2007 Submission

Jay Kirby
Department of Geography

1) Background
In August 2006, as a student in the Geology 390 course, I embarked on an eleven day field trip through the Canadian Cordillera to observe and study the geology, and the five geomorphological terranes of the Insular Belt, the Coast Belt, the Intermontane Belt, the Omeneca Belt, and the Foreland Belt. The trip covered more than 4,000 km of southern British Columbia and Alberta. A final project for the course was originally to be a report on our observations. However, our instructor Steve Earle approached the students if any of us would like to create a geoscape poster of the trip. I had already taken the Geography 328 course–an introduction to GIS, and had discovered a new career path. With ambitions to enrol in the ADGIS, and having some basic GIS skills, I decided to take on the project. This project would solidify my decision to become a GIS professional. One major benefit to this project was that I was actively involved in the field trip, and was able to incorporate my photographs from the trip.

2) Design
The idea behind the design of the geoscape poster was to incorporate a base map of the Canadian Cordillera. I had several ideas including the use of a Digital Elevation Model (DEM). However, due to the limited time frame, and to my limited experience, I decided to use satellite imagery. This step of obtaining the correct satellite imagery was one of the many labour intensive stages of designing the poster. After a week of research I found suitable data on www.landcover.org (Global Land Cover Facility). Once the suitable base map data was obtained I had a better idea of what the output could look like. I wanted to incorporate the four geomorphological belts, the most significant geological sites we studied, and the pictures that correlate to the sites. The focus of the poster is the base map with picture and text insets around the perimeter of the poster. I wanted the poster to be as large as possible due to the fourteen most significant geologic stops we made. Many of the pictures I took were panoramic scenes, which require more space on the poster. I decided that 48"x36" would be a sufficient size, and would be compatible with our local print shops. One aspect of the design that I did not want was a cluttered or confusing poster. I wanted it to be easy to read, and I did not want to obstruct the appearance of the satellite imagery. I decided to use symbols instead of lines to match the insets to the stops along the route.

3) Data
The focal point of the poster is the Landsat7 ETM+ images from www.landcover.org. I merged seven Landsat7 images nearly seamlessly to create the mosaic. Using the rotate tool I was able to eliminate most of the distortion created by Earth’s curvature. I also decided on a scale of 1:700,000, which would display as much detail as possible without adding unwanted distortion when the images were merged. The Landsat images were geo–referenced using WGS 1984. Digitizing the geomorphological belts was difficult as I had to refer to numerous sources including www.MapPlace.ca.  I referred to Google Earth to guide the digitizing of the US/Canada border, and for the correct scale. Google Earth allowed me to pinpoint landmarks as a guide. All of the photos were taken by me, which allowed me to easily correlate the photos with the stops. The text for each inset was obtained from my field notes, and numerous geological sources. The textbook Roadside Geology of BC was the greatest source for cross–referencing my notes.

4) Personal Experience with the Project
Using ArcView 9.0, I had to work with seven data frames, which made the project very labour intensive. The most difficult task was digitizing the route, and the geomorphological belts. I had to digitize one data frame at a time while constantly switching between layout and data frame views to visualize the accuracy of the polylines. Not only were the data frames merged, but where all seven data frames joined I had to merge the polylines so there was no overshoot or undershoot. This process was painstakingly slow. One stressful problem with this project was making sure I did not accidentally move the data frames. I could not lock down the Landsat images, I could only lock the scale as to not zoom in or zoom out. All of the photographs were meticulously enhanced, cropped, resized, and where applicable labelled. All of the panoramic photos were merged and enhanced using Adobe Photoshop. Steve Earl proofed the text for geologic accuracy, and Tim Naegele edited for grammar, and punctuation. Once the poster was completed I was ready to take it to print. However, a major problem occurred when I went to save the poster to disc. The graphic output of the Landsat images created a spiking effect using all file formats. ArcView was not capturing the seven Landsat images. This was probably caused by the original images. It appears that the original images were snapped over top a frame, and when I rotated the images, ArcView only rotated the image, and not the entire frame. What I was left with was a triangular spiking effect. Tim Naegele and I decided to perform 24 screen captures, and crop each image using Adobe Photoshop 24 times. Then I pasted each screen capture into Paint, realigned them, and finally had a complete poster, which I could save as a BMP, JPG, and TIFF. This project was the largest single project I have ever encountered, and took nearly 3 months and well over 200 hours in labour. I am very proud of the output, and I am especially proud that all the photographs were taken by myself.