Ina Anderson
Intro GIS -- Assignment 5
1). The Providence School Department’s budget has been cut and they are unable to provide after-school activities for students. Unfortunately, many children are left unattended after school for several hours before their parents return home from work and are in need of after-school care. This “project” seeks to identify appropriate resources in the community where after-school programs could take place that are in close proximity to Providence schools. These include libraries, outdoor recreation and open space areas and churches. Ideally, these activities should be within walking distance ( ¼ mile ) from the school. Positional accuracy will be important as children will be walking together unsupervised and should be able to locate their programmatic activities within +/- 15 feet.
2). The three different road centerline data sets that I compared for this project are Street Map USA, Census 2000 RI TIGER Lines, and RI GIS Road centerlines. The RI GIS centerlines are the most accurate when compared to the orthophoto. The Street Maps USA and the TIGER data are both far off when compared to the orthophoto and the RI GIS centerlines. They run through buildings and show curves where there are none in the road. At some points the TIGER lines are so far off that it is difficult to tell which street they represent. The illustration below shows the differences in
the three sets of data for the same street, Power Street in Providence. The RI GIS layer (Green) accurately shows Power Street. The Street Map layer (Blue) is off by 114’ from Power Street. The RI TIGER line layer (Yellow)is even more off, 180’ from Power Street. There is even a difference of 60’between the Street Map line and the TIGER line. For the purposes of my project, the best data would be the RI GIS data, as it most accurately represents the streets.
3). The hydrography layers that I used are the RI GIS Streams and the Census 2000 TIGER hydrography data. Rhode Island does not provide vector polygon files for streams, just line data. The TIGER water bodies are off when compared to the RI orthophoto and the RIGIS data. There are areas where the TIGER data does not continue to follow the river when the river narrows and the polygon is incomplete. (see illustration).
This illustration shows the RI Streams data (blue line) and the TIGER waterbodies data (aqua blue polygon). The two sets of data are not comparable as there is no place where they overlap. The RI Streams data is not very accurate when compared to the orthophoto, the arrow shows where the line is 150’ from the water. Another line goes straight through a parking lot and over roads. The TIGER data fills the polygon to a certain point, and then is incomplete when compared to the orthophoto. For the purposes of this proposed project, the accuracy of the waterbodies is not critical.
4). Positional Accuracy: I am able to provide quantitative positional accuracy for the optional data layers for the project (schools, libraries, open space). Positional accuracy is provided by RI GIS as reported in the metadata. The RI GIS Open space layer is on a scale of 1:24,000 and provides inherent accuracies of +/- 50 ft. The RI GIS schools layer is based on 1:5000 scale with assumed accuracy of +/- 10-15 ft. The RI libraries layer provides positional accuracy of 10-15ft. I am not able to provide positional accuracy for the fourth layer, churches, as this layer was geocoded from data provided by Reference USA and I do not know the positional accuracy.
5). Qualitative assessment of the positional accuracy of the other layers (open space, libraries, schools and churches): The open space layer from RI GIS represents open space in varying ways – in some cases the polygon is filled in, and follows the actual contours of the park or field. However, in other places, the open space is simply represented by a large dot that covers the space, which is confusing as it looks as though it should signify some meaning that the user is unaware of. Additionally, in some places where there is open space next to or surrounding a building, the entire polygon is filled in making it look as though the entire area is open space, when in reality it is a school complex near a field.
In the illustration above, I made the open space polygons hollow so that it is possible to compare to the orthophoto. Note that the rectangular tennis field is represented by a circle and that there is a playing field that is not included in the data. Circles are used to cover areas where there is a small lot or other open space feature. This would be problematic for my project as it is hard to understand the actual position of the recreational area.
The libraries, schools and churches appear to be positionally accurate when compared to the orthophoto, which makes sense as the RI GIS metadata indicates that 1:5000 orthophotos were used as the source material to create the school and library layers. (see illustration below).
When compared to orthophoto and RI GIS streets, schools, libraries and churches are accurately placed.
6). Are the optional layers appropriate for your project?
The school and library layers are appropriate for this project as they provide positional accuracy within 10-15 feet. The open space layer is more problematic as it provides accuracy within 50 feet, which would make a real difference in planning activities for children, where it would be important for them to be able to find their programs easily. The church layer is also appropriate for the project.
7). Currency: I was able to determine the currency of the the data from the metadata. Some of the data is more up to date than others. The RI GIS State Conservation, Park Lands and Open Space layer is based on conditions prior to 1990 and was updated in 2002. The RI GIS schools data layer is quite current, derived from data collected in the 2008 school year. The library data is from 2007. The RI Roads is the most up to date, considered current from June 2009. The RI streams layer is from 1997. The TIGER census data is from 2000.
8). Completeness: The Rhode Island hydrogaphy layers appear to be incomplete, as they do not follow the narrow rivers and streams throughout the city. The open space layer also is incomplete in many instances.
9) Attribute Accuracy: The attribute tables for the schools was quite complete, all fields filled in and address and zip data complete. I checked 5 schools that I know and their information was correct. The Library layer was also very complete, out of 106 records, only 6 were missing the address # but still had the street address. I checked several that I am familiar with, and the information was correct. For the churches, the information from Reference USA appears to be complete. Aside from the misspelling of Benevolent street, several other churches that I looked at were listed correctly.
Above is the attribute table for the RI open space layer. There are no values entered in most fields that indicate type of open space. Although not shown in this picture, there were many names missing from the name field.
The attribute information for the schools, libraries and churches is adequate for this project as it provides address locations and names for those entities. In terms of open space, the attribute tables are missing crucial information including names and type of open space to be useful in planning activities for children.