GEOG 337 LAB 2

Goals and Background

     The goal of lab 2 was to re-familiarize ourselves with georeferencing and data creation. Understanding these two basic concepts is essential to learn before diving into deeper processes of ArcGIS. Georeferencing is the act of assigning a coordinate system to data that has no spatial orientation in GIS. For part one of this lab, our class georeferenced a map created in 1878 of Eau Claire with a basemap and shapefile from more recent years (2017 and 2014). Part two focused on creating data through digitizing, which is also a skill that is used quite often in ArcMap for many reasons. It allows for new information to be collected so that a spatial issue can be resolved. In part two, we digitized the river and lake of each map to compare the polygons from each year and how they've changed over time.

Methods

           Part one was all about georeferencing, so we first had to add on each layer before georeferencing. I added the topographic basemap of Eau Claire before adding a road centerline shapefile and lastly the 1878 map. It is important to add these in a specific order so that the visibility of each could be seen properly. The 1878 map is going to be based off of the other maps, so it has to be laid over the other two to reference. This way, the transformation between the coordinate systems of each layer was properly done. For this lab, we wanted the coordinate system of the 1878 map to be transformed into the same system of the centerline and basemap data. Before georeferencing, control points have to be made in order to connect the reference raster to the new layer. Creating the control points was easy using the road centerlines and intersections to reference to the 1878 map. I created ~30 points to reference since many things had changed over the years, seen in Figure 1. Having this many control points allowed me to gather data across the entire dataset. Though I realize control points do not always create a more accurate map, this seemed like an appropriate amount to still receive a fairly low RMS error of 1.4. Though it is by no means perfect or insanely accurate, it would be seen as acceptable with combining the outdated map with the newer one. Since there is such a time gap between the two, there has been new infrastructure created and some of the old has been rebuilt. Additionally, a large part of the map surrounds the Chippewa River, which itself has changed a significant amount. This adds difficulty to creating a precise map. The last step was to choose a transformation to occur between the reference raster and the new layer, in which I chose a spline transformation. I chose this because the map was a local and small, which can prove to be the most accurate for a spline transformation. The spline transformation is useful with using a lot of control points, which I did. The final map for part one can be seen as Figure 2 in the results section. 

 
Figure 1. Screenshot of control points used. 

           Part two of the lab involved creating data through digitizing. We still needed both maps in the same coordinate system, which they were after georeferencing. I had to create two polygons on each of the layers, one representing Half Moon Lake and the other the Chippewa River. After creating these polygons, we could turn on both layers and compare how each feature has changed since 2017 and 1878. To create these feature classes, I first had to create a file geodatabase for them to be stored in. When creating a geodatabase you must ensure it is in the same coordinate system as its reference layer. After making the geodatabase, a new feature class can be created for a polygon. Then the fields must be defined, so that attributes can be stored. After turning off the transparencies for each layer, I outlined the lake and river for each map. Then I was able to compare and analyze the differences. 

Results

Figure 2. Map for part 1.

Figure 3. Map that shows the comparison between water features in Eau Claire.

      The river seems to be getting wider in a few areas but mostly thinner in others, while Half-moon Lake is expanding its total area. In most parts, the  Half-moon Lake had an overall general growth with almost every part of the lake wider. The shrinking river over the years was most notable in its big curves, specifically the one near campus. It seems that rocks and land were eventually deposited at the point bar of the river, making it smaller and thinner. By looking in the attribute table, I found the shape lengths and areas of the river and lake. In 1878, the river’s area was: 11,646,299 units while in 2018, the area is: 9,585,666. Half-moon Lake’s area has grown, with it being 5,681,031 in 1878, and 6,523,224 in 2018. A potential cause for my water features to be different from my classmates is human error during the digitizing. The points made to create the polygon can be extremely detailed or not at all, depending on how many were created. Another cause for error is that people georeferenced the 1878 map differently, making it more or less distorted than another classmates.

Sources

1.     David Rumsey Map Collection https://www.davidrumsey.com/luna/servlet/detail/RUMSEY~8~1~4181~480085#

2.     Master_Centerlines feature class (clipped), Eau Claire County

3.     World Topographic Map, ESRI, 2018: https://www.arcgis.com/home/item.html?id=30e5fe3149c34df1ba922e6f5bbf808f