and does would not give the city of Somerville an accurate picture of current bike paths and trails in the city. The attribute table shown below (figure 4) demonstrates the incompleteness of the data.

The BRA open space layer is a relatively complete dataset, but contains fewer attribute fields than the MassGIS open space layer.

Whereas the MassGIS open space layer is less complete, but has more attribute fields. You can also see that the MassGIS attribute layer is inconsistent in its use of capitalization

There are very few attributes for the BRA hydrography layer, which makes me think that it is incomplete.

Conclusions

Many of the data layers discussed above are not appropriate for the precision required by the city of Somerville for its bicycle path extension project. If the city had to choose from among these layers, I would suggest that it utilize the BRA street centerlines and the BRA open space data layers. It should do field work to confirm and improve the values for the bike paths and try to find a more complete and positionally accurate layer for hydrography. It may also want to conduct a more current inventory of businesses such as grocery stores and convenience stores (especially because of the Recession) to create a more complete picture of errand and traffic patterns for bike paths.