Author: evincent@uoregon.edu

[embeddoc url=”https://blogs.uoregon.edu/giswinter2016/files/2016/03/Eugene_Schools-15cbc8k.pdf” download=”all” viewer=”google”][embeddoc url=”https://blogs.uoregon.edu/giswinter2016/files/2016/03/Demographics-2dkspe9.pdf” download=”all” viewer=”google”][embeddoc url=”https://blogs.uoregon.edu/giswinter2016/files/2016/03/Part3FINAL-18035g5.pdf” download=”all” viewer=”google”][embeddoc url=”https://blogs.uoregon.edu/giswinter2016/files/2016/03/GEOG-482-Final-Lab2-276gzi3.pdf” download=”all” viewer=”google”]

[embeddoc url=”https://blogs.uoregon.edu/giswinter2016/files/2016/02/GEOG-482-Lab-4-1uo0upj.pdf” download=”all” viewer=”google”]

1. The DEM layer shows elevation of the area of interest. The cell size is 32.79768284, 32.79768284. The cell size is important because helps identify x and y coordinates to locate information.
2. It is helpful to reclassify this way because water is either there or it isn’t. There is no intermediate value for the presence of a lake or a stream. If an area has water present (value of 100), the resulting cost surface will be higher in that area. If there is no water present (value 0) the cost surface will be less.
3. Land cover could also be used to calculate cost distance. Including this information would create a more thorough and accurate representation of what is actually there than the slope/water cost surface we have generated. Adding terrain type or land cover to the analysis would make the representation more realistic.
4. The attribute table for the Observer Points raster layer tells us whether the area the mushroom hunter may be in is visible from the lookouts. It gives this information as binary 1 and 0 for yes or no. Cells assigned a 1 are visible and cells assigned a 0 are not visible.
5. I recommend the forest service use lookout 1 for their search. More area is visible from lookout 1, including the known planned final destination of the mushroom hunter.

6.
Looking at the basemap, I would prioritize searching in the area between the start point and Nash Lake that have lower elevations and easy access to water. Since the mushroom hunter is lost, they are likely looking for water and staying low. On the other hand, it may be beneficial to search at higher elevations if the mushroom hunter decided to go to higher ground for a potentially better view of where they are. In reference to the basemap, this would be in areas just southwest of the startpoint or closer to Nash Lake and the immediate surrounding area.
The least cost distance map shows the level of difficulty to cross or maneuver an area with lighter hues representing an easier level of difficulty and darker hues representing more difficulty. This layer was generated using slope and water body information. In creating this layer, I encountered an error in reclassifying the slope and too many classes were the result, making my map of the cost distance more difficult to interpret. Instead of showing clear areas with different difficulty levels, mine shows a gradual gradient of difficulty, which makes it more difficult to determine which areas are harder to maneuver than others. Unfortunately, because of this my map wouldn’t be as helpful in finding the lost mushroom hunter.
The least cost distance map also includes the least cost path. This would be an appropriate area to search for the hunter if it is determined that they chose to remain at lower elevation. Adding a buffer to the path would help in searching for the hunter as well, in order to broaden the extent of the search with some direction and constraints. This would make searching the wilderness less daunting for the forest service while being able to search a large area.
Finally, the map showing the lookouts and the visible areas from them shows that the search party should look in the areas that are not visible by either of the lookouts. This would be the area west of the start point and not immediately around Nash Lake. This approach coincides with the leas cost path, as it is an area with relatively lower elevation. However, if somebody sees something of interest from Lookout 1, the searchers should look in the area surrounding Nash Lake, since that is the only area visible from Lookout 1. The searchers should use Lookout 1 for their search almost exclusively at first. If they are unsuccessful in finding the hunter from there, they should move their looking to Lookout 2 to cover a larger area that is of lower elevation.
While this method of evaluating the level of difficulty involved in maneuvering an area is useful to an extent, there are other factors that would make a better analysis. For example, including land cover, vegetation type, and terrain type would provide a much more extensive and perhaps accurate representation of “cost”. Another limitation of this method is the way water bodies are dealt with. In this analysis, water was assigned a class system that simply indicates whether or not it is present. Depth and dimensions of the water bodies were not taken into account in calculating the cost. This would be difficult to do, since water bodies change over time and are not necessarily discrete in their nature. This kind of information would be very helpful in calculating a more accurate cost if it were possible to obtain.

[embeddoc url=”https://blogs.uoregon.edu/giswinter2016/files/2016/02/Lab-3-1rnpoxn.pdf” download=”all” viewer=”google”]

Areas of safety concern are the yellow-colored walk ways. Areas of high safety concern based on the lack of accessibility to a callbox and low streetlamp coverage AND the occurrence of crimes is within the yellow circles.

1. The datum matters because the coordinate system must be consistent in order for the data to be compatible and meaningful. The coordinate system must be consistent in order to compare data sets and use them for analysis.
2. The coordinate system for the campus data is “+proj=lcc +lat_1=42.33333333333334 +lat_2=44 +lat_0=41.66666666666666 +lon_0=-120.5 +x_0=1500000 +y_0=0 +ellps=GRS80 +towgs84=0,0,0,0,0,0,0 +units=ft +no_defs”. The GPS coordinate system is latitude-longitude.
3. For the call boxes I used a 60 feet buffer distance. This is twice the buffer distance of the lights buffer distance. I chose to make the buffer for the callboxes larger than the lights buffers because I personally think that the blue lights on the callboxes are visible from a greater distance and easy to spot in the dark. If someone needed to get to one, they would definitely be able to identify it and run to it in a reasonable amount of time from 60 feet away, even if they were injured.
4. To identify areas outside of the callbox and lights buffer, I used the intersect tool with the no access to call box layer as the input and the no lights layer as the intersect, making a no lights or callbox layer as the output. This created a layer that shows portions of the sidewalks that are not covered by the light buffer or the emergency call box layer.
5. Areas that seem most adequately covered by street lights and emergency call boxes are areas around buildings and along streets. Most parking lots are adequately covered, as well as areas by the dorms and the education building. More lights could be used on the quad in front of Lillis, and near the Knight library and Gerlinger Hall.
6. Some issues with the crime data set include the different kinds of theft that won’t be represented in the map (for example, Theft 1, Theft 3, Theft 3, Bicycle Theft, etc). Also, there were some crimes identified as “assault” which I was unsure whether were a type of harassment. For this problem, I decided to only include crimes that actually were described using the word “harassment” or “theft”.
7. A more complete report would include the nature of the crimes and perhaps an analysis on patterns in locations of crime. It would also include better information about the distances between callboxes and streetlamps with more research done to understand how far away from a callbox is too far to reach it in case of an emergency. It might also be useful to include data about how often emergency callboxes are used on a weekly, monthly, or yearly basis. This information could be used to determine how useful and important the callboxes are on campus.

Simply using buffers as areas of concern does not take into account terrain which has the potential to make accessing a callbox more difficult. In this particular analysis, the buffer distance for the emergency callboxes was estimated with little knowledge about from how far away they are visible, what kind of terrain they are surrounded by, and if there is some kind of other obstacle preventing access to them.

[embeddoc url=”https://blogs.uoregon.edu/giswinter2016/files/2016/01/GEOG-482-Lab-2-u7kmeq.pdf” download=”all” viewer=”google”]

1. The source tab explains the data type, where it came from, and some of its properties.
2. This layer is in GCS North American 1983 geographic coordinate system. Geographic coordinate systems use a grid on a sphere, describing absolute location but not good for measurements of for small scale uses. Projected coordinate systems are on flat, two-dimensional surfaces with constant lengths, angles, and areas.
3. The shape of the continental United States is distorted differently in each projection. The Albers Equal Area Conic is a projection that preserves area well and shows little distortion. Mercator projections are based on the cylindrical map projection. Mercator projections show reasonably true shapes and distances, with minimal distortion at the equator and increased distortion at the poles. The Plate Carree projection has less distortion at the poles than Mercator and is a cylindrical projection. Shape and area are most distorted in Plate Carree while distance and direction remain true. The Robinson projection has strong distortion at the poles. The longitude lines are concaved and the latitude lines are parallel and straight. The Robinson is best used to show the whole world at once and not for smaller scale uses.
4. The distances between cities seems to vary with different projections. For example, on the Plate Carree projection Philadelphia and Chicago seem to have a greater distance between them than on the Albers Equal Area Conic projection.
5. Albers Equal Area Conic projection distorts distance and size. Mercator distorts area and distance. Robinson distorts size, distance, and shape. Plate Carree distorts size, distance and shape.

Projection	Distance between Philadelphia and Chicago (meters)
Plate Carree	1,412,328.793957
Robinson	1,037,244.857367
Mercator	1,423,156.657775
Albers Equal Area Conic	1,067,745.692009

 

7. The data set I chose is from the US Energy Information Administration. It contains data about industrial CO2 emissions in the United States in millions of metric tons for the years 1980 to 2013 and is organized by state. For my maps, I chose to show emissions for the year 2012.
8. I used equal interval, quantile, and standard deviation classification systems. The equal interval makes it seem that there is little CO2 emitted in most states while Texas, Louisiana, and California emit the most, with Texas having the highest. It puts the rest of the US all together, even though emissions may vary significantly between them.

The quantile method gives a more even distributed look to CO2 emissions. It breaks it up more, but the section with the highest emissions shows a range of more than 150 while the other colors show ranges much smaller.

The standard deviation method may be not the best for this purpose, as there are outliers in the data that can greatly influence the mean and not provide an accurate depiction of a representative average.