The learning objectives for this lab assignment include the following:

1. To learn about the issues related to performing a change detection analysis on multi-date multi-spectral satellite imagery.
2. To design and implement a small research project, research information on how to address the assigned problem, and to apply a customized solution.
3. To bring together several of the concepts that you have learned throughout this course in a single assignment.

I assume that, at this point in the course, you have a good working knowledge of the WhiteboxTools and general image processing techniques. Therefore, this final lab is more free-form than the previous assignments. You will be given some background information on what the general topics and requirements for the lab are, but it will be up to you as to how to carry out the analysis. One of the fundamental skills that you will need to take from this course is how to search for and find the information that you need to solve a remote-sensing problem.

In this assignment, you are to complete a land-use/land-cover change detection analysis between July 15, 2013 and June 11, 2018. You have been provided with Landsat 8 imagery for both of these dates on the CourseLink site. These data comprise of the first 7 bands of a Landsat 8 scene (actually a clipped sub-scene of the path 18, row 30 image) that covers part of the Greater Toronto Area (GTA), Guelph, Cambridge, and Kitchener-Waterloo. Note that the image metadata, providing details of the image acquisition, is also distributed along with the raw imagery. This is the same area (but different images) that we have been using for the lab assignments throughout the semester. As such, you should have a good level of familiarity with the data and the study area.

For this assignment, you are a remote sensing consultant, hired by me. Don't get too excited, I’m hiring you to work gratis! I am an urban planner tasked with developing a plan for Ontario's growing population over the next decade. As such, I am interested in documenting and evaluating the urban expansion that has taken place in the GTA and surroundings over the recent past, such that I may work with municipalities to plan for the anticipated growth to their communities. Information about the previous development in the region will allow me to better characterize the capacity for growth in each community. Therefore, I am interested in documenting the recent development in the GTA and surrounding communities and have contacted you, a remote-sensing specialist working for a private consulting company, to aid with documenting urban expansion over the 5-year period between image acquisition dates. I am particularly interested in documenting which areas of the region have experienced the greatest changes (urban expansion into surrounding rural areas) and documenting what type of urban development these past changes have been associated with (e.g. industrial growth, suburban development, etc.). My primary concern is document expansion of residential neighbourhoods. To answer these questions you will have to perform automated change detection analysis on the provided satellite imagery. It is up to you to research how to go about conducting this analysis but I have provided you with a short list of resources to help guide you. You should supplement this list with any additional resources on change detection analysis that you can find during your research.

Prepare a short report of no more than 4 to 5 pages (not including maps and references) documenting the data and procedure you followed, as well as the results of your analysis. Your final report should include an introduction, a brief description of your methodology, the results of your analyses, your recommendations, and a references cited (see the Marking Format, below). Remember, this is a consultant's client report and we are expecting a level of professionalism. Marks will be assigned for originality, demonstration of sound remote sensing principles and practice, and innovation.

The methodology section of your final report should include a Python script(s) that documents your approaches to performing the change detection analysis.

You will need to download a fresh copy of the latest version of the WhiteboxTools library before you begin this assignment. Changes may have been made to the library since you completed Lab 4 and use of an older version will likely result in incorrect results. In addition, you will need to download the data associated with this lab assignment from the GEOG*3420 CourseLink site. These data, as usual, are quite large and you will need to consider data storage solutions (e.g. a dedicated USB memory stick for the course).

When you download the data for this lab exercise, notice that a sub-directory called Change detection resources is also created. In this folder, you will find a number of resources to help guide you through your change detection analysis.

Jensen, J.R. (2016) Introductory Digital Image Processing, Chapter 12. This is a set of lecture slides associated with Jensen's excellent digital image processing textbook. It provides quite detailed information and contains a wealth of background material on the topic of change detection analysis.

Mas, J.F. (1999) Monitoring land-cover changes: a comparison of change detection techniques, International Journal of Remote Sensing, 20(1): 139-152. This is an excellent journal article comparing some of the common techniques for performing change detection analysis. I'd certainly recommend reading this source carefully.

You will also find the short section in your textbook (Chapter 6) provides some useful background information on change detection analysis:

Mather, P.M. and Koch, M. (2011) Computer Processing of Remotely-Sensed Images, Chapter 6 pp. 187-195.

And a short list of some other useful sites:

Natural Resources Canada (NRCan) (2013) Land Use Change (Rural / Urban), online resource: https://www.nrcan.gc.ca/earth-sciences/geomatics/satellite-imagery-air-photos/satellite-imagery-products/educational-resources/9357.

O'Neil-Dunne, J (2019) Change Detection Considerations, online resource: https://www.e-education.psu.edu/geog883/node/496.

The WhiteboxTools library contains several tools that can be useful in the workflow of performing change detection analysis on two-date, multi-spectral satellite imagery. Some of the more common tools include, but are not limited to:

• ChangeVectorAnalysis
• CrossTabulation
• Divide
• ImageRegression
• KMeansClustering
• ModifiedKMeansClustering
• NormalizedDifferenceIndex
• PrincipalComponentAnalysis
• Subtract
• WriteFunctionMemoryInsertion

Please refer to the WhiteboxTools User Manual for details on the operation of each of the tools.

The following grading scheme will be used to assess your report:

• Introduction (5 marks)

  • A brief introduction to the problem.

• Methods (20 marks)

  • Describe the data and the study site.
  • Description of the analysis approach.
  • Include the Python script that you used to help document your approach the the change detection analysis.
  • Be sure to use at least two different change detection approaches.
  • Describe the approaches that you used to address the problem. For example, what pre-processing techniques (e.g. radiometric calibration), if any, did you need use (and you will need to use some, believe me!)? Did you use pre- or post-classification approaches, or perhaps a combination of both? In other words, exactly what techniques did you select?
  • Be sure to provide justifications for the approaches that you chose.

• Results – Graphics (10 marks)

  • Results of analysis (images, maps, graphs, and/or tables where relevant).

• Results – Documentation (20 marks)

  • Interpretation of results.
  • Statement of limitations of your results (i.e. because of limitations of your methodology).

• General presentation (10 marks)

  • Neat and concise reporting
  • Spelling and grammar (be sure to proofread your work)
  • Organization (report is well laid-out using appropriate sections)
  • Consistent in-text citations and reference list.

The total is 65 marks.