GIS 5027 Module 2 - Land Use / Land Classification

 

This week's lab exercise consisted of analyzing an image of Pascagoula, Mississippi according to its land use / land classification. After the initial analysis, feature classes were created for each type of land classification [USGS Land Classifications, Levels I and II] and polygons were overlaid throughout the entirety of the aerial photograph according to the determined classifications. Once the entire map had been categorized, Google Maps Street View was used as a replacement for "in situ" ground truthing, and each sample point was visited using Google Earth. The accuracy for each of these sample points was determined as either correct or incorrect. As the map states above, the overall accuracy of my map was 70%. However, a certain percentage of my inaccurate classifications were due to the sample points falling in areas outside the confines of my minimum mapped unit, so I am pleased with this percentage rate. Overall, this exercise took a great deal of time and a great deal of decision making but was a good opportunity to learn the LU/LC process. I was content with the accuracy percentage, and I am happy with the aesthetic quality of the map itself. More information on the USGS Land classification system can be found in PDF form at the address listed below:

report.pdf (usgs.gov)

GIS 5027 Module 1 - Visual Interpretation

This week's lab assignment was the first of our Photo Interpretation and Remote Sensing class. This module focused on visual interpretation of aerial photographs that were provided by UWF [originally obtained from the United States Geological Survey]. The map above analyses an image according to various tones and textures, with a created feature class to distinguish the 5 classes of each.  The map below contains an image that identified objects according to [1] size/shape, [2] shadows, [3] patterns, and [4] associations of nearby objects. The images were then placed on layouts and the cartographic theories learned in GIS 5050 were applied to these created maps. The third exercise in the module was to compare the output of a true-color aerial photograph to the output of a false-color near infrared aerial photograph. No map was created for this exercise, but it was the most interesting portion [in my opinion]. Overall, Module 1 was very straightforward and I did not encounter any major issues while navigating through this lab assignment; I was also very happy with the quality of my maps, so I would say that this module was a success.


 

GIS 5050 - Final Project

 

For the Final Project of this course, we performed a Geospatial analysis of the proposed location of the Bobwhite-Manatee Transmission Line; this included analyzing its effect on environmentally sensitive lands in the area, the number of homes affected by the transmission line, and the number of schools in proximity to the transmission line. Finally, we attempted to calculate the length and the cost of this 230-volt electrical line that spans between two counties in western Florida [Manatee and Sarasota County]. The scope of this project was very extensive, covering almost all skillsets we have learned throughout the past eight weeks. 

I was happy with the outcome of ArcGIS Story Map [embedded above], but not entirely happy with my video presentation of this slideshow. I pondered re-doing the entire presentation, but then I gave myself a little grace and decided that I am my own worst critic and that I would probably critique the second rendition worse than the first. Overall, this was a great introduction to the science of Geographic Information Systems and a great opportunity to put my knowledge to the test. I enjoyed this project thoroughly and I hope you find it interesting as well. The video presentation [and transcript of the presentation] can be viewed at the following link:

GIS Final Presentation - Marc Wright

The ArcGIS Story Map can also be view without the presentation at the following link:

Bobwhite-Manatee Transmission Line ArcGIS Story Map

GIS 5050 Module 6 - 3D Scenic Map

The second portion of this lab assignment was to take a LIDAR dataset and convert it into a DEM [Digital Elevation Model]. This workflow is essentially creating a 3D terrain from information obtained from LIDAR equipment and importing it into ArcGIS Pro. Once this was accomplished, we were able to take the raster images and layers created in the Georeferencing portion of the lab and put them on this digital terrain. The map above was created using the DEM [as opposed to the digital terrains that come in ArcGIS Pro. I chose to use this because the topology of the land looked more accurate than the standard 3D basemap. This portion of the lab slowed my computer down immensely, so I had to be very intentional on the changes I made to the view / layout. I am extremely happy with the final product of this map. The aerial images overlaid very nicely on the DEM, and the orange / red colors used on the buildings / roads provides a nice contrast that draws the viewers eyes straight to the 3D terrain. This portion of the lab was very interesting, but working on this through a virtual desktop was a little frustrating at times [I believe ArcGIS spontaneously shut down four times].
 

GIS 5050 Module 6 - Georeferencing

 This week's lab assignment was based on Georeferencing. Georeferencing is basically taking an aerial [or raster] image and overlaying it on a basemap in ArcGIS. Once the image has been imported, points on the image can be selected and then matched up with control points on the basemap; this gives the aerial image coordinates where it lies on the Earth. No image is going to be perfect, so this lab assignment focused on Root Mean Square error which is a value that measures the distance between predicted calculated values and the actual, measured value. Generally speaking, the lower the RMSE the better. This is not always true [especially on images that have higher distortion levels], as was the case in the southern image. On this image, 2nd and 3rd Order Polynomial Transformations were used; these transformations bend / warp the image to fit the control points set in the Georeferencing process. During this process, a sacrifice was made on the RMSE value in order to get a more accurate image. After the Georeferencing process was complete, we then created features and added them to existing feature classes. In this case, I made a personal decision to export them to their own feature class so I could make them a different colors and label only those specific objects. Lastly, we created a multiple ring buffer around a Bald Eagle's nest that is located in proximity to the UWF campus. 

The final product of this lab is a map of the University of West Florida campus with three raster images Georeferenced and overlayed on the basemap. The buildings and roads [orange] were used as control points to get the best fit possible. The map also includes the UWF Gym and a portion of Campus Ln [which did not exist at the creation of the shapefiles] and the location of the Bald Eagle's nest with a 330ft and 660ft buffer ring encircling it. I chose this particular basemap because I felt that it provided much contrast against the raster images, but did not clash with the color scheme either. I went with an orange color as a warning of being within 660ft of the eagle's nest, and a red color as a warning of being within 330ft. I used these colors again for the buildings and roads to tie them into a common color scheme, and I used a bright purple color to make the digitized Gym and road [Campus Ln] to really stand out on the map. Overall, I am happy with the quality of the map, and I did not encounter any issues that prolonged the duration of this lab assignment.