Lab 32 - Final Project

Objective: Design and execute a remote sensing analysis using Google Earth Engine that addresses a real-world question.

What You'll Demonstrate

  • Apply remote sensing concepts to a real-world problem of your choosing
  • Select appropriate datasets and methods from the course
  • Create reproducible analysis workflows in Earth Engine
  • Communicate results through maps, figures, and written interpretation
  • Demonstrate mastery of GEE skills learned throughout the course

Why This Matters

The final project is your capstone experience:

  • Portfolio piece: Demonstrate your skills to future employers
  • Integration: Combine multiple techniques into a complete workflow
  • Independence: Practice making analysis decisions on your own
  • Communication: Present technical work to broader audiences

Before You Start

  • Prerequisites: Complete all prior labs and get your project proposal approved.
  • Estimated time: 120+ minutes (multiple work sessions recommended)
  • Materials: Earth Engine account, project plan, and reference sources.

Project Overview

Your project can be a research paper or an interactive app that demonstrates a complete workflow: problem framing, data selection, method, results, and discussion.

Key principles:

  • Clarity - Can someone else understand what you did and why?
  • Reproducibility - Can someone else run your code and get the same results?
  • Technical rigor - Are your methods appropriate for your question?

Requirements

Component Requirement
Topic Any domain covered in class (land cover, rivers, wildfires, health, agriculture, urban, etc.)
Data Use at least one satellite image collection; cite all sources
Methods Apply 2+ core methods (indices, reduction, classification, zonal stats, time series, etc.)
Outputs Maps/figures and a table of key metrics
Reproducibility Share a viewable Earth Engine script URL with documentation

Project Ideas

  • Change detection: Deforestation, urban expansion, or glacier retreat over 10+ years
  • Disaster response: Flood mapping, fire severity, or hurricane damage assessment
  • Agriculture: Crop classification, irrigation mapping, or drought monitoring
  • Health: Heat exposure, air quality proxies, or disease risk factors
  • Conservation: Wetland mapping, coral reef monitoring, or protected area analysis
  • Urban: Green space access, urban heat islands, or informal settlement growth

Deliverables

  1. Report (4-6 pages)
    • Title, author, and abstract (150-250 words)
    • Introduction and research question
    • Data and study area description
    • Methods (include key code snippets and workflow diagram)
    • Results (maps, charts, tables)
    • Discussion (interpretation, limitations, future work)
    • References (at least 3 peer-reviewed sources)
  2. Appendix: Earth Engine script URL(s) with viewer access and reproduction instructions
  3. Optional GEE App: Interactive application demonstrating your workflow

Grading Rubric (100 points)

Criterion Points Description
Problem Framing 15 Clear question with real-world relevance
Data Selection 15 Appropriate datasets, properly preprocessed
Methods 25 Correct application and clear justification
Results Quality 25 Clear maps, figures, and meaningful metrics
Reproducibility 10 Documented, runnable code with comments
Writing Quality 10 Clear prose, proper citations, professional format

Suggested Timeline

Week Milestone Activities
1 Topic Selection Choose topic, scout data, submit proposal
2 Prototype Develop initial analysis, create early figures
3 Full Analysis Complete methods, generate all results
4 Polish & Submit Write report, refine code, create app (optional)

Before Submitting, Ask Yourself

  1. Is my research question clearly stated and answerable with my data?
  2. Can someone else run my code and understand what it does?
  3. Do my maps and figures have proper titles, legends, and scale bars?
  4. Have I discussed limitations and potential improvements?
  5. Are all my data sources properly cited?

Common Mistakes to Avoid

  • Scope too broad: Focus on one specific question rather than many
  • No preprocessing: Always apply cloud masking and quality filters
  • Missing context: Explain WHY your results matter, not just what they are
  • Uncommented code: Add comments explaining each major step
  • No error handling: Test with different AOIs and date ranges

?? Final Project Submission

Subject: Final Project - [Your Name]

Submit:

  • PDF report (4-6 pages)
  • Earth Engine script URL(s) with viewer access enabled
  • Link to GEE App (if created)
  • Any exported data files (CSV, GeoTIFF) as appendix