For this lab, I used a geographic information system to properly reference spatial data that was misaligned (not all of the layers of data conformed to the same coordinate system). When sorting through different layers of data for GIS use, it is likely that not all of the files will contain the same coordinate system information. When spatial data files are uploaded into ArcGIS, they are automatically displayed as if each data layer has the same coordinate system, so visually, they align. For example, in my lab, the spatial data consisted of a satellite image of North America, a shapefile of Canada, the major cities of Canada, Canada’s lakes, rivers, and national parks. Corresponding to each of these “layers” (the image is a layer, the cities are a layer etc.) is a specific coordinate system that is subsequently used to project the data and visualize it on ArcGIS. When I added each layer to the dataframe (the interface that allows you to view the layers simultaneously), the system “projected on-the-fly” so that despite the difference in projection, I was still able to view cities in their relative location to lakes, provincial boundaries, and rivers. In order to accurately work with the data, however, it is necessary to change the coordinate systems of each layer so that they match the coordinate system of the entire data frame. Sometimes this involves using tools from ArcToolbox to modify the spatial data itself, giving it an entirely new coordinate system altogether. I took the rivers layer and used the “Project” tool so that it aligned with the rest of the data.
For accuracy purposes in spatial analysis, it is important to use the right projection. In this lab I measured the distance between Halifax and Vancouver using two different coordinate systems for the cities data layer. Using the 1983 Geographic Coordinate System, the distance was 4,441.73 km, and using the Canadian Lambert Conformal Conic projection, it was 4,446.53 km. The “minor” difference of 4.79 metres may not seem like a large margin of error, but for large scale (zoomed-in) mapping and surveying, it matters quite a bit.
Secondly, I used Landsat images of Mt. St. Helens before and after the 1980 volcanic eruption to compare the physiography of the region in 1979 vs. 2002. Landsat imagery is very useful when comparing regions before and after natural disasters, flooding, and other large-scale events change the physical features of the earth. For example, Landsat data would be helpful for the geographic analysis of New Orleans before and after Hurricane Katrina in August/September of 2005. A potential research question is “which neighbourhoods in the New Orleans – Metairie metropolitan area were most affected by flooding?” The geographic location would focus on the Mississippi Delta and Lake Ponchartrain, specifically the urban centre of New Orleans. The analysis would outline the importance of the 17th Street Canal and the failure of its east dike which broke and allowed lakewater to spill into low-lying neighbourhoods, as from comparison of the Aug. 22, 2005 image and the Sept. 7, 2005 image (16 days apart), the darker areas in the Sept. 7 image represent flooding, which affected nearly all of the northern city, eastern city, St. Bernard’s Parish and the Lower 9th Ward. Further imagery from October 2005 shows lighter pink/grey raster images of the city, which represent the upheaval of buildings and entire districts due to water damage. Clearly, the time interval would begin from the first image before the storm (Aug. 22) to images after the storm which monitor the governments’ rebuilding efforts. The spatial analysis of environmental change in New Orleans, given by the US Geological Survey, can be found at https://earthshots.usgs.gov/earthshots/node/83#ad-image-11. In this instance, Landsat images that monitor reconstruction should be selected from spring or fall when New Orleans on average receives a relatively low amount of precipitation and cloud coverage.
Other examples of where Landsat data would be useful would be for spatial analyses of deforestation, glacier retreats, environmental impact assessments of hydroelectric dams, and city growth.
Accomplishments:
- Repaired vector files by using “project” and “tranform” tools, aligning all data to the correct projected and geographic coordinate system so they can be analyzed properly and accurately.
- Repaired satellite imagery by fixing RGB channels, resulting in a false-colour image analysis that successfully compared 1979 and 2002 Landsat imagery of the Mt. St. Helens eruption. Also resulted in a visual assessment of the damage done to a nearby waterbody.
Follow
