Final Project: Land Use Land Cover Providence, RI

 

This project is inspired by an article from the United States Geological Survey’s Earth Resources Observation and Science Center called “Characterizing Heat Trends in 50 Cities Across the U.S.”. Research physical scientist George Z. Xian authored a fact sheet including maps of mean surface urban heat island intensities of 50 cities. 

Being aware Land Use Land Cover within a city can help its leaders and planners understand how they can move toward minimizing effects of urban heat islands that are harmful to residents. This project explores Land Use Land Cover in Providence, Rhode Island using a Level II classification scheme, represented by a two-digit code in the map’s legend. 

To complete this project, I downloaded true color aerial photographs from RIGIS.org. Municipal data was also downloaded from RIGIS to place the city boundaries on the aerial photographs in ArcGIS Pro. Using the Raster Functions Tool, the City of Providence boundaries were clipped from the rest of the state so only Land Use Land Cover analysis would be completed within the city boundary. 

I used visual interpretation techniques that we learned and practiced in previous labs in the beginning of the course. These include using the recognition elements: tone, texture, shape and size, shadows, pattern, association, and true color. Each classification had its own description to keep consistency throughout the area. If a classification made up over 50% of a section, it was labeled as that majority classification. Where it was harder to determine the percentage of each land feature, sections were broken into smaller polygons, especially in cases where it seemed to compromise accuracy too greatly.

Residential areas were classified based on a grid-like pattern of homes in a neighborhood. These areas were mostly distinguished from commercial areas based on the size of the roofs on the buildings. Commercial buildings were generally longer and more rectangular shaped and were not necessarily in an identifiable pattern. Commercial buildings were also associated with parking lots adjacent to these longer rectangular buildings. They also tended to be located on the side of a main road where most of the residential neighborhood patterns began at least one or two blocks in from a larger road.

The large industrial area in Providence was determined based on the size of the area. Industry generally requires a lot of space for its production. Because Providence is so densely developed, there is not a lot of space available for big industry. This made the industrial area stand out from the rest of the city. This area had the largest structures along with vast areas of impervious surfaces or barren land seemingly existing for space to move around large machinery. There were also docks in the river within the industrial area, making it a likely place for cargo ships to distribute product in large quantities.

This project includes major highways within the Transportation, Communication, and Utilities classification. The other less significant roads were included in Residential and Commercial classifications. Major highways were identified by a larger break between residential or commercial buildings and by the pattern of cars in lines driving on the road.

The large grass areas in the aerial photograph were classified as Cropland and Pasture. Some of the grass areas were too large to be considered for another classification, especially when considering the desire to identify impervious surfaces located around the city. These grassy fields were identified as a break in forested, residential, or commercial areas. The color was a lighter green than the forested areas and had a finer texture. The Deciduous Forest classification was determined by a darker green color and a coarser texture.

Water was classified based on the dark blue/green color and the irregular shape compared to the more grid-like patterns of the city. The water features were also the finest in terms of texture. The Streams and Canals were classified as longer bodies of the water while the Lakes had a more defined area in the image.

There was one section of barren land in this image that was too large to include in the adjacent residential or commercial and services classification. There was tan color similar to sand where there were likely cleared trees or old buildings in preparation for newer construction. 

This project was great practice for my Land Use Land Cover mapping skills. I also struggled at first to find imagery that was easiest to work with. First, I tried using Landsat imagery on Glovis and Earth Explorer which was provided to us as options in the project's instructions but found that multiple downloaded images had a lot of cloud cover and the scale was so large it didn't make sense. Then, I looked on state and local websites and found GIS data that felt much less intimidating to use. It was a great reminder that local government resources can be a great option for municipal level studies!

First semester towards my GIS Graduate Certificate at UWF is complete!