ESLARP East St. Louis Action Research Project
University of Illinois at Urbana-Champaign


Sanitation Code Violation in East St. Louis

Methodology

The first problem of data manipulation is that these two data sets have different map projections. Since LRMS does not contain the Metadata, I changed it to Lambert which is the projection of DNR data. I used 'projectdefine' and 'project' command in Arc/Info to change the projection of LRMS Data. After changing the projection, however, these two data were not perfectly matched but it seemed to look okay. It might be because of digitizing done by different person. Then, I made a base map for this project. I made my own coverage by using Census90 coverage data. This coverage was originally for whole St. Clair County but I selected the areas within the boundary of East St. Louis.

The next step was the mapping of D.J.'s data of waste disposal into the base map. The data was collected at parcel level, too much to handle in short time. DJ's data indicated that there were 793 parcels which had code violation. Actually, I tried to develop this project at parcel level but geocoding was not possible because addresses were not given. It just indicated that the corner of street such as 'Tudor & Main St.' Another problem of parcel level GIS was that it needed too much disk space and time to load. It took more than thirty minutes to bring all the parcels in the East St. Louis area in either GMS Lab's computer or DURP Lab's one. It seemed impossible to handle the parcel level data to this project. So I made a decision to manage the data at block level. There was still a lot of dumping blocks even when I aggregated my data to block level.

I met a most difficult problem in my project here, determining a key by which I could relate the various tables. The fastest way was to join the attribute table of map, including census data, and my violations data spreadsheet. However, there was no common field. The block IDs did not match with each other. Instead, I developed a new table manually. It took a long time, but I created my own way. That is creating the new column for applying to both the attribute table and our original data spreadsheet. The new item was the combination of Census Tract Number and Block Id. As a result, my foundation map now has violation information. However, there still were some blocks missing the violation information. Hence, I decipher the Census Tract numbers from the violation data manually. Because it was based on neighborhood, and not census tract, I overlaid neighborhood and census tract and manually noted and added census tract data to the violations table. Finally, I made two types of violation maps. One is the thematic map of graduated color and the other is dot map of the number of violation which I used a lot for overlaying with other themes representing statistical analysis for this project.

For the spatial analysis, I used the analysis tool in the Arc View. I made a 1/8 mile interval buffer. Since the unit of this map is feet, I gave the feet value to the buffer. I use 'find distance' function for visualizing the buffers, and 'selected by theme' functions for constructing many tables for statistical analysis. I made a lot of themes in Arc View for this project. I made each land use and each class of road as a different theme for conforming 'selected by theme' function which took most part of my query. In addition, I digitized the highway exit cover, based on Illinois Road Map (Rand McNally 1994).

The final step was the Georeferencing. Since I though that actual site photograph would help to understand the condition of real site, I tried to register aerial photo of actual site to my violation map. I used 'register' and 'rectify' command of Arc/Info for this process. 'Register' allows aerial photo to have real map coordinate, and 'rectify' rotates the aerial photo to be matched with the map.

In this project, both Arc/Info and Arc view were used. Many technical lessons were learned during this project. Arc View is better to use visual or conceptual task because it has a dominant ability in visualization. Arc/Info is a great tool for handling geographical data. I found many functions which required Arc/Info. If these two programs can be used in conjunction with the same project, a much better GIS is possible.

Because of the nature of this project, I had to export the data into a different software to perform some statistical calculating. I chose Excel because of its ease of use. Excel could handle most simple statistics, and served my purposes well. For each land use category, I applied the sum statistical tool to determine both total land area of that land use, as well as the total number of violations and total number of blocks. I could then make some probable assertions, such as which land use categories contained more case of violations.

For land use analysis, the process involved a simple sort (by land use ID) and summation the numbers for each category. I summarized land use into five categories. Population densities for all parcels were sorted and divided into five areas of equal count. For transportation issues, I selected areas within the various buffers which have 1/8 mile interval and exported them into separate tables. I tried to make a spatial relation between the major routes and the violation, and highway exit and the violation.

Document author : Yong Wook Kim
HTML by : Yong Wook Kim
Last modified: May 16, 1997


Methodology

East St. Louis Action Research Project
Refer to the ESLARP mailing address on the ESLARP home page